Professional Documents
Culture Documents
Solar Energy Technology Handbook
Solar Energy Technology Handbook
Solar Energy Technology Handbook
ENERGY
TECHNOLOGY
HANDBOOK
ENERGY, POWER, AND ENVIRONMENT
Editor
PHILIP N. POWERS
Professor Emeritus of Nuclear Engineering
Purdue University
West Lafayette, Indiana
Consulting Editor
Energy Management and Conservation
PROFESSOR WILBUR MEIER, JR.
Head, School of Industrial Engineering
Purdue University
West Lafayette, Indiana
Editors
WILLIAM C. DICKINSON
PAUL N. CHEREMISINOFF
CRC Press
Taylor & Francis Group
Boca Raton London New York
This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to publish
reliable data and information, but the author and publisher cannot assume responsibility for the validity of all materials or the
consequences of their use. The authors and publishers have attempted to trace the copyright holders of all material reproduced in this
publication and apologize to copyright holders if permission to publish in this form has not been obtained. If any copyright material has
not been acknowledged please write and let us know so we may rectify in any future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form
by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording,
or in any information storage or retrieval system, without written permission from the publishers.
For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.
com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a
not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a
photocopy license by the CCC, a separate system of payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and
explanation without intent to infringe.
Publisher’s Note
The publisher has gone to great lengths to ensure the quality of this reprint but points out that some imperfections in the original copies
may be apparent.
Disclaimer
The publisher has made every effort to trace copyright holders and welcomes correspondence from those they have been unable to
contact.
Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the
CRC Press Web site at http://www.crcpress.com
The American Section of the International Solar Energy Society, Inc.
sponsors this Solar Energy Technology Handbook as a contribution to
the foundation of a most important segment of our future energy supply.
The society recognizes the concerted effort of numerous experts to pro-
vide basic information and supports the most valuable intent to present a
carefully assembled comprehensive handbook which undoubtedly will
become a permanent reference for many of us who are working in the
field of solar energy conversion as scientists, engineers, and users.
Karl W. 13ber
Chairman, Publications Committee
American Section of the International Solar Energy Society, Inc.
Headquarters: American Technological University
Killeen, Texas
Publications: University of Delaware
Newark, Delaware
PREFACE
Solar energy, as generally defined, includes energy derived directly from sunlight as well
as indirectly in the form of wind, waves, tides, ocean thermal gradients, or as fuel from
biomass and other photochemical reaction products. Over the past several years there
has been an explosive growth in solar energy research, development, and demonstration,
particularly in the United States.
In 1972 a solar energy panel, organized by the National Science Foundation and
the National Aeronautics and Space Administration, made the first comprehensive
assessment of the potential of solar energy as a national energy resource.* They also
examined the state of the technology in the various solar energy application areas. The
total U.S. budget for solar R&D in that year was $1.7 million. In 1979 the annual U.S.
solar budget had increased to $550 million and is expected to be more than $700 million
in 1980.
This tremendous increase in government funding, not only in this country but in
several other countries, has resulted in a proliferation of new ideas and concepts as well
as a large increase in available information and data in all of the solar technologies.
Hence there is a need for a comprehensive handbook describing the present state of
knowledge and offering the best available information and data in each solar energy
technology. It is hoped that this Solar Energy Technology Handbook will fulfill this
requirement.
Although there may, indeed, be "nothing new under the sun," it is highly probable
that in the coming years there will be technical and economic "breakthroughs" in almost
all of the solar technologies covered in this handbook. New materials and new measure-
ment techniques will be developed. There will be a continued advancement and refine-
ment of theoretical understanding. Although a serious effort has been made by each of
our contributing specialists to present the fundamentals of theory and experiment that
will have enduring value, this handbook can represent only the present state of knowl-
edge. The handbook is intended to supply the practicing engineer/scientist and student
with an authoritative reference work that covers the field of solar engineering as well as
peripherally related fields. References and primary citations are given to the extensive
solar literature for those who wish to dig deeper.
*Solar Energy as a National Energy Resource, Prepared by the NSF/NASA Solar Energy Panel,
December 1972. Co-chairmen: Dr. Paul Donovan and Mr. William Woodward; Executive Secretary:
Mr. William R. Cherry; Technical Coordinator: Dr. Frederick H. Morse; Executive Committee:
Dr. Lloyd 0. Herwig.
vi Preface
The handbook, for convenient use, is divided into eight main units: (1) The Solar
Resource ; (2) Solar Thermal Collectors; (3) Photovoltaics; (4) Bioconversion; (5) Wind
Energy; (6) Solar Energy Storage Systems; (7) Applications of Solar Energy; (8) Non-
technical Issues. In addition there are three Appendixes containing unit-conversion tables
and useful solar data. It became obvious early in this project that if proper coverage were
to be given each of these areas it would be necessary to divide the handbook into two
volumes. The first six units constitute Part A, Engineering Fundamentals and the last two
units constitute Part B, Applications, Systems Design, and Economics. These volumes
have been prepared primarily as reference books, but it is felt that many of the sections
will prove useful for practicing engineers, scientists, and students.
The subject of units has been a troublesome one in assembling this handbook. We
were tempted to take a purist approach and insist on the strict and exclusive usage of
SI units throughout. However, this did not seem practical or desirable. Since solar
energy is an applied engineering technology, the use of English units (feet, horsepower,
Btu, psig) is still deeply entrenched in the United States. However, the change to metric
units (meters, kilowatts, joules, pascals) is well underway in all technical areas. We have
attempted to soften the transition by asking our contributors to give the equivalent value
in English units parenthetically after the metric value, except for the simpler units where
metric is used alone. We do not claim 100 percent success in this effort, particularly in
some of the tables and graphs. To make life easier for confirmed users of either set of
units, a comprehensive set of conversion tables is included as Appendix A at the back of
each volume.
To ensure the highest degree of reliability the cooperation of a large number of
specialists has been necessary, and this handbook presents their efforts. Our heartfelt
thanks go to the 58 contributors, each of whom has endeavored to present an authorita•
tive and up-to-date overview of his/her area of solar expertise and has given willingly of
very valuable time and knowledge. The editors also wish to thank Marcel and Mau
Dekker, the publishers, and Graham Garratt, for their encouragement and constructive
suggestions.
William C. Dickinson
Paul N. Cheremisinoff
CONTRIBUTORS
vi
viii Contributors
Preface
Contributors vii
Contents of Part A xi
ix
x Contents
unit 3 Photovoltaics
Principles of Photovoltaic Conversion Charles E. Backus
Solar Cell Fabrication Martin Wolf
Photovoltaic Materials Joseph J. Loferski
unit 4 Bioconversion
Biological Energy Conversion: The Photosynthetic Process Raymond L. Ward
Energy Farming Robert Hodam and Virginia Lew
xi
xii Contents of Part A
Appendixes A-C
unit 7
SUBRATO CHANDRA
Florida Solar Energy Center
Cape Canaveral, Florida
Solar domestic hot water (DHW) heaters have had a long and varied history. In the
United States, solar water heaters were first used in the warm and sunny climates of
southern California and Florida. Interesting tales of the early commercialization of solar
energy are portrayed by Butti and Perlin [1] and Scott [2]. The following two para-
graphs are adapted from these two references, respectively.
In southern California, the solar water heater industry started in about 1891,
flourished during 1909-1920, and died around 1930. The early solar water heaters were
of the combined collector-storage type, in which shallow, blackened metal water tanks
are enclosed in pine boxes, covered with glass, and exposed to the sun. Hot water from
these units had to be used up during the day, otherwise the heat would be lost during the
night. In 1909, W. J. Bailey invented the Day and Night Solar Water Heater, which
separated the collector from the storage tank and which was very similar to the present-
day Thermosiphon solar water heater. These units sold well. By the end of World War I
about 4000 units had been sold, and sales peaked in 1920 with more than 1000 units
sold. The units sold for about $100 installed for a 40-ft2 collector and a 40-gal storage
tank (1910-1920 dollars). The units sold for purely economic reasons, as the backup
3
4 Applications of Solar Energy
natural gas cost about $3.20/MCF in 1900.* Solar water heater sales decreased in the late
1920s with the advent of automatic storage-type natural gas water heaters and cheap
natural gas. This history is told in an interesting illustrated article by Butti and Perlin [I].
In southern Florida, the early solar water heater industry flourished during the
1940s and 1950s in the Miami area. It is estimated that as many as 60,000 solar water
heaters may have been installed during this period. These solar water heaters were of the
thermosiphon type shown in Fig. 27.1. During the peak sales period of the late 1940s,
an 82-gal unit with a 48-ft2 collector cost about $350 installed, and the backup electric-
ity price was about 40/kWh [2]. The demise of the early solar water heater industry in
Florida can be attributed to the rising installed cost of solar water heaters, the declining
cost of electricity, and the convenience of automatic electric water heaters.
In today's age of renewed interest in solar energy, solar domestic hot water heaters
are again prominent in the sales of solar systems. This is to be expected, since solar
domestic water heaters, when compared to solar space heating and cooling systems, (a)
have maximum utilization due to the near-constant hot water load in residences through-
out the year, (b) are the simplest mechanically, (c) are the lowest in first cost, (d) are
retrofittable to existing houses and, (c) are most economical because of the above reasons.
Present-day solar water heaters cost between $1000 and $3000 (1980 dollars) installed
depending on the collector size (4 to 8 m2 ) and the system type. Direct water heaters
with minimal freeze-protection devices suitable for mild climates cost between $1000 and
$2000 and closed-loop antifreeze systems with heat exchangers cost between $2000 and
$3000. There is a wide variety of system types and configurations marketed by a large
(more than 100) number of companies today. The market potential is excellent, because
domestic hot water accounts for about 15 percent of U.S. residential energy use. A
recent survey [3] shows that more than 2000 solar water heaters were installed in Florida
residences alone in the first 6 months of 1977. Fifty-nine percent of the installations
were retrofit and 41 percent new. Average installed cost was $375/m 2 for residential and
$300/m2 for commercial installations. Solar water heaters are popular worldwide today,
particularly in Israel, Australia, Japan, and South Africa.
As of 1978, a wide variety of solar DHW systems are being sold. The proliferation of
system types is due to the necessity of tailoring system configurations to climatic regions
as well as to the fact that the reborn solar industry is still not mature. In the material
that follows, the more common types of solar DHW systems are described and their oper-
ating principles examined. Freeze proofing solar systems is very important, as a single
freeze may destroy the costly collectors. Indeed, the variety of system configurations is
due mainly to the different ways the systems are protected against freezing. Two excel-
lent references, one by McCabe et al. [4] and by ITT design manual [5], should be
consulted for details of system hardware integration and various other topics.
The primitive combined collector-storage type of solar water heater is rarely used
today and will not be discussed further. The systems used today have the collectors and
storage tanks as separate units interconnected by plumbing. More than 90 percent of the
solar DHW heaters currently being installed in the United States are of the "pumped"
variety, in which the collector fluid is circulated by a small electric circulating pump
operated by a controller and the storage tank is installed within the house to protect it
from freezing. A wide variety of pumped systems are described later. The other type of
solar water heater is the thermosiphon water heater, which does not use a pump or a
controller.
The following system types will be described:
A. Thermosiphon System
Figure 27.1 shows a typical thermosiphon solar water heater. The entire system is under
city water pressure, and no pumps or controllers are required for system operation. The
solar preheat tank is located above the collectors. In the morning as the sun heats the
collectors, the hot water inside rises by natural convection and the colder storage-tank
water flows into the collectors by gravity. Thus the circulation loop is automatically
established whenever there is sufficient sunshine, and circulation automatically stops
during insufficient insolation when the upward buoyancy force is unable to overcome
the fluid friction losses inside the plumbing. In order to prevent reverse circulation at
night, the tank should be located above the collectors. During bright sunshine, collection
flow rates between 40 and 601/h • m2 (1 and 1.5 g/h • ft2 ) of collector can be achieved
in a well-designed system. The plumbing must continuously slope upward with smooth
bends to prevent air pockets which, if formed, can cause flow stoppage. To minimize
fluid friction losses, the collector piping must be at least 1.9 cm (3/4 in.) I.D.
As shown, the cold water enters the bottom of the preheat tank and the collector
return is located at the top of the tank to promote stratification and collection efficiency.
An auxiliary heater provides hot water during inclement weather. A mixing valve (M),
as shown, keeps the household hot water supply at the desired temperature by mixing
cold water with the auxiliary tank water if it is too hot. The check valve (C) on the cold
water inlet is necessary to satisfy building codes and prevents backflow to the city mains
if the system fails. The contents of this paragraph on stratification, mixing valve, and
backflow preventer are applicable to all the system configurations to be described later,
as can be seen in Figs. 27.2 through 27.8.
It is inconvenient to freezeproof thermosiphon systems. The system shown can be
drained manually only by closing the collector isolation valves V and opening the
draincocks D. It is also possible to employ a heat exchanger inside the preheat tank and
to use antifreeze in the collectors to permanently freezeproof the system.
6 Applications of Solar Energy
KEY
PUMP
VALVE
ELECTRICALLY CONTROLLED—
VALVE 113"min
MIXING VALVE
C.W.
PRESSURE RELIEF VALVE —
H.W.
VENT 4
TEMPERATURE SENSOR
MANUAL DRAINCOCKS
Figure 27.1 A thermosiphon solar water heater. * Denotes that either auxiliary energy
source is acceptable. (From NBSIR 77-1272.)
B. Pumped Systems
The remainder of this subsection discusses pumped solar DHW systems. Some general re-
marks on pumped systems are made before proceeding to individual system descriptions.
All pumped systems have at least four basic solar components, the collector(s), the
storage tank(s), the control system, and the circulation pump(s). See Fig. 27.2 for an
example. In pumped systems the absorbed energy in the collector is transferred to the
storage tank by forced circulation of the collector fluid by a pump. The pump is con-
trolled (turned on and off) by an electronic differential controller. The controller has
two sensors, one mounted at the collector outlet and the other near the bottom of the
tank. In the morning, as the sun heats up the stagnant collector fluid to a temperature
about 6 to 10°C above the storage tank bottom temperature, the controller turns the
pump on. This 6 to 10°C differential temperature needed to activate the pump is called
the turn-on AT. As the circulation is established, AT between collector and storage drops
to about 3 to 5°C. If the sunshine is insufficient to maintain a AT of 0.5 to 1.5°C (the
turn-off AT), the controller stops the pump. The turn-on and turn-off AT's can be
adjusted on some controllers. So far the above discussion assumes fixed flow-rate pumps,
which are the most widely used pumps for solar DHW systems. Recently some have
started using proportional flow controls which vary the pump speed to collect a greater
amount of energy. See Ref. 6 for a discussion of proportional flow controllers.
Another control scheme uses thermostatic snap switches. The pump is turned on
whenever the collector switch senses a temperature of 55 to 60°C and turned off when
the temperature drops below 45°C. To prevent hot tank water from circulating through
the collector, another snap switch mounted near the tank bottom prevents circulation if
the tank bottom temperature is greater than 45°C. Although less efficient, snap switch
controllers may be more reliable.
In order to maximize collector efficiency, one needs to supply the collectors with
the lowest possible fluid temperature. This is the reason for taking the collector supply
from the bottom of the tank in a direct water system as shown in Fig. 27.2. In order to
promote stratification, the collector return is connected to the top of the solar storage
tank if it does not contain an electric backup element. If it does, as in Fig. 27.2, the
collector return should be connected below the element. This small detail can improve
system performance by up to 40% by preventing backup electric use during mornings.
8 Applications of Solar Energy
It is also a good idea to have the backup element operated by a timer so that it never
comes on during the day. This will be even more important when time-of-day pricing
is introduced by the utilities.
The various pumped systems will now be discussed. The basic energy-collection
mechanism is similar to all as discussed above. The differences arise in the various ways
the collected heat is transferred to the hot water tap and in the different freeze-protection
methods. There are two principal types, the direct heating systems (Figs. 27.2 through
27.4), in which the tap water is circulated through the collectors via a storage tank, and
the indirect heating systems, in which a different fluid is employed in the collectors
(Figs. 27.5 through 27.8). Direct systems are more efficient and more suitable for milder
climates with, say, less than 45 freezing nights per year. Indirect systems, although about
10 to 15 percent less efficient, are better protected against freezing and thus more suit-
able for colder climates.
Figure 27.2 shows the simplest direct solar water heater with a single tank. The entire
system is under city pressure. The single tank serves as both the solar storage and
auxiliary tank. This type of a system must have an electric backup. Fossil fuel backup
will keep the entire tank water at least at the water set-point temperature. This will
greatly reduce collector efficiency. The electric elements heat water above the element
and therefore the water at the tank bottom gets a chance to cool off. Freeze protection
in this system is achieved by automatic recirculation of hot tank water during cold nights.
As the ambient night temperature drops below about 1.5°C (35°F), a separate freezestat
sensor in the collector (or in some designs the same collector sensor can serve a dual
cN"
purpose with another lead) turns on the pump. A check valve located in the collector
return line prevents thermosiphon of the hot tank water through the cold collectors at
night. Check valves in solar systems are known to fail frequently. Therefore many
systems today use a solenoid-operated ball valve which opens only when the pump is on.
Figure 27.3 shows a two-tank design based on the same principles. Note that the
solar preheat tank does not contain any auxiliary, and this increases solar collection
efficiency. The auxiliary tank can be the existing house hot water tank for retrofit
situations. The preheat tank should be sized for the entire solar system. In other words,
the auxiliary tank storage volume should not be counted as partly meeting the storage
size requirements. This is because the auxiliary tank does not normally store any solar
heat. The solar-heated water goes to the auxiliary tank only when there is a draw from
the house hot water faucets. As before, hot water is recirculated on cold nights for freeze
protection. Two-tank systems are a necessity if conventional oil or gas water heaters are
to be used as a backup. For systems with electrical backup a one-tank system is prefer-
able due to its superior performance and reduced cost. The substantial tank heat losses
from two-tank systems decrease their performance when compared with one-tank systems.
Figure 27.4 shows an automatic draindown system very similar to that sketched in
Fig. 27.3 except that the collectors are automatically drained during cold nights by
thermostatically controlled electric solenoid valves. During cold nights or during power
failure, solenoid valve S1 shuts off, and S2 and S3 open to drain the collectors. In
solenoid valve terminology, S, is a "normally closed" and S2 S3 are "normally open"
valves. The word "normal" in solenoid valve terminology refers to the unenergized state
of the valve (e.g. when it is in a packing box). Further information on direct systems,
including installation procedures, may be found in Ref. 7.
10 Applications of Solar Energy
Indirect heaters employ a heat exchanger between the collector loop and the storage-
demand loop. They are freezeproof in all climates. Because of low pressure requirements
the collectors can be built more cheaply and copper tubes need not be used. (In direct
systems 1-cm or 1.25-cm I.D. copper lines are essential for corrosion resistance.) On the
other hand, the system performance is slightly poorer due to the heat exchanger perform-
ance penalty.
FILL
DOUBLE
WALL HEAT
EXCHANGER
Figure 27.5 shows a particularly interesting system in which the collector fluid is
still water but the water drains down into the tank whenever the pump stops. This
requires the collector return lines to be larger than the supply lines and vacuum breakers
for assured draining. A heat exchanger immersed in the preheat tank transfers heat into
the DHW loop. A variation that would reduce tank corrosion, etc., due to atmospheric
venting (in Fig. 27.5) would entail filling up the collector solar tank with an inert gas
(e.g., nitrogen) and making it a closed system.
A. Component Types
The hot water delivery temperatures required of solar DHW systems are moderate and
usually below 60°C (140°F). Therefore, flat-plate collectors are universally used, with
more than 95 percent of the systems using liquid collectors. Single-cover collectors are
used in moderate climates and either double-glazed or single-glazed with selectively
coated collectors are used in colder climates. Direct systems usually require copper
absorber tubing and plumbing. Indirect systems using nonpotable collector fluids can
use steel, aluminum, or other material with suitable corrosion inhibitors.
The storage tanks should be well insulated (R-9 or more)* to minimize standby
losses. It is convenient to have storage tanks with four plumbing ports, one each for
collector supply, collector return, cold water in, and hot water out. The tanks should
be corrosion resistant, with glass lining and sacrificial anode rods.
The circulation pumps should be magnetically coupled or have a stainless steel
casing or otherwise be corrosion resistant. They are normally of 1/20th or 1/12th HP
rating for domestic applications. Note that the pump merely needs to overcome the
fluid friction head for most applications.
The controls are usually of the on-off variety, with a high-temperature cutoff.
The relief valves on the tank and collectors prevent excessive pressure/temperature
buildup.
Solar system sizing and performance prediction is an involved process, usually requiring
the use of computers. The first step is to calculate the average daily hot water use.
*See Appendix C for definition of R value and values for common insulating materials.
Domestic Water Heating 13
Generally the demand in residences is more or less uniform throughout the year and
ranges between 50 and 100-1 (15 to 25 gal) per person per day. The higher figure is for
a household with an automatic clothes washer and dishwasher. The storage tank should
be sized to hold between one and two times the daily demand.
Usually solar DHW system annual efficiencies range between 30 and 50 percent for
a well-designed system, with smaller-area collector systems being more efficient because
of the increased storage-to-collector area ratio, which results in a cooler collector inlet
temperature.
The collector area should be optimized based on a combination of system perform-
ance and economic analysis. Based on the F-chart method, solar DHW system perform-
ance graphs for 11 eastern states is given in Ref. 4. Balcomb and co-workers provide
similar graphs for residences in the Pacific West region in Ref. 8 and for commercial
applications in Ref. 9. To give an estimate of system performance, consider an average
demand of 265-1(70 gal) per day of hot water at 60°C. A 6-m2 system will supply about
55 percent of the energy requirement in Boston and about 85 percent in Miami.
C. Costs
At the present time, system costs vary widely depending on system type. System hard-
ware costs range from between $700 and $800 for a small direct system to about $2000
for a larger indirect system. Installation costs range between $350 and $750 depending
on system type, local labor rates, and on whether it is a retrofit or new application.
Putting it differently, the current cost varies between $200 and $400/m2 of installed
collector area. Since the solar hardware is material intensive, it is unlikely that the
hardware costs will decrease more than about 15 percent in the future. However, as
system packaging improves and installation procedures are standardized, the installation
costs may be reduced by a factor of 2 or more.
Currently, solar DHW systems are economic in many areas of the country when
compared against electricity on a 20-year life-cycle basis. See the paper by Schulze et al.
[10] for an economic feasibility analysis of solar DHW heaters in every state of the
United States.
REFERENCES
1, K. Butti and J. Perlin, "Solar Water Heaters in California 1890-1930," The Co-
Evolution Quarterly, 4-13, (Fall 1977), published by the Whole Earth Catalog.
2. J. E. Scott, "The Solar Water Heater Industry in South Florida: History and
Projections," Journal of Solar Energy 18, no. 5 (1976).
3. R. Talwar et al., "New and Retrofit Solar Hot Water Installations in Florida,
Jan.-June, 1977," Florida Solar Energy Center Report No. EA-77-3, December
1977.
4. McCabe et al. Intermediate Standards for Solar Domestic Hot Water Systems/HUD
Initiative, NBSIR 77-1272, July 1977.
5. Solar Heating Systems Design Manual, Bulletin TESE-576, 1976, ITT, Fluid
Handling Division, Morton Grove, Ill., $2.50.
6. R. J. Schlesinger, "Preliminary Comparison of Proportional and Full On-Off
14 Applications of Solar Energy
Control Systems for Solar Energy Applications," pp. 9-15 to 9-18, Proceedings
ISES 1977 Annual Meeting, Orlando, Florida, June 1977.
7. S. Chandra et al., "Solar Water and Pool Heating Installation and Operation,"
FSEC, Cape Canaveral, Florida, 1979.
8. ERDA 's Pacific Regional Solar Heating Handbook, November 1976, San Francisco
Operations Office in conjunction with Los Alamos Scientific Laboratory (available
from Superintendent of Documents).
9. ERDA Facilities Solar Design Handbook, August 1977, ERDA Division of Con-
struction Planning and Support and Division of Solar Energy. Prepared by Los
Alamos Scientific Laboratory, ERDA 77-65 (available from NTIS, Springfield,
Va., $7.50).
10. W. D. Schulze, S. Ben David, and J. D. Balcomb, "The Economics of Solar Home
Heating," paper prepared for the Joint Economic Committee of the U.S. Congress,
January 1977 (available from Superintendent of Documents, $1.35).
chapter 28
RAJ TALWAR
Mid-American Solar Energy Complex
Minneapolis, Minnesota
28.1 Introduction 15
28.2 Technical Considerations in Solar Pool Heating 16
A. Heat Gains 17
B. Heat Losses 17
C. Overall Energy Balance 18
28.3 System Considerations for a Solar Pool Heater 19
A. How It Works 19
28.4 Sizing Considerations 24
A. Example 24
B. Effect of Pool Covers 25
C. Sizing for Plastic Pipes 29
28.5 General Information 30
A. Costs of Solar Pool Heaters 30
B. Installation Angles 30
C. Orientation 30
References 31
28.1 INTRODUCTION
Federal government programs and funds have been directed toward promoting the use of
solar energy for heating and cooling of buildings and providing domestic hot water.
Surprisingly, however, industry activity in terms of the manufacture and sale of solar
collectors has been largely in swimming pool heating. Thus, according to U.S. Depart-
ment of Energy (DOE) estimates of national solar industry activity during the first
6 months of 1977, approximately 320,000 m2 of collectors were used for pool heating
as compared with 65,000 m2 for domestic hot water and 75,000 m2 for space heating
and domestic hot water.
Despite this major industry activity in pool heating, there is little available informa-
tion on this application of solar energy. This chapter is designed to provide that assist-
ance, and it is oriented specifically toward commercially available systems. Since pool
15
16 Applications of Solar Energy
1. The desire to swim normally is phased with a warmer trend in the weather.
2. The preferred water temperature usually is only a few degrees above ambient
(surrounding) temperature.
3. The amount of water that must be heated and circulated through the solar
energy collector panels is relatively large, the recommended flow rate being
a circulation of the entire pool water through the filter about once every 8 to
12 h. This means a flow rate in excess of 110-1(30 gal) per minute for an
average 57,000-1(15,000-gal) residential pool, compared with an average flow
rate of 1.9-1(0.5 gal) per minute for domestic water heating.
4. The pool itself, if exposed to the sun, acts as a collector, since water absorbs
about 75 percent of the solar energy striking its surface.
5. The volume of water in the pool being relatively large, its temperature is
affected only slightly by transient changes in atmospheric weather conditions.
6. The need for a storage tank is eliminated, since the pool itself serves as the
storage.
The goal of any performance analysis of swimming pool solar heating is an estimate
of the pool's bulk temperature. Because of its large mass, the pool's water temperature
Swimming Pool Heating 17
usually changes little during a daily 24-h period (through a range of 2.5°C). Therefore,
the pool temperature can be estimated by calculating an energy balance for a typical day
for each month of the year. This may be done by equating the heat gains to the heat
losses from the pool surface at the equilibrium pool temperature. The reader is referred
to Ref. 2 for a more detailed discussion on pool heat gains and heat losses.
A. Heat Gains
Direct absorption of solar energy striking the pool surface may be expressed as aw Ap Fl,
where
aw = average absorptivity of water, 0.75
F1 = monthly averaged daily insolation values on a horizontal surface,
MJ/m2 • d (Btu/ft 2 • d)
A P = surface area of pool, m2
Heat supplied by the collector panel may be expressed as
where
B. Heat Losses
1. Long-wave infrared radiation exchange between the pool and the sky, R,
Wm-2 (Btu/ft2 • h). This exchange depends on the emissivity of water as well
as ambient weather conditions such as relative humidity, dew point, cloud
cover, and temperature. Based on a 0.9 value for emissivity of water and
national daily averaged weather conditions, Bliss [3] estimates this exchange
value to range from 69 Wm-2 (22 Btu/ft2 • h) in the summer to 88 Wm-2
(28 Btu/ft2 • h) in the winter.
18 Applications of Solar Energy
NAp [eR + hr (Twd - Tav) + he(Twd - Tav) + 0.0229 (Pw - Pa)] + vol. X pc
)] (28.2)
X (Twd - Tav) = aw Ap FI + Ae [FR - NeR - NUL (Tw - Tay
where vol. = volume of water, m3 ; p = density of water, 1000 kg/m3 ; c = specific heat
capacity, 4.184 J/kg °K. Equations (28.1) and (28.2) are programmed easily on an HP
or TI programmable calculator. Some results of the program are shown in Sec. 28.4,
and a step-by-step hand calculation showing the procedure has also been performed.
Swimming Pool Heating 19
A diagram of a typical solar pool heater is presented in Fig. 28.1. As shown, it is con-
nected by means of tees to the water line between the filter and the pool.
A. How It Works
When adequate sunshine is available, the filtered pool water is circulated through the
tubes in the collectors, where it is heated by the available solar radiation and then
returned to the pool.
Two collector designs, both commercially available, are in general use for pool
heating.
1. Metal tubes soldered to metal plates, both painted black, covered with plastic
or glass, insulated and housed in a box. This design is widely employed for
Bank of Collectors
House
Warm Water
Collector Bypass To Pool
• =ZS,..
Pump F,Iter •G,,, -..
.':,
Cool Water
From Pool 0 —.',,— - .-.—
_.-,-4,---<----
- ---7
.;,& •,,'''''.-
___.------. ..:;--
----_
_ ._,,, ....._,__ ,...''
.=,... ----- —"`"-
- ,,,=,,-• _ .—.,-=. .1----
_.--, .....- '''-...,--,.
-,— .--- ,...
Sunshine
Transparent Cover
Wale( In
Water Out
Tubing
Absorber
Header System
Insulation
Box
Drain Hole
Sunshine
Plastic Pipe
PVC /
Coupling
Continuous Plate
(a)
Water Out
Viatei In
(b)
10 to 11°C above ambient temperature and therefore, depending on the local climate
data, they may not be suitable for all-year heating of swimming pools. In all-year heating
applications, one might wish to choose an MTC. However, since the area of collector
panel required for winter heating is relatively large, the investment becomes prohibitively
high. An alternate solution, which combines LTCs with pool covers, is probably the
optimum method of providing all-year swimming (see Sec. 28.4).
Besides the bank of collectors, there are several other important components that
should be installed to assure proper functioning of the pool heater. These components
are shown in the schematic diagram of Fig. 28.5, and a brief discussion of each follows.
Check valve. When the pump is shut off, water may flow backwards through the
skimmer and/or filter drain. A check valve is recommended to prevent such backflow.
Gate valves. Gate valves 1 and 2 provide manual control over the pool heater.
When sunshine is available, gate valve 2 is closed and gate valves 1 and 3 are opened, thus
diverting the pool water through the collectors. At night or during cloudy days, when
the pump is on, gate valve 2 should be opened. Simultaneously, gate valve 1 should be
closed for ground-mounted panels and may be left open for roof installations. In
22 Applications of Solar Energy
1.0
0.9
0.8
0.7
0.6
EFFIC IENCY
0.3
0.2
0.1
0.0
0 10 20 30 70
addition, gate valves 1 and 3 also serve to isolate the collectors from the piping loop con-
necting the pool, filter, and pump. Isolation of the collectors may be necessary for
regular inspection and/or maintenance.
Vacuum breaker and drain cock. When the drain cock is opened, the air vent
vacuum breaker allows air to enter the solar panels, allowing them to be drained. Drain-
ing may be necessary to free the system of air bubbles or to prevent freezing during cold
weather. The vacuum breaker is typically installed at the highest point in the return
piping from the collector.
Pressure-relief valve. This valve protects the solar panels against excessive pressure
buildup, which might occur under conditions of water stagnation, blockages from equip-
ment malfunction, or improper maintenance procedures.
Piping. The recommended piping is PVC Schedule 40, 1.5-in. to 2-in. nominal
diameter, depending on existing pool plumbing.
Optional pump. The decision to install an additional pump depends entirely on the
added friction loss created by introducing the collector panel loop into an existing system.
Swimming Pool Heating 23
Usually, the average pool installation of 57,000-1(15,000 gal) is equipped with a 3/4-hp
pump motor. This pump supplies the pressure that is required to overcome the total
pressure loss in the system at the recommended flow rate of the entire pool volume to
be recirculated once every 8 to 12 h. The principal pressure loss in the system is 70,000
to 100,000 Nm-2 (10 to 15 psi) drop through the filter, depending on the filter's condi-
tion, and a recommended practice is that the pressure drop through the collector panel
be small (30,000 to 35,000 Nm-2 or 4 to 5 psi) compared to the loss through the filter.
Pool owners are advised to consult with the manufacturer regarding the pressure loss
through the collector panel and whether the added loss may be large enough to warrant
installation of an additional pump.
Optional automatic control. As explained previously, gate valves 1 and 2 provide
manual control of the pool heating system. An optional automatic control is available
with most commercial systems. A typical arrangement is shown in schematic form in
Fig. 28.5. When adequate sunshine is available as determined by the temperature sensed
at sensor 2, exceeding that sensed at sensor 1 by a predetermined amount (usually 2 to
3°C), the normally open electric valve 1 is closed, and the pool water flows through the
collectors. When sunshine is not available the control box switches off and the electric
valve is opened, allowing the collectors to be bypassed, which is also the normal mode
of operation.
Bank of Collectors
Bank of Collectors
Vacuum
Sensor 2
Vacuum Breaker Breaker Pressure
Pressure Relief Relief
Valve Valve
Io Control Box
Drain Cock Gate Valve 3 Drain Cock
Pool
Gate Valve 3
From Pool To Pool
& Filter Gate Valve 1
From
Pool
Check
& Filter /
Valve Gate Valve 2
Sensor 1
Electric Valve
(a)
(b)
Figure 28.5 Schematic arrangement of a solar pool heater. (a) Manual control. (b)
Optional automatic control.
24 Applications of Solar Energy
The above discussion on components is with reference to Fig. 28.5, which shows
their arrangement in the schematic diagram of a solar pool heater. Many manufacturers
offer alternative arrangements and different designs. Details on both MTC and LTC
systems that are commercially available throughout the country may be obtained from
the National Solar Heating and Cooling Information Center, P.O. Box 1607, Rockville,
Maryland 20850.
Equations (28.1) and (28.2) are programmable relatively easily. Before they can be used
for sizing, several weather data parameters have to be determined. The numerical example
below will illustrate the step-by-step sizing procedure.
A. Example
For a 37-m2 , 60,000-1 pool located in Pensacola, Florida, determine the square meters of
bare collector panel required to assure a 24-h minimum of 24.5°C pool water temperature
during March through November.
1. Heat losses
(a) Long-wave infrared radiation
24A eR = 24 X 37 X 0.9 X 25 X 3.152 X 3600 = 226.7 MJ
(b) Radiation with surroundings
24Aphr (Tw - Ta) = 24 X 37 X 5.68 X 3600 (24.5 - 15) = 172.5 MJ
(c) Convection with surroundings
he = 5.68 + 3.78V = 5.68 + 3.78 3.415
(d) Evaporation
P„, = 4.148T,,2 - 11.358T,, + 878.49 = 3090.1 Nm -2
Evaporation loss = 24Ap X 0.0229 hc (Pw - Pa) = 24 X 37 X 0.0229
X 8.65 X 3600 X (3090.1 - 1379) = 1084 MJ
Total heat loss = 1747.6 MJ
2, Heat gains
(a) Direct pool absorption
cxw FlAp = 0.75 X 16.95 X 37 = 470.37 MJ
(b) Panel gain
Ac [FR (ce)fit — NeR - N UL (T„, - Tav)] = A,(0.9 X 18.96 X 106
- 11.04 X 0.9 X 25 X 3.152 X 3600 - 11.04 X 14.33 X 3600 X 5)
= 11.4Ac MJ
3. Energy balance
11.4Ac + 470.4 = 1747.6
Ac = 112.0 m2
1. Heat losses
(a) Infrared radiation
4 20
226.7 X — + 0.9 X — X 226.7 = 207.8 MJ
2424
26 Applications of Solar Energy
100
95
IN
90
85
80
75
70
65
60
55
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
15 15 15 15 15 5 15 15 15 15 15
MONTHS
LIMO
100
_ Frapil.."11ONE
gio INEOEFIIERE
95
I _MiliMMEIaNm
n IR _PMERMIN_QSNekimin
90
ERMI a.4.001141163111i.
0
Y‘,,0
wil-A \°°, =M NNI ST
85
NEW °° v
'N-ML
80
70 n
65
60
55
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
15 15 15 15 15 15 15 15 15 5 15
MONTHS
2. Heat gains
(a) Direct pool absorption
0.85 X 470.37 = 399.8 MJ
(b) Panel gain
11.4Ac MJ
3. Energy balance
11.4Ac + 399.8 = 801.8
Ac = 35.3 m2
28 Applications of Solar Energy
wira...
FROPME2 Mak
AVERAGEPOO LT EM PERATUR E DURI NG DA YLI GHT HOU RSIN D EG REES F
PERM
100
95 n
90 Al AMPIP—Ne...__
pp.P--
, , '0/
85
n '°r
, ,,
rav
80
75
1111PEORWA It
1111111WAWF
70
ErigrAMII
EMI%
65
60 id r n MIL
55
PPP i
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
15 15 15 15 15 15 15 15 15 5 15 15
MONTHS
It can be seen that the effect of covering the pool has been to decrease the collector
area required almost by a factor of 3. Hence it is essential that before a pool is solarized,
it must first be covered. An excellent discussion on pool covers is presented in Ref. 8.
Although it is recommended that a single, continuous pool cover be used, it might
present a handling problem in larger pools. In such cases, several modular units may be
the answer.
Continuous covers and modular units are commercially available at an average cost
of $5/m2 and normally have a lifetime of 2 or 3 years. The most commonly used cover
Swimming Pool Heating 29
material is transparent PVC (polyvinylchloride), which has a high solar radiation trans-
missivity. For the do-it-yourself pool owner, commercially available PVC film of 4 to
6 mils thickness is recommended as cover material. Since it is relatively inexpensive—
about $30/100 m2 —replacement cost is a minor consideration. Water-resistant tape
can be used to secure the film to several modular frames to cover the pool surface.
The inconvenience of covering and uncovering a pool is more than offset by the
significantly longer swimming season. Covers also help keep pool water clean, thereby
reducing the cost of chemicals and filter maintenance.
In order to conclude the section on sizing, information is presented below for converting
the panel area to an equivalent lineal meters of pipe using Fig. 28.9, which has been de-
rived from radiation shape factors obtained by Hottel [9]. This is done using Eq. (28.3):
12
I0
N
0.8
0
I-- 0.6
0.4
0.?
0 •
2 3 4 5 7 8
CENTER-TO-CENTER DISTANCE
TUBE DIAMETER 0
Figure 28.9 Factor of comparison relating plastic tube solar heaters to those consisting
of continuous plastic panels.
30 Applications of Solar Energy
AXN
L (28.3)
d(N - 1) + D
where L is lineal meters of pipe, A is "transformed collector area" in square meters, N is
the number of rows of pipes, d is the center-to-center distance between two rows of pipes
(meters), and D is the diameter of the pipe in meters. The "transformed collector area"
may be obtained from the panel area using Fig. 28.9. Knowing the ratio d/D, one obtains
"factor" and dividing the collector area by "factor" will give the "transformed collector
area."
The major cost component of a solar pool heating system is the collector panel. Typical
LTC costs range from $30 to $40/m2 of collector area. MTC costs are generally three to
four times higher. An automatic control system of the type described in Sec. 28.3 will
cost about $350. Gate valves cost $12 to $15, and an extra $75 should be budgeted for
the vacuum breaker, pressure-relief valve, draincock, and miscellaneous requirements.
A typical figure for installation costs for an average-sized pool will be about $400.
These cost figures are only estimates. Ideally, after going through the sizing proce-
dure, the reader should contact several manufacturers of pool heating systems to obtain
their installation prices. Knowing his or her own estimates, the reader should then be
better able to decide exactly what and how much can be obtained for his or her money.
B. Installation Angles
The tilt angle, that is, the angle the collector makes with the horizontal, should be close
to latitude. However, in certain instances, depending on the slope of the roof or personal
tastes, it may not be practial to install the collectors at their optimum tilt angle. Fortu-
nately, this does not become a critical factor, since as long as the tilt angle is within about
15° of the optimum, the overall effect on the heat collected is not significant.
C. Orientation
The orientation of the collectors normally should be due south. However, in many parts
of the country, summer afternoon thundershowers and early winter morning haze are
common occurrences. Three situations arise:
REFERENCES
1. R. Talwar, A Guide to Collector Sizing and System Design Considerations for Solar
Pool Heating in Florida, FSEC 77-9, Florida Solar Energy Center, October 1977.
2. F. de Winter, How to Design and Build a Solar Swimming Pool Heater, Copper
Development Association, 1975.
3. R. W. Bliss, Atmospheric Radiation Near the Surface of the Ground: A Summary for
Engineers, Solar Energy Laboratory, The University of Arizona, Tucson, Ariz.
4. F. Kreith, Principles of Heat Transfer, 3rd ed., Intext Educational Publishers, New
York and London, 1973, p. 283.
5. Solar Engineering Magazine, June 1977 issue.
6. Mean Solar Radiation for Florida Cities, FSEC 77-3, Florida Solar Energy Center,
April 1977.
7. A. Whillier, "Solar Radiation Graphs," Solar Energy 9, 164-165 (1965).
8. G. 0. G. Lof, "Performance of Solar Swimming Pool Heaters—Transparent Cover
Type," Proceedings of the 1977 Annual Meeting, American Section of the Inter-
national Solar Energy Society, Orlando, Fla., June 6-9,1977.
9. H. C. Hottel, "Radiant Heat Transmission," Mechanical Engineering July 1930,52,
no. 7,699-704 (July 1930).
chapter 29
29.1 Introduction 33
29.2 Building Design Considerations 35
A. Reduction of Building Thermal Load 35
B. Site and Climate 35
C. Passive Solar Elements 35
D. Design Approach to Optimal Solar System Cost
Versus Fuel Savings 36
E. Maintenance and Lifetime Considerations 37
29.3 Space Heating Methods 37
29.4 Solar Space Heating Systems 38
A. Liquid Heating Systems 39
B. Air Heating Systems 48
C. Solar-Heat Pump Combinations 60
29.5 System Performance Prediction 64
A. Solar System Characteristics and Performance Prediction 66
B. Simulation Methodology 66
C. Design Parameters, Liquid Systems 67
D. Design Parameters, Air Systems 83
29.6 Performance Results of Installed Systems 89
A. CSU I and CSU II at Colorado State University 89
B. NSRSC at the Los Alamos Scientific Laboratory 96
References 100
29.1 INTRODUCTION
The technology of space heating with solar energy has been developing rapidly during
the past 5 years. The fabrication and installation of factory-built hardware is a growth
industry driven primarily by the increasing scarcity of fossil fuels and their resulting
high cost. This solar industry growth has necessitated the development of system
standards issued by U.S. government agencies [1, 2] .
Because the solar radiation received on the roof of a residence during the heating
season is several times its winter heating requirement, a substantial portion of the energy
33
34 Applications of Solar Energy
needs may be met by solar energy. Space heating with solar collectors in which a liquid is
heated for transport to a hot water storage tank is the most widely used system. However,
solar collection using air as the transport fluid to rock-bed storage is becoming important.
Several thousand residential and commercial buildings in the United States have been
provided with one of these two types of solar heating systems.
Most of the heat needed in a residence must be supplied at night because of the
lower ambient temperatures usually prevailing during that period. The duration of night
hours is also greater than sunlit hours in winter, and there is no direct solar heating
through windows and other surfaces. It is therefore essential, for practical and economic
reasons, that solar heat collected in the daytime for residential use be stored for delivery
as needed at night. This generalization may not apply to certain commercial and
industrial buildings where there is little or no occupancy at night when lower interior
temperatures may be acceptable.
A primary consideration in the use of several components in the heating of
buildings is their integration into a functional system. It is therefore a purpose of this
chapter to show how these components can be effectively and economically combined
with the other elements in a space heating system.
The greatest need for the solar industry and system designers is the development
of cost-effective components and systems essential for the widespread use of solar heating.
This means improved performance at lower overall system cost (capital plus operating).
Systems analysis, which includes system simulation and testing against laboratory and/or
field measurements, is essential to this development and is emphasized in this chapter.
Using systems analysis, the system performance can be predicted for varying system
configurations, building types, load characteristics, climates, and competing energy costs.
Through such analysis, the sensitivity of each variation, in terms of both performance
and life-cycle cost, can be predicted. Performance can be mapped over the expected
operating ranges of the significant parameters.
Based on system simulation, and coupled with parametric studies, solar heating
system performance can be satisfactorily predicted for virtually any climate. Recently
developed simplified methods, which use correlations based on multiple simulations,
have made cost-optimal system design practical. Standardized methods for performance
optimization, which are becoming widely used, are also discussed in this chapter.
The simplest design tools are "rules of thumb." A number of these are discussed
in this chapter, but should be used with caution because of their approximate nature.
Rules of thumb are useful for preliminary estimates, but should be followed with more
detailed analysis.
In the presentation that follows, some general design considerations are presented,
followed by a description of solar heating systems and their operating characteristics.
Those aspects of component and system-type characteristics that affect system perfor-
mance and design are then discussed. System simulation, as a basis for performance
prediction and collector sizing, is presented next. Finally, operating experience gained
from installed heating systems is presented. Collector sizing is discussed in Chaps. 43
and 44.
Building Space Heating: Active Systems 35
Site and climate may strongly affect the performance or maintenance of a solar-heated
building and should be considered during the site selection and planning studies. Such
factors should be considered at the earliest conceptual planning stage to integrate solar
use concepts into the total building design.
Obscuration by buildings, trees, hills, or mountains should be studied to ascertain
when and how much solar radiation is lost. Until legal sun rights are established by
legislative or judicial processes, the designer should anticipate whether the site might
fall at some future time in the shadow of a nearby high-rise construction or a growth of
trees on adjoining land. Such considerations, as well as those of cost and technological
limitation, suggest that solar space heating installations are best suited for residential or
commercial buildings in rural and suburban areas, and not yet for densely populated,
high-rise, inner-city areas.
The natural topography and microclimate have significant effects on solar per-
formance. Prevailing winds, frost and drifting snow, site slope, and ground reflectivity
should be carefully considered. These items are discussed in Refs. 3 and 4 and elsewhere.
Thermal storage of solar radiation is required in all heating systems intended to provide
heat during night-time hours and periods of cloudiness. In stone, cement block,
concrete, or adobe structures, some useful thermal storage mass is already in the structure.
This mass is effective in smoothing out variations in inside temperature at locations
36 Applications of Solar Energy
where the ambient day-night temperature changes are large. Building mass is particularly
advantageous for storing solar energy that enters the structure through windows. Thus,
the placement of thermal mass with respect to window position is important.
Placement of the majority of windows on the south side of the building in the
northern hemisphere (and on the north side in the southern hemisphere) maximizes
winter solar gains that contribute to solar heating. Control of solar penetration, called
"sun tempering," through south and west windows is most important in summer. Devices
can be used to shade walls and windows selectively during the summer, while admitting
radiation in the winter. Movable, insulating window coverings are required in most
climates to reduce thermal loss on winter nights.
Normal building spaces can be designed to be an effective part of passive solar
energy collection, storage, distribution, or control. A combination of active and passive
elements (hybrid systems) may be the most effective design.
For a complete discussion of passive heating and cooling, see Chap. 31.
Economic evaluation of solar energy systems involves a trade-off between the added
capital and operating costs of the solar system and the resulting fuel savings. This must
be determined by a life-cycle cost analysis. Design optimization, life-cycle costing, and
economic assessment procedures are discussed in Chap. 45.
A life-cycle cost analysis permits determination of the collector size that results
in the lowest combined solar and conventional heating costs through the life of the
system. If this optimum size is not zero (that is, if 100 percent fuel heating is not the
lowest life-cycle cost system), the economic feasibility of solar heating and/or cooling
can then be assessed. Because the payback time for the solar system investment is
sensitive to assumed system costs and the cost escalation of the displaced fuel, the
relative economic standing of the solar system can only be approximated. Besides exer-
cising caution in drawing conclusions from this economic assessment, the designer should
bear in mind the decreasing availability of conventional fuels and the savings of these
fuels resulting from use of solar energy.
Any design optimization involves trade-offs between cost and performance. At
some point, the extra performance that can be achieved by adding more insulation to a
wall or ceiling or increasing a solar collector area will exceed the savings incurred. In
some cases, performance can be maximized without affecting costs, such as by use of
ideal collector tilt and orientation.
For most locations, it is uneconomical to design a solar heating system to provide
for 100 percent of the heating requirement, because collector areas and storage volumes
needed to collect and store enough heat for the coldest days or during cloudy weather
would be large and unwieldly. Thus, in general, a solar heating system should be designed
with a full-capacity auxiliary heating unit for carrying the load when the solar system
cannot.
Building Space Heating: Active Systems 37
Once the type of solar heating system has been decided upon, questions of maintenance
and lifetime have a direct bearing on the selection and construction of the system com-
ponents. Because solar equipment usually represents from 5 to 20 percent of the building
investment, amortization of its cost normally requires a useful life equal to that of the
building (at least 15 to 30 years). Important factors in the maintenance and durability of
solar heating systems are discussed in Refs. 1 and 2.
Most buildings are heated by delivery of warm air to the occupied space. The air may be
heated by contact with hot surfaces in a furnace in which fuel or electricity is used. In
some buildings, particularly large commercial, industrial, and multifamily apartments, air
is heated by exchange with hot water or steam, which is in turn provided by combustion
of fuel in a water or steam boiler.
Hot water or steam may also be used for heating solid surfaces in the living space
using "radiators" or a baseboard heating strip, comprised of one or more finned tubes.
Hot water or steam circulates through the tubes and heat is delivered to the room mainly
by natural convection to air moving between the fins. Less common are the radiant
systems, involving warm water supplied from a boiler to extended lengths of tubing
embedded in floors, walls, or ceilings of the living space. The moderately warmed surfaces
transfer heat to the contents of the room by convection and radiation, the latter mechan-
ism predominating particularly when heated ceiling panels are used.
In regions where central heating is not a necessity, separate heaters may be used in
individual rooms. Transfer of heat from these units is mainly by radiation when electric
resistance coils or fuel-heated tubes are mounted in wall units. Natural convection to
room air is usually involved in electrically heated and fuel-heated floor furnaces.
The use of solar energy as a primary source of heat for buildings involves con-
sideration of means for transfer of solar heat to the living space, the medium in which
transfer is effected, the manner of final delivery, the temperature requirements, and the
capability of the solar system to meet those requirements.
The purpose of the space heating system in a building is the delivery of heat to the
occupied space at a rate equal to the rate of heat loss from the building by conduction
through its exterior surfaces and by air exchange between the interior and the surround-
ings. Under these conditions, temperatures in the living space may be held nearly at a
constant value. In most residential heating systems, heated fluid is supplied to the living
space or to heat transfer units in the living space at a temperature varying only to the
extent imposed by transients and heat capacity effects in the systems. Fluid circulation
rate, whether air or liquid, is usually constant. Modulation of supply to meet varying
demands (primarily because of outdoor temperature changes) is usually effected by
controlling the duration of fluid delivery, thereby providing intermittent flow of the
heating fluid. The on-off control method results in small excursions of temperature in
38 Applications of Solar Energy
the living space. Although warm air systems operate at essentially constant heat delivery
temperature (except during short startup and shutdown periods), hydronic (hot water)
systems show considerable temperature variation, commensurate with load, because of
larger heat capacity and resulting dependence of temperature on the duration of the on-
off cycle. In larger systems, particularly in commercial buildings, modulation may be in
the form of fluid temperature control, flow rate control, or combinations of the two,
known as the variable air volume (VAV) system. Increased comfort and decreased fuel
use are possible with such designs.
The temperature at which the fluid is delivered to the heated space or to the heat
exchange area in or adjacent to the space is an important consideration. In warm air
systems, air is commonly heated in a furnace to 50 to 65°C (120 to 150° F), then
delivered through ducts to the rooms. The warm air rapidly mixes with air in the living
space, normally at 21°C (70° F). An equal quantity of air at slightly less than 21°C is
returned from the room to the furnace or other heat source for reheating. Outdoor air
may also be supplied by mixing with the return air. The temperature and flow rate of
the warm air heated in the furnace and delivered to the rooms are sufficient to provide
the maximum heat requirements of the living space (equivalent to the demand on the
coldest, or "design" day), necessitating continuous circulation during such periods.
Thus, a house with a maximum heat demand of 19 kW (65,000 Btu/h) would require
about 400 1/s (860 cfm) of air at a temperature of 60°C (140°F). If air were supplied at
50°C (120° F), 5701/s (1200 cfm) would have to be circulated.
Other than in the warm air system, where the heating medium is actually supplied
directly to the occupied space, some type of heat exchange surface must be provided
for supply of heat to the rooms from the heat transfer fluid. The temperature at which
the fluid must be supplied to the heat exchanger is dependent on the amount of heat
transfer surface available. Baseboard heaters, usually employing hot water as the trans-
port medium, have comparatively small surface areas, so adequate heat-delivery capacity
requires considerably higher temperatures, normally about 82°C (180°F). If larger
radiator areas are used, as in the older hot water systems, temperatures of 50°C (120° F)
can be adequate. For very large heat transfer surfaces, such as entire floors, walls, and
ceilings heated by hot liquid circulating in tubing embedded in those surfaces, still lower
temperatures are satisfactory, typically 40 to 50°C (100 to 120° F). The temperature of
heat delivery in a building is thus highly dependent on the type of heating system
employed.
Solar space heating systems can be divided into two main types. The first type involves
the heating of a fluid in a solar collector, its delivery to storage, and the supply of heat
from storage to a conventional heating system through use of a fluid circulated by a
pump or fan. The collector and storage are separate units, connected by conduits through
which a fluid is circulated. This type is called an active solar space heating system. A
second type eliminates fluid circulation by incorporating the collector and storage into
the building envelope itself. The stored heat is supplied to the occupied space by radia-
Building Space Heating: Active Systems 39
tion and/or natural convection. This latter type, called a passive system, is described in
detail in Chap. 31.
In active systems, either air or a liquid may be used for collection of solar heat,
and the same medium may be employed in the supply of heat to the occupied space.
Heat exchange with another fluid, however, may be provided, in which case the manner
of heating the building is not dependent on the nature of the heat-collection medium.
For example, a solar collector of the liquid heating type may be used in a system in which
heat storage is accomplished by use of a hot water tank. The stored hot water may then
be used as the heat source for one of the hydronic heating systems (under-the-window
radiators, baseboard heating strips, radiant floor or ceiling panels, etc.). Alternatively, by
means of a central heat exchanger (a water-to-air coil), warm air may be used as the
medium of heat transfer to the rooms. If air is the original solar collection medium, its
direct delivery to the rooms in a warm-air heating system is practical; storage may be
provided by heat transfer to a pebble bed and subsequent delivery to air circulated
through the heated pebbles to the occupied space. Warm air is normally supplied to
rooms at temperatures of 50 to 65°C (120 to 150° F), a comparatively moderate require-
ment for air collectors.
The temperature at which heated fluid must be supplied to the rooms in a building
is dictated primarily by the amount of heat transfer surface available. With liquid systems,
the smaller the heat exchange surface, the higher the temperature must be. Because solar
space heating systems using flat-plate collectors are seldom operated at delivery tempera-
tures above 65°C (150° F), use of solar-heated water is usually impractical in baseboard
heating installations, whereas it is suitable for radiant floor or ceiling systems and in
finned-coil warm air systems. The performance of a solar heating facility employing
flat-plate collectors will be considerably better in those systems that do not require the
higher temperatures. If available at acceptable cost, concentrating or evacuated-tube
collectors can be used when higher temperatures are required.
Fluid temperature requirements are thus seen to be dependent on the character-
istics of the specific heating system in the building. These requirements affect, in turn,
the operating conditions in the solar heat supply and storage system.
A solar collector of the liquid heating type such as one of those described in Chap. 9,
in combination with an insulated storage tank or reservoir as described in Chap. 23,
with associated pumps, piping, and an auxiliary heater, are the components necessary
for solar heating of a dwelling or other type of building.
Solar
Collectors
Warm Air
to House
Hot Water
System
House Rooms 4
\ Air Returns
Blower
Cool
Water
from
Pump Return
Cold Water
Supply
Figure 29.1 Solar heating system schematic, water used as collection medium.
discharge valve and an air venting valve, both actuated by ambient temperature or by
solar radiation. Alternatively, an unpressurized, vented collector may be designed to
drain into storage whenever the pump (centrifugal) stops. Air then enters the collector
through the open vent. On starting, the pump forces air out through the vent. In a third
type of drainable system, the air that replaces water drained from the collector rises into
it from the top of the storage tank through an amply sized downflow pipe or a separate
air bypass. Pumping head and power use are substantially higher in most of the drainable
systems than in a permanently filled system because of unrecoverable static head loss.
If a self-draining collector is not used and if the climate is characterized by sub-
freezing temperatures, a nonfreezing liquid should be used in the collector and a heat
exchanger provided for transfer of heat from the collector loop to the water storage
tank. A design employing such a system is shown in Fig. 29.2. Fluids suitable for use
in the collector are common antifreeze solutions of ethylene or propylene glycol in
water and certain organic heat transfer liquids, such as silicone oil or Dowtherm J. When
glycol-water solutions are used, corrosion and toxicity must be dealt with.
Protection against boiling can be provided by draining the collectors, by pressur-
izing the piping system, or by "dumping" excess collected heat through an air-cooled
fan coil unit.
The penalties in the use of a heat exchanger are the temperature loss between two
liquids, the cost of the exchanger and the pump circulating water between the exchanger
and the storage tank, and the cost of pump operation. These disadvantages can be
weighed against questions as to the reliability of a self-draining system and the often
prohibitive cost of using a nonfreezing liquid throughout the collector-storage system.
Collector-to-storage heat exchangers are generally of the shell-and-tube type,
either single or multiple pass, although in some cases finned-tube exchangers are used.
Building Space Heating: Active Systems 41
The high flow rates required for efficient heat exchange in single-pass exchangers cause
pumping power expense and reduced storage tank stratification because of mixing.
However, efficient heat exchange is more important to overall system performance than
is establishment of storage tank stratification.
The heat exchanger between the collector and storage is usually sized so that its
effectiveness is not less than 0.70. This design provides an approach temperature
difference (the difference between the incoming heating fluid and the outgoing heated
fluid) of 3 to 8°C (5 to 15°F) [5]. Exchangers sized to less than a 3°C (5° F) approach
will improve collector performance only very slightly for a comparatively large increase
in exchanger size and cost. Heat exchangers sized for greater than 8°C (15°F) approach
will reduce collector performance excessively relative to the decreased exchanger cost.
If the heat exchanger is immersed in the storage tank, a much greater heat transfer
surface area is required because of the lower heat transfer coefficient resulting from the
very low water velocity induced by natural-convection heat transfer on the storage
tank side. Continuous-pipe coils or finned-tube exchangers may be used, but scaling
can result and pressure loss may be excessive. Locating and repairing leaks may be
difficult. Details of heat exchanger design are found in Refs. 5 and 6.
Pumps used in the collector-storage loop are nearly always of the centrifugal type
with direct-coupled motors. Centrifugal pumps offer advantages over positive displace-
ment types. A safety feature is that they will develop a pressure only slightly above
rated pressure if the fluid loop becomes blocked. Also, they provide increased flow
rate as the fluid temperature increases. This is a result of viscosity changes of the fluid
and improves the collector efficiency at high temperatures.
Valves and air vents must be included in a liquid system. A purpose of the valves
is to provide the proper rate and uniformity of flow through the collectors, as well as
proper drainage. If valves are used for flow balancing, they should be adjusted after the
system is installed to ensure that the flow rates are close to those required by the design.
Building
Air
Nonfreezing
Liquid
Water Duct
Collectors Coils
Solar
WIT—Auxiliary
Figure 29.2 Solar heating system schematic, nonfreezing liquid used as collection
medium.
42 Applications of Solar Energy
Aluminum
Upper Collector Manifold Pipe
—1E— Manual Valve
— Normal Flow Path Solar
Alternate Flow Path Col lector
Panels
Open Neoprene Hos
Vent Connectors
Dielectric
Coupler
Sight Expansion Tank
This Portion Located at Gauge
Level of Attic Floor
Check
Aluminum Pipe Copper Pipe Valve
Copper
Getter Filter Pipe
Because it is virtually impossible to exclude air from an unpressurized system, air vents
are necessary. The air vents and valves should be of the same material as the piping so
that corrosion is minimized.
Additional components that are usually included in appropriate applications are
shown in Fig. 29.3. These include filters, check valves, expansion tank, isolation valves,
and ion getters and/or dielectric couplers, if necessary for corrosion prevention. Methods
for selection and placement of these components are discussed in Refs. 4 and 5.
A few solar heating systems have used two or more heat storage tanks, arranged so
that different temperatures may be obtained in each. The controller selects the tank to
be heated and the tank to be used for heating the building in such a way that it maxi-
mizes solar energy collection while maintaining the highest practical heat delivery
temperatures. The highest-temperature tank is heated when solar energy collectibility is
high, a cooler tank being heated when solar intensity is low. Water may be supplied to
the collector from a cool tank and returned to a warmer tank, with series flow between
them. The load is carried by the tank at the temperature nearest that required to meet
the demand. The added cost of a considerably more complex control system and of
multiple insulated tanks must be weighed against the performance improvement. It is
unlikely that the benefit outweighs the costs in residential-size installations, but there
may be economic advantages in large commercial systems.
In addition to the collector and storage facilities, a solar heating system involves
an auxiliary heater employing conventional energy and means for transferring heat
Building Space Heating: Active Systems 43
from storage to the occupied space. In most of the solar heating systems involving liquid
collection and storage, heat is supplied to the building by warm air circulated through a
central blower and a water-to-air heat exchanger. Alternatively, warm water from the
storage tank can be circulated through heat exchange surfaces in the occupied space.
Practical design temperatures are about 60°C (140° F) for the air exchange system and
about 50°C (120°F) in panel ("radiant") heating systems. Unlike that in conventional
systems, the temperature of the solar-heated water available to the heating system may
vary considerably, so the components must be able to provide good heat exchange at
the lower supply-water temperatures.
For maximum efficiency, the collector should be supplied with the coolest
available liquid, whereas the warmest should be supplied to the load. Water from the
bottom of the storage tank is therefore supplied to the collector (or to the collector
heat exchanger), and heated water from the collector or exchanger is delivered to the
top of the tank. Similarly, the load loop is supplied with hot water from the top of the
tank, and the return from the load is admitted to the bottom. Under normal conditions
of operation, temperature gained through the collector loop is 6 to 11°C (10 to 20° F)
near noon on a sunny day, and the load exchange system is designed for approximately
the same water temperature decrease. Although storage temperatures could be 6 to
11°C higher at the top than at the bottom if effective baffles are used, stratification is
difficult to achieve because of comparatively high pumping rates, turbulence at entrance
and exit tank openings, and the variation in temperature rise through the collector
during the day. Practical design therefore should normally be based on assumptions
of uniform storage tank temperature.
Of several locations for the addition of auxiliary heat, maximum advantage is
gained by its supply to the load loop rather than to the collector loop or storage unit.
Such a design maximizes the use of auxiliary energy by employing it only when neces-
sary and only to augment or replace solar heat being supplied to use. Any other design
forces the collector to operate at a higher temperature (and corresponding lower
efficiency) than necessary, and also results in some of the heat storage capacity being
used for auxiliary heat rather than the solar heat for which it is designed.
Auxiliary heat may be supplied by a conventional hot water boiler in one of two
modes. Although not usual, solar heated water can be pumped through the auxiliary
heater on its way to the load (auxiliary in series), as shown in Fig. 29.4a, with a tem-
perature increase provided by auxiliary energy if needed. Water is then returned to the
storage tank. In this mode, auxiliary heat is used as a "booster" so that solar heated
water is further heated in the boiler, and the temperature of the water returning to
storage from the load may be higher than the storage temperature, thereby adding part
of the auxiliary energy to storage. Continued operation would gradually drive the solar
storage temperature up, thereby using storage capacity for auxiliary rather than solar
and reducing collector efficiency.
A preferable arrangement, minimizing auxiliary fuel use, in which the auxiliary
is in parallel with the load, uses solar heated water exclusively whenever the storage
temperature exceeds the required heating coil, convector, or radiant panel temperature
(Fig. 29.4b). When the storage temperature is too low, circulation of solar heated water
is discontinued and auxiliary heat is used exclusively. Piping and valving must be arranged
so that the water bypasses the storage tank when auxiliary heat is in use; an automatic
44 Applications of Solar Energy
Auxiliary
Heat
Automatic
Valve
( a) ( b)
Figure 29.4 (a) Series and (b) parallel auxiliary heat supply.
valve and single pump provide heat from the appropriate source. Under severe conditions,
comparatively warm water in storage may go unused for a time while the auxiliary
supply is meeting a high demand in the building. However, stored solar heat may be
called upon later, when the load is less severe or when storage temperature has been
increased by addition of solar heat. Solar heat is therefore not wasted.
Where distribution is by air heated by exchange, a better strategy is to use a two-
coil arrangement. In this mode, solar heated water supplies heat to the air in a preheater
coil whenever the storage temperature is above a minimum, usually about 27°C (80°F).
A second coil, immediately downstream of the solar preheater coil, is heated by auxiliary
energy, increasing the air temperature to the required level. This method ensures maxi-
mum use of solar heat and minimum use of fuel, because even low-temperature solar heat
is applied usefully whenever it is available. The cost of a second heating coil is incurred,
a small expense for improved performance. This method is used in the system shown in
Fig. 29.2.
An alternative design of the warm air system in which an auxiliary hot water
boiler is used involves only one water-to-air heat exchanger. Solar heated water is cir-
culated to the coil when it is at a temperature sufficient to meet the heating demand.
If not adequate, the solar supply is discontinued and hot water from the auxiliary boiler
is circulated through the coil.
A variation on this last strategy uses a conventional warm air furnace just down-
stream of the solar preheater coil (Fig. 29.5). The furnace, designed to meet peak heating
requirements, boosts the air temperature when necessary, or can furnish all the heat if
none is available in solar storage. The furnace can be gas fired, oil fired, electrically
supplied, or the condenser of an air-to-air heat pump. In warm air systems, this is the
most economical means of auxiliary supply.
Another location for the furnace is in a bypass air circuit designed for use ex-
clusively with auxiliary energy. In this arrangement either solar or auxiliary is used, but
not both together. The temperature "boosting" advantage is sacrificed for better furnace
efficiency, but an additional controlled damper is needed.
Building Space Heating: Active Systems 45
Warm Air
Supply
f From
Flue Cold Air Collector
Return
Fuel
To
Collector
Return
Air
Figure 29.5 Solar heating system with water storage and warm air furnace.
Valve No I
Auxiliary
Storage Boiler
Tank
-
Pump No I Pump No.2 Pump No 3
Pump No 4
Several other methods for heat exchange and storage, such as with storage in fused
salts or waxes, are specific to certain experimental systems. These are discussed in
Chap. 25.
is below this value, another circuit actuates another pump (or the same pump with re-
positioning of automatic valves), and auxiliary heat is supplied to the load. If a warm air
system is involved, the room thermostat also actuates the air blower.
This simple controller fails to provide any solar heat when the storage tempera-
ture is somewhat lower than the set point, even though a moderate heating demand
could be met by storage. But if the set point is lowered, stored solar heat will fail to
meet the demand when ambient temperatures are lowest. An improvement on this con-
trol strategy is the resetting of the control point in the storage tank. This could be done
manually every few weeks, or an automatic load sensor could be used to improve the
solar utilization. Such a system involves an external ambient temperature sensor that
reflects the load that the heating system must meet. If, for example, the ambient tem-
perature sensor measures 4°C (40°F), the storage controller set point could be, say,
43°C (110°F), whereas with an ambient temperature of -18°C (0°F), the controller
set point would have to be 60°C (140°F). Although this control system is an improve-
ment over those above, it fails to use any solar heat unless solar can meet the entire load.
The most practical control strategy for residential heating involves a room thermo-
stat with a double set point, thereby eliminating the need for sensing the storage tem-
perature. Whenever the rooms require heat, water from the solar storage tank is pumped
to the load, regardless of storage tank temperature. [A low limit near 27°C (80°F) may,
however, be useful in reducing temporary decreases in room temperature.] If storage
is warm enough, room temperature rises, and heat supply is subsequently discontinued.
If, however, the stored water is not hot enough to meet the load, the room temperature
will continue to decrease until another contact in the thermostat, set to operate at
0.5 to 1°C (1 to 2° F) below the initial or higher temperature contact, turns on the
auxiliary system and discontinues operation of the solar circuit. The thermostat may
be designed so that the room temperature increases all the way to the point at which all
heating ceases, or the solar set point can be regained and the solar system restarted. In
the first case, the entire system will shut off until the next demand for heat, solar first
being tried and then auxiliary used if the solar is inadequate. In the second mode, the
system operates alternately between solar heat (which fails to meet the demand until
either solar heat has been supplied to the tank or the weather moderates) and auxiliary
heat. This control system prevents use of auxiliary when solar heat can carry the entire
load and provides intermittent deliveries of solar heat when that source is not hot enough
to meet the full demand.
If the building is heated by air from a solar water-to-air coil and auxiliary heat is
supplied to the air either by a hot water boiler and separate water-to-air coil or by a
warm air furnace, the most practical control system involves the double-contact thermo-
stat described above. This thermostat first actuates the air blower and hot water pump
for solar heat supply. If that source is not adequate, the second thermostat contact
actuates the auxiliary heat to supplement (boost) the solar. Solar heat, even when at
low temperature, is thus fully utilized.
a fully comparable development stage. In fact, the oldest solar heating system in con-
tinuous operation to the present time is an air type [7] . Various types of air collectors
are described in Chap. 9, and the use of a pebble bed for storage of heat delivered by
solar-heated air is outlined in Chap. 23.
In addition to the obvious differences between heat collection in liquids and in
air, the following technical and operational factors may be noted.
Solar air systems involve the same medium for solar collection and space heating;
solar heated air can be delivered directly to the building without heat
exchange or storage.
Heat storage can be accomplished in a bed of loose solids, typically 2.5- to 5.0-cm
(1- to 2-in.) gravel, which also serves as the heat exchanger.
Temperature stratification in a pebble bed and the return of air to the collector
directly from the living space both provide low-temperature 21°C (70°F)
air to the collector with resulting favorable efficiency.
The combination of air density, specific heat, and practical flow rates provides
a considerably higher temperature rise through the collector, typically 33
to 50°C (60 to 90° F), than in a liquid type.
As with liquid heating systems, there are numerous options for integrating the
solar collector and pebble-bed storage into the complete building heating assembly.
A schematic of one of the practical designs is shown in Fig. 29.7. This system employs
a solar air collector, pebble-bed storage, an air-to-water heat exchanger for service hot
water supply, a single blower, and a conventional gas-fired furnace and distribution
Pebble- bed
Storage
Damper to
Direct Flow L
duct system. There are three primary modes of operation, use of which depends on the
availability of solar energy and the heating requirement. There are also two modes of
auxiliary use for space heating and one for service hot water supply.
.)11
FIEATED SPAT'
Solar
7 —1
Heated
1,< Return
Hander
a
Supply Ar
Air
Air
I<
Domestic
Auxiliary Water Coil
Heating
HEAT STORAGE UNIT
Temperature Stratifies
in the Rock
(a) (b)
(c) (d)
Figure 29.8 Solar air system operating modes. (a) Heating from collectors, (b) Storing
heat, (c) Heating from storage, (d) Summer water heating with optional air conditioning.
When heat is needed in the building at the same time as the collectors are on, a
room thermostat signals the control unit to move dampers and direct the flow of heated
air from the collector directly to the zones requiring heat, bypassing storage, as shown in
Fig. 29.8a. In this mode, hot air passes from the collector through the blower, through
the furnace, and into the warm-air distribution system. Air from the rooms circulates
52 Applications of Solar Energy
70 60 40
A 5:130
1000
A MIIMILW21 -41411117
fr0800
._1IF0600
z 0400 A/‘
Alik‘Mh.NMINIMINN.E.7 0200
AM; Mn&l.g&
2400
7
AMMIIMMIL=W1111111111,/
egiMMI2 MENNE V 2200
d IMMIMM1111111 .1
2000
EarilM1
MINMAMM0111.11Mr/ 1800 .
70
60 A4=15.0.11117
1 I&DO
Te mpera ture (°C)
50
mimainimpy 1400A
40 A 1200
A
A
30 1000 c•
Jc)
20
'rMilIMMAIMIIIIIIIIP
51.111111 0800
0 0.3 0.6 0.9 1.2 1.51.7
Top Depth (m) Bottom
Figure 29.9 Typical pebble-bed temperature profiles.
back to the collector through the conventional cold-air return ducts. Either a motorized
damper or a check damper (operated by slight pressure difference) is in this return
duct. When room temperature requirements are satisfied, the thermostat breaks contact,
and the storing mode is resumed.
When there is a high heat demand and when the temperature of the air being
delivered from the collector is insufficient to meet the demand, room temperature
will continue to decline. A lower thermostat set point then turns on the auxiliary heater,
which increases the temperature of the air in the distribution system. The full design
capacity of the furnace will always provide sufficient heat to meet any demand, so the
building temperature will be restored to the preset value.
position history in the pebble bed then proceeds in reverse to the storing cycle as shown
by the night-time curves in Fig. 29.9, a relatively steep temperature gradient being main-
tained during a gradual movement of this profile toward the hot end of the bed. In
effect, the zone of heated rocks is driven toward the hot end of the bed and is reduced in
volume. The air leaving the hot end of the bed is only a few degrees below the rock
temperature at that point.
If the pebble-bed discharge temperature is sufficient, air entering the rooms will
provide enough heat to satisfy the thermostat, and the blower will cease operation.
If, however, room temperature continues to drop, the auxiliary heat supply will be
actuated by the lower thermostat set point, and auxiliary heat will also be supplied.
Operation then alternates between stored heat and stored heat plus auxiliary, until
stored heat is again sufficient to maintain the load by itself or until direct solar heat is
available. Alternatively, the system may run until the upper set point is regained and
shutoff occurs.
It can be seen from the above description that the use of solar heat is maximized
by (a) collecting solar heat whenever moderate temperature delivery of 27 to 32°C
(80 to 90° F) is possible; (b) utilizing even such low-temperature heat, supplemented if
necessary with auxiliary; (c) providing, by means of temperature stratification, high-
temperature storage even when the storage unit is only partially heated; (d) bypassing
storage when heat is needed during sunny hours; (e) using auxiliary energy only as a
supplement, not as a replacement for solar.
blower, auxiliary furnace, and other system components, usually in the basement, may
minimize duct lengths if the hot end is at the bottom of the pebble bed. But unless there
is some practical reason to do otherwise, heated air should be supplied to the top of the
bed so that there is an absolute minimum loss of temperature stratification.
Horizontal flow pebble beds have also been used, but the top cover of the bed
must be in close contact with the pebbles so that air does not channel through an open
space above the packing. There is some evidence that channeling of warm air through
the upper part of the bed and of cool air through the lower part may occur. If a horizon-
tal position cannot be avoided, baffles should be provided to prevent channeling.
With the optimal storage size recommended in Sec. 29.5.D, 0.15 to 0.25 m3
rock/m2 (0.50 to 0.75 ft3 rock/ft2 ) of collector, and if all the heat collected during the
day were stored, a typical mid-winter collection of 9 MJ/m2 (800 Btu/ft2 ) of collector
would result in a mean storage temperature rise from 20 to 65°C (70 to 150°F). During
most of the winter, in most practical systems, half to two-thirds of this heat would be
placed in storage, the balance being used directly in the daytime. Essentially all of the
stored heat would then usually be delivered to the heated space during the following
night.
Studies have shown that the performance of a pebble-bed heat storage unit is
dependent primarily on the mass of material and is comparatively insensitive to type of
rock, dimensions of the bed, pebble size, and air flow rate. As a heat exchanger,
effectiveness tends to increase with length of bed and with decreases in air flow rate
and pebble size. But pressure loss and blower power requirements also increase with
bed length and pebble fineness. Optimum design is thus an economic combination of
these factors. A typical practical design would use a near-cubic bin containing 2.5 cm
nominal pebbles, 2 cm to 4 cm range, with air delivered at a superficial velocity of
0.15 m/s (0.5 ft/s).
Additional quantitative aspects of rock-bed design are discussed in Sec. 29.5.
The designer of a solar air system has the decision as to air flow rate through the
collector. Delivery temperature, hence absorber plate temperature, is strongly dependent
on air circulation rate; in a specific design, the coefficient of heat transfer between plate
and fluid depends on air velocity. But fan power requirement is also a function of air
flow, and there are practical limits to air circulation rates in the occupied space of a
building. As discussed in Chap. 9 and in Ref. 9, the efficiency of a solar air heater is
dependent not only on volumetric air rate, but also on air velocity. Velocity, in turn, is
affected by manifolding and length of travel of air in the collector as well as on the
width of the air passages.
At a practical level of 101/s•m2 (2 cfm/ft2 ) of collector and a velocity of about
3 m/s (10 ft/s), the performance of a solar air heater of commercial design is shown in
Chap. 9. Variation in air velocity causes delivery temperature and collector efficiency
to change as indicated by the two curves in Fig. 29.10 [9] . It is seen that efficiency rises
and exit temperature falls as air velocity is increased. At velocities lower than 101/s/m 2
(2 cfm/ft2 ), considerably higher temperatures can be realized, but efficiency declines.
Building Space Heating: Active Systems 55
Qsolar
180
F'=0.83
at 2 cfm/ft
FR =0 '701
100 160
exit
Collection Efficiency, %
80 140
a)
Measured Efficiency 0.
E
60 120
40 100
0 60
0 2 3 4 5
At an air flow rate of 2 cfm with a 4-m air path through a solar collector having a 1.3-cm
air passage, a pressure drop of approximately 62 Pa (0.25 in. of water) is typical. Power
requirements at this point of operation are moderate, less than 750 W (1 hp) for circu-
lating 4701/s (1000 cfm) of air through the collector and pebble bed.
Additional data on air circulation rates are given in Sec. 29.5.
System Efficiency
There is an important difference in the factors affecting efficiency in air collection
and liquid collection systems. Were it not for the fact that an air collector, in combina-
tion with a temperature-stratified pebble-bed storage, is normally supplied with 21°C
(70°F) fluid (rather than a warm fluid from an isothermal storage), air collector
efficiency would be substantially lower. Efficiencies in air collector systems, therefore,
are strongly dependent on the type of storage used. The higher temperature difference
between the absorber plate and the fluid in an air collector is thus compensated for by
the lower temperature of the fluid supply to the collector. The net result of these effects
is an average difference between plate temperature and ambient temperature approxi-
mately equal in the two systems, so collection efficiencies are comparable.
56 Applications of Solar Energy
Storage Options
In addition to pebble-bed storage for use in a solar air system, storage in water, with
heat exchanger, and in phase-change materials has been suggested. System efficiency is
reduced, however, with either of these storage media, because of the lack of temperature
stratification. Removal of heat from an air stream coming from the collector would be
possible only to the stored water temperature, which for practical heating use would
normally be above 38°C (100° F). Thus, accounting for heat exchange, air would have to
be delivered to the collector at a temperature above 43°C (110°F). The isothermal
character of storage as latent heat would be even more detrimental to collector efficiency
because air being returned to the collector would usually be at a temperature significantly
above the fixed temperature of the phase change.
One other type of storage may be comparable to a pebble bed in its capability to
store heat and produce low recirculation temperature. A bin filled with stacked or
dumped plastic spheres, small metal cans, or other small containers of water or other
high-heat-capacity liquid can be employed [10] . Small jars of water have been used
successfully in a full-scale experiment at the Los Alamos Scientific Laboratory [11] .
Temperature stratification can be achieved within a substantially smaller volume. Whether
the cost of the containers can be competitive with the very low price of gravel, even
including the larger space it requires, has yet to be demonstrated. Also, whether water is
a reliable material for this use is not certain because of the possibility of temperatures
reaching levels at which pressures might rupture the containers.
BC, Filter
Back Draft Dampers
Hot Air
to or from
Manual Damper Storage
Manual Damper
(Closed in Summer)
((Open in Summer)
Air to Vent
Hot Air
to Furnace
(Summer Only)
and Rooms
I r
Ai
Cabinet
Blower
0 Hot Air from
Col lector
r
Automatic Dampers A l and A2 Actuated by Operator A
Automatic Dampers Bi and B2 Actuated by Operator B
Al A2 13 1 B2
cost, the damage that can result from accidental leakage from a solar system, and the
problems associated with boiling under occasional conditions. These problems and
hazards can all be handled satisfactorily, but at the cost disadvantages imposed by heat
exchangers, self-draining and -venting collector arrangements, pumping energy increases,
corrosion-resistant metals, leakproof fittings, and control requirements.
Air systems have advantages and disadvantages essentially the reverse of those
associated with liquid systems. Their drawbacks are the bulkier conduits involved in
moving heated air between collector and storage, a storage volume about three times
that of the water storage required, and the lack of a commercial air conditioner that
can be operated with solar heated air. Advantages of the air system are freedom from
hazards associated with corrosion, freezing, boiling, and liquid leakage. Additional
advantages are the direct association with warm-air heating systems, and the supply of
heat at temperatures usable in those systems.
Although performance data for operating systems are limited (summarized in
Sec. 29.6), the seasonal output of solar heating systems employing air collectors approxi-
mates that of systems in which liquid collectors of comparable transmissivity, absorp-
tivity, and heat loss coefficient are used. Comparisons of typical performance by use of
the F-chart design method (Chap. 43) show a slight air system advantage in certain
climates [12] . Electric power requirements of the two system types are comparable,
but efforts to minimize power use in both systems have been limited. In mid-winter, net
energy supplied (solar heat recovery minus electric power use for fluid circulation) by
typical air systems is slightly higher than the net supply by liquid systems employing
collectors of equal size and similar optical and thermal characteristics.
II
Hot Air from IPy Hot Air to Auxiliary Heater-Blower
Col lectors Blower
I I and to Rooms
/!, 0
P_, lop-
Pi py
P P
PRIMIZIM
Finned Coil —
Alternate Location for Water Heating
for Hot Water Coil
-JI
Summer Bypass Duct
Hot Air to Collector
to and from
Storage
I
Manual Dampers
Dampers for Seasonal
1 F
Figure 29.13 Air-handling module, two-blower design.
60 Applications of Solar Energy
source temperature increases, a combination of these two systems appears to have advan-
tages. During winter, the energy collected at temperatures too low for direct solar heating
may be used as a source for the heat pump. Because this energy usually would be at a
temperature above that of the ambient outdoor air, the capacity and COP of the heat
pump would increase over that for the heat pump alone. At these low temperatures,
collector heat losses would decrease and higher collection efficiency would result.
Alternately, less expensive collectors might be used without efficiency sacrifice. As shown
below, however, use of an air-to-air heat pump in parallel with the solar system, with out-
door air rather than solar heat as the heat pump source, may require less electric energy
than does the series system.
System Descriptions
There are three basic combined solar-heat pump system types: parallel, series, and
dual source [13, 14] . The simplest combined system is a conventional solar system with
an air-to-air heat pump as an auxiliary energy source. This arrangement is called a parallel
system (see Fig. 29.14). Direct solar heating is used whenever possible, and the heat pump
operates whenever there is insufficient solar energy. Electric resistance heat may be used
when neither source can meet the load. This arrangement does not benefit from the use
of solar energy as a source for the heat pump.
In the series solar-heat pump system, the heat pump is placed between the solar
system and the load, so that the heat pump "boosts" the temperature of the solar energy
by using solar as the source for the heat pump (see Fig. 29.15). The condenser supplies
heat to the fluid being circulated to the building. The heat pump uses stored solar energy
whenever storage is above a set minimum temperature. Provision is also made for direct
solar heating by bypassing the heat pump when the storage temperature is high enough
to deliver heat directly to the load. The series system has the advantages of raising both
the heat pump COP and the collector efficiency, but it has the disadvantage of depleting
storage in mid-winter so that direct solar heating is seldom possible.
Heated
Space
Figure 29.14 Parallel solar-heat pump system schematic. Adapted from Vol. 2.1 of the
Proceedings of the 1978 Annual Meeting of the American Section of the International
Solar Energy Society, Inc., Denver, Colorado, August 28-31, 1978.
62 Applications of Solar Energy
Figure 29.15 Series solar-heat pump system schematic. Adapted from Vol. 2.1 of the
Proceedings of the 1978 Annual Meeting of the American Section of the International
Solar Energy Society, Inc., Denver, Colorado, August 28-31, 1978.
In the dual-source system, the heat pump has two evaporators, one placed in the
storage tank and the other outdoors. This arrangement allows the heat pump to use
either the collected solar energy or ambient air as the source, depending on which results
in a higher COP. Direct heating is possible when the tank is at a high temperature. When
the tank temperature drops below the control point, the heating mode is the same as in
the series system. When the tank temperature is either below the minimum (usually just
above the freezing point of the liquid in the tank) or less than the outdoor air tempera-
ture, operation is like that of the parallel system. Thus the dual-source system appears to
take advantage of the best features of the parallel and series systems, but the equipment
is more expensive and control is more complex.
System Performance
size necessary for the solar system (with electric resistance auxiliary) to consume less
auxiliary energy than the conventional heat pump system is between 15 and 20 m2 . If a
parallel system is substituted for electric resistance auxiliary, with a 20-m2 solar collector
system, the fraction of demand met by "free" energy rises from 40 to 60 percent. As
collector area is increased, the "free" energy supplied by the system rises, about 80
percent at 60 m2 , but more gradually than the fraction carried by the conventional solar
system (67 percent at 60 m2 ). With larger-area collectors, the solar system meets a larger
fraction of the load, but the heat pump operates under less favorable conditions at a
reduced COP. As a result, a large combined system performs more like the conventional
solar system when large-area collectors are used. Under such conditions, the additional
cost of a heat pump does not appear justified.
The series system performance is better than that of the conventional solar system,
but somewhat below that of the parallel system. Below about 30 m2 , the small collector
area starves the heat pump by not being able to meet the energy demand. The tank
temperature remains very near the minimum usable temperature and the heat pump
operates at nearly its lowest COP. Because of the continuously low storage temperature
in the colder months, very little direct solar heating is possible, and heavy heat pump
usage is required. As collector area is increased above 30 m2 , the small difference
between the series and conventional solar system performance remains nearly constant
at about 10 percent. There appears to be no advantage in series heat pump use with a
solar system capable of supplying more than 60 to 70 percent of the total heat require-
ments in northern U.S. climates.
The analysis shows further that the dual-source system has very nearly the same
performance as the parallel system.
1.00
10
r Conventional Furnace
20 30 40
Collector Area, m2
Figure 29.16 Fraction of annual load met by "free" energy for solar-heat pump systems.
Adapted from Vol. 2.1 of the Proceedings of the 1978 Annual Meeting of the American
Section of the International Solar Energy Society, Inc., Denver, Colorado, August 28-31,
1978.
64 Applications of Solar Energy
It should be noted that the parallel system has a better overall performance than
the series and dual-source systems, even with a lower collector efficiency and a lower
heat pump COP. In the dual-source and series systems, heat pump work is expended
to boost the temperature of the solar energy that is collected and supplied directly to
the building without heat pump work in the conventional solar and parallel systems.
Utilizable solar energy in the series and dual systems must, essentially, be delivered
twice: once by the solar system and again by the heat pump. The temperature of storage
in the series and dual-source systems is rarely high enough to supply heat directly (by-
passing the heat pump) during the heating season, whereas the higher storage tempera-
tures in conventional solar and parallel systems frequently permit direct heating.
The series and dual-source systems have been investigated with larger storage-to-
collector-area ratios [13] . When the storage size is increased from the value normally
recommended for conventional solar systems, 75 kg water/m2c (15 lb/ft22), to values
nearly 10 times this size, the system performance is only slightly improved.
In summary, a properly designed combined solar-heat pump system can be built
that requires less auxiliary heating energy than either a strictly solar system or a heat
pump system. However, the energy saved by each of the separate systems is relatively
large, and combining the two does not produce great additional saving in energy use.
From an economic standpoint, the solar source heat pump capability would appear to
be a cost-effective addition to a solar system only if either electricity is very expensive,
cooling (using the heat pump equipment) is needed, or it can improve collector efficiency
drastically. Drastic improvement of collector efficiency will not occur with conventional
flat-plate collectors. However, if low-temperature collectors could be produced at low
cost, the series heat pump design could be used with them to economic advantage (as is
indicated in Ref. 18). As discussed in Chap. 9, the availability of a durable, trouble-free
collector, even though designed for low-temperature operation, is doubtful.
Thus, the cost effectiveness for conventional solar systems and conventional heat
pump systems separately does not ensure the cost effectiveness of a combined solar and
heat pump system. Based on heating season performance alone, assuming characteristics
of commercially available air-source heat pumps, the parallel system appears to be the
best combined system, but the economic choice may still be a conventional heat pump
or a conventional solar system. This conclusion is echoed in Ref. 17, where it is con-
cluded from simulations of air-source solar heat pump systems that the parallel system
saves the greatest amount of electricity. On the other hand, the development of special
liquid-source heat pumps, combined with ground-coupled storage and special low-cost
collectors, may make the series system the best competitor with the conventional heat
pumps [18] .
One of the main purposes of this chapter is to enable the designer to determine the
performance of the collectors and associated solar components in an active space heating
system.
System performance can be predicted by hourly computer simulations for a variety
of system configurations. The results of many such simulations can be used to develop
Building Space Heating: Active Systems 65
generalized performance data that correlate the heat delivered by a particular type of
system with its design parameters and the weather. Furthermore, these results can be
used to develop simplified methods that do not require the further aid of a computer
to estimate solar system performance (see Chap. 44).
The first detailed simulation of solar heating systems [19] , published in 1970,
shows the annual fraction of total heating requirements supplied by solar energy to
residential buildings in eight U.S. cities. Hour-by-hour computations of the effects on
performance of varying collector and system parameters were made.
Another simulation model, developed at the University of Wisconsin in 1973-1974
[20] , includes thermal capacity in the building (ignored in the earlier study referenced
above), but is of limited utility because it is specific to a particular building and heating
system design. It was used in delineating the effects of collector area and storage size
on monthly and annual solar heat utilization in a hypothetical Albuquerque, New Mexico,
residence.
Many hour-by-hour simulation computer programs for analysis and design of
active solar heating and/or cooling systems are in use today. The most widely used,
comprehensive, versatile, and well documented of these is the TRNSYS program
developed at the University of Wisconsin in 1974 [21] . Models of individual components
are mathematically assembled to simulate complete systems that can then be run with
varying design parameters to determine designs that provide maximum solar heat
utilization.
Many TRNSYS runs, for a wide range of weather conditions, have been used to
correlate both liquid and air system performance. The result is the simplified performance
prediction and collector sizing procedure called F-chart [22] , also developed by the
University of Wisconsin. F-chart is the most notable of several simplified design and
sizing procedures that recently have been developed [4, 22-29] . Another of these pro-
cedures, sponsored by the U.S. Department of Energy, is the SOLCOST program [23] ,
which uses a "typical day" simulation for each month to predict the monthly system
performance. The use of the F-chart method is discussed in Chap. 43; the reader is
referred to this chapter for the information needed to size solar collector arrays.
There are several factors involved in the prediction of the performance of a solar
space heating system. One is the heat-delivery capability of the system. This is analyzed
by varying system design parameters and noting the effect on annual system performance
and/or cost. The earliest complete studies of residential solar system performance and
economics were conducted by Lof and Tybout [30,31] . Results of recent parametric
studies based only on the performance (no economics) of residential systems, are dis-
cussed below. The tables and graphs accompanying these parametric results are based on
reference liquid and air system designs. Note that these results allow the designer to
maximize results with respect to performance only; economic factors should be included
where appropriate to decision making in the design process (see Chap. 45).
The next important factor is the building thermal load, that is, the building heating
requirement. Because the sensitivity of performance results to design parameter changes
depends on the magnitude of the load, results are presented below for representative
moderate and high solar fractions of the load met.
The geographic location of the building, in terms of solar radiation and ambient
temperature, is the third factor. Using hourly computer simulations, the performance of
66 Applications of Solar Energy
the reference systems has been calculated for selected cities where sufficient weather
and solar data have been available. Examination of the results of many such simulations
indicates that the results can be generalized if a range of representative climate types is
studied. The trends of solar performance as a function of parametric variation thus are
shown below for cities representing a range of climates.
Solar energy systems differ from conventional heating systems in three key respects that
relate to performance and collector sizing requirements.
First, incident solar energy is not available continuously, and may not be available
at all during peak demand periods at night or in cloudy weather. Because solar energy
may not be able to meet all demands economically, auxiliary energy supply systems must
be provided. Sizing of solar system components is therefore based on annual system per-
formance, rather than on peak load calculations. The auxiliary system is still sized for
peak loads.
Second, both the energy demand and the thermal storage mass influence the
operating temperature of the solar collectors, and therefore influence the amount of
incident solar energy that a given collector array can collect. Consequently, seasonal
building heating and cooling load variations strongly influence the annual performance of
a solar energy system.
Third, solar energy is typically collected and delivered over a large temperature
range. This variation influences the performance of heating coils and absorption chillers.
For these reasons, conventional "peak load" design methods are not acceptable.
A reliable and accurate method for estimating the annual performance of a given
solar energy system must be used to select solar system components properly, to establish
optimum collector area, storage tank volume, and collector tilt and azimuth angle, and to
determine the economic feasibility of a solar heating system. Such a method is provided
by hourly computer simulation techniques.
B. Simulation Methodology
The performance of a solar heating system depends on the solar radiation available to the
collector, the outside ambient air temperature and wind condition, the collector design,
the collector inlet fluid temperature and flow rate, and the thermal load on the system.
A general understanding of how the above factors affect system performance in various
locations is most useful. Predicting system performance is difficult because of the un-
predictability of the weather and the complex relationships among system components.
However, average long-term performance can be satisfactorily predicted by use of
historical average weather and solar data in a digital computer program on an hour-by-
hour basis. The change in performance resulting from system design modifications can
then be determined by repeating the calculations in a series of simulation runs.
The net energy that can be extracted from the collector is calculated each hour.
This is determined from the collector system characteristics and the conditions prevailing
during that hour. If this energy is positive, it is added to storage. The thermal load is
Building Space Heating: Active Systems 67
calculated from the outside temperature, the envelope and ventilation loads, and the
water heating and internal heat generation (people, lights, and equipment) profiles. All
or part of the collected energy is extracted from storage by the heat distribution system
to meet the load. If the load cannot be satisfied from storage, auxiliary heat is added to
make up the difference. The change in storage temperature over the hour is computed
from an energy balance on storage. This calculation is repeated for each of the 8760
hours of the year, and energy flows are tabulated hourly, monthly, and yearly.
*The mathematical models for the solar collector and other system components have been developed
and used exclusively by LASL and have not been documented or published in detail. However, agree-
ment of computed results with those of published models is satisfactory (see Chap. 44), and the
analyses and results presented here are considered reliable.
68 Applications of Solar Energy
Solar collector
1. Orientation Due south
2. Tilt (from horizontal) Latitude + 10°
3. Number of glazings 1
4. Glass transmissivity (at 0.86 (6% absorption, 8% reflection)b
normal incidence)
5. Surface absorptance (solar) 0.98
6. Surface emittance (IR) 0.89
7. Coolant flow capacity rate 115 W/°C.m! (20 Btu/11.°F.fq =
(mCp/m2 ) 0.04 gpm/ftD
8. Effective heat transfer 170 W/°C.m (30 Btu/h.°F.ft!)
coefficient (U,A/m2c )
9. Back insulation U value 0.5 W/°C-m! (0.083 Btu/11.° F-ft2c )
(UbA/m2c )
10. Heat capacity 20 k.J/°C.ni (1 Btur F.fq)
Collector piping
11. Heat loss coefficient, to 0.23 \NT C.in (0.04 Btu/h.°F.ft2c =
ambient (U p A/m2c ) U = 0.3 Btu/h• °ft 2 pipe or 1 in.
insulation of k = 0.075 Btu/11.°FM
Collector-to-storage heat exchanger
12. Heat transfer coefficient 57 W/°C.rn (10 Btu/11.° F1q)
(Uhe A/m2e )
Thermal storage
13. Heat capacity 305 kJ/°C.rn (15 Btu/° F.fq = 1.8 gal
water/ft2c )
14. Heat loss coefficient 0 W/°C.m, (0 Btu/11-° F.ftD
(Us A/mc2 ) (i.e., assuming all heat loss is to heated space)
Heat distribution system
15. Design water distribution >56°C(133°F)c
temperature
Controls
Residential building system and
load profile assumed
Building maintained at 20°C (68°1-)d
Collectors on when advantageous
a Values of parameters used for the "reference" solar heating system using liquid-heating solar
collectors, a collector-to-storage heat exchanger, water-tank thermal storage, and forced-air heat
distribution system to the building. The values are normalized to collector area.
b
These values apply for normal incidence on ordinary double-strength glass, 0.32 cm (1/8 in.). For
other angles of incidence the Fresnel equation is used.
c
The coil and air circulation are sized to meet the building load with an outside temperature of
-19° C (-2° F) with 56°C (133° F) water and an air flow rate adequate to make up the space heat
losses at an air discharge temperature of 49°C (120° F). This corresponds to a finned-tube coil
effectiveness of 80%.
dIt is assumed that internal sources such as people, lights, and equipment will be sufficient to raise
the internal temperature by 2°C (4° F), i.e., up to the comfort standard of 22°C (72° F).
Building Space Heating: Active Systems 69
For each calculation, only the parameter under study was varied; therefore, all of
the complex system interactions that result from changing that parameter have been
taken into account.
The results of these calculations are intended to show trends in solar performance
as design features are varied, rather than be taken as absolute results. No actual solar
heating system will perform exactly as predicted. Furthermore, because the design
parameters were varied singly to determine their individual effect on the overall per-
formance of the system, care should be used in applying the results to simultaneous
variation of two or more parameters because some of them may be coupled nonlinearly.
If one simultaneously varies two or more design parameters to any significant extent, the
net change in performance (compared to the reference system) may be different than the
performance estimated by changing the parameters individually and summing the effect
of all changes. A separate computer simulation may be needed for an accurate perfor-
mance prediction for this case.
Although only residential systems were simulated, the parameter variation results
shown should also apply to commercial buildings maintained at the same temperatures.
The results of the computer simulations are shown below in two forms: first, in
summary form for each parameter in Table 29.3, and, second, in detail in Sec. 29.5,
with selected results plotted in Figs. 29.17-29.30. Graphical results for the remainder of
the parametric runs are found in Ref. 32.
Since there are 12 curves on some of the graphs (that is, two curves for each of the
six cities, one for the nominal 75 percent solar heating system, and one for the 40 percent
system), their order from top to bottom on the graph is indicated by the order of the
symbols of the cities on the graphs. Fifteen parameters have been studied for the
reference liquid system. Parameters 1 through 6 and 9 are the same as for the reference
air system, a discussion of which is found below.
1. Collector Orientation
The orientation for maximum solar heat utilization is usually due south. Variations
east or west of up to 30° reduce performance by only 2.4 to 5%, but larger variations
can reduce performance substantially. (Same for reference air system.)
2. Collector Tilt
The tilt angle for maximum solar heat utilization will range from latitude plus 10°
up to latitude plus 25° , depending on climate. Like orientation, small variations
from the ideal have a small effect. Variations of more than about 20° can substan-
tially reduce performance. (Same for reference air system.)
3. Number of Glazings
Justification of the extra cost of double glazing depends on climate. Collector area
can be reduced by only 5% by the use of double glazing in Phoenix, but by 25% in
Bismarck. (Same for reference air system.)
4. Collector Glass Transmissivity
Improving glass transmissivity by 6% by the use of "water white" glass should
increase the annual solar heat collected by 2 to 5% depending on the site. Improving
transmissivity by reducing glass reflectance should have an even greater effect.
(Same for reference air system.)
5. Collector Surface Absorptance
Nonselective and selective surfaces having an absorptance greater than 0.90 are
available. (Same for reference air system.)
6. Collector Surface Emittance
A high selective surface can reduce the required collector area by 15% in the warmer
climates and by up to nearly 40% in the colder climates such as Bismarck. Surface
durability may be a problem. (Same for reference air system.)
7. Collector Coolant Flow Capacity Rate
Although not a critical design parameter, the flow rate is usually designed to obtain a
coolant temperature rise of about 11°C (20° F) under peak conditions.
8. Effective Heat Transfer Coefficient of Collector
A net effective heat transfer coefficient equal to or greater than about 60 W/°C•rn2
(10 Btu/h.°F.ft2 ) of collector is adequate to achieve near maximum performance.
9. Collector Back and Side Insulation
U values of less than 0.6 W/°C.m2c (0.1 Btu/h.°F•ft,2) represent good design practice.
(Same for reference air system.)
10. Collector Heat Capacity
Maximum solar collection can be achieved by minimizing the mass of metal and the
fluid inventory in the collector.
11. Distribution Pipe Insulation
Collector-to-storage distribution pipe heat losses can be kept acceptably low by
insulating to a U value of 1 W/°C•m2, (0.2 Btu/h.°F-ft2c ).
12. Heat Exchanger Effectiveness
A heat transfer coefficient of 60 W/°C•rri2 (10 Btu/11.°F•ftD results in a practical
temperature difference in the heat exchanger of 5 to 10°C (9 to 18° F) under normal
operation.
Building Space Heating: Active Systems 71
from due south is somewhat greater for the colder northerly cities than for the warmer
southerly ones, but variations of up to 30° east or west reduce performance by only
2.4 percent on the average, or 5 percent at the most.
Collector Tilt (Fig. 29.18). Collector tilt angle is an important design consideration.
Common practice is the use of a tilt equal to the latitude of the site plus 10 or 15° for
100
SM PH
Percent ofTota l Hea t from Solar
- FR SM
PH FR/SE
80 _ SE/MD MD
BI 81
60
40
SM SE
20 FR FR/MD -
SE/MD PH/SM
PH BI -
BI
0 "
-90 -60 -30 0 30 60 90
WEST SOUTH EAST
Figure 29.17 Percent of annual load met by solar versus collector orientation.
72 Applications of Solar Energy
• •
e„,..--"" 37°
90° 90°
Tilt 0 Tilt
100% 100%
Fresno 37°N Medford 42°N
•
62°
52° 50°
0 . 90° 90°
Tilt Tilt
100% 100%
Seattle 48°N Bismarck 47°N
67° 68°
•
51° 63°
90° 0 90°
Tilt Tilt
Figure 29.18 Percent of annual load met by solar versus collector tilt.
space heating applications. Clearly, latitude is important; however, the tilt for maximum
solar utilization is also affected by the monthly distribution of the heating load—whether
it is concentrated in a short period of 2 or 3 months or spread out over half the year—and
by the solar heating fraction. It is important to note that the curves have a relatively flat
maximum. This means that major deviations from the ideal tilt angle of 10 to 20° above
the latitude have only a minor effect on the solar heating fraction, usually less than 10
percent.
Many other considerations may play a more important role than maximizing
performance, such as ease of assembly and repair, shedding of snow and rain, architec-
tural integration, and potential overheating problems in summer. A vertical collector
may well be best in some situations.
Building Space Heating: Active Systems 73
Collector Glazing (Fig. 29.19). At least one sheet of glazing is essential in space
heating applications. Changing one or another characteristic of the glazing produces a
net yearly effect on performance as the result of complicated interactions that depend
not only on the glazing but on the rest of the system and its application, which together
determine the operating regime of the collector. Intuitive judgments of performance
based on collector performance charts alone are often misleading or wrong as they fail
to take into account the interaction of all the other factors. In this and the following
three sections, the effects of the parameter changes on the net yearly performance of
space heating applications are discussed.
The decision to use one or two sheets of glass depends on a trade-off between the
added cost of the extra sheet and the increase in performance to be gained. The perfor-
mance to be gained by an additional sheet of glass is predominantly the result of heat loss
reductions that depend on the ambient temperatures where the collectors are operated.
Co llector Area Ratio; Dou ble Glazing / Single
I.0
• PH
SM
FR•
MD
0.9
SE•
• B1 -
0.8
40% Solar Fraction
0.7
0 5000 10000 (65°F Base)
Figure 29.19 Ratio of collector area required with single- and double-glazed
collectors.
74 Applications of Solar Energy
This dependence suggests a correlation with heating degree-days. In Fig. 29.19 it can be
seen that the ratio of double-glazed collector area to single-glazed collector area, for the
same annual solar heat utilization, in a 75 percent solar system correlates well with the
January degree-days, whereas a 40 percent solar system correlates better with annual
degree-days.
To illustrate how these graphs may be used, suppose that a solar heating system
with a single glazing will cost 10 percent less than one using double-glazed collectors of
the same area. It can be seen that it would be more economical to install a 75 percent
solar heating system with double-glazed collectors at sites where the January heating
load exceeds about 360°C • days (650°F • days) such as Medford, Seattle, and Bismarck,
where the saving in collector area is greater than 10 percent. It is more economical to
install single-glazed collectors in Pheonix, Santa Maria, and Fresno.
Glass Transmissivity. It is customary to lump the reflection and absorption effects
into a single variable called the transmissivity for solar radiation, which is the trans-
mitted fraction of the total solar energy falling perpendicularly on one or more trans-
parent surfaces. For one sheet of ordinary window glass, reflection accounts for a
reduction in transmissivity of about 8 percent and absorption for about 6% at normal
incidence, resulting in a transmissivity of about 86 percent. Glass with low impurity
content, especially iron, shows less absorption, and transmissivity as high as 90 percent
is available.
Performance curves resulting from a variation in glass transmissivity [32] show
that the relative performance decrease is less than the relative reduction in transmissi-
vity, principally because energy absorbed in the glass is not all lost from the collector—
it raises the glass temperature and thereby reduces the convective and radiative losses
from the absorber surface. A 1 percent decrease in glass transmissivity reduces the yearly
solar energy collected by 0.4 to 0.9 percent, depending on the site, the number of glaz-
ings, and the solar fraction.
Collector Surface Absoiptance. Solar energy that is not absorbed by the surface
is not available as heat energy in the rest of the system. Therefore, collector surface
absorptance is a very important factor. Nonselective absorbers usually are painted black,
but even among black paints there will be variations in absorptance from 92 to 98 percent.
Nearly all collectors, both selective and nonselective, have a surface absorptance above
90 percent.
It is interesting to note that a 1 percent decrease in surface absorptance decreases
the overall annual system performance by only 0.4 to 0.7 percent [32] . This may be a
result of the fact that with the lower absorptance the system runs somewhat cooler with
correspondingly lower heat losses throughout.
Collector Surface Emittance (Fig. 29.20). A major heat loss in flat-plate collectors
is long-wave radiation emitted by the absorber surface. Painted surfaces have a high
effective hemispheric emittance for long-wave or infrared radiation; a value of 0.9 is
typical for a good black paint with a high absorptance.
Selective surfaces currently available have effective emittances near 0.1 and ab sorp-
tances near 0.95. Figure 29.20 shows that a single-glazed collector with a selective surface
with an absorptance of 0.95 and an emittance of 0.10 needs to be only 60 to 85 percent
as large as a collector with a matte black surface (absorptance, 0.98; emittance, 0.89).
Building Space Heating: Active Systems 75
1.0
09
Collector Area Ratio; Selective / Non-selective Surface
0.8
0.7
Single Glazing
75% Solar Fraction
060
1000 2000 (65°F Base)
0 500 1000 1111 (18°C Base)
January Degree Days
I0
09
•PH
SM
FR
08 MD
• SE
BI
0.7
... Single Glazing
40% Solar Fraction
0.60
5000 10000 ( 65°F Base)
0 2000 4000 5556 ( 18 °C Base)
Annual Degree Days
Figure 29.20 Ratio of collector area required with selective and nonselective
absorber surfaces.
In all cases studied, solar heat utilization was actually reduced by the use of double
glazing over a selective surface compared to single glazing over the selective surface. One
of the purposes of double glazing is its absorption of infrared radiation emitted from the
collector surface, but with a selective surface this effect is small; the increased absorption
and reflection of solar radiation caused by the second layer of glass is a greater loss than
the gain resulting from decreased convective losses.
The simulation results show a reasonable correlation with heating degree-days
indicated on the graphs of Fig. 29.20. The improvement in performance is almost linear
with changes in emittance, as is discussed in Ref. 32.
However, the choice of a selective surface depends on several other factors besides
solar heat delivery. Cost is one. For example, assuming the absorptance and emittance
76 Applications of Solar Energy
figures above, a selective surface will reduce the required collector area in Medford by
about 30 percent over a collector with a nonselective surface. The question then be-
comes whether an array of collectors with a normal black surface will cost more than
an array of selective surface collectors 30 percent smaller in area. Durability is a factor—
some selective surfaces have shown a tendency to break down both mechanically and
chemically over a period of time.
Collector Coolant Flow Capacity Rate (Fig. 29.21). Collector coolant flow
capacity is important in a liquid collector primarily because increases in the flow rate
reduce the collector temperature rise, which, in turn, reduces the amount of heat lost
from the collector at a given inlet fluid temperature.
Simulation results presented in Fig. 3.21 show that solar heat utilization is
relatively insensitive to the product of collector coolant flow rate and specific heat above
a value of about 10 W/° C • m2 (2 Btu/h • °F • ft2).
The proper choice of coolant flow capacity rate depends on the pressure drop
through the collector, the viscosity of the fluid, the size and cost of available pumps,
and the price of electricity. Although higher coolant flow rates increase the performance
of the system, they also require more pumping energy. Therefore, increases in the flow
rate may not result in a net increase in solar heat utilization or cost savings for the
system.
100
Percent ofTota l Hea t from So lar
80
60
40
20
00 2 4 6 8 10 (8tu/h • °F• ft 2 )
c
0 10 20 30 40 50 568 (W/ °C• mc 2 )
Collector Coolant Flow Capacity Rate
Figure 29.21 Percent of annual load met by solar versus collector coolant flow rate.
Building Space Heating: Active Systems 77
100
SE
MD
FR
BI
Percent ofTota l Hea t from Solar
PH
80
SM
60
40
20 _ SE/MD/FR
81/PH
SM
0-
0 10 20 30 40 50 (Btu/h • °F • ftc 2 /
Figure 29.22 Percent of annual load met by solar versus collector heat transfer
coefficient.
Collector Heat Transfer (Fig. 29.22). The average collector heat transfer co-
efficient is the effective heat transfer coefficient between the average surface tempera-
ture of the collector and the average coolant temperature. As a result of the heat transfer
properties of liquids, the heat transfer coefficient in liquid collectors is relatively high.
An effective heat transfer coefficient equal to or greater than about 60 W/°C • m2
(10 Btu/h • ° F • ft2 ) of collector is adequate to achieve near maximum performance
from the solar heating system.
Collector Back and Side Insulation. Heat can be lost from the absorber surface by
conduction through the back and sides of the collector housing. Typical design practice
is the use of the equivalent of 8 to 10 cm of Fiberglass insulation. A U value of 0.6
W/°C • m2 (0.1 Btuth • ° F • ft2) provides nearly the maximum possible solar heat utiliza-
tion; doubling the back and side heat loss has only minor (less than 2 percent) effect on
the solar performance [32].
Collector Heat Capacity. A certain amount of solar energy is required to heat the
collector to operating temperature every morning—energy that is not fully recovered at
the end of the day. The same effect occurs each time the sun goes behind a large cloud
and causes the collector temperature to drop below storage temperature.
The contributors to collector heat capacity are the fluid it contains and the mass of
metal, insulation, and of all components of the heating system that cycle in temperature.
78 Applications of Solar Energy
The magnitude of heat capacity depends on collector design. In certain cases in liquid
collector designs, such as with large fluid inventories, collector heat capacity may have
a significant effect on the performance of the overall system [32] . This is generally not
the case with air collectors.
Distribution Pipe Insulation. The piping that connects the collector array to
storage should be well insulated. Minimizing heat losses from the pipe on the hot side of
the collector is more important than minimizing losses from the collector itself, since
the hot-side piping may be the highest temperature of the entire system.
Because distribution pipes are usually 5 cm in diameter or less, heat losses from
them can be kept relatively low. Also, this heat loss is usually to the building interior.
Therefore a nominal value for distribution pipe heat losses of only 0.23 W/°C • m (0.04
Btu/h • ° F • ft2 ) of collector was chosen for the reference system calculation [32].
Heat Exchanger Heat Transfer Effectiveness (Fig. 29,23). In the reference liquid
system, the heat from the collector fluid is transferred to water by a heat exchanger.
A net heat exchanger heat transfer coefficient of about 60 WI°C • m (10 Btu/h ° F • ft2c)
results in a practical temperature difference in the exchanger of 5 to 10°C (9 to 18°F)
in normal operation. For example, if the total amount of energy being collected at
a particular time is equal to 300 W/m2e (100 Btu/h • fee), the average temperature
difference between the fluid on one side of the heat exchanger and the other is about
6°C (10° F). Collector liquids with poor heat transfer characteristics, such as
paraffinic oils, require either larger heat exchanger surface area or higher temperature
difference. A higher temperature difference results in a degradation of system
performance.
Thermal Storage Heat Capacity (Figure 29.24). When used with the reference
liquid system, the nominal thermal storage capacity of 85 W/°C • m2, (15 Btu/° F • ft2c =
1.8 gal of water/ftD would be sufficient to heat a 75 percent solar building in Medford
for 20 h, assuming an initial storage temperature of 66°C (150°F) and an outside tem-
perature of -7°C (20° F). After this time, heat would continue to be extracted from
storage, but an increasing amount of auxiliary heat would be required to maintain an
inside temperature of 20°C (68°F).
The simulation analysis indicated that thermal storage in excess of 85 W/°C • m2,
(15 Btu/°F • ftD does not improve the yearly performance significantly. However, fairly
severe performance losses are predicted if the storage mass is less than about 60 W/°C •
m (10 Btu/°F • ft = 1.2 gal of water/ft2c ). Thus the primary function of thermal storage
is to carry over heat from the daytime hours to the night. The reason that only a small
performance improvement is obtained with use of considerably larger storage is that,
on a typical clear mid-winter day, 40 to 80 kg of water have ample heat capacity to store
all the heat deliverable from 1 m2 of collector (or 1 to 2 gal/ft2 ). With the collector
area normally designed for a specific building, the stored heat would usually be totally
depleted overnight, so that even if considerably larger, the storage would have no addi-
tional capability to supply the load. The collector is therefore undersized and the
storage oversized for operation during the coldest months; that is, collection is the
limiting factor. Note, however, that the effect of changing storage mass is strongly
location dependent.
Building Space Heating: Active Systems 79
100
SE
MD
FR/BI/PH
Percent of Tota lHea t from Solar
SM
80
osormowaimmip
60
40
FR/MD/SE
20 PH/BI
SM
1
5 10 15 20 25 (Btu/h .°F• ftc 2)
20 40 60 80 100 120 140 (W/°C• mc2 )
Figure 29.23. Percent of annual load met by solar versus collector-to-storage heat
exchanger heat transfer coefficient.
100
SE
MD
BI
FR
80 SM
Percent of Tota l Heatfrom So lar
PH
60
40
20 FR
SE
MD/BI
PH
SM
o
o 5 10 15 I 20 I 25 (Btu/h •°F •ft 2)
2 c
0 100 200 300 400 500 (kJ/°C•m I
c
Thermal Storage Heat Capacity
Figure 29.24 Percent of annual load met by solar versus thermal storage heat capacity.
80 Applications of Solar Energy
The condition under which larger storage capacity could be of slight advantage in
mid-winter would be a series of mild, sunny days. The use of solar heat might be less
each day than collection, so heat could be stored. Additional stored heat in the larger
system could then provide a larger fraction of the load on a subsequent cold, cloudy
day than could the smaller storage unit. The frequency of these occasions, however, is
sufficiently low in most climates that annual performance is only slightly increased.
There is also some gain in solar heat supply with larger storage in the spring and
fall when loads are lower and solar radiation is normally higher than in mid-winter.
The effect is similar to that encountered on the mild, sunny winter days described above.
There is a possibility, however, that large storage units may have a slight technical
disadvantage. In mid-winter, even though more heat may be in storage in the large sys-
tem, it will usually be at a lower temperature than in the small systems. Because the
lower temperature may be insufficient to meet the load, auxiliary heat will have to be
used. In the smaller system, on the other hand, the higher temperature may be adequate
to meet the load. Also, heat losses from a larger tank (if outside the heated space) may
more than offset the gain from larger heat capacity. Thus, the portion of the load carried
by solar actually may decline as storage size is further increased.
Choice of proper storage capacity is, however, an economic decision. Optimum
size, that which provides solar heat at the least annual cost, has been shown to be about
80 kg/m2 (2 gal/ft2 ) of collector [9] , even in adverse solar climates [33] .
Arguments have been advanced for providing sufficient storage for 1 or 2 weeks of
overcast winter weather or even for seasonal storage where solar heat collected in
summer and fall could be used in winter and spring. Weekly storage is not cost effective
because there would not be sufficient solar energy collection in the winter months to
provide carryover of heat from one day to the next. Moreover, in mid-winter, storage
temperatures would tend to be too low to meet the load unless large and expensive heat
transfer surfaces or multiple tanks were used. The concept of seasonal storage has not
been explored sufficiently to yield definitive answers, but the possibility of practical
use in some climates merits consideration.
Heat Losses from Storage (Fig. 29.25). If the heat storage tank is within the
space to be heated, heat lost from storage simply goes into the heated space, and there
is only slight penalty to the system performance. The tank should be insulated for times
when the space does not require heat and when, therefore, heat losses from the tank
would contribute to overheating the building. These losses reduce capability to meet
subsequent heating needs.
Where the storage tank is located outside the space, performance degradation can
be very severe, as shown in Fig. 29.25.
To use Fig. 29.25, it is necessary to calculate the surface area of the tank, multiply
that by the effective U value of the tank insulation, and divide by the collector area. The
result is the storage heat loss coefficient based on unit collector area.
Design Water Temperature* (Figs. 29.26 and 29.27). In the forced-air distribution
system, a finned tube coil is used to transfer heat from the storage-water tank to the air
*Design water temperature is, by definition, the water temperature required for meeting the design
(maximum) heating load. Actual water temperature is highly variable, meeting various fractions of
the load under different conditions.
Building Space Heating: Active Systems 81
Percent of Tota l Heat from Solar
Figure 29.25 Percent of annual load met by solar versus storage heat loss coefficient.
100
Percent ofTota l Heat from Solar
80
27 30
Figure 29.26 Percent of annual load met by solar versus design water temperature,
forced-air distribution.
82 Applications of Solar Energy
100
Percent ofTota l Heat from Solar
80
60
40
20
0
80 100 120 140 160 180 ( °F )
2; 30 40 50 60 70 80 82 (°C)
Figure 29.27 Percent of annual load met by solar versus design water temperature,
baseboard distribution.
circulating to the space. The air flow requirement is a function of the design water
temperature. As the design water temperature is decreased, the required air flow is
increased, resulting in an increase in solar heating system performance, but at higher
capital and operating costs. However, up to a design water temperature of about 60°C
(140°F), the annual solar fraction shows only slight variation in all locations. Determina-
tion of the building air flow is discussed in Ref. 32.
Calculations were also made with a baseboard hot water distribution system (Fig.
29.27). Normal baseboard systems are designed with inlet water temperatures of 70 to
90°C (160 to 200°F). For a solar system, the baseboard convectors are usually over-
sized so that they can supply sufficient heat at the lower temperatures normally available
from the thermal storage tank.
The baseboard curves on the graphs show the design water temperature needed to
meet the indicated annual load fraction at the design AT. At higher outside air tempera-
tures, a lower baseboard temperature could provide the load. The heat output in the
baseboard simulation is assumed to vary as the 1.176 power of the water-to-room tem-
perature difference and is based on manufacturer's data.
When the heating load is higher than can be supplied at the solar storage tank
temperature, hot water from the auxiliary source (boiler) is used in the baseboard
convectors.
Building Space Heating: Active Systems 83
The general remarks concerning the reference liquid system apply to the reference
air system as well. Parameters for the reference air system are listed in Table 29.4. The
performance simulations were based on the same locations and weather data, and the
same cautions apply in using the data presented in the pages that follow. They are meant
to indicate trends, rather than absolute results, and the design parameters may be, in
some cases, coupled nonlinearly.
The principal differences in the performance characteristics of liquid and air
systems are a result of the characteristics of the heat transfer and heat storage media.
Water storage tanks have commonly been used in solar heating applications, and, as a
result, their characteristics are now well known. The pebble beds used in air systems
have also been successful, but their characteristics have not been as extensively analyzed.
The design parameters for the reference liquid and air systems, and the effects of
varying them, are essentially the same except for those listed in Table 29.5. Four param-
eters are unique to the reference air system. They are discussed in detail below. The
effects of changing the parameters are summarized as follows:
Effective Heat Transfer Coefficient of Collector (Fig. 29.28). The effective heat
transfer coefficient of the collector (HA/Ac) is the product of the heat transfer coeffi-
cient, times the heat transfer area, divided by the collector area. The heat transfer area
A can be increased by adding fins or by making all sides of the flow passage effective for
heat transfer. The heat transfer coefficient can be raised by increasing the flow velocity
and by increasing the flow channel size.
Note that in this analysis the variation in HA/Ac is made assuming that the air flow
rate is constant at the nominal value of 101/s • (2 cfm/ftc2).
Building Space Heating: Active Systems 85
100 ,
SE/MD
FR/PH
81
Pe rcen t of Tota l Hea t from Solar
80- SM
60
40 -
FR/SE
20- MD/BI
PH
SM
, i ,
00 ,2 4 6 8 10 (Btu/h •°F• ft 2)
2 c
0 10 20 30 40 50 57 (W/°C• m c )
Figure 29.28 Percent of annual load met by solar versus collector effective heat transfer
coefficient, air system.
Below the nominal heat transfer coefficient (HA/As) value of 23 WI°C • m2, (4 Btu/
h • °F • ft2,), solar heat utilization is significantly reduced; however, above this value per-
formance increases only slightly. With the nominal HA/A, value, the difference between
the temperature of the collector surface and the average air temperature would be 14°C
(25°F) when solar heat is being collected at a rate of 300 W/m2, (100 Btu/h • ft2c).
Air Flow Rate (Fig. 29.29). As the air flow rate and heat transfer coefficient are
interdependent, the effect of air flow rate is considered in two ways in this section. Use
of the data presented depends on the problem the designer faces—whether, for example,
he is designing a collector module or is selecting an air flow rate for use in a chosen
collector array.
Effective heat transfer coefficient held constant. As air flow decreases, the
collector AT increases and collector efficiency decreases. A severe solar heat utilization
penalty will result at air flow rates below 5 standard 1/s • m2, (1 standard cfm/ft2,); how-
ever, performance is only slightly improved if the nominal flow rate, 10 standard 1/s • m2,
(2 standard cfm/ft2,), is doubled. Since H is dependent on flow rate in a collector of fixed
geometry, holding HA/A, constant implies changing collector geometry as the air flow
rate is changed.
Effect of air flow rate, collector geometry fixed. In a collector of fixed geometry,
the effective heat transfer coefficient HA/A, is dependent on volumetric flow rate. As
86 Applications of Solar Energy
100
MD
SE
FR HA/Ac = 22.7 W/°Crrq (4 BTU/h°F fq)
PH/61
Heat from Solar
80 SM
60
Percent ofTota l
40 -
20 SE
MD/FR
BI
PH
SM
0
2 3 4 5 (cfm/ftc 2)
L 2
10 20 25 (lis •mc )
Figure 29.29 Percent of annual load met by solar versus collector air flow rate, air
system.
air flow rate is changed, HA/A, varies as (flow rate)" for a fixed geometry. If air flow
rate through a given collector is doubled, HA/A, is increased approximately by a factor
of 2" = 1.7. In addition, the improvement in efficiency resulting from lowered
operating temperature is as indicated in Sec. 29.5. The two effects combined are approXi-
mately additive, so annual solar heat deliveries would increase by 6 to 20% relative to the
nominal values, depending on location and collector size.
Effect of Thermal Storage Heat Capacity (Fig. 29.30). As in the liquid system,
the effect of varying thermal storage capacity is seen to be independent of climatic
conditions except in unusually low capacity ranges. Above storage capacities of 200
kJ/°C • me (10 Btu/°F • ft2c ), there is only slight improvement in annual solar heat
delivery. This capacity corresponds to 230 kg of rock/in (50 lb of rock/ft2c ). The nomi-
nal value of 300 kJ/°C • me (15 Btu/°F • fte2 ) requires 350 kg of rock/m,2 (70 lb of rock/
ft2c ). Most rock has a specific heat of about 0.9 kJ/kg • °C (0.2 Btu/lb • °F), a density of
about 2600 kg/m3 (160 lb/ft3 ), and packs with a void fraction of about 0.40 if the rocks
are all roughly the same size. A practical storage volume is 0.15 to 0.25 m3 of rock/m2
(0.50 to 0.75 ft3 of rock/ft2 ) of collector. This is roughly three times the volume
required to achieve the same heat capacity in a water tank. With a 46-m2 (500 • ft2 )
collector, 13 tons of rock would normally be used that would occupy a volume of 7.7 m3
(270 ft3 ), equivalent to a 2.0-m cube.
Building Space Heating: Active Systems 87
Heat from Solar
Percent ofTota l
5 10 15 20 25 (Btu/h .°F • f tc 2 )
0 100 200 300 400 500 (kJ/°C• rnc 2)
Figure 29.30 Percent of annual load met by solar versus thermal storage heat capacity,
air system.
pebble bed that has a short air flow path (and a large frontal area) requires less fan
power than one with a long air flow path (and small frontal area) because of pressure
drop differences.
Pebble-bed heat capacity and air flow rate can be related to parameters of real
interest to the designer—rock size, bed length, pressure drop—through a performance
map, such as the one in Fig. 29.31. A figure of merit that balances the heat transfer and
pressure drop is the ratio of the number of transfer units to the bed pressure drop,
that is, NTU/AP. For the design value of G given above, a rock diameter of about 2.5 cm
optimizes NTU/AP [35] . This corresponds to an NTU of about 50 for a bed length of
2 m. However, Refs. 32 and 34 indicate that annual system performance is not severely
impaired as long as NTU 5, which corresponds to a bed length of about 16 rock
diameters. If the bed length is decreased below this value, the temperature stratification
becomes ineffective. Thus a rock diameter of 5 to 7.5 cm, which results in a much lower
pressure drop but does not optimize NTU/AP, would be appropriate for applications
designed for high solar heating fractions [35] . In considering a mix of rock sizes, one
should use the smallest diameter in the mix to estimate pressure drop, since small rocks
will fill the interstices between the larger rocks, thereby increasing the air flow resistance
of the bed.
CN .
o (Inches of water )
•
U-
O rsi-
0
E
- 10' 31.5
2500
2 2
Air Flow = 10 liters/s•m (2 cfm/ft )
c c
Mass heat capacity = 305 kJ/°C mc2 (15 Btu/h °F • ft 2)
100 100
250
6•01
Bed Pressure Drop
10°
25
10-2
2.5 - 10-2
10-2
10-3 -0.00315
0.25 - 10-3
2 5 10 20 50 100
As an example of the use of Fig. 29.31, consider the case of a 1.8-m-long pebble
bed, in the air flow direction, with 2.5-cm rocks. The number of transfer units is 42,
the bed length-to-rock diameter ratio is 72, and the pressure drop is 17 Pa (0.07 in. of
water). The corresponding theoretical air pumping power requirement to overcome
pebble bed friction is 0.16 W/m (0.05 Btu/h • fte2 ). With a typical residential collector
area of 50 m2 and a motor/fan efficiency of 20 percent, only 40 W are required for
circulation through the pebble bed.
In summary, a 1- to 2-m length of air travel through a bed of 2- to 4-cm pebbles
(length-to-diameter ratio of 25-100) provides a satisfactory heat transfer and pressure
loss balance.
Detailed heat transfer characteristics of pebble beds are discussed in Chap. 23,
and in Refs. 34 and 35.
A multitude of solar space heating systems are in operation throughout the world. How-
ever, only a few of these are adequately instrumented, mostly under the sponsorship of
the U.S. Department of Energy, so that their performance can be assessed. Three of these
buildings, which have been carefully analyzed, are the CSU I and CSU II residences at
Colorado State University and the National Security and Resources Study Center
(NSRSC) at the Los Alamos Scientific Laboratory. Results of the performance of these
heating systems are summarized below.
The Buildings
CSU I is a two-story residence of 128 m2 heated area with a liquid solar heating
and cooling system [36-38] . The design heating load was computed to be 17.5 kW at
-23°C (59,600 Btu/h at -10° F; corresponding to 19,000 Btu/°F • day). This house was
heated and cooled from August 1974 to November 1976 with a system in which flat-
plate collectors were used. Since January 1977, the system has used evacuated tube
collectors of two types.
CSU II is a nearly identical residence located next to CSU I at Colorado State
University in Ft. Collins, Colorado (latitude 40.6°N) [8, 36] , where the mean annual
heating degree-days (18.3°C or 65°F base) are 3500°C • days (6300°F • days). Its pri-
mary heat supply is a flat-plate air collector and pebble-bed storage system. The heat
load in CSU II is somewhat greater than in CSU I because of higher ventilation require-
ments. Both buildings, although designed as residences, have been used as laboratories
and offices. These two buildings provide an opportunity to compare air and liquid solar
systems under nearly identical conditions.
Heating/Cooling Valve
Air
to Floor
ilegistersi to Auxiliary
Main Pump
DH VP DHW Htr
Heat DHW
Preheat Tank
to Load 4-- Main
\ \ \
Solar/Auxiliary Valve
P imary
Air Water Heat
Chiller Ex.
Mover
Cooling Tower Heat
(Outdoors) 3 Storage
Collect Expansion
V-40-•
A I. Tank
Pump Aux
House Hot
Return •\ Storage
I I ., \1 \11\i Chilled 9 Water
Pump
— — 1/ Water N Boiler
Air
Pump 111\\ 1 \ \
Forced Air
C
Blower Cooling Tower Pump
Figure 29.32 Liquid solar system schematic CSU I, Fort Collins, Colorado.
A2Jau3 J Pl os jo suoppon dde
Building Space Heating: Active Systems 91
a nonselective absorber surface, two layers of glass, and fiberglass insulation. Fluid is
pumped through the collectors at the rate of 0.015 1/s • m2 (0.022 gal/min • ft2 ) of col-
lector area. The first of the two types of evacuated tube collectors consists of 216 Pyrex
tubes each enclosing a thin, flat copper absorber plate with a black chrome selective sur-
face. Liquid passes through copper tubing bonded to the absorber plate and sealed into
one end of the glass cylinder. The absorber surface totals 39.9 m2 (429 ft2 ), aperture
area of the 36 modules is 50.0 m2 (538 ft2 ), and gross occupied "roof" area is 75.2 m2
(809 ft2 ). Both collector arrays face due south at a tilt angle of 45°.
As shown in Fig. 29.32, heat is transferred in a shell-and-tube heat exchanger from
the collector fluid to water circulated from and to a 4280-1(1130-gal) storage tank
located in an insulated basement room. Hot water from storage is pumped to a heat
exchanger in the air distribution duct for space heating, or to the generator of a lithium
bromide absorption chiller for space cooling, and to a heat exchanger loop for domestic
hot water supply. For part of the several-year operation, the domestic hot water
exchanger was used in the collector liquid loop rather than in a storage loop. A gas-fired
boiler provides auxiliary space heating and cooling and an automatic gas water heater
supplements the solar hot water supply. Warm or cool air is distributed to the rooms of
the building in a conventional manner.
Air is used in CSU II as the heat collection and heat transport fluid. The site-built
collector, of 64 m2 net absorber area and 67.1 m2 (722 ft2 ) gross area, is similar to the
liquid heating flat-plate collector, with a black-painted steel absorbing surface covered by
two glass panes and insulated beneath with Fiberglass. The absorber is a flat air duct
about 1 cm (0.4 in.) high through which air at 9 1/s • m (1.8 cfm/ftD flows. Air passes
in parallel through 24 collector units at a velocity of 4 m/s in the absorber passages.
Hot air from the collectors (Fig. 29.33) is passed through the pebble-bed heat
storage unit, or is supplied directly to the heated space when required. Air from the
collector and air from the house are passed through the pebble bed in opposite directions.
This arrangement establishes temperature stratification in the pebble bed, which is an
important requirement in a solar air heating system. The pebble bed consists of 18,200
kg (20 tons) of rock of 2 to 4 cm diameter. Domestic hot water in CSU II is solar heated
in a duct coil located in the hot air stream leaving the collectors. Auxiliary heat is
supplied by a gas furnace.
Cold
Water
MD- Motorized Damper Supply
0 Hot
Water
Supply
Heat
Exchanger
0 I
W/S
Damper
1
Summer
Blower
MD MD Auxil ary DHW
MD Heater
Air Filter DHW Preheater
Evaporative Main Blower
Cooler
MD MD
to
Rooms
MD
MD Auxiliary
Furnace
Outside Air
Pebble- bed
Storage
MD MD
-0- .0— from
Rooms
Figure 29.33 Air solar system schematic, CSU II, Fort Collins, Colorado.
During the 5-month heating data collection period, the solar system in CSU I
provided 43,660 MJ (41.4 X 106 Btu) for space and DHW heating, which was 71.1 per-
cent of the load. Solar supply during November 1976 was unusually low because of
severe weather and poor solar conditions during portions of the month.
Also shown in Fig. 29.34 is the performance of the flat-plate collector in October-
November 1976, and the evacuated tube collector from January to April 1977. During
the January-April period, the evacuated tube collector delivered 57 percent of the total
solar energy falling on the aperture area, which was equivalent to 63 percent of the solar
energy received during collector pump operation.
A summary of monthly energy use for space and DHW heating in CSU II is shown
in Fig. 29.35. The data are for the air collector system during the seven-month period
from November 1976 through May 1977. During the data collection period, the solar
Building Space Heating: Active Systems 93
Energy Supply Rate (MJ/day)
Ei Auxiliary Supplied
lectors Installed
600-
559
12E3Solar Supplied
500 476
7104
400 - 396 379
Col
343 338 330
300 84%
2 285
69%
/// 227
200
170
100 0
>
0
Nov Dec Jan Feb Mar Apr*
19.7 1500
20 1000
18.6
2
500 •.•••
6 300 - ....... 200
4 250 200-
100
2 :.14% 100
.13%
0 0
Oct Nov Jan Feb Mar Apr*
* Collector area reduced during part of April, for average of 35.0 m2.
Figure 29.34 Solar heat utilization and collector performance for liquid system in CSU I,
October 1976-April 1977.
94 Applications of Solar Energy
71Auxiliary Supplied
ED Solar Supplied
510
500
400
300
200
100
LL
400
rn
a) 4 200
• 2
0
Nov Dec Jan Feb Mar Apr May
Figure 29.35 Solar heat utilization and collector performance for air system in CSU II,
November 1976-May 1977.
system provided 55,400 MJ (52.5 X 106 Btu) for space and DHW heating, which was
73.6 percent of the load.
In addition, Fig. 29.35 shows the collector performance for CSU II during the
same collection period. The collector efficiency, based on the solar energy incident on
the total collector area is 27 percent. Based on the insolation when the collector is
operating, the efficiency is 31 percent.
Figure 2936 shows the substantial performance improvements achieved by replac-
ing site-built flat-plate collectors with well-designed factory-built collectors of air and
Building Space Heating: Active Systems 95
19.1 19.3
18.7
Days of
Usable Data 30 30 31
Dec. Jan.
1978 1979
Figure 29.36 Solar collection in air and liquid systems with commercial collector in
CSU Solar Houses II and III.
liquid types. Both collectors have black chrome absorber surfaces and single glass covers.
The system in CSU III, very similar to that in CSU I, was operated with antifreeze liquid
and heat exchange to water storage, whereas air and pebble-bed storage were used in
CSU II. By comparing Figures 29.34, 29.35, and 29.36, it may be observed that heat
collection and efficiency of systems employing commercial flat-plate air and liquid col-
lectors were considerably higher than those in which site-built collectors were used. A
four-month solar collection efficiency of 34.3 percent was obtained with the system in
which commercial air collectors were used, and in the same four months of an earlier
year, the site-built collectors operated at only 29.7 percent efficiency. Approximately
15 percent more heat was supplied by the new collectors, on a normalized solar basis.
An even greater improvement was obtained with the commercial liquid collector, the
four-month efficiency being 34.2 percent as compared with efficiencies obtained, in
previous years with the site-built flat-plate collector in CSU I of about 25 percent.
It is seen in Figure 29.36 that the air system delivered about 10 percent more heat
than the liquid type per unit collector area in midwinter (December and January), nearly
the same in February, and about 14 percent less in March. These differences are charac-
teristic of the system types, air being more efficient than liquid in colder and cloudier
weather, mainly because of the higher utilization possible in air collector/pebble-bed
systems. An important reason for the superior midwinter perfrmance is the capability
96 Applications of Solar Energy
of the air collector, when supplied with relatively cool air from the rooms or from a
stratified pebble-bed, to collect heat over a greater portion of the daytime hours.
Data for the preceding winter, with the flat-plate collector system in CSU I, were
obtained before the system was provided with DHW heating in the collector loop and
with improved control features [36] . Data were also collected for the air system in
CSU II during this winter [36]. Performance of these two systems is compared below for
the period January 14 through May 15, 1976. Solar preheating of domestic hot water
had not commenced in CSU II, so only the space heating performance is compared.
During the data acquisition period from January 14 to May 15, the liquid system
in CSU I provided 18,100 MJ (17.2 X 106 Btu) of solar energy for space heating, which
was 71.7 percent of the load. During the same period, the solar air heating system pro-
vided 29,380 MJ (27.9 X 106 Btu) for space heating in CSU II, which was 69.8 percent
of the considerably larger load.
Electrical energy consumption by the solar equipment in the two systems was also
compared. Electrical energy use can be presented as the useful solar heat delivered for
space heating divided by total electricity consumption. This electrical "coefficient of
performance" for solar space heating was 11 for CSU I and 15 for CSU II. The electrical
energy for distribution of heat in both buildings has been included so a portion of this
usage is equivalent to that in a conventional heating system.
The liquid system in CSU I offers the advantage of compact size within the building
space. The water storage tank requires only one-third the volume and about half the
floor space of the pebble-bed storage unit, and has the same heat capacity. The air
handling equipment occupies more than twice the space of liquid-handling equipment
and pipes, but this difference can be virtually eliminated by consolidation of dampers,
ducts, and blower into a modular unit.
Repair and maintenance have been higher on the liquid system. Antifreeze concen-
tration and pH of the collector fluid must be checked at least twice a year. A collector
fluid filter has had to be replaced every 3 months because of accumulation of trapped
solids. Liquid leakage has occurred on a few occasions. The leaks have developed at the
clamped rubber connections. More reliable connections are available, but leakage remains
a factor to be considered.
The NSRSC building, located in Los Alamos, New Mexico (latitude 35.8°N), has a heated
and air conditioned floor area of 5574 m2 on three floors [39] . Long-term average
heating degree-days at the site are 3500°C-days (6300° F-days) per year. The lower two
floors house a library and the upper floor houses meeting rooms, offices, and conference
areas. Energy conservation was a primary objective of the building design. The building
shell is heavily insulated, and the HVAC system makes use of air recirculation, heat
recovery, and complete shutdown at night.
The HVAC system is a two-zone (perimeter and interior) variable-air-volume
system with separate supply fans and cooling coils for each zone. The terminal boxes
Solar Collectors
Air Washer
Heat
Recovery Exhaust
Heat Cooling •
Tower • Unit Heating
Exchanger Cooling Coil
* Coils
4
Exhaust
••
is:••••-4-;.•.:9
4
Chiller
1
Exhaust
n- •
Fresh Air 00 •
L-4.----.1.-...- Filter
1 I
1 4
o —48)- —1 I
• '4 iI Air
i ( 4)
Storage 1 T 11
I Washer
s 4- -I
• Tank - 4
1 rok
1 1
1 ..... I Steam Heat Exchanger Interior Perimeter
Zone Zone
4•1 11=1.
Storage Tank
Figure 29.37 Solar system schematic for National Security and Resources Study Center, Los Alamos, New Mexico.
98 Applications of Solar Energy
in the perimeter zone only have hot water reheat coils. Most of the heat in the building
is provided by the people and the lights, so that the interior zone, being isolated from
the outside walls, requires only cooling throughout the year. The air-handling system
features recirculation of inside air and a heat-pipe heat recovery unit in the perimeter
zone system. Return air cools the light fixtures. A mechanical system schematic is shown
in Fig. 29 37.
A planar array of site-built liquid collectors forms the roof of the equipment room.
Solar energy heats two thermal storage tanks that, in turn, supply hot water to the
perimeter reheat coils and to either an absorption chiller or an experimental solar
Rankine cycle chiller.
Although the system has been in continuous operation since March 1977, only
heating season performance data from November 1977 through April 1978 are reported
here.
Month
Figure 29.38 Solar heat utilization for the National Security and Resources Study
Center, November 1977-April 1978.
Building Space Heating: Active Systems 99
30
Energy, MJ /m Day
2
Month
Figure 29.39 Collector performance for the National Security and Resources Study
Center, November 1977-April 1978.
The hot water temperature required for reheat is scheduled based on outside air
temperature, with 60°C (140° F) water required for -18°C (0° F) outside temperature,
and 38°C (100°F) water for 21°C (70° F) outside air temperature. If this schedule cannot
be met by solar storage, the auxiliary steam heat exchanger system provides water at the
required temperature.
About 160 channels of instrumentation are installed in the building for evaluating
system performance. Basic energy consumption measurements are made on all of the
energy subsystems. These measurements consist of flow rates and temperature differ-
ences. In addition, electrical power consumptions of all of the pumps and fans in the
system are measured. Solar flux is measured in the plane of the collector and in the
horizontal plane. Data acquisition is controlled by a PDP-11/34 computer.
Solar heat delivery during the 1977-1978 heating season is shown in Fig. 29.38;
collector performance is shown in Fig. 29.39 [39] . During this period, the heating
system delivered 680 GJ (645 X 106 Btu), which was 79 percent of the heating load.
Average collector efficiency during this period, based on total incident solar energy, was
32 percent.
The normal maximum temperature for hot water storage was 82° C (180°F). Hot
water flow from the tank varies with the heating demand, but was nominally 3.81/s
(60 gal/min). Pumping power for hot water distribution was 1.2 kW.
100 Applications of Solar Energy
Electricity consumption by the solar system pumps was measured to be 5.4 per-
cent of the solar energy delivered to the load [39] . Pumping power for the oil coolant
was 9.6 kW.
REFERENCES
1. Intermediate Minimum Property Standards for Solar Heating and Domestic Hot
Water Systems, U.S. Department of Commerce, National Bureau of Standards
report NBSIR 76-1059, April 1976.
2. Interim Performance Criteria for Solar Heating and Cooling Systems in Commercial
Buildings, U.S. Department of Commerce, National Bureau of Standards report
NBSIR 76-1187, November 1976.
3. P. Steadman, Energy, Environment, and Building, Cambridge University Press,
New York, 1975.
4. DOE Facilities Solar Design Handbook, U.S. Department of Energy report DOE/
AD-0006/1, January 1978.
5. Solar Heating Systems Design Manual, ITT Training and Education Department,
Fluid Handling Division, Bulletin TESE-576, 1976.
6. W. M. Kays and A. L. London, Compact Heat Exchangers, 2nd ed., McGraw-Hill,
New York, 1964.
7. G. 0. G. Lof, M. M. El Wakil, and J. P. Chiou, "Design and Performance of
Domestic Heating System Employing Solar Heated Air-The Colorado Solar House,"
Proceedings of the United Nations Conference on New Sources of Energy, Vol. 5,
p. 185, Rome, 1961.
8. S. Karaki, P. R. Armstrong, and T. N. Bechtel, "Evaluation of a Residential Solar
Air Heating and Cooling System," report COO-2868-3, prepared for the U.S.
Department of Energy by the Solar Energy Applications Laboratory, Colorado
State University, December 1977.
9. G. 0. G. Lof, "Systems for Space Heating with Solar Energy," Chapter XII in
R. C. Jordan and B. Y. H. Liu, eds., Applications of Solar Energy for Heating and
Cooling of Buildings, ASHRAE GRP 170, 1977.
10. H. Saha, "Heat Transfer Characteristics of Water Filled Cans as Solar Thermal
Storage Medium: A Comparative Test Data Analysis," Proceedings of the 1978
Annual Meeting of the American Section of the International Solar Energy Society,
Vol. 2.1, pp. 664-670, August 1978.
11. J. D. Balcomb and J. C. Hedstrom, "Thermal Performance of the LASL Solar Heated
Mobile/Modular Home," Proceedings of the 1977 Annual Meeting of the American
Section of the International Solar Energy Society, Vol. 1, pp. 13-22, June 1977.
12. R. L. Oonk, L. E. Shaw, B. E. Cole-Appel, and G. 0. G. Lof, "A Method of Com-
paring Flat-Plate Air and Liquid Solar Collectors for Use in Space Heating Appli-
cations," Proceedings of the Joint Conference of the American Section of the
International Solar Energy Society and the Solar Energy Society of Canada, Inc.,
Vol. 2, pp. 83-93, Winnipeg, Manitoba, August 1976.
13. T. L. Freeman, J. W. Mitchell, and T. E. Audit, "Performance of Combined Solar-
Heat Pump Systems," Proceedings of the 1978 Annual Meeting of the American
Section of the International Solar Energy Society, Vol. 2.1, pp. 293-298, August
1978.
Building Space Heating: Active Systems 101
SOLAR COOLING
30.1 INTRODUCTION
103
104 Applications of Solar Energy
the process, review some of the major recent published work, and note what we consider
to be important development areas that may make the processes technically successful
and economically practical. Many of the important references in each of the three classes
of processes are cited.
The future of many of the possible methods will depend on developments beyond
the cooling process. The limitations in operating temperature of collectors clearly limits
what can be expected of solar cooling processes. As collector operating temperatures are
pushed upward, storage may become a critical problem.
From the point of view of use of a conventional energy source, there is a single
index of performance for rating cooling processes, that is, the COP (coefficient of per-
formance), the ratio of the amount of cooling to the energy required. For solar operation
there are two additional factors, the temperature level required in the solar collector to
drive the process and the ratio of cooling produced to solar energy incident on the
collector (a system COP). As solar processes are inevitably transient in their operation,
the energy ratios and temperatures will vary with time, and COP based on long-term
integrated performance provides an appropriate index of performance.
Operation of absorption air conditioners with energy from flat-plate collector and storage
systems is the most common approach to solar cooling today. A schematic of a solar
absorption cooler is shown in Fig. 30.1. In essence, cooling is accomplished as the
generator of the absorption cooler is supplied with heat by a fluid pumped from the
collector-storage system or from auxiliary. Heat is supplied to a solution of refrigerant
in absorbent in the generator, where refigerant is distilled off. The refrigerant is con-
densed, and goes through a pressure-reducing valve to the evaporator where it evaporates
and cools air or water. The refrigerant vapor goes to the absorber where it comes in con-
tact with the solution which is weak in refrigerant and which flows from the generator.
The vapor is absorbed in the solution, which is then returned to the generator. A heat
exchanger is used for sensible heat recovery and greatly improves cooler COP. Pumping
to move absorbent solution may be by mechanical means or by vapor-lift pumping in the
generator for low-pressure systems. Low-pressure (LiBr-H2 0) systems require water
cooling of absorber and condenser. Systems of the type shown in Fig. 30.1 (or variations
Solar Cooling 105
To
cooled
space
Cooling
water
in
on it using other methods of energy storage, auxiliary energy input, multiple stage
coolers, etc.) have been the basis of most of the experience to date with solar air
conditioning.
A. LiBr-H20 Coolers
The coolers used in most experiments to date are LiBr-H2 0 machines with water-cooled
absorber and condenser. A pressure-temperature-concentration equilibrium diagram for
LiBr and H2 0 is shown [1] in Fig. 30.2. The ideal steps in the solution cycle are indi-
cated on the cycle shown on the diagram. The pressure in the condenser and generator
is fixed largely by the condenser fluid coolant temperature, and also by the temperature
drops across heat transfer surfaces in the generator and condenser. The pressure in the
evaporator and absorber is fixed by the temperature of the cooling fluid to the absorber
and by the temperature drop across the heat transfer surfaces in the evaporator and
absorber. Thus, to keep the generator temperatures within the limits imposed by the
characteristics of flat-plate collectors, the critical design factors and operational parame-
ters include effectiveness of the heat exchangers and coolant temperature. Common
practice in solar experiments has been to use water-cooled absorbers and condensers,
which in turn requires a cooling tower.
The pressure differences between the high- and low-pressure sides of LiBr-H2 0
systems are low enough that these systems can use a vapor-lift pump and gravity return
from absorber to generator, thus obviating the need for a mechanical pump to move the
solution from the low-pressure to the high-pressure side.
Many of the machines used to date have nearly constant COP (ratio of cooling
produced to energy supplied to the generator) as the generator temperature varies over
the operating range, as long as it is above a minimum value. The COP of LiBr-H2 O
coolers is often in the range 0.6 to 0.8. The effect of variation in the solar energy supply
106 Applications of Solar Energy
200
co
E too -
E
CC
rn ,
U)
Twto GE1s4=190
CC 10
a_ AO"
cr -(2
co TGEN
a.
TABS 85
1 .t." TCP0,1-4NT-T5,
50 icio 150 200 250 300
TEMPERATURE ,°F
Figure 30.2 Pressure-temperature-concentration diagram for LiBr-1-12 0, with idealized
cooling cycle. Critical temperatures are shown. Adapted from Ref. 1.
temperature to the generator is to vary the capacity of the cooler. Water is used as a
coolant, and the generator temperatures may be in the range 75 to 95°C (167 to 203°F).
Major problem areas of systems using solar-operated LiBr-H2 0 coolers include the
following. Supply temperatures of the fluid to the generator must be higher than the
generator temperatures. There is a "squeeze" between the supply temperatures required
and the upper temperature limits of 100° C for unpressurized water storage tanks. Opera-
tion at 100°C is difficult with many collector types, particularly flat-plate collectors.
Cooling towers are needed in these systems.
Early experiments on solar operation of LiBr-H2 0 coolers [2, 3] used commercial
3- to 5-ton machines without significant modification for the solar experiments. More
recently, coolers designed for solar energy supply to generators have been developed [4].
Heat exchangers have been modified to minimize temperature drops, submergence in
generators is reduced to minimize rise in solution boiling point due to static heads of
solution, and mechanical pumps have been used rather than vapor-lift pumps, These
developments all have the effect of maintaining or improving COP while reducing genera-
tor temperature requirements.
Solar-operated LiBr-H2 0 absorption coolers are now installed and being installed
in both residences and large buildings. Coolers designed for solar operation are available
in 2.5-ton [5], 3-ton, and 25 ton [6] capacities. Experiments on residential-scale appli-
cations [7] and on institutional buildings [8] are now being made.
Simulation studies of solar air conditioning have been used to explore the long-term
performance of solar cooling system performance, to assist in developing designs, and to
provide the basis of economic studies. One of the earliest of these studies [9] dealt with
the costs of solar heating and cooling systems in eight U.S. cities and concluded that,
particularly in the southern parts of the country, the combined functions of solar heating
and cooling were more economical than either heating or cooling alone. Another simula-
tion study [10] showed monthly performance for a system on a building in Albuquerque,
New Mexico, with four collector areas. A follow-on study applied these techniques to
Solar Cooling 107
B. NH 3 -H 2 0 Coolers
The schematic diagram of an ammonia-water cooler is similar to that of Fig. 30.1, except
that a rectifying section must be added to the top of the generator to reduce the water
vapor content of the vapor stream going to the condenser. Pressure-temperature-concen-
tration data for the ammonia-water system are shown in Fig. 30.4. The basic solution
processes are similar to those of the LiBr-H2 0 system, except that the pressures and
pressure differences are much higher. Mechanical pumps are required to return solutions
110
COP = 0.67
100
C3
W 90
I-
4
Ir
6
1' 80
aR
,-
I— 70
U
*4
CL
6%
U 60
of I
70 80 90 Figure 30.3 An empirical model of a LiBr-
WATER TEMP TO GENERATOR ,°C H2 0 absorption cooler. Adapted from Ref. 11
108 Applications of Solar Energy
P-T-X,AMMONIA-WATER
350
300
250
200
150
p
N
0. 100
iu
cc
n 75
N
co
LAJ
cr 50
o_
30
20
10
0 80 240 320
TEMPERATURE °F
from the absorber to the generator. In many applications the condenser and absorber are
air cooled, with generator temperatures in a range of 125 to 175°C (257 to 338°F). In
applications where water cooling is used, generator temperature may be in the range 95
to 120°C (203 to 248°F). Idealized cycles for both air-cooled and water cooled systems
are shown on Fig. 30.4. The condensing temperatures for air-cooled condensers
correspond to much higher generator temperatures than do those for the liquid-cooled
systems.
There has been relatively little done on experimental operation of NH3 -H2 0
systems with solar energy. The generator temperatures required in today's commercial
coolers using air-cooled condensers and absorbers are too high for present flat-plate
collectors. Two approaches can be taken to this problem: Increase the temperature
range of operation of collectors or reduce the generator temperature of the cooler. Work
on NH3 -H2 0 systems has been directed at development of cycles using higher concentra-
tions of NH3 [15, 16] and high effectiveness components to reduce generator tempera-
ture requirements. Collector developments such as evacuated tubular collectors, CPC
(compound parabolic concentrators) or other focusing collectors, or well-designed flat-
plate collectors with selective surface absorbers and low-reflectance covers may provide
means of supplying energy at higher generator temperatures. In this case, new means of
energy storage in the range 100 to 175°C (212 to 347°F) would have to be available.
Solar Cooling 109
VAPOR MAKE- UP
WATER
EN FLAT- /got
PLATE COOLING -
COLLECTOR-4 WATER
ABSORBER -1 EVAPORATOR-5
/1! 2
TRAP 6
7
Evaporative Exhaust
Coolers
(a)
Ventilation
Mode
EXHAUST
Humidity
OO
Figure 30.6 (a) Schematic of solar-
Temperature MEC system in vent mode. (b) Solar-
(b) MEC states on psychrometric diagram.
Solar Cooling 111
For maximum capacity and COP, the components should be ideal. That is, the
evaporative coolers saturate to the saturation line, and the effectiveness of the mass and
heat exchangers are unity. In reality, these components are not ideal. The effect of less
than ideal components is to increase the amount of thermal energy required to meet the
load.
Evaporative coolers and rotary heat exchangers are well understood, and their
performance can be predicted accurately. The analysis of rotary components in which
heat and mass transfer occur simultaneously is more difficult (see Refs. 30-33). The
analysis calculates the performance by analogy with that for heat transfer in rotary heat
exchange wheels. Characteristic potentials and capacity ratios analogous to temperature
and specific heat are formulated that allow use of rotary heat exchanger theory.
The development of desiccant systems for solar energy applications is in a very
early stage, and few experimental data are available. Performance predictions show that
a seasonal COP of 0.3 to 0.75 is possible, depending on the system and climate. An
interesting result of this study is the development of a control strategy that meets various
combinations of sensible and latent loads with minimum energy expenditure. It was
found necessary to vary the degree of humidification of both evaporative coolers and the
regenerating stream temperature in order to achieve the desired comfort conditions in
the room.
Open-cycle cooling systems using liquid triethylene glycol as the desiccant are in
commercial manufacture and use [34]; their use in systems in which regeneration is
accomplished by solar energy has been explored in preliminary experiments [22, 35].
A solar cooling system that has received some attention in recent years couples a solar-
powered Rankine-cycle engine with a more or less conventional air conditioning system.
The problems associated with this system are basically the problems of generating
mechanical energy from solar energy, and coupling of the air conditioning equipment
to the solar energy source for part-load operation.
Studies on Rankine-cycle solar air conditioning systems to date have concentrated
on theoretical analysis [36-40]. The experimental system in the Honeywell mobile
laboratory [41, 42] has been in operation for nearly 2 years, but no long-term operating
data are yet available. United Technology [43] has constructed and tested a laboratory
unit, and some performance data are available.
The problems associated with solar-Rankine air conditioning systems are substan-
tial. Conversion of solar energy to mechanical energy has been a long-term aim. Genera-
tion of solar power in large-scale solar energy systems is not yet competitive with power
generated from conventional fuels; it is difficult to envision a household-scale system that
will convert solar to mechanical energy at less cost per unit power output than a large-
scale solar system. Consequently, at the present state of development, solar-Rankine air
conditioning does not appear to be cost competitive, and the utility of the method will
depend on future developments.
A simplified schematic of a Rankine-cycle cooling system is shown in Fig. 30.7a.
Energy is transferred from the storage tank to a heat exchanger and thence to a heat
112 Applications of Solar Energy
engine. The heat engine exchanges energy with the surroundings and produces work.
As shown in Fig. 30.7b, the efficiency of the solar collector decreases as the operating
temperature increases, whereas the efficiency of the heat engine will increase with
operating temperature. Figure 30.7c shows the overall system efficiency for converting
solar energy to mechanical work and indicates that an optimum operating temperature
will exist for steady-state operation.
A schematic diagram of a Rankine heat engine is shown in Fig. 30.8a, and the
corresponding temperature-entropy diagram is shown in Fig. 30.8b. This cycle differs
from conventional power plant-cycle practice of superheating and extracting steam to
prevent moisture erosion at turbine blades and to increase the cycle efficiency. In a
small solar energy system, superheating may not be desirable because of the increased
temperature requirements of the collector, and extraction is not economical. In order
to avoid moisture in the turbine, a proper fluid must be selected; this is usually a fluori-
nated hydrocarbon refrigerant. For the particular fluid shown in Fig. 30.8b, the outlet
temperature of the turbine is signficantly higher than in the condenser and a regenerator
has been used to preheat the fluid before it enters the boiler.
The thermodynamic analysis of such a cycle at a design point is not difficult.
However, the analysis under operation with varying off-design conditions is difficult.
Control to meet time-varying loads such as would exist in solar energy applications is
complicated.
With a constant-speed air conditioner, the output of the engine will seldom match
the required power input. When the engine output is greater than needed the matching
5
-BOILER
TEMPE RATU RE
3-
EXPANDER
REGENERATOR
7
CONDENSOR Figure 30.8 (a) Rankine power cycle
ENTROPY
with regeneration. (b) Temperature-
PUMP
entropy diagram of Rankine power
(a) (b) cycle.
Solar Cooling 113
can be accomplished either by throttling the energy supply to the engine, which means
that the engine will work at an off-design condition and available energy will be wasted,
or by producing excess electrical energy for other purposes. When the engine output is
less than that required by the air conditioner, auxiliary energy must be supplied either
in the form of heat to the boiler or electricity to the air conditioner electric motor.
Because of the low cycle thermal efficiency, it is probably better to add energy directly
as electrical power to run the compressor. The system could be designed to operate at
variable speed. However, the operation of the air conditioner will be off-design with
subsequent reduction in its output. Thus, systems cannot be designed with confidence
at the steady-state condition, but instead, integrated yearly performance must be
estimated.
A preliminary study [36] of a solar-source Rankine-cycle air conditioning system
has explored some of the important design features. Seasonal simulations were carried
out to determine performance in Albuquerque, New Mexico, and Miami, Florida. It was
found that for any given collector area there was an optimal engine size, and that the
fraction of the load met by solar decreased if the engine was either under- or oversized.
Storage tank size was also found to affect performance. Larger tank sizes allowed the
system to produce more power than smaller sizes, mainly during periods of low demand.
The study points out the many unknown factors involved in the design and control of
these systems. Further research is needed into the control variables and the evaluation
of off-design performance. It appears that such systems may be too complex for
residences but may be feasible for commercial buildings.
When solar energy equipment is installed for the purpose of heating a building, some of
this equipment can be used to cool the building. However, operation is not carried out
with energy supplied by the sun. We note three such examples: (a) heat pump systems;
(b) sky radiation systems; and (c) night cooling storage systems. The objective here is not
to discuss these systems in detail, but to point out the features that allow cooling to be
obtained.
In a solar-heat pump system designed to supply heating for a residence, the heat
pump can be used to supply cooling during the summer [44, 45]. In this case cooling
is supplied in a conventional manner by an electrically operated heat pump, and solar
energy is not utilized. The proceedings of a workshop conducted by ERDA [46]
describes research on solar energy heat-pump systems. Although a combined solar-heat
pump system may not be cost effective in itself, the use of the heat pump as an air
conditioner allows part of the costs of the heat pump to be charged to summertime use.
The second method relies on dissipating energy to the night sky by radiation, using
the same surface that is used for solar energy collection. A few such systems have been
built, for example, Bliss [47] and Hay [48] . From careful experiments in Arizona, Bliss
found that on a monthly average his uncovered collectors could dissipate 4.1 MJ/m2
(360 Btu/ft2 ) per night to the night sky. Such a system normally does not provide
114 Applications of Solar Energy
dehumidification, and dry climates are necessary in order to provide sufficiently low
effective sky temperatures for these systems to function.
Night cooling of the pebble-bed storage has been discussed by Close et al. [49,50]
and applied experimentally by Colorado State University [51]. During the summer
months, when the rock bed is not used for storing solar heat, cool night air can be drawn
into an evaporative cooler and then used to chill the rock pile. At the end of the night
the rock pile will be near to the wet-bulb temperature. During the day, dry outside air
can be drawn into the rock pile and cooled. As the air emerges from the rock pile, its
temperature can be further reduced in an evaporative cooler. Low atmospheric moisture
is needed in order for these systems to function.
30.7 CONCLUSIONS
It is clear that there is a wide spectrum of potentially interesting methods for using solar
energy for air conditioning, including many not mentioned in this brief chapter (see, for
example, Refs. 18 and 52. Some of these have been studied extensively, whereas some
are still completely undeveloped. It is not clear at this time which of these various
processes will find the widest application, or whether several of them will succeed in
various climatic zones. The future of cooling is linked to component (collector and
storage) development. It is also closely interrelated with developments in solar heating,
as the economics of the two processes appear to be better than that of either one alone.
ACKNOWLEDGMENTS
This chapter is a revision and update of a review paper presented at the Winnipeg meeting
of the American Section of the International Solar Energy Society and the Solar Energy
Society of Canada, August 15-20,1976.
REFERENCES
*Reprinted with permission from Annual Review of Energy, Vol. 3. Copyright © 1978 by Annual
Reviews, Inc. All rights reserved.
119
120 Applications of Solar Energy
Most definitions and discussions of solar heating and cooling are restricted to system
designs that use manufactured hardware and components that are either off-the-shelf
(e.g., pumps, fans, heat exchangers) or can quickly be mass produced, marketed, and
serviced (e.g., collectors, fluids, controls). "Solar systems" are usually assumed to require
an assemblage of components that includes collectors, thermal storage, a thermal energy
transport system between the collector array and storage, and another between storage
and the heated or cooled building. Both transport systems are assumed to use pumps or
fans. Further, the components usually are attached to or installed in a building without
having much effect on the building's architectural fabric (roof, walls, floor, etc.). Such
systems are often referred to as "active."
The terms "passive" and "active" solar energy systems represent the extremes in
a continuum of building and mechanical systems used to convert the sun's energy to
useful thermal energy for the heating and cooling of buildings.
Meanwhile, energy conservation in buildings is usually restricted to meaning the
reduction of energy consumption, whether the conserved energy is renewable (e.g.,
solar) or nonrenewable (e.g., fossil fuel). Thus, although solar energy can be used to
reduce consumption of fossil fuels, it is not generally regarded as an energy-conservation
feature. Conversely, many simple architectural features, such as the combination of
south-facing glass and thermal mass, are not usually considered energy-conservation
techniques.
There is a growing awareness that these definitions of solar heating and cooling
systems and of energy conservation exclude dozens of simple methods for using solar
and other natural forms of energy. The use of the term "passive" was introduced several
years ago to describe those methods of solar energy that do not use mechanical power
(e.g., pumps, fans) but instead use natural energy flows for the transfer of thermal energy
into, out of, and through a building.
As discussed at the Definition of Passive Systems Workshop at the Passive Solar
Heating and Cooling Conference in Albuquerque, New Mexico, May 18-19, 1976, the
nature of the components that comprise the various systems is important in distinguish-
ing between active and passive systems. Quoting from p. 168 of the Conference
Proceedings:
Although these distinctions are generally believed applicable, they are not by them-
selves sufficient to distinguish between active and passive systems. For example, attached
solar greenhouses and convective loops are not greatly affected by the buildings to which
they are applied.
Passive Solar Design 121
A lack of consensus has existed over the definition of passive systems, primarily
because of this "architectural, inherent, and intrinsic" approach. Because passive solar
energy components may also serve functions other than that of the conversion and stor-
age of the sun's energy, the classic question is "When is a window a window, and when is
it a solar collector?" Partly because active solar energy systems can be separated from the
fabric of the house and, in fact, operated independently of the building's thermal require-
ments, active systems have lent themselves to more rigorous study and acceptance;
passive systems have not been afforded the same rigorous study and acceptance.
If we are required to define passive solar design formally in a way analogous to
"active" solar, the following would find general acceptance: Designs that consist of
architectural features, components, and/or assemblages thereof that make use of the
natural transfer of solar-generated thermal energy (i.e., without the use of fans or pumps)
for the purpose of water heating, space heating, and/or space cooling. Thermal energy
transfer in and out of buildings, in and out of thermal energy storage, and around and
through a conditioned space is by natural means. Control elements and other comfort-
regulating devices can be incorporated into passive solar designs to suit the needs of the
building's occupants. Thermal energy storage and control elements are not considered
necessary components of many passive designs.
This chapter, of course, discusses passive solar design. As may become apparent
here, passive's most advantageous characteristic may at the same time be its greatest
shortcoming. There are a wide variety of established approaches from which to select
the most appropriate for a given climate, site, and building type. Although this variety
gives passive considerable versatility, it requires that the selection of the proper system
type be based on more than just a passing interest in solar heating and cooling. A com-
pensating factor is that whereas any given system design may be optimized for thermal
and economic performance, passive system performance is usually fairly stable over many
design variations. In general, passive solar designs are simple concepts to understand and
to build. However, anlayzing, predicting and evaluating their thermal performance is a
relatively complex task.
Knowledge about passive solar systems has not been greatly developed outside the main-
stream of the active solar field. In part this is due to the lack of a common interest base.
The cause-oriented promotional and lobbying groups who have spurred public acceptance
of solar energy for heating and cooling have often disregarded passive systems. More
seriously, government-funded research and information-dissemination efforts have, at the
time of this writing, consistently ignored passive solar technologies. The one exception
has been the Los Alamos Scientific Laboratories, which has, through the work of Douglas
Balcomb and others, significantly advanced our understanding of passive solar systems.
This history stems in part from the following:
1. The fact that many passive design techniques are extensions of and closely
interwined with, energy conserving design, which is seen as separate from the
solar energy field.
122 Applications of Solar Energy
In order to speed the implementation of passive solar heating and cooling, general
public recognition of the additional fact that these techniques are often more cost-
effective ways to reduce nonrenewable energy use is required. There is the need to ensure
that passive design and engineering finds its appropriate niche in the existing spectrum of
professional responsibilities. A distribution of skills and knowledge in the design, engi-
neering, and construction fields paralleling that being established for active solar systems
is needed for passive heating and cooling as well.
This knowledge is developing as techniques for computer simulation and "rules of
thumb" for design purposes are. For building purposes, specific construction details and
materials specifications accepted by practitioners in the field can be expected to emerge
in manuals of accepted practice.
The societal impacts of passive solar heating and cooling are varied. Currently
taxation policies and building codes are affected by federal and state governments' desire
to establish certain incentives for widespread solar use, in pursuit of overall national
energy goals. Issues of financing are likely to have a longer-term impact, as the precise
cost differentials between conventional and passive solar construction are hard to clarify.
Public acceptance of passive solar heating and cooling will come with an increased
awareness of the true first cost and life-cycle cost of passive design, paralleling wide-
spread costs and performance comparisons with alternative approaches to reduced
nonrenewable energy use.
The majority of passive systems for space conditioning (both heating and cooling) can be
placed within one of five generic system types described by J. Douglas Balcomb of Los
Alamos Scientific Labs in his summary of the Passive Solar Heating and Cooling Confer-
ence, May 1976. Variations within each generic class have to do with differences in the
passive system aperture and storage media and with variations in the level of control
occupants have over the heat transfer between collection and/or storage and the condi-
tioned space. The five generic systems are
Direct gain
Convective loops
Thermal storage walls
Roof ponds
Attached greenhouses
As is the case with active systems, passive systems have certain key components—
glazing/insulation systems, absorber, storage, and the space to be heated/cooled. The
Passive Solar Design 123
manner in which these components are arranged determines the generic passive system
type. The permutations within each generic type are innumerable. In describing each
of these generic system types, a number of these permutations are illustrated.
A. Direct Gain
A direct gain system admits sunlight directly into the space to be heated (Fig. 31.1).
The system aperture is usually double-pane glass, always located on the south face of
a building. The interior materials of the building are capable of absorbing large amounts
of energy through radiation and convection. Although not required as part of the generic
type, some method of controlling solar heat gain is common. Shutters, reflectors, and
roof overhangs are methods for increasing or decreasing solar heat gain at varying times
of the day and year.
A common example of a direct gain system is a south wall with a large number of
windows admitting solar radiation into the building space and onto a concrete floor
absorber. The walls are either heavy masonry materials or several layers of gypsum board.
Variations in the aperture include the number of layers of glazing, the type of glazing
material (glass, plastics, etc.) and their location (south walls, clerestories, roofs) such that
sunlight falls on the floor or a back wall. The storage material may be concrete, masonry,
slate, brick, or plaster in the floors or walls; several layers of gypsum board, adobe, or
water containers scattered around the space; or other heat-absorbing material. The
storage material is spread throughout the space but concentrated where it may be directly
irradiated. Again, this is usually a floor in front of south-facing windows but may well be
in a north wall if the fenestration plan provides for its direct irradiation.
Examples of direct-gain systems applied to nonresidential structures are the Annex
of St. George's School (Wallasey, England), showing heavy masonry mass interiors and
directly irradiated floors, and the Cambridge School (Weston, Massachusetts), showing
an irradiated north wall. The Cambridge School uses a roof overhang, whereas the Annex
of St. George's School does not (Figs. 31.2 and 31.3).
SUMMER
SOLAR
KYLIGHT
SUMMER
WINTER HEAT
at% STORAGE
WALL
SOLAR WALL
B. Convective Loops
Convective loops utilize an absorber surface, remote from the building space, to absorb
incident radiation and then convect warmed air or water into the building space (or water
preheat tank, in the case of domestic hot water). The classic thermosiphon water heater
is a good example of this generic type (Fig. 31.4).
Another example is an air thermosiphon attached to the vertical south wall of the
building (Fig. 31.5). Cool air is sucked in at floor level and discharged as warm air at
ceiling level. Other than the interior surfaces of the building, storage is not included in
this circuit. Even more important than in the direct-gain system, large quantities of
thermal mass dispersed to provide high heat transfer rates should be provided to absorb
the convected energy. Thermosiphon air loops cannot take advantage of directly
irradiated storage mass.
Passive Solar Design 125
HOT WATER
OUT
SHUT OFF
VALVE
t
COLO WATER IN
-DRAIN
Figure 31.4 A thermosiphoning water heater with separate collector and storage tank.
---
COLLECTOR
METAL
LATH -
The Davis House in Albuquerque (Fig. 31.6) is an air thermosiphon with storage
remote from the living space. Here a rock-bed storage mass is provided in the convective
loop between the absorber and the building interior space.
DOUBLE GLAZING
BLACK PAINTED SURFACE
SOLID CONCRETE BLOCK WALL
PLASTER
GLAZING
THERMAL
CURTAIN
WATER
STORAGE
GLASS
BLACK
SURFACE
111
WATER
INSULATING DRUMS
SHUTTER
~1 HAND
CRANK
Figure 31.10 Hand-operated insulating shutter with water drum thermal storage.
128 Applications of Solar Energy
D. Roof Ponds
Roof ponds provide thermal storage in the ceiling rather than on south-facing walls.
The most widely accepted version of this system employs a shallow pond of water
(actually, several bags of water constitute the pond) in thermal contact with a strong
but highly conductive flat roof and ceiling structure. The water bags, which serve as
the thermal storage, are exposed to solar heat gain and protected against heat loss at
appropriate times by controlled movable insulating panels (Fig. 31.11). Cooling is
accomplished with combinations of night-sky radiation, night-air convection, and, in
some cases, evaporative cooling of water flooding the outer surface of the bags. In
this case, the movable insulation covers the mass during the day and is opened at night.
Of the five generic passive types, roof ponds appear to be most applicable to cooling,
especially where low humidity and clear summer night skies prevail. Roof ponds, as
described above, will generally be limited to southern latitudes with relatively mild
day-time temperatures (greater than 0°C). Their applicability to colder northern climates
ereAvdrossisiV."
I
N
ii
SUMMER COOLING
can be increased through the use of reflectors and by sheltering the roof pond from
convective losses with enclosed, translucent roof structures.
Although roof ponds are the commonly accepted way of providing thermal storage
in the ceiling, they may not be the only method. As this interesting idea is given more
attention, possible new concepts will necessitate the broadening of this category to
"thermal storage roofs."
Roof pond applications include the Atascadero Skytherm Southwest House
(Atascadero, California), the Winters House (Winters, California), and the Cool Pool by
Living Systems (Figs. 31.12-31.14).
INSULATING LID
DRUMS
DRUMS
Figure 31.13 Roof pond heating and cooling system in the Winters house.
130 Applications of Solar Energy
at,
"v°2-
i'41/1 Av vs
÷ NORTH
'
1_mr.VENTILATION
ho* elnIr
r 1
,„„,,,0
FA
E. Attached Greenhouses
Attached greenhouses are a mix between direct gain and thermal storage wall systems.
In the simplest case, a lean-to type greenhouse is built on the south side of a building,
and excess heat from the greenhouse is allowed to enter the house. As more control
over the temperature in the greenhouse is desired, large quantities of thermal mass are
added to the greenhouse. More sophisticated attached greenhouses place the mass in the
wall between the greenhouse and the living space (Fig. 31.15).
Location of the greenhouse, placement and quantity of mass, and shutters and
shades are variables in the attached greenhouse generic type. An important, but difficult
to quantify, aspect of attached greenhouses is their esthetic and food functions.
PARTIAL SHADE
FOR SUMMER
1
./1r • 7. ,; •
CORRUGATED FILON
OR LASCOLITE
-EXISTING
WALL
COOL AIR
ROCKS
PUMICE
Figure 31.17 The Hofman house: a concrete south wall combined with a greenhouse.
132 Applications of Solar Energy
Examples of the solar greenhouse approach include the Solar Sustenance Project,
Albuquerque, New Mexico, and the Hofman House, Canterbury, New Hampshire
(Figs. 31.16 and 31.17). A November 1977 solar greenhouse conference at Marlboro
College, Marlboro, Vermont, provides many more examples.
F. Heat Storage
Variations in the medium used for thermal storage create a number of possibilities within
each class of system; they account for the greatest variation in the architectural quality,
construction methods, and, consequently, in system costs.
Storage media used include heat storage materials such as
Not controllable by the users. (Nautral heat flows to and from the building are
completely unregulated. A fixed window with a fixed summer sunshade is
in this category.)
Controlled by natural forces under the user's direction. (Movable window insula-
tion, reflectors, awnings, and shades. Devices such as Skylids, which use the
natural properties of Freon gas to open and close them, and one-way vents,
used in convective loop systems to prevent reverse thermosiphoning, also fall
into this category.)
Passive Solar Design 133
Hundreds of thousands of passive solar water heaters have been used throughout the
world. In a domestic hot water system, the solar collector is at an elevation below that
of a hot water tank. Warmed liquid rises through the collector and up to the water tank
where it is stored. In some systems, an electric heating element is located at the top
third of the tank to boost the temperature of the water prior to use. In other systems,
the solar tank serves to preheat the water prior to final heating in a conventional water
heater.
Passive solar water heating can also be accomplished through the use of black
uninsulated water tanks directly exposed to sunlight. Efficiency is usually enhanced
through the use of one or two layers of glass or plastic between the black storage con-
tainer and the outdoors. In cold climates, insulation covers the tank at night.
Plastic water bags on a level platform can be used in mild climates where heat loss
is not a severe problem. Reflecting surfaces can increase the amount of heat the water
absorbs, resulting in higher temperatures.
B. Heating Only
Probably the simplest "heating only" passive systems are thermosiphoning air collectors.
The simplest form of a thermosiphoning air collector is illustrated in Figure 31 18. As
the air in the space between glass and the blackened absorber surface is heated expands
and becomes lighter, rises through the collector, and flows into the room from the vent at
the top. Cooled room air is drawn into the collector through another vent at the base of
the wall, replacing the warm air leaving the collector. It is heated and subsequently
expelled from the top of the collector into the room. This process continues as long as
there is enough sunlight to raise the temperature of the collector above the temperature
of the room.
The simplicity of direct gain systems is apparent, and they are already used to some
extent on nearly all of our 70 million buildings. Thermal mass on the order of 600 kJ/
m' • °C (30 Btu/ft' • °F) of south-facing window is considered the requirement if the
mass is directly exposed to the sun. If the mass is not struck by direct sunlight, approxi-
mately four times as much thermal mass, or 2400 kJ/m2 • °C (120 Btu/ft' • ° F) is
required for best performance.
Direct gain can be increased by reflective surfaces used outside of the building.
Nighttime insulation for reducing heat loss in combination with reflectors can double
and even triple the net effective energy output of direct gain systems. Beadwall, designed
by Zomeworks Corporation, uses polystyrene beads blown between two panes of glass or
134 Applications of Solar Energy
WARM AIR
illi'lli ►. 'III
i 7'
41,,,,i,
_
C001- AIR I : '
161 .991
BLACK AfiSORES
GLASS OP PLASTI
plastic at night to reduce heat loss. Three inches of beads provide an R value of about 10
(see Appendix C). The beads are drawn out during sunny winter weather by small
vacuum cleaner blowers which operate only a few minutes per day.
Figure 31.19 combines the Beadwall with a concrete thermal storage wall. The
economic value of movable insulation in passive systems increases with increasing
numbers of degree days. However, most concrete storage wall systems to date have not
used movable insulation because of its relative inconvenience and increased first cost.
Figure 31.10 shows an example of water as a thermal storage wall. This system,
first developed by Steve Baer of Zomeworks Corporation, uses 55-gal drums filled with
water. Insulating panels hinged at the base of each wall can cover the single layer of glass
at night to reduce heat loss. With the shutters open and lying flat on the ground, the
aluminum surface reflects additional sunlight onto the drums. During the summer, the
shutters, in closed position, shade the glass.
Another example of a water wall (Fig. 31.9) uses vertical tubes of water, produced
by Kalwall Corporation. In this example, the tubes are separated from the living space
by a wall through which air from the room can pass and contact the warm tubes. A fan
controls the movement of heat from the "collector" to the living space. A thermal
curtain closes between the tubes and the glass at night to reduce heat loss.
Attics can be used to advantage for passive heating. One method, for example, is
to glaze south-facing attic roofs, circulating the overheated attic air to the building below.
Passive Solar Design 135
DAMPER
GLAZING
MOVABLE INSULATION
ALTERNATE
INSULATION
LOCATIONS
A variation of this scheme includes heat storage. In some cases, heat storage
consists of black plastic waterbed-like containers lying on the floor of the attic. The
warm attic air is then circulated through the use of fans to spaces below, or the heat
radiates from the bags of water directly through a heat-conducting ceiling in a manner
similar to Harold Hay's Skytherm system.
Solar heat from a convective loop collector can be stored in rock bins located inside
the house. Little or no mechanical power may be needed. The design by Jonathon
Hammond shown in Figure 31.20 shows the heated air rising through the collector, then
through a vertical rock storage bin in the house. Air flow passes through the living space
between the storage bin and the collector. In this particular design, the wood stove flue
is imbedded in the storage bin to provide supplementary heat during periods of cold,
cloudy weather.
SOLAR COLLECTOR
OCK BIN
WOOD
STOVE --,••
C. Heating/Cooling
Figure 31.21 shows a thermal storage wall for heating also being used for natural ventila-
tion. Dampers are positioned as shown, so the solar heated air between the glass and the
warm concrete creates a "chimney effect," drawing warm room air to the base of the
collector and cool outdoor air into the house through vents in other exterior walls.
In addition to the basic Skytherm system described earlier for both heating and
cooling, variations include flooding the bags of water to use evaporative cooling during
the day in addition to nocturnal radiational cooling at night. This is particularly valuable
in excessively hot climates.
To increase performance during the winter, a thin layer of transparent plastic can
be inflated slightly above the roof ponds; the result is higher collection efficiency.
In direct gain systems, the thermal mass used for storing heat during the winter
can be used to reduce cooling loads in the summer. The building can be opened up
during cool nights. The resulting cool thermal mass will help keep the buiding cool
during the day.
D. Cooling Only
Although often considered an energy conservation measure, the best means of "solar
cooling" is to keep the sun from hitting and/or penetrating the building during the
cooling season.
So also, cross-ventilation and the use of the chimney effect to increase natural
ventilation can in many cases be used in lieu of other forms of air conditioning.
In many cooling-only climates, the Skytherm-based system of roof ponds can
provide nocturnal radiational cooling without necessarily being used for heating.
There is direct gain through the south side of David Wright's semicylindrical house, which
has 36 m2 of insulating glass. There are few windows elsewhere. Adobe walls 43 cm
thick and an adobe floor 60 cm thick are insulated on the outside by 5 cm of poly-
urethane foam. The house loses about 24,700 kJ/°C • day (13,000 Btur F • day). On a
clear January day, as much as 530 MJ (500,000 Btu) enter the house through the south
window. The temperature of the house is permitted to fluctuate between 15 and 25°C
(60 and 80°F). This fluctuation in combination with the large expanse of glass and
considerable volume of thermal mass permits 90 percent of the heating needs to be
satisfied by solar in the 3400 °C • day (6200 ° F • day) climate of Santa Fe, New Mexico.
At night, heat loss through the south glass is reduced by folding, insulating shutters,
made of 5-cm-thick foam insulation covered with canvas (Fig. 31.22).
COLLECTOR
METAL
LATH
becoming heated in the process. From there it travels through the rock bed located
below the porch. Heat rises through floor registers, and cool house air returns to the
collector. Approximately 75 percent of the heat for this 93 m2 house is provided by
the passive movement of air. Reverse thermosiphoning on winter nights is prevented by
manual dampers, which also control the amount of heat rising into the house.
The main office building of the National Scientific Research Center in Odeillo, France,
uses thermosiphoning wall panels in combination with windows. Together, the windows
and wall panels supply about 50 percent of the building's heat (Fig. 31.28).
Passive Solar Design 141
DAMPERS
t.
Tit tiv1-1%1-;',.lik:
- at. ILI.
WINDOW
WARM
AIR
GLAS
r .COOL AIR
INSULATION
Figure 31.28 Odeillo office building.
thermosiphoning wall panel.
No provision has been made to store the heat other than the thermal mass of the
building, which in this case is reinforced poured concrete. The panels are designed to be
easily closed and "turned off" during warm weather. Their design allows cool air to
settle to the bottom of the air passages, and prevents cold air in the panels from entering
the building at night.
A breadbox solar water heater usually consists of a tank (or tanks) of water covered by
glass or plastic and enclosed in an insulated box. The tank is painted flat black and lies
horizontally inside the insulated box, which is oriented south with its long axis in the
142 Applications of Solar Energy
A. Trombe Walls
SOLAR
RADIATION
RADIATION
FROM STORAGE
NATURAL
CONVECTION
TROMBE WALL
Decorative grills are installed over these openings on the interior face. The lower grill
includes the backdraft damper.
The exterior glazing system is mounted 7 to 10 cm away from the outer blackened
concrete surface. Where glazing supports are attached to the wall, wood or other insulat-
ing material should be used as a separator. The glazing should extend above and below
the face of the storage wall, fully exposing it to the sun. Since glazing is the weather skin
of the building, it must be airtight and water resistant.
Construction and Installation. Building the Trombe wall described here normally
requires only general contracting skills. Depending on contractor preference, the installa-
tion of the glazing system can usually be handled by the manufacturer's representative.
This usually enables the building owner to obtain a better warrantee on its weather tight-
ness. The storage wall itself should be constructed at the lowest cost possible given the
thermal, structural, and interior finish requirements outlined above. If the contractor or
subcontractor normally uses poured concrete only in foundation work, or if multistory
installations are planned, the solid masonry unit wall is usually preferred.
Scheduling of the work presents no special problem if the contractor carefully
reviews the construction requirements in advance. The glazing system is usually fabri-
cated to site dimensions; therefore, to avoid delays in closing the building, these dimen-
sions should be fixed early in the construction whenever possible, and orders placed for
the glazing. Concrete finishing work may require the appropriate trades on the job site
at other than the normal times.
System Selection Criteria
Building Integration. Trombe walls are appropriate to a variety of building types.
Their temperature stability means that the overheating associated with some other passive
systems is not a problem. The thermosiphoning air vents provide a significant level of
control over heat delivery to the space. The opaque wall eliminates the excessive glare
associated with direct sun. Sun damage to goods and furnishings is avoided, an advantage
if used in retail stores and commercial buildings. The structural wall is fire resitant and
provides perfect security for warehouses and manufacturing plants as well as structural
stability in high-rise construction. Finish details can be very rough to suit manufacturing
and industrial applications, yet the Trombe wall can be detailed and constructed to fit
the finest of residential designs.
Costs. First costs for Trombe walls vary with differences in construction and
detailing of the mass storage wall and the exterior glazing. In areas where concrete and
concrete masonry construction above grade in residential structures is common, the
storage wall is generally inexpensive. Similarly, if an experienced subconstractor is
available, or if materials can be obtained cheaply through local suppliers, the exterior
glazing can be low in cost.
Cost estimates prepared for the design described here (with plaster interior finish)
vary from a low of $120/m2 to a high of $140/m2 in 1977 dollars. To obtain a true
net cost differential for passive solar heating, the cost of conventional construction
replaced by the Trombe wall should be subtracted. Since the most expensive conven-
tional exterior wall construction, including insulation and interior finish, usually runs
146 Applications of Solar Energy
between $20 and $30/m2 , the true first cost of the Trombe wall is estimated at about
$100/m2 .
Operating costs for the Trombe wall are zero, and little or no maintenance is
required. Maintenance in many climates is comparable to that for vinyl siding: An
occasional (every 2 to 4 years) washing of the exterior glazing is advised. Harsh indus-
trial environments may degrade the plastic glazing; "refinishing" coatings are available
from leading manufacturers and may be applied on a 3- to 5-year basis.
Thermal Performance. Thermally, the Trombe wall performs reliably. Heat
losses, even under worst conditions, are not very different from much conventional
commercial construction. The overall U factor of 1.3 W/m2 • °C (0.23 Btu/ft2 • h • °F)
(back thermosiphoning prevented) enables the wall to meet ASHRAE 90-75 standards
for single-family residences located in climates not exceeding 2900°C • days (5200°F •
days). If solar gains are considered, the effective U factor is of course negative, as it is a
net heat producer.
Solar energy collection takes place at low to moderate temperatures at the surface
of a Trombe wall, generally not exceeding 65°C (150°F). This provides good instantane-
ous efficiencies (generally on a par with active system flat-plate collectors), whereas the
vertical south wall orientation enables good winter heating performance and minimizes
summer overheating concerns except in the deep south. Air is delivered to the room
through the vents at low temperature, generally not exceeding 32°C (90°F). Normal
air flow is approximately 0.3 m3 /min • m2 (1 cfm/ft2 ).
The interior surface of the wall also delivers heat to the space by radiant means,
fluctuating in temperature on a daily basis. Temperature swings on the interior wall face
are most often recorded between 4 and 8°C (8 and 15°F). The maximum delivered air
temperature and the maximum interior wall surface temperature occur approximately
8 hours apart, with the latter generally in the late evening. Peak heat transfer rates are
normally lower than 340 kJ/m2 • h (30 Btu/ft2 • h) for combined radiant and convec-
tive heat transfer from the interior surface. Total convective and radiant heat transfer
from the interior wall is usually not more than 400 kJ/m2 • h (35 Btu/ft2 • h).
Over the course of the heating season, the Trombe wall absorbs enough solar
energy to cancel all thermal losses from the building interior through the wall itself
and delivers excess heat to the remaining building load. Simulations at Los Alamos
Scientific Laboratories support estimates that each square meter of Trombe wall provides
500 to 800 MJ (45 to 70,000 Btu) to the building heating load each season. If the pre-
dictable seasonal losses that would have occurred through a conventional wall are taken
into account, each square meter of Trombe wall replacing conventional construction will
effectively reduce building oil consumption by 36-1(9.7 gal), or electrical consumption
by 19 kWh.
Economics. Estimates for delivered energy and the cost per GJ (or MBtu) delivery
are shown for representative cities in Table 31.1. For comparison with current conven-
tional energy costs, consider the following: electricity at 4.4 0/kWh is equivalent to an
energy cost of $12 20/GJ; number 2 fuel oil burned at a seasonal efficiency of 0.50 at
13 0/1(50 0/gal) results in an energy cost of $6.75/GJ. Actual performance will vary
with the ratio of system size to building load. Like all solar systems, oversized Trombe
Passive Solar Design 147
Table 31.1 Delivered Energya and Associated Costsb for Trombe Walls
City GJ/M2 • yrc $/GJ $/MBtu)
Fort Worth, Texas 0.434 $23.60
Madison, Wisconsin 0.510 $20.00
Boston, Massachusetts 0.535 $19.10
Medford Oregon 0.538 $19.00
Los Angeles, California 0.610 $16.80
Boulder, Colorado 0.710 $14.40
a
These estimates were prepared by J. Douglas Balcomb at Los Alamos Scientific Laboratories for a
particular double-glazed Trombe wall, and a fixed area-to-building load ratio: 0.1 m2 /(kJr C • h)
or 2 ft2 /(Btur F • h).
bThese figures are the effective energy cost obtained by applying a capital recovery factor of 10
percent (corresponding to a 7.5 percent interest rate and a 20 year term) to estimated first costs.
cMultiply by 0.0881 to convert to MBtu/ft2 • yr.
wall passive systems will be less cost effective. Within the constraints of normal building
designs, however, it is difficult to oversize south-facing Trombe wall systems, and the per-
formance estimates provided here can be taken as representative for initial project
development.
Alternative System Configurations
Variations in Trombe wall design reflect the nature of local materials, builder skills,
and architectural approaches. Walls have been constructed from almost every masonry
material, from architectural-grade brick to adobe. Thicknesses vary, reflecting each
designer's attempts to work with the time lag and thermal storage characteristics of the
wall material. (Recent work on this subject suggests that dense concrete between 0.2 to
0.5 m thick is "proper.")
Glazing systems can be of one, two, or even three layers; selections of one- or three-
layer options should reflect a careful analysis of the cost effectiveness of the particular
design.
As mentioned previously, the height and width of a Trombe wall can be changed to
suit architectural requirements without any loss in performance.
New Materials. One of the most important characteristics of the Trombe wall as
it is currently built is its use of existing, available construction materials for solar heating.
These materials do not behave as well, thermally, as some would wish. A close relative
of the Trombe wall, the waterwall (epitimized by the Zomework's drumwall) is adopted
by some to obtain faster heat transfer into storage and cooler surface temperatures. New
techniques and materials could be developed along these lines. Heat pipes could link a
storage space and an absorbing surface, separated by a thin but effective insulating
barrier. The resulting heat transfer would be "oneway," reducing to a minimum the
losses from the system, especially those that occur at night. (Insulating shades and
movable insulation systems, including Beadwall, have been tried to reduce these heat
losses, but the cost effectiveness of such approaches has not been demonstrated.) Such
one-way flow of energy through a thermal mass wall is accomplished by the "thermic
148 Applications of Solar Energy
The roof pond is unique in that it is the only passive solar heating system that
provides substantial cooling effects as well. The most widely accepted version of this
system employs a shallow pond of water in thermal contact with a strong but highly
conductive flat roof and ceiling structure. Water bags, which serve as thermal storage, are
exposed to solar heat gain and protected against heat loss at appropriate times by con-
trolled movable insulating panels. Cooling is accomplished with combinations of night-
sky radiation, night-air convection, and, in some cases, evaporative cooling of water
flooding the outer surface of the bags.
The design of solar roof ponds varies substantially from region to region, as efforts
are made to improve winter heating performance in climates harsher than those where
the original installations are located. All are adaptations of the original Skytherm, now
a proprietary design (U.S. Patent Numbers 3,299,589; 3,450,192; and 3,563,305),
marketed by Skytherm Processes and Engineering, Los Angeles, California.
The adaptation illustrated in Figure 31.31 follows the approach taken on a retrofit
roof pond application in Concord, New Hampshire: the Freese House (1976). It uses
another proprietary system, Beadwall (by Zomeworks, Inc., Albuquerque, New Mexico)
to replace the horizontal movable panel insulation used by Harold Hay. The alteration
increases the weatherability of the system in climates where snow and ice are common
winter conditions, but decreases the system's natural cooling potential and its cost
effectiveness. For installations in climatic regions characterized by significant cooling
loads and mild (usually above-freezing) winters, the original system, known as Skytherm-
Southwest, is recommended. (See also Alternative System Configurations.)
Materials. The solar roof pond attic replaces the ceiling and roof in conventional
construction. It is a modular design based on a 12-ft span. Major subcomponents are
Passive Solar Design 149
SOLAR
RADIATION
RADIATION
FROM
STORAGE
ROOF POND
Construction and Installation. The solar roof pond design illustrated requires
careful scheduling by both contractor and owner. The two key items are the installation
of the metal deck and of the Kalwall Sun-Lite panels. The deck must be detailed to drain
away leakage from the waterbags, and therefore provides a reasonable weather skin during
Passive Solar Design 151
construction. It should be tested with a water spray for leaks before finish work is
started on the floor below. The glazed roof must be completed early, however, to close
the building in. Since the Kalwall glazing incorporating Beadwall is a special-order item,
requiring close coordination between two manufacturers, the architect or builder must
arrange for this well in advance of installation time.
Installing and filling the water bags is the last step in construction. To service the
Beadwall system and allow periodic inspection of the water bags, a hatch or door into
the attic space must be provided.
System Selection Criteria
Building Integration. The solar roof pond is the largest passive system type
described (in architectural volume). The 3.6-m module is based on the span capability
of both the steel deck and the wood rafters used. Larger spans are possible, but since an
8:12 or greater pitch must be used for the Beadwall/Kalwall roof, the system would
become excessively large in volume with little increase in floor area coverage. The 3.6-m
dimension is characteristic of residential room sizes and commercial structural grid sizes.
Since the system distributes heat by direct radiation to the occupied space, it can
only serve the floor directly beneath it. The system cannot be ranked in sawtooth
fashion across a large flat-roofed structure without causing self-shading of adjacent units.
On larger structures it is logical to place the solar roof pond above zones along the north
side, since the south zones can usually benefit by the more direct passive heating system
described elsewhere in this chapter. Multilevel construction is possible when the building
takes a terraced form.
Costs. The solar roof pond used for the prototype Skytherm House in Atasca-
dero, California had an in-place cost (1973) of approximately $73/m2 of bag area. Cost
reductions would result from experience with this prototype, but given the rise in
material and labor prices in the interim, the first cost of a similar system (1977) might
be the same. For the modified design described above, first costs will be considerably
higher. Estimates of the component costs include the deck at $38/m2 , the north wall at
$27/m2 , the Beadwall system at $90/m2 , and the Kalwall roof glazing at $70/m2 (all per
unit surface area). Translated to cost per square meter of roof pond area, this is about
$240/m2 , to which should be added $16/m2 for structural integration. This first-cost
estimate of $256/m2 of pond area is for an assembly that replaces the entire ceiling and
roof in that portion of the building where it is used. A 1976 study at Princeton Univer-
sity estimated the system cost of a solar roof pond (of very different design) at $150/m2 .
This wide range of cost estimates is created in part by the relative scarcity and current
high price of acceptable automatic movable insulation systems. If the Beadwall estimate
used above were cut in half, and the Kalwall price reduced by one-third, the estimated
system cost would drop to just under $200/m2 .
Running costs for this solar roof pond design are essentially those for the Beadwall
system. Beads become charged with static electricity, and the antistatic agent used to
keep the beads flowing freely through the system must be replaced. This is in reality a
simple maintenance task, but it is necessary perhaps every 3 months. Physical abrasion
of the beads can occur in poorly designed systems, and ultimately UV breakdown of the
polystyrene requires the replacement of the bead stock. The life of the bead is not
152 Applications of Solar Energy
known at this time. Electrical power consumption is small, for the blowers, which may
draw 7 A at 120 V, are used less than 1 hour a day.
Thermal Performance. Simulations of roof pond performance by Philip Niles of
the Atascadero House, and work by Harrison Fraker at Princeton Unviersity on roof
ponds in a temperate climate, are the basis for the following observations.
Solar roof ponds are characterized by very low-temperature operation. The season-
al variation in storage temperature may never exceed 45°C; the daily swing in winter may
average 3°C. The average mid-winter temperature of the pond is seldom greater than 6°C
over room temperature.
Under these conditions, the average daily heating contribution of a 25-cm-deep
pond is in the range of 1 to 3 MJ/m2 (100 to 250 Btu/ft 2 ), delivered primarily as
radiation at rates usually between 30 and 60 W/m2 (10 and 20 Btu/ft2 • h), depending
on both the surface temperature of the pond-heated ceiling and the average unheated
room surface temperature. The seasonal contribution to the heating load from the roof
pond is difficult to assess; in instrumented buildings the contribution has been 100
percent of the demand. The Atascadero House had a demand equal to 175 kJ/°C • day •
m2 (8.6 Btu/°F • day • ft2 ) of roof pond; to provide 100 percent heating in a 2200°C •
day climate the pond supplied 390 MJ/m2 (34,400 Btu/ft2 ) over 5 months.
Given the average daily contribution estimated above, a pond would supply 500
MJ/m2 (44,000 Btu/ft2 ) to a space heating load over a 9-month period. The perform-
ance of the pond is characterized by almost constant heating at low rates. That portion
of a building's external skin that is a roof pond will generally never contribute to the
auxiliary system's load. This means that space conditioning equipment can be sized
smaller than in conventional buidings. For a building load of 150 kJ/°C • day • m 2
(7.5 Btu/° F • day • ft2 ) of pond area, the pond should provide all heating needs down to
about 1.5°C outside temperature, and about half the load at -17.8°C (0°F).
Economics. The performance and cost estimates for the particular solar roof pond
presented above can be translated into an energy cost of between $34 to $46/GJ ($36
to $49/MBtu). At present prices, this is far above other conventional energy sources,
and almost four times more expensive than the solar energy obtained from the thermo-
siphoning air heater. The pond, however, unlike the thermosiphoning heater, can provide
significant reductions in peak demand. Where significant savings in operating or first
costs can be obtained by lowering peak space-heating demand, this extra cost may be
justified.
concept as developed by Harold Hay (Skytherm) includes cooling, and natural cooling
strategies rely heavily on climatic variables such as relative humidity as well as nighttime
air temperatures, a great deal of regional variation can be expected as additional work is
completed on these systems.
New Materials. Developments that could radically alter the design characteristics
of solar roof ponds include materials with reverse thermosiphoning characteristics such as
the "reverse juices" proposed by William Mingenbach of The Architects Taos, Taos, New
Mexico. Solar salt gradient ponds are examples of large-scale systems where the cooler
liquid rises to the top. This effect creates a naturally insulating layer at the top of the
pond; this could replace the most troublesome part of the solar roof pond concept in all
its current versions: the movable insulation.
Improved Engineering. Failing developments along these material lines, improved
engineering of movable insulation subsystems, emphasizing lower first cost, and trouble-
free operation in all weather conditions is needed to broaden the applicability of the solar
roof pond concept in the space heating mode. Engineering approaches that take full
advantage of the system's peak-reducing characteristics must be clarified for builders and
designers.
Ni
2 3 4 5 6
.„.
.. ....__,,„,..,,..
. ...:‘,
41%--
SOLAR
RADIATION
NATURAL
CONVECTION
(behind collector plate)
1 INSULATING GLASS
2 WALL FRAMING
3 METAL ABSORBER PLATE
4 INSULATING CORE
5 INTERIOR FINISH
6 CONTINUOUS AIR VENTS
In any construction, the glazing supports should minimize the thermal connection
to the thermosiphoning panel itself, which is built behind the glazing. The design recom-
mended here doubles as a structural wall in lightweight wood-frame construction, and the
glazing details are carried out in wood, a naturally insulating material. The core panel
consists of a urethane and mineral board roof insulation sandwiched between the black-
painted absorber and an interior finish panel of half-inch fire-code gypsum board. The
interior gypsum surface can be used as the base for a variety of conventional interior
finishes, or can be supplied prefinished with fabric or vinyl coverings, as are a variety of
modular partition systems.
Passive Solar Design 155
The insulating core in the design described here is fabricated from a roof insulation
supplied by a variety of manufacturers. These boards are usually 0.9 by 1.2 m, with a
nominal thickness of 3.8 cm, providing a nominal R value of 6.7* The mineral board side
of the urethane/mineral composite faces the absorber panel; the flammable urethane
foam is protected by the gypsum board.
The absorber panel is corrugated metal siding, a building material readily available
complete with compatible fasteners and preformed EDPM or Neoprene closure strips,
simplifying the construction of an airtight, durable thermosiphoning heater. The ribs
give such an absorber self-spacing characteristics and structural stability. Alcoa ribbed
20-cm siding, of 0.8-mm aluminum selected for this design, has a cross-sectional free
area for the thermocirculation of 58 cm2 per 30 cm of width.
The absorber is obtained as mill-finished aluminum and must be prepared for
painting with an etching-cleaner. The recommended finish is a thin covering coat of flat
black enamel. Nextal brand Black Velvet by 3M Company, over a suitable primer, is
commonly used.
Venting of the panel to the heated space takes place through a full-width register
that provides manual control of the air flow. Continuous linear diffusers such as the
Hart & Cooley Royalaire 3.8-cm series should be used.
Backdrafts at night or during very poor weather are prevented by the use of very
lightweight 0.025-mm plastic film one-way dampers. These dampers must he specially
fabricated. In the suggested design the plastic film is attached to a punched or die-cut
24-gauge galvanized sheet with double-sided adhesive tape. See Fig. 31.33.
Design. The dimensions of the thermosiphoning air heater are based on its use as
an infill panel in a wood-framed bearing wall. Doubled 5- by 10-cm (2- by 4-in.) studs,
0.91 m (3 ft) on center replace conventional framing. The panel should not be relied
upon to provide overall wind bracing of the building frame. Good building practice
requires the use of methods such as plywood shear panels or diagonal bracing.
This 0.91-m module permits the use of 0.86- by 2.29-m insulating glass lights.
The larger 2.29-m size permits the design to be used as a full-height infill panel in
between-floor construction. The addition of a spandrel glass element (not described
here) would allow the design to be incorporated into curtain-wall construction. The
core panel is dimensioned to fit the 0.91-m module created by the 0.86-m glazing.
Overall depth of the panel from black absorber to interior finish is a nominal 7.5 cm,
providing, when installed, a 1.27-cm clearance to the exterior stud surface to which the
glazing system is attached.
Construction and Installation. The modular, lightweight (3.6-kg/m2 ) panels are
best shop-fabricated, with glazing and interior trim (vent grills and dampers) applied
after installation in the field. If the seal made between panel sides and bottom and
abutting construction is adequately caulked, the panel provides a temporary weather-
skin, and installation of the exterior glazing can be scheduled independently. Care
must be taken, however, not to let the foam core or interior gypsum finish be damaged
by weather.
SHEET-METAL FRAME
1.5-MIL POLYETHYLENE
TAPED TO FRAME
Unlike most active systems, there are no startup procedures to follow, and no
precautions to take against freezing or overheating during installation. Electrical wiring
should be handled with surface-mounted systems.
System Selection Criteria
Building Integration. Thermosiphoning air heaters (natural convection collectors)
are well suited to structures where the heating load is large when compared to the pro-
posed panel area. (This is because all of the panel output can then be absorbed immedi-
ately by the building load, so that the lack of thermal storage does not decrease system
performance.) Schools and low-rise office buidings are such structures. The intermittent
use of these buildings matches well with the thermosiphoning air heater's daily cycle.
The air heaters can be used in both single-family and multifamily houses, but unless
separate thermal storage designs are also incorporated, the low loads (especially in charac-
teristically well-insulated housing) and 24-hour use mean that a small fraction of the
heater's potential output can be usefully employed, decreasing its applicability.
Passive Solar Design 157
Architecturally, the modular air heater, with its curtain-wall characteristics and
glass-surfaced appearance, is well suited for integration into most new commercial
construction. With minor modifications in the design, the heater can also be a cost-
effective retrofit application over existing exterior walls.
Costs. First costs for thermosiphoning air heaters are dependent primarily on
variations in labor costs. The major materials used are supplied by a number of competi-
tive manufacturers and are widely available. Setup charges for shop fabrication may be
high given the unconventionality of the product, but the contractor will save a portion
of the normal markup sequence for manufactured products.
Materials costs for designs similar to the one shown here have been in the range of
$38 to $43/m2 (1976). Estimates of installed cost for the modular air heater panel in
eight cities were consistently near $110/m2 . Since the panel replaces the conventional
wall in light-frame construction, usually $28/m2 , or the infill in steel-framed construc-
tion, usually $60/m2 , the net cost applicable to the solar aspects of the panel are $82
and $50/m2 , respectively. Operating costs are nonexistent.
Thermal Performance. Performance in the thermosiphoning air heater depends
entirely on the delicate natural convection currents set up in the system. Air flow is low
to nonexistent (reverse flow being prevented by one-way dampers) at times of little or no
sun, but establishes itself rapidly under sunny conditions. Flow rates are generally higher
than that in Trombe walls; flows of 0.9 m3 /min • m2 (3 cfm/ft2 ) of collector have been
noted. The design shown here has a probable maximum flow rate of 0.6 m3 /min • m2
(2 cfm/ft2 ) with an average temperature rise from an inlet to outlet of 53°C. The result-
ing 1 MJ/m2 • h (90 Btu/ft2 • h) represents an average collection efficiency similar to
that of low-temperature flat-plate collectors used in standard active system designs.
The chief determinant of overall performance is the ratio of load to collector area.
Effective performance deteriorates rapidly with increasing system size. Estimates made
of useful delivered energy, using simplified analyses, range from 340 to 1360 MJ/m2 per
heating season (30,000 to 120,000 Btu/ft2 ). The high numbers in this range are typically
associated with very poorly insulated buildings in cold climates (Madison, Wisconsin),
the low numbers with low-load buildings in warm climates (Atlanta, Georgia, Ft. Worth,
Texas); 570 to 910 MJ/m2 each heating season in moderate climates (Boston, Massachu-
setts) (for total solar contributions between 60 and 40 percent of total load, respectively)
are reasonable estimates for preliminary design.
This wide fluctuation in estimated performance is due primarily to the lack of
stabilizing thermal mass which can effectively store excess heat as it is produced. To
maximize system performance, especially in well-insulated (low-load) buildings, modifica-
tions to increase the building heat storage, such as doubling the thickness of gypsum
board, should be considered. Potential overheating with resulting lack of both perform-
ance and comfort must be assessed whenever systems are estimated to provide over 30
percent of the seasonal load.
Economics. Combined performance and first-cost estimates establish a probable
energy cost between a high of $14.40/GJ ($15 .20/MBtu) and a low of $5 .70/GJ ($6/
Mbtu), based on a capital recovery factor of 10 percent This is an extremely economical
range when compared to the current $12.20/GJ ($12 .90/MBtu) for 4.4 0/kWh electricity.
158 Applications of Solar Energy
Table 31.2 Delivered Energya and Associated Costsb for Thermosiphoning Air Panels
City GJ/m2 • yrc $/GJ (R-- $/MBtu)
Ft. Worth, Texas 0.58 $12.25
Medford, Oregon 0.69 $10.25
Boston, Massachusetts 0.72 $9.90
Madison, Wisconsin 0.91 $7.80
Boulder, Colorado 1.1 $6.50
a Delivered energy figures are averages of estimates prepared for four different building types and
multiple building load/system size ratio. (Source: Passive Systems, a TEA report prepared for Booz-
Allen-Hamilton, 1977.)
bThese figures are the effective energy costs obtained by applying a capital recovery factor of 10
percent to the average ($8.25) of the high and low net first-cost estimated.
cMultiply by 0.0881 to convert to MBtu/ft2 • yr.
Energy cost estimates prepared for specific cities (based on other work by TEA, 1977)
are shown in Table 31.2.
Alternative System Configurations
The thermosiphoning air heater could easily be incorporated into the existing
building products manufacturing industry. A large number of system types, providing a
choice of construction and installation techniques, a range of prefabrication, and a variety
of sizes, could be developed.
New Materials. Differences in materials use and design details in alternative con-
figurations center around the absorber plate, and the type and number of glazings. The
use of mesh or multiple-screen absorbers is favored by some researchers. Site-fabricated
projects where labor is substituted for materials costs usually emphasize multiple-lite
untempered glass (often recycled) or plastic glazing, and absorbers fabricated from
cheap materials such as discarded aluminum printing plates.
Improved Engineering. Further development of the air-heating thermosiphoning
panel should identify those dimensional characteristics, such as height, air gap, and outlet
and inlet free areas, that affect collector efficiency. Confirmation of the average operat-
ing conditions discussed above, and identification of the probable operating extremes (in
a variety of climates) are necessary tasks. Most important for this particular passive
system, engineering methods that will provide good matches to particular building load
and thermal storage characteristics are yet to be developed.
31.7 AFTERWORD
This chapter is but a brief outline of passive solar systems: their definition, applications,
performance and costs. In preparing this chapter we have drawn heavily from the work
of many others. This work is described in the literature, and anyone working with passive
Passive Solar Design 159
solar systems will profit from the exposure to the experience of others that may be
gained by reading this material.
In particular, very good information can be found in the following:
1. Conference and Workshop Proceedings, First Passive Solar Heating & Cooling
Conference, May 18-19, 1976, Albuquerque, New Mexico (available from
NTIS).
2. Proceedings of the 1977 Annual Meeting, American Section of ISES, June
6-10, 1977, Orlando, Florida (available from ISES).
3. Bulletin of the New Mexico Solar Energy Association (available from NMSEA,
P.O. Box 2004, Santa Fe, N.M. 87501).
4. Decision Making in Solar Technology, First Annual Conference Proceedings of
the New England Solar Energy Association, June 24-27, 1976, Amherst, Massa-
chusetts (available from NESEA, Box 541, Brattleboro, Vt. 05301).
5. Passive Solar Buildings: A Compilation of Data and Results, Report by the
Sandia Laboratories, August 1977 (available from NTIS).
Since a good background in basic solar principles and heat transfer phenomena is
useful when evaluating alternative system designs, the following books are also recom-
mended:
ACKNOWLEDGMENTS
Figures 31.2, 31.4, 31.16, and 31.18 have been reprinted from The Solar Home Book, by
Bruce Anderson with Michael Riordan, illustrated by Edward A. Wong. Copyright ©
1976 by Brick House Publishing Co.
Figures 31.3, 31.14, and 31.18 have been reprinted from Passive Solar Heating &
Cooling Research & Development Program Status Report and Plan, U.S. Department of
Energy, Conservation & Solar Applications, Michael Maybaum, Program Manager,
October 1977.
All other figures have been prepared by Barbara Putnam, Total Environmental
Action, Inc.
chapter 32
RAYMOND W. HARRIGAN
Sandia Laboratories
Albuquerque, New Mexico
32.1 INTRODUCTION
The term total energy, as used here, is meant to represent any power-generation system
capable of utilizing some or all of the energy normally rejected during the power conver-
sion process to satisfy some thermal energy requirement for heating or cooling. Power
161
162 Applications of Solar Energy
SWEDEN
OUT
OIL 92.5% WASTE HEAT 47%
UNITED STATES
OIL 17%
Total: 3
GAS 28% ELECTRICITY 31.5%
20. 5x10 kWhtl pe rson
NUCLEAR 3%
Figure 32.1 Comparison of energy usage in Sweden and the United States. From Ref. 1.
generation, although normally construed to mean electric power generation, may in the
more general sense include mechanical power, such as a direct-driven fluid circulation
pump. Historically, a characteristic of total energy plants has been that they are small
(in terms of utility power plants) and located on-site, close to the demand for thermal
energy. U.S. total energy systems accounted for approximately 0.2 to 0.3 percent of
the electrical energy capacity in 1971.
At a time of constrained energy supply, the reason for interest in total energy
systems is clear: the potential for increased energy utilization. Figure 32.1 indicates the
quantity of "waste heat" from electric power generation in Sweden and the United States
in 1971 [1]. Figure 32.1 further shows the potential for utilization of this so-called
waste heat by illustrating Sweden's ability to employ large quantities of the waste heat
(more properly called reject heat) from their electrical generating plants.
The following discussion recounts briefly the recent historical development of
total energy systems and then proceeds with an examination of the design of total energy
systems fueled by solar energy.
32.2 SCOPE
This chapter deals primarily with the development of solar total energy (STE) system
designs to the degree useful in evaluating concept feasibility. Most potential users initially
have only a general idea of their energy requirements, and the objective here is to provide
a basis for formulating designs using this information. In addition, since in a developing
technology such as solar, costs are very ill defined, a design strategy stressing maximized
utilization of solar equipment will be employed without detailed economic evaluation.
The philosophy is that any solar equipment used to its maximum extent will be the most
cost effective. Some rough guidelines on solar equipment costs will be outlined, but any
in-depth economic evaluations will be delayed until execution of more detailed designs.
Total Energy Systems Design 163
This chapter, it is hoped, will allow the user to determine if there is enough poten-
tial gain to warrant investing the time and money required to execute a detailed design
of a STE facility for the particular application. A comprehensive discourse on STE
design is beyond the scope of a single chapter.
After a brief review of the history of total energy and an introduction of the
concept of solar total energy, a hypothetical application will be defined to serve as an
example to illustrate the various design considerations to be discussed. The flow of
the chapter from that point will be basically as follows:
In summary, the design philosophy adopted for this chapter will be that concept
evaluation should start with a rough pencil-and-paper design using average performance
values and generally employed design conditions, if possible. This is then followed by
design performance validation with a simplified, energy-flow, computer simulation using
hourly weather data and then initiation of an in-depth detailed design. At all stages the
design is iterative and, for example, concept evaluations may be performed many times
as the design changes during detailed design.
Total energy systems have been in existence for quite some time, but significant interest
in actually developing the concept into a major power source in the United States is
rather recent. The distribution of the number of new total energy plants with time
164 Applications of Solar Energy
cr) 40
i- INDUSTRIAL
z
a
-I
a_
f2a0 ,-
I- -I
U- -I
Or-r-
du) ,I i
OZ
zz 1965 1970
1960
YEAR
COMMERCIAL
1
I I
Figure 32.2 Construction of new
1 I
1960 1965 1970 total energy plants since 1960.
within the commercial and industrial sectors is shown in Fig. 32.2. The sharp increase in
new plant installations beginning in the early 1960s can be linked to the establishment by
the gas utilities of GATE (Group to Advance Total Energy). The subsequent decline in
new plant installations in the late 1960s and early 1970s has been assigned to many
factors [21. Perhaps most important of these was decreased availability of natural gas
followed by a decrease in the activities of GATE to promote on-site total energy systems.
Other technical factors contributing to the decrease in total energy activities included
high maintenance cost due to low reliability of some plant components, few truly com-
petent commercial total energy designers, and poor fuel economy due to adapting large-
peaking turbines to smaller continuous applications, resulting in severe off-design
performance penalties. The fact that total energy systems have been largely custom
designed for each application retarded standardization, which resulted in high initial cost
for the systems and contributed to reduced system reliability.
A very important institutional factor strongly influencing technical considerations
has been the isolation of on-site power-generation facilities from the electric utility grid.
On-site total energy power generation facilities are clear competitors with local or
regional utility power companies. Historically, utility companies have been given exclu-
sive license to supply energy across public throughfares while operating under public
regulation to supply energy to all regions (even those marginally economic) at regulated
rates of return on investment. This has limited the size of total energy plants, and in
many cases, prohibited the servicing of a mixed application load that may have been
particularly suited to total energy. In effect, the total energy plant and application must,
in many cases, be located on the same city block.
In addition, stand-alone total energy systems suffer from high expense for capacity
to meet nondiversified peak loads as well as equipment redundancy to provide system
reliability. In those cases where total energy systems have been interfaced with the utility
grid for backup, the charges for this service have been very high, adversely affecting the
economics of such total energy systems. Also, until recently, there has been little desire
Total Energy Systems Design 165
on the part of the electric utilities to purchase excess power from on-site plants. There
are current indications, however, that utility attitudes may be changing in this area.
In a memo [3] dated September 6, 1977, Southern California Edison has advised its
Rate Book Holders that "Edison will accept energy generated back into its system at
times that customer generation exceeds customer needs." This experimental tariff
applies to customers having 20 kW demand or less.
Total energy may be envisioned as a process that cascades energy of high quality to
lower quality in discrete steps. At each step, characterized by temperature, energy is
withdrawn from the stream and utilized to meet requirements particular to that given
temperature. Thus, low-temperature, low-quality energy is utilized to meet heating and
cooling requirements, whereas high-temperature energy is utilized for conversion to high-
quality electrical energy. An equivalent system in which the fuel input is displaced by
solar energy is termed a solar total energy (STE) system. Such a STE system incorporat-
ing solar-specific technologies, as energy storage, and solar collectors, not normally
found in conventional total energy systems, is represented schematically in Fig. 32.3.
or
!j // ,!/
SOLAR i ' ! ' !
THERMAL/ i i
COLLECTOR/ i 0
HIGH -TEMPERATUR
( STORAGE
HEAT
EXCHANGER
II \
II I \ n
II 1 POWER — I1
il il l CONVERSION II
II
1l1 ,; SYSTEM I
,.,.. ........ ri
.. ,--:-... 10
11
HEAT
. .. ..-• i1 0 EXCHANGER
.-..
.:::.\.:.' .....,4 , \
..-- ....... . ,
LOW — ..... .....,
TEMPERATURE \. .. . -,
\J
THERMAL ENERGY .. ':
LOW-TEMPERATURE
STORAGE
The major components are the solar collector field, thermal storage, and the power
conversion system.*
A. Energy Conservation
The first thing that anyone interested in evaluating the application of any solar technol-
ogy to supplying energy needs should do is evaluate the potential for energy conservation.
Without doubt, a potential STE user will find that it is considerably cheaper to save
energy than to generate it using solar technologies. During visits to many industrial
plants as part of a recent STE application study [4], McDonnell Douglas concluded that,
in every case, significant conservation could be easily achieved through minor process
modifications. Indeed, in some cases, conservation may displace more conventional
energy than installation of a STE facility.
*Although this sytem shows electric power generation via a thermodynamic cycle, photovoltaic cells
could be used. Discussion in this chapter will be restricted to solar thermal total energy concepts with
the knowledge that the technologies discussed in Chap. 37 can be combined with the reject heat-
reclamation techniques discussed here to construct photovoltaic solar total energy systems.
Table 32.2 shows some of the factors that should be considered in evaluating the
potential for energy conservation in an application serviced by a STE facility. Basically,
these factors consist of making the building weathertight, better use of internal lighting,
and process modifications. The energy conserved and resultant cost savings by evaluating
such considerations in actual solar installations are reflected in Tables 32.3 and 32.4.
In each case the cost of the STE facility could be reduced 30 to 50 percent through
energy conservation. The loads listed in Table 32.1, which are the design loads, are meant
to represent the energy-conservative loads obtained after load engineering.
Table 32.4 Solar Plant Cost Savings Resulting From Energy Conservation
Facility Design
Annual Cost Above Capital
Energy Conventional Cost Reduction
Project Reduction (%) Design (%) Solar Plant (%)
Sandia Laboratories 40 17 50
Building 832
Dallas County 62 5 33
Community College
Bleyle Knitwear 37 Some cost 25-30
Manufacturing, savings
Shenandoah, Georgia
168 Applications of Solar Energy
Some applications are better suited to the employment of STE technology than others,
and it is worthwhile at this point to check the load characteristics of the application to
help decide design strategy. Basically, the objective of a cascaded total energy system is
to maximize the utilization of energy input to the power conversion system (PCS). The
ratio of the quantities of thermal to electrical energy leaving the PCS is defined by the
efficiency of power generation. A well-suited application will have thermal and electrical
energy requirements in a ratio similar to that ratio defined by the PCS. Thus, the first
step in our design definition will be to look at expected power-conversion efficiencies
followed by load redefinition, if possible, to make the application's energy demands
more compatible with the PCS energy outputs.
A. Power Conversion
Perhaps the area that most distinguishes solar total energy from other solar technologies
is the generation of electrical energy at decreased power-conversion efficiencies in order
to have resultant reject heat at temperatures hot enough for beneficial use. The desired
temperatures of the reject heat will vary depending on the application, with resulting
variations in the efficiency of power generation. In the case of the example STE system,
process steam at 135°C is desired.
A method of approximating the efficiency of a Rankine cycle [5] is given here that
eliminates the requirement for a detailed cycle analysis while yielding results sufficiently
accurate for baseline design considerations.* The heart of the method consists of a
modification of the normal Carnot efficiency [6] definition.
Thigh Tlow
Carnot efficiency — (32.1)
Thigh
where Thigh (basically the turbine inlet temperature) is replaced by Thigh average and
this efficiency is then modified by empirical factors. The form of the equation is given
by
where the temperatures are in degrees absolute. For saturation cycles with boiling
occurring at constant temperature, T high average Tboiling • For superheat cycles,
Thigh average (Tboiling + Thigh)/2. For supercriticalcycles , Thigh average (Thigh +
Tiow)/2. Tiow is the condensing or reject temperature. EG and ET are the generator
and turbine efficiencies, respectively, which can be taken from Fig. 32.4. The factor K
is an empirical factor, and a value of K = 0.8 seems to work well for everything but
supercritical cycles, where K = 1.1 seems to work better. If the cycle being analyzed
*Sec. 32.11 goes into more depth on selection of the proper PCS.
Total Energy Systems Design 169
1.0
EG
0.9
CORRECTIONFACTOR
0.8
0.7 ET
0.6
0.5
Figure 32.4 Variation of turbine efficiency (ET) and generator efficiency (EG) with
PCS capacity.
is receiving heat from a sensible heat source fluid entering at T i and exiting at T2 then
there are restrictions on T1 and T2 that are different for each cycle as summarized in
Table 32.5. The restrictions on T1 and T2 arise from heat transfer limitations in driving
a Rankine cycle with a sensible heating fluid. This is the so-called pinch-point problem,
and will be discussed further in the detailed design section.
The method given above seems to work with acceptable accuracy (±10 percent in
most cases) and appears appropriate for broad-scope system analysis, but care should be
taken to do detailed analyses at points identified as being of interest.
As an example of the use of Eq. (32.1), consider a superheat toluene cycle serving
the example STE system with Thigh = 340°C, Tboiling = 246°C, and T iow = 135°C. The
temperature enthalpy (T-H) diagram for this cycle is shown in Fig. 32.5. T high = 340°C
was chosen because at temperatures much in excess of 340°C, line-focusing solar collec-
tor efficiencies tend to degrade significantly. Tboiling = 246°C is also somewhat arbitrary,
and was chosen as representative of a typical cycle. Sec. 32.11 discusses variations in
these temperatures. From Fig. 32.4 we have EG = 0.94 and ET = 0.71 for a turbine
with 600 kWe capacity. Table 32.5 yields a value of 0.8 for K in a superheat cycle.
300
250
200
2
w 150
I-
We thus find:
340 + 246
Thigh average = + 273 = 566
566° K
2
= 0.15
Thus, in our example system utilizing the above-defined toluene PCS, the ratio of thermal
to electric energy leaving the PCS is 85/15 = 5.7. An application with a thermal demand
to electric demand ratio less than this would either have to discard excess thermal energy
while satisfying the electric demand or be content with meeting the thermal demand
while generating only part of the electric demand (which would necessitate purchasing
makeup energy from the electric utility).
Total Energy Systems Design 171
Most thermal STE systems are driven to the use of high-temperature concentrating
collectors by thermodynamics. In order to generate electric energy with some sort of
reasonable efficiency while maintaining the reject heat at a useful temperature, the inlet
temperature to the PCS must be reasonably high. Most current state-of-the-art STE
designs employ PCS inlet temperatures near 300 to 350°C. Actually, this temperature
is set more by the current limitations on thermal storage (see Sec. 32.8) than collector
technology. Removal of storage temperature limitations (either through development of
advanced concepts or using a design that eliminates the need for thermal storage) would
strongly encourage the generation of higher PCS inlet temperatures. This would result
in more electrical output with a decreased low-temperature thermal output per unit of
thermal energy input to the PCS.* As discussed above, 340°C will be used as the tem-
perature of the thermal energy out of the collector field in the example STE system.
An in-depth discussion of the performance and characteristics of solar thermal
concentrating collectors is presented in Chap. 10 of this handbook. The objective here is
not to reiterate that discussion but to provide a somewhat simplified method for quickly
sizing collector fields for various locations. In accordance with the stated objectives of
this chapter, such sizings are of a preliminary nature and useful in determining if there is
sufficient feasibility to warrant pursuit of a more complete design.
As part of its role in constructing the Solar Total Energy Systems Test Facility,
Sandia Laboratories has evaluated several distributed collector concepts for inclusion in
that facility [8] . A summary of the results of that evaluation are presented in Table 32.6.
The cost of construction, collector efficiency, and quantity of energy collected through-
out the year is reported for each collector along with a "figure of merit" reflecting the
resultant cost of collecting energy. The numbers in this table will no doubt change as
collector designs change, but they are exceedingly useful to the engineer doing baseline
design and may easily be adjusted to incorporate design and cost changes (see below).
Basically, Table 32.6 was generated using manufacturer-supplied design and cost infor-
mation, which was then adjusted to reflect developments foreseen by Sandia Labora-
tories as reasonable design improvements. Thus, for example, the reflectivity of all
reflective surfaces was set at 0.92 (the specular reflectivity of back-surfaced silvered glass)
regardless of the reflector material in the current design. Receiver efficiency is seen to be
most sensitive to type, with linear, noncavity receiver efficiency at about 70 percent,
linear cavity receiver efficiency at about 75 to 80 percent, and the point-focusing receiver
at 93 percent. The product of the concentrator and receiver efficiencies given in Table
32.6 yields the overall efficiency defined as the amount of solar energy passing through
its aperture that a collector converts to thermal energy in the heat transfer fluid at 290°C.
It will be assumed here that 290°C is close to the average collector field tempera-
ture associated with a baseline design delivering 340°C out of the collector field. Note
that to obtain the daily net energy figures, the time-dependent cosine of the angle
between the sun and the collector aperture must be considered in order to compute the
solar energy entering the collector aperture from knowledge of the direct normal solar
insolation available. The net energy figures also consider the shadowing of one collector
by another as well as heat loss from insulated connecting piping, which occurs in the
*The selection of the proper operating temperature for the solar collectors involves a study of the
interactions among the collector field, storage, and the PCS, as is discussed in Sec. 32.11. Because of
the decreasing efficiency of line-focusing collectors with increasing temperatures, high-temperature
collector fields may be dominated by point-focusing devices.
Total Energy Systems Design 173
actual STE Systems Test Facility Collector field. These effects are, of course, important
in the trade-off studies of the detailed design stage and are discussed further in Sec. 32.11.
It is also important to point out that the net energy values of Table 32.6 are for
average clear days during the seasons listed. Thus, these values are good for sizing collec-
tor fields but not for determining annual performance. (See Sec. 32.10.)
Use of the numbers in Table 32.6 as applied to the example STE plant is illustrated
in Table 32.7, where a collector area of 15,000 m2 and a land requirement of 45,000 m2
for a collector field using E/W parabolic troughs are computed.
Note that the summer net energy value was used in computing the collector area.
This was done to minimize the amount of collected energy that is wasted—the net energy
value from Table 32.6 for the summer is greatest of the four seasons and will thus yield
the smallest collector field. Since the demand is constant throughout the year, a STE
174 Applications of Solar Energy
collector field sized for yearly average net energy would collect more energy than
required during the summer. The excess energy would have to be dissipated, thus lower-
ing the cost effectiveness of the collector field. A collector field sized for the summer,
however, would not result in wasted collected energy but would be somewhat undersized
during the rest of the year, resulting in the need to purchase makeup energy. Given the
two alternatives, it is normally most cost effective to design for minimized waste of
collected solar energy. In this case that means sizing for summer conditions. It is pos-
sible too that, because of weather conditions in some locations, summer may not be the
period of highest daily solar insolation, in which case a different sizing season would be
chosen.
The proper sizing season may also vary depending on the load demand and its
seasonal variation (e.g., schools). Basically, the objective is to design for maximum
energy displacement while minimizing the waste of collected energy. The seasonal
variation in collected energy is not very large for the E/W parabolic trough chosen for the
example, but is can be quite large for other types of collectors such as the parabolic dish
(see Table 32.6). The problem of seasonal collector performance is discussed further
in Sec. 32.11.
Table 32.7 also computes the amount of land required for the deployment of the
collector area defined above. A packing factor of 33 percent was assumed. Land utiliza-
tion and the effect of shadowing should be carefully considered during the detailed design
stage (see Sec. 32.11).
90
80
OVERALLCOLLECTOREFFICIENCY
70
60
50
40
30
400 600 800 1000
SOLAR INSOLATION , W/m2
fixed. If the solar intensity input to the receiver changes, however, the collector efficien-
cy will change. Table 32.6 was constructed for collectors located in Albuquerque. If the
collectors were moved to a location with lower solar insolation, such as Boston, solar
collection would suffer for two reasons—there would be less solar energy to collect, and
the collector would perform less efficiently. Another way of saying this is that the
receiver efficiencies reported in Table 32.6 are insolation dependent.
Such dependency is normally handled in the detailed design stage by computer
modeling of the collector characteristics including heat loss. For our purposes here,
however, it is desirable to be able to simply correct the "figure of merit" numbers in
Table 32.6 for solar insolation. Figure 32.6 shows the insolation dependence of collector
efficiency for the collector types listed in Table 32.6. This figure was constructed using
predicted heat loss data as computed from Table 32.6. For example, given that the
concentrator efficiency of a parabolic trough is 0.85 while the receiver jacket transmit-
tance is 0.93 and receiver absorptance is 0.95, 701 W/m2 is absorbed by the receiver tube.
To achieve the reported 70 percent receiver efficiency at 290°C and 936 W/m2 solar
insolation, the receiver heat loss must be 150 W/m2 based on the collector aperture area.*
The receiver efficiency was recomputed at 290°C as the solar insolation entering the
collector aperture was varied. The resultant collector efficiencies were computed using
the following expressions:
_ Rinsol Rhl
Erec D
(32.4)
"insol
*Recent preliminary test results for parabolic trough collectors indicates that heat loss at 290°C may
be somewhat less than 150 W/m2 . The results of this and other testing will be published in the future
as described in Sec. 32.12.
7.0 6.0
5.0
SAUL1 51E. MARIE
6.0
BROWNSVILIE
6.0
A2J ou3 JPios Jo suo!lpon d de
Figure 32.7 Mean daily direct normal solar insolation available during the summer season (kWh/m2 • day).
Total Energy Systems Design 177
where
Since the receiver temperature remains constant, the heat losses are constant while the
light incident on the receiver varies.
As the average direct normal insolation available to a collector changes, so does its
average collection efficiency.* As an example, consider the problem of designing a STE
facility to meet the loads of the hypothetical application but located in Boston,
Massachusetts. This involves resizing the collector field for Boston summer conditions.
Figure 32.7 is a map of the continental United States from Boes et al. [9], which shows
summer mean daily direct solar insolation in the United States with Boston somewhat
less than 5.0 kWh/m2 and Albuquerque somewhat greater than 8.0. Therefore, as a rough
estimate, Boston receives approximately one-half the summer direct normal solar insola-
tion of Albuquerque. Thus, the collector area required to collect 2.03 X 108 kJ/day
would be about twice the Albuquerque area or 30,000 m2 . This collector area should
also be increased by a factor of about 57/49 (the ratio of the respective parabolic trough
efficiencies assuming 10-hour days from Fig. 32.6), due to decreased average collector
efficiency. The resultant new collector area is approximately 35,000 m2 .
At this point we have defined an energy-conservative load compatible with the PCS
of the STE system and sized the solar collector field to meet that load. The next step
will be to look at phasing of the load. If the energy demanded by an application is not
coincident with the availability of solar insolation, energy storage must be contemplated.
The next section deals with determining the quantity of high-temperature storage needed.
Storage requirements arise from the desire to decouple the energy demand from a fluctu-
ating energy source, the sun. A typical requirement is for nighttime energy. Even if no
energy demand were present at night, storage is frequently required to flatten out the
daytime solar energy supply profile, which peaks at midday and tends to fall off in the
morning and afternoon. This is frequently done in order to provide the PCS with a
reasonably constant load to avoid serious off-design performance penalties (see Sec.
32.11). Although detailed analysis of the load requirements is needed to determine
accurately the proper quantity of storage, it is reasonably straightforward to estimate
storage size for concept feasibility assessment.
*Note that the collector efficiency values reported in Table 32.6 are not annual efficiencies, but
instead reflect the collection efficiency after the light enters the collector aperture. Only the net
energy values have the cosine effect included.
178 Applications of Solar Energy
There are many options open to the STE designer for storage of thermal energy.
Unit 6 of this handbook discusses the various energy storage concepts in detail. At
present, the leading candidate for energy storage in STE systems is thermal energy storage
and in particular sensible heat thermal energy storage. Although heat-of-fusion energy
storage holds much promise for high-density heat storage at competitive costs, the
more highly developed sensible heat storage concepts are best understood and are current-
ly the concepts most in use for high-temperature storage in STE systems operating in the
range 300 to 400°C. Many high-temperature sensible heat storage concepts employ an
oil heat transfer fluid. The high vapor pressure of water above 240°C limits the use of
water as the storage medium. The use of oil limits the operational temperature of these
storage concepts to about 400°C due to thermal decomposition at higher temperatures
[10]. Some silicon oils, potentially stable at temperatures above 400°C [11], are under
consideration for collector and storage heat transfer media. Additional drawbacks of oil
media include their relatively high cost (current prices range from approximately $0.75/
gal to $15/gal depending on the type of oil) and their flammability and potential adverse
health affects. Chapter 24 discusses high-temperature sensible heat energy storage in
detail.
If, in the example STE system, it were desired that the STE system satisfy the daily
demand during the hours of 8 a.m. to midnight (see Table 32.1), approximately one-half
of the total energy demand would, on the average, be required outside daylight hours
resulting in a daily storage of about 108 kJ. Additional storage would not be advisable
due to the decreased utilization associated with the extra storage capacity and the extra
solar collectors associated with it. Each unit of storage requires not only the expense of
the storage unit but also the expense of the solar collectors required to collect the energy
needed to charge the storage. It is very difficult to justify the cost of any collectors plus
storage capacity that would not be utilized to their maximum extent throughout the
year. The question of proper storage size will be reopened in Sec. 32.10.
One case in which storage capacity in excess of that needed to meet daily operating
needs appears possibly economic is when excess storage leads to increased utilization of
existing collectors [4]. Such a situation might arise if decreased loads were encountered
on weekends. In this situation, additional storage capacity that stored energy during the
weekend (when there was reduced demand) for the upcoming week might prove econom-
ic. Such storage capacity would increase the amount of useful energy produced by the
collector field on a weekly basis without increasing the collector area.
The designs performed up to this point have utilized average, steady-state data. Solar
systems are, however, subjected to fluctuating solar conditions as well as variations in the
load demand. Such variations can only be evaluated properly by computer simulation.
It is strongly recommended that potential STE designers perform simplified computer
simulations prior to executing a detailed design in order to verify baseline design concepts.
A. Computer Simulation
It is of interest to know how representative the baseline design executed to this point is.
To evaluate the design for the example STE, an hourly computer simulation of the per-
formance of the baseline design was performed. The logic flow of the computer routine
used to do this is illustrated in Fig. 32.8.* The routine computes the energy leaving the
collector field based on available solar insolation and unshadowed collector area. This
energy production is compared with the demand for thermal energy, with excess thermal
energy being stored in either high- or low-temperature storage. The thermal demand by
the turbine is determined by dividing the input electric demand by the input cycle
efficiency . Furnaces are included in the logic to keep track of when and to what extent
the solar design does not meet the demand.
The results of the above simulation are shown in Table 32.8. The average solar
displacement is obtained using solar insolation data from Boes et al. [9] . A page from
that work is reproduced at Table 32.9, which gives the average daily solar insolation
incident upon an E/W tracking surface for Albuquerque. An annual average insolation
can be computed by averaging the four seasonal figures. This average annual insolation,
when multiplied by the collector efficiency reported in Table 32.6 and the collector area
provides the annual solar energy produced.t All the collected energy was assumed to be
*This computer routine is an unpublished Sandia STE system design performance evaluation code. It
requires 5 to 10 seconds for the hourly simulation of a year's performance on a CDC 6600 computer.
tFor locations other than Albuquerque, a correction to the collector efficiency for insolation would
have to be obtained from Fig. 32.6 as described in Sec. 32.7.
COOLING
TOWER
A YES
G.T. = Greater Than
H.T.
CSOLA1D STORAGE H.T.
DATA FULL STORAGE
IYES (AS NEEDED)
SOLAR SOLAR G T SOLAR + NO
COLLECTOR (ELEC DEM)/n, H.T. STORAGE
?(AS NEEDED
G.T.
1' (ELEC. DEM)/n 'FURNACE
I
COLLECTOR YES
CHARACTERISTIC) TURBINE
(EFF=n )
T
COOLING
TOWER
YES
NO THERM. REJECT
L.T. L.T. YES G T.
STORAGE STORAGE
THERM. DEM.
FULL
FURNACE (AS NEEDED) NO
(AS NEEDED) THERM. REJECT +
L.T. STORAGE G.T.
DEM.
r 1 THERM. DEM.
4- JYES
mos Jo sucqpo!i ddv
Table 32.9 Direct Normal (D) and Total (T) Energy Available to a Tracking E/W
Surface (kWh/m2 • day)
Spring Summer Fall Winter
(M, A, M) (J, J, A) (S, 0, N) (D, J, F)
D T D T D T D T
Albuquerque 5.4 7.6 6 8.1 5.3 6.4 4.8 5.6
Apalachicola 3.8 5.9 3.4 5.9 3.4 4.8 2.8 3.9
Bismarck 3.3 5.4 4.6 6.9 3 4.1 2.3 3.1
Blue Hill 2.6 4.6 3.1 5.5 2.4 3.6 1.9 2.6
Boston 2.6 4.6 3.1 5.4 2.4 3.5 1.9 2.7
Brownsville 3.1 5.6 4.3 6.9 3.3 4.9 2.6 3.9
Cape Hatteras 3.5 5.6 3.6 6 3.1 4.5 2.6 3.5
Caribou 4.2 5.6 4.3 6 2.6 3.5 3.4 3.9
Charleston 3.2 5.3 3 5.6 2.9 4.4 2.6 3.6
Columbia 3.2 5.1 4.2 6.5 3.2 4.5 2.5 3.3
Dodge City 4.3 6.1 5.2 7.3 4.2 5.4 3.9 4.8
El Paso 5.7 7.6 6 8 5.2 6.5 4.9 5.8
Ely 4.9 6.9 5.8 8 5.1 6.2 4 4.8
Fort Worth 3.3 5.5 4.4 6.8 3.5 4.9 3.1 4
Great Falls 3.2 5.4 4.8 7.1 3.1 4.2 2 2.7
Lake Charles 2.8 5.2 3.1 5.9 2.9 4.5 2.2 3.4
Madison 2.9 4.9 3.7 6.1 2.5 3.8 2.1 3
Medford 3.4 5.4 5.5 7.5 2.9 4.1 1.3 2.1
Miami 3.2 5.7 2.6 5.5 2.7 4.7 3.1 4.6
Nashville 2.9 4.9 3.5 6.1 2.8 4.2 2 2.9
New Yrok 2.5 4.6 2.9 5.3 2.2 3.4 1.7 2.5
Omaha 3.2 5.4 4.3 6.6 3.1 4.3 2.6 3.4
Phoenix 5.7 7.7 5.9 8.1 5.1 6.4 4.5 5.3
Santa Maria 4.2 6.2 5.4 7.2 4.2 5.4 3.5 4.4
Seattle 3.5 5 5.3 7 2.5 3.4 1.3 1.9
Washington, D.C. 2.8 4.9 3.2 5.7 2.6 3.8 2 2.8
182 Applications of Solar Energy
OMAHA )110r410SE::ORK
CHARLESTON
FORT WORTH
116.19 07
AR73
Figure 32.9 Mean daily direct normal solar insolation—annual average (kWh/m2 • day).
usefully applied against the demand without dissipation. The displacement obtained
through simulation was obtained from the hourly summation of the computed solar
contribution to the load. The computer simulated performance showed a 74 percent
displacement value, whereas the average performance evaluation indicated a 79 percent
displacement of the load demand by solar—reasonably good agreement.*
In order to compute the annual average daily insolation incident on an E/W track-
ing surface at a location not listed in Table 32.9, a city with similar annual average daily
normal insolation can be used. Figure 32.9 can be used to identify which cities in
Table 32.9 have annual direct normal averages similar to the actual location.
In addition, the computer simulation pointed to the need for about 35 percent less
storage capacity than indicated by the average-value analysis. The reduction in storage
can be evaluated by looking at the effect of the additional storage on solar displacement
as shown in Fig. 32.10. At a storage capacity of 108 J, the percentage of the load dis-
placed by increasing storage, while holding the collector area constant, has reached a
maximum. This reflects that, by itself, increased storage does not lead to increased load
displacement if there is not enough collector area to charge the storage. Additionally,
Fig. 32.10 graphically shows the decreasing effect of storage on displacement as storage
*The agreement is not too surprising, since the redefined loads for the hypothetical application are
constant and have a good thermal-to-electric ratio. Had there been fluctuating loads, for example,
agreement would not have been as good, and computer simulation would have been more crucial.
Total Energy Systems Design
PERCENT UTILIZATION/ DISPLACEMENT(%)
183
100
80
60
40
20
0
0 0.5 1.0 1.5 2.0
Figure 32.10 Variation in solar displacement and utilization with storage capacity.
capacity approaches 108 J. This is the result of the collector field being sized for a clear
summer day and thus undersized at all other times. Proper sizing of storage can really
be achieved only by contrasting the value of additional increments of storage (starting
at no storage) with the value of the resultant annual energy cost savings.
A similar trade-off study can be performed evaluating the cost of additional
collector area versus the quantity of energy displaced. Figure 32.11 illustrates this
contrast for the example STE system. Near 15,000 m2 , selected above as the proper
i- 100
Z
1•
.1-1 90
w PERCENT UTILIZATION OF
0 80
a COLLECTED SOLAR ENERGY
_1
cl- 70
(n
PERCENT LOAD
▪ 60 DISPLACED BY SOLAR
o
R 50
a
NJ 40
7J
D 30
D
1—
z 20
Li
0
cr
w 10
Q-
00
15 20 25 30 35 40
Figure 32.11 Variation in solar displacement and utilization with collector area.
00
Table 32.10 Preliminary Cost Algorithms for Candidate Capital Equipment Items (Installed)
Unit Size Range Algorithm
Storage
Hot storage 150-150,000 ft3 Cost = $352 (vol., ft3 ) 0.515 + $3 (vol., ft3 )
Cold storage 150-150,000 ft3 Cost = $352 (vol., ft3 ) 0.515
Battery Cost = $50 (size, kWh) + $75 (system size, kWh)
Packaged steam boilers 104 -105 lb steam/h-unit Cost/unit = $3690 (lb/steam/h unit) 0.32
Refrigeration
Centrifugal chillers (electric) 250-2,000 tons Cost/unit = $4,240 (tons/unit) 0.61
Centrifugal chillers (engine) 500-2,000 tons Cost/unit = $2,270 (tons/unit) 0.745
Absorption chillers (steam) 250-1,200 tons Cost/unit = $2,194 (tons/unit) 0.7
Power-conversion systems
Sunstrand organic Rankine cycle 100-1,000 kW Cost = $3,815 (kW/unit) 0.66
Rankine cycle-steam 1-1,000 kW Cost = $22,700 + $350 (kW)
Rankine cycle-steam 1,000-1,000,000 kW Cost = $1,250 (kW) 0.825
Shell-and-tube heat exchangers 20-2,000 ft2 Cost/unit = $300 (area, ft2 ) 0.56
Cooling towers-dry Cost/unit = $40 (area, ft2 ) 0.6
A2Jau] RI GS JO SU OpP OilddV
Total Energy Systems Design 185
collector area for the load demand defined in Table 32.1, the fraction of the load dis-
placed by solar starts to progressively decrease with added collector area. In addition,
above 15,000 m2 of collector area the utilization of collected solar energy begins to
drop, indicating wasting of collected solar energy and decreased cost effectiveness.
Since the design philosophy was to maximize utilization of solar energy, it is not surpris-
ing that 15,000 m2 of collector area results in 98 percent solar utilization. Notice that
above about 30,000 m2 , almost no additional demand can be displaced by adding extra
collector area. This is a result of not sizing storage for much more than one day's duty
cycle.
The above economic trade-off studies should be continued until the proper collec-
tor and storage sizing are obtained. For purposes of concept evaluations, however, sizing
for maximum utilization is reasonable, since, as in Figs. 32.10 and 32.11 show, the cost
effectiveness of the STE system starts to decrease above these sizes.
B. Economics
There is not much to be said about STE economics that can be considered firm. The
technology is developing and as a result cost estimates can vary considerably—there is
virtually no experience with operations and maintenance of a STE facility, for example.
Whenever possible, vendor quotes should be used in assessing capital costs. The method-
ology for computing life-cycle costs for solar thermal systems is discussed in detail in
Chap. 46.
Some estimates developed by Sandia Laboratories for solar collector costs are given
in Table 32.6. In addition, Atomics International [12] has developed capital-cost
algorithms for some of the major components in a STE system. These algorithms are
listed in Table 32.10 together with the resultant capital costs for the example STE system
as an illustration. The Atomics International storage-cost algorithms have been slightly
modified to reflect mixed media rock/oil storage. The storage cost in Table 32.10 is
referenced to a volume and hence a temperature difference across storage, as well as a
heat capacity, is required in order to compute the cost. For concept evaluation, it is
suggested that 80°C for the temperature difference and 1.88 X 106 J/m3 • °C (29.5
Btu/ft3 • °F) for the volumetric heat capacity be used in order to get an idea of storage
costs [13]. The determination of the actual temperature difference is discussed in the
next section on detailed design. It must be stressed that use of such cost algorithms are
estimations at best and that vendor quotations should be used, if possible, especially
during detailed design.
If, at this point, the applicability of solar total energy to a user's application seems
promising enough to execute a detailed design and performance evaluation, it is worth
considering what the characteristics of such a design should be and what information
will be required by the user to allow a designer to execute a first-rate design. The list
is not meant to be all inclusive, but instead points out some of the more important
186 Applications of Solar Energy
aspects of the detailed design process. The term detailed design is used here to mean
those activities that lead to the decision actually to construct a facility and is not meant
to have a formal technical meaning. Indeed, if during the detailed design stage significant
changes to the baseline design occur, the designer should return to the baseline design
stages above to see if this changes the potential of STE for the application.
Perhaps the most important step prior to initiating any action toward executing a detailed
design of a STE facility is to visit an existing solar facility. Several solar facilities utilizing
technology applicable to STE that are currently under design and/or construction and
operation are listed in Table 32.11.
The design and operation of such facilities has yielded practical experience that
impacts system design. A partial list of some of the more important practical lessons
learned is given below and forms the basic outline for the remainder of the chapter.
It should be stressed once again that the STE design process is iterative. Design
constraint imposed by one consideration must be evaluated in light of how this constraint
affects other aspects of the design. This is perhaps best illustrated in the Solar Subsystem
Trade-Off Studies section discussed below. To start, however, let us consider the input
data required to begin the detailed design.
for comparison with any long-term climatological data to be used in the design process.
A weather station that measures such data as direct normal solar insolation, total horizon-
tal solar insolation, dry-bulb and wet-bulb temperatures, and wind conditions would be
adequate. A schematic of the weather station located at the site of the Bleyle Plant in
Shenandoah, Georgia, is given in Fig. 32.12.
As an example of the impact of solar insolation on STE design, consider the season-
al variations in net energy collected for each collector, reported in Table 32.6. The differ-
ences arise mainly from the seasonal variation of the angle of incidence of the sun with a
given collector (which is a function of tracking capability) and the number of daylight
hours. The actual solar intensity also varies seasonally, and its magnitude must be known.
The seasonal variation of solar output from a collector field can have an important effect
on design.
In the comparison of a two-axis tracking parabolic dish with a one-axis tracking
E/W trough, the dish has an average net energy value about 27 percent higher than the
E/W trough (see Table 32.6). At first glance, this would say that the dish is 27 percent
more cost effective than the trough. This conclusion, however, is valid only if you can
JUNC
SOLAR
DATA - 7 BOX
SYSTEM
WEATHER
AC WIND DIRECTION
SERVICE WIND SPEED
BAROMETER PRESSURE
HUMI DITY
TEMPERATURE
D RECT
NORMAL
INSOLATION
(DUAL
P RHE LI OMETE R)
-EQUATORIAL
_ -
MOUNT
TRACKER
POWER
TOTAL SOLAR
INSOLATION
( PYRANOMETER)
- - - POWER
DATA
use all the energy generated by the dish. A look at the net energy values for the dish and
trough in Table 32.6 reveals that the output from the dish peaks very strongly in the
summer, whereas the trough presents a much flatter seasonal profile. Thus, if the applica-
tion cannot utilize the excess energy generated in the summer by the dish, assigning a cost
advantage to the dish is invalid. Alternately, a STE designer may wish to install a two-
axis tracking dish system to service a seasonally flat load with the knowledge that such a
system, if sized so as not to discard energy in the summer, would require purchase of
utility energy in the winter. On the other hand, an E/W trough system may be chosen
even though its average "figure of merit" is listed as poorer than the dish because of its
relative insensitivity to seasons.
In summary, a detailed design exercise may require that design considerations,
established in the conceptual evaluation stage described earlier, be changed on detailed
analysis of collector and other component performance as well as knowledge of the
variations of the energy demand and solar resource.
COLLECTOR
EFFICIENCY
HEAT CYCLE
EFFICIENCY
u.
W
COMBINED SYSTEM
EFFICIENCY
OPERATING TEMPERATURE
regeneration process in the schematic. As can be seen, it is the temperature on either side
of the toluene boiler (i.e., the turbine inlet temperature Thigh and the regenerator outlet
temperature T reg) that basically define the temperature requirements placed on the col-
lector field. Tent and Tin refer, in Fig. 32.14, to the collector field outlet and inlet
temperatures respectively, and Tour - Tin is essentially the temperature drop across the
high-temperature storage unit.
The addition of sensible heat from the collector field heat-transfer fluid to the
power-conversion cycle is illustrated in Fig. 32.15. Basically the process consists of
preheating the PCS working fluid from the regenerator outlet temperature to the boiling
temperature, followed by vaporization of the fluid and superheating at constant pressure.
Note that the temperature of the heat transfer fluid must, at all times, be hotter than the
PCS working fluid to which it is transferring energy. Since vaporization occurs at con-
stant temperature, point A in Fig. 32.15 tends to act as a pivot for the sensible heat
addition curve. This effect, called the pinch point,* often results in Tin being consider-
ably different from Treg and the temperature drop across the collector field being con-
siderably different than the difference between Thigh and T reg depending on the cycle.
These effects must be considered in designing a STE system.
The temperatures used in the example STE system were Thigh = 340°C, T boiling
246°C, and Tin, = 135°C. Figure 32.16 shows the T-H diagram of Fig. 32.5 for this
toluene cycle (now denoted cycle 1), but includes, as well, the sensible heat input to the
cycle. In addition, a second cycle (denoted cycle 2) illustrates a second possible cycle
with T b = 288°C, along with its sensible heat input. T reg for each cycle is noted on
the T-H diagram. According to Eq. (32.2), the cycle with the higher Tbelling (i.e., cycle
2) is more efficient than cycle 1 (17 percent versus 15 percent) and, at first glance, may
T
TOUT HIGH
SOLAR TURB
011" THERYAL
COLLECT ENERGY
•
HT
REGEN
STOR EXCHG .
BOILER
Il
T T
Figure 32.14
IN
. T REE
'- TOUT
COLLECTOR
o•FIELD AT
HEAT-
TRANSFER
FLUID
rTIN
TEMPERATURE
REGENER- I
4-ATOR
TEMPER-
ATURE PREHEAT VAPORIZATION SUPERH-
EAT
HEAT SUPPLIED FROM COLLECTORS
LOW
appear to be a better system. However, this increase in cycle efficiency penalizes the
performance of the solar components.
The penalty can be seen in the sensible heat input curves. In raising the boiling
temperature from 246 to 288°C, the average temperature of the heat input [essentially
the average collector field temperature, (Tin + Tout)/21 increased from 298 to 304°C.
Figure 32.17 shows the temperature dependence of collector efficiency for the parabolic
trough reported in Table 32.6 and shows that the slight increase in average collector field
temperature experienced in going from cycle 1 to cycle 2 has negligible effect on efficien-
cy. On the other hand, the temperature difference across the collector field, as defined
by the sensible heat input curves of Fig. 32.16, decreased from 103 to 89°C in elevating
Tboil•ing to 288°C. This would result in a 16 percent increase in the volume of sensible
heat storage required to store a given quantity of energy.
The trade-off is thus between the increase in displaced conventional energy cost
savings generated by going from a 15 to a 17 percent efficient turbine system versus the
extra cost in storage resulting from the respective 16 percent increase in storage volume.
If, for example, electric energy is valued five times as much as thermal energy and there
is a use for the thermal energy, such an increase in turbine efficiency would result in a
5 percent increase in overall system cost savings. Thus, without changing the turbine
inlet or condenser temperatures, significant changes in the efficiency of operation of the
power-conversion system and cost of high-temperature thermal storage have been
observed.
194 Applications of Solar Energy
The effect of decreasing Thigh from 340 to 260°C while maintaining Thoiiing at
246°C is shown in Fig. 32.18. Cycle 1 is the same as above, whereas cycle 3 and cycle 4
refer to those systems with T high equal to 300 and 260°C, respectively. Treg for each
cycle is also indicated in Fig. 32.18. As seen by Eq. (32.2), cycle efficiency decreases
uniformly as Thigh is lowered. A possible strategy for lowering T high might be to reduce
the operating temperature of the collector field. Considering the pinch point, lowering
Thigh results in the reduction of average temperature of the sensible heat input from 298
to 254°C. As shown in Fig. 32.17, this corresponds to an increase in collector efficiency
from 59 to 62 percent. In addition, the temperature difference across the collector field
(and thus storage) decreases rapidly with decreasing Tin, and it is doubtful that the cost
penalties combined with the effect of decreased turbine efficiency and increased storage
volume would be offset by the small increase in collector efficiency. However, the
collector field is invariably the major cost item in a STE system and small reductions
Total Energy Systems Design 195
OVERALL COLL ECTOR EFFICIENCY
40
0 50 100 150 200 250 300 350
TEMPERATURE (CC)
in collector field size can result in large cost savings. An economic evaluation of each
cycle would have to be performed before it was clear which of the three cycles was best.
An additional type of turbine cycle not yet addressed is the supercritical cycle.
Figure 32.19 contrasts two supercritical cycles (cycle 5 and cycle 6), operating at T high
= 340 and 399°C, respectively (cycle 1 is not shown for clarity). The respective cycle
efficiencies computed for cycle 5 and cycle 6 are 16.2 and 18.6 percent. Again, these
cycle efficiencies and the energy cost savings they represent must be contrasted to the
resultant changes in solar collector and storage costs resulting from the corresponding
changes in collector field average temperature and temperature drop across the high-
temperature storage.
Although the above discussion concerned the power-conversion system and solar
collector field, similar trade-off studies should be performed for all subsystems. One
example would be the effect of absorption chiller inlet temperature on power-conversion
efficiency and chiller coefficient of performance.
It is hoped that the above discussion on the interaction of the solar collector field,
high-temperature storage, and the power-conversion system has given some idea of the
important design considerations that must be considered in a detailed design. Let us
now turn our attention to a discussion of the importance of off-design performance and
parasitic losses.
The rest of this section will briefly outline several additional areas of concern that
should be addressed during the detailed design.
196 Applications of Solar Energy
450
400
350
HEAT ADDITION
— 300 CYCLE I
0—
0 CYCLE 3
CYCLE 4
cc 250
cc ...- --,••
D e
I—
CC
Lii 200
0
2
LL.1
I-
150
100
50
0
0 250 500 750 1000 1250 1500
The above discussion on turbine cycle efficiency inherently assumed that the turbine was
operating at its design conditions. Frequently, in a STE system, where load demand may
fluctuate throughout the duty cycle, equipment such as the turbine may operate under
off-design conditions for considerable time. The resultant performance degradation can
be very significant. As an example, Table 32.12 illustrates the performance penalties
associated with operating a General Electric 600-kW marine turbine generator [15] at
two different power levels. This is a steam-extraction turbine of the type under consider-
ation at the preliminary design stage for use in the Shenandoah Solar Total Energy
installation. Turbine inlet, extraction, and exhaust temperatures are 299, 169, and
110°C, respectively. As can be seen, significant degradation in turbine efficiency can be
encountered at reduced loads, especially in the second stage. Recall that historically,
reduced power-conversion system performance, due to off-design operations, was cited
Total Energy Systems Design 197
450
THIGH
400— /CYCLE 6
/
/
/
350— HEAT CYCLE 5
ADDITION _---'
CYCLE 6
300— CYCLE z 4ti
il
f
TREG
7
, CYCLE 6
(-3 250
2
w
ct
o 200—
i_
/cyCLE 5
a
ce
w
a
w 150—
I-
100—
50—
Figure 32.19 Supercritical
toluene power-conversion
250 500 750 1000 1250 1500 cycles.
ENTHALPY IN JOULES PER GRAM TOLUENE
by HUD [2] as a strong contributor to the decline in fossil fuel-fired total energy systems
(see Sec. 32.3).
In addition to determining the quantity of energy produced by a STE facility, a
designer should pay careful attention to the quantity of energy required to run the
facility as well. The energy (called parasitics) required to operate the facility includes
pumps to circulate fluid through the various system components and the drive mechan-
isms of tracking collector fields. As an indication of the magnitude of the parasitics that
can be encountered in a STE facility, Table 32.13 reproduces some of the parasitics
associated with a conceptual design executed by Acurex Corporation for a STE plant
servicing the light-industrial Bleyle plant located in Shenandoah, Georgia. This design
utilized a parabolic trough concept and contained approximately 10,000 m2 of collector
aperture area. A large amount of parasitics is associated with absorption chilling, as can
be seen. Clearly parasitic losses are not negligible, and any detailed design must address
these power requirements explicitly.
Another type of energy loss requiring consideration with respect to collector field
layout is heat loss from the interconnecting piping within the collector field. Such losses
can be significant. In their anlaysis of the parabolic dish concept, Raytheon Company,
estimated [16] that about 16 percent of the energy absorbed by the receivers in a field
of 440 dishes is lost through interconnecting piping heat losses and night cooldown on an
annual average. Similarly, in an analysis to evaluate heat loss in a field of parabolic
trough collector [17] , a field of N/S tilted troughs lost 13 percent of the collected energy
through heat loss in the interconnecting piping, while an equivalent E/W field lost about
4 percent. The difference in losses was due to the extra interconnecting piping required
for a field of N/S tilted collector versus the E/W field. In an E/W field, troughs can be
butted end to end to provide long rows of collectors with little interconnecting piping,
whereas tilted N/S fields cannot because of the need to raise collectors off the ground at
one end. A N/S horizontal field would have piping losses similar to an E/W field.
The above discussion points out the necessity of carefully evaluating not only the
STE system's design performance but also how much energy is required for the STE
system to operate under normal conditions. The following section deals with another
area that can also lead to severe degradation of design performance.
Operations and maintenance (O&M) costs are a virtual unknown. About the only way
to accurately predict O&M characteristics of equipment is to look at the historical long-
term performance of similar equipment. However, there is no long-term performance
data available for most solar components, especially the collectors and storage. For this
reason, the operating solar facilities listed in Table 32.11 are extremely important for
providing indications of O&M even though these facilities have been in operation only
a very short time. Unfortunately, the rather short-term O&M data available from actual
collectors indicate deterioration of the reflective surface and selective coatings upon
exposure to the environment and operation at temperatures above 285°C.
Table 32.14 outlines several specific areas where significant O&M problems have
been uncovered. Cleaning without scratching of metallized reflector films, for example,
has proven to be such a difficult problem that harder surfaces, such as glass, may wind up
as the surface of choice simply for ease of cleaning. Similarly, selective receiver coatings
used at temperatures above 280°C tend to deteriorate rapidly, and dirt tends quickly to
build up inside receiver tube envelopes that are not sealed against the environment. All
these effects quickly degrade STE system performance.
Total Energy Systems Design 199
Most of the O&M problems being uncovered are typical of first-generation equip-
ment and will be solved with time. However, the problems are indicative of the immatur-
ity of the technology and unless carefully considered, will destroy any chance of building
a cost-effective STE system. Once again, anyone contemplating constructing a STE
facility is urged to visit existing solar facilities to get a current update on O&M
experiences.
F. Land Utilization
Perhaps one of the most serious constraints on energy systems located close to the point
of application is land availability. This constraint was clearly evident in a study of com-
mercial applications for STE executed by Atomics International [12] . Land costs near
commercial establishments, such as shopping centers, are very high and usually very
valuable for further commercial development. Atomics International found that the
average shopping center, for example, had at most about one-third the land area
(counting roof and parking lot area) necessary for a STE system to power the shopping
center, given current energy-consumption patterns. Conversion to energy-conservative
practices consistent with ASHRAE 90-75 could, however, reduce the demand by 50
percent in most cases.
Thus, any in-depth design of an on-site STE facility must carefully consider utiliza-
tion of available land. Utilization considerations include, among other things, esthetics,
health and safety to personnel and the general public, environmental impact, and local
codes as well as technical considerations such as the effect of shadowing from nearby
structures and other collectors within the collector field itself. As an example of the
impact of shadowing within the collector field on land area, Fig. 32.20 gives an illustration
200 Applications of Solar Energy
P
0
CC
w 25
z
w
0
Li 20-
i-
0
I.J.1
_J
-J
15 PHOENIX
o
J
4
o 10-
z
z BLUE HILL
4
5-
z
0
F
o 0
0 10 20 30 40 50 60 70 Figure 32.20 Effect of shadowing
w
x REDUCTION IN LAND AREA (%) on land utilization.
of some of the possible effects of reducing the land area on a field of two-axis tracking
linear Fresnel [4] collectors. The curves show that reducing the land area up to 25 per-
cent in Blue Hill leads to less than a 5 percent reduction in annual collector field output
over the zero shading case. In Phoenix land area could be reduced almost 40 percent
with similar results.
The effect of land reduction on the output from any collector concept considered
during the detailed design stage must be evaluated to determine optimized land utiliza-
tion. Some collector concepts may collect considerably more energy from a given land
area than other concepts, and the trade-off between collector cost and land utilization
must be evaluated.
G. Utility Interface
The manner in which an on-site energy system interfaces with the existing utility grid
system may seem subtle to the individual user, but it may represent one of the more
severe potential constraints on establishment of a broad based STE market. Whereas
an individual energy user may be concerned only about how STE affects his or her own
energy usage, the utilities must be concerned about how a multiple of such STE facilities
affect the utility grid with respect to such things as peaking, reliability, and cost of
supplied energy. A small number of on-site STE plants will probably not affect a large
utility grid very much, but installation of a large number of plants will, if they are inter-
faced with the grid.
If it is deemed desirable to interface with the utility grid, for either backup or
makeup power or both, consideration must be given to the type of demand profile the
STE facility presents to the grid and the reliability of maintaining that profile. Perhaps
the worst kind of demand a utility grid can have is an unpredictable one, and the rates
a utility would charge for such service are understandably very high. Rates for makeup
power would probably depend on the time of day such power makeup was needed and
Total Energy Systems Design 201
If one thing should be clear by now, it is that one chapter cannot do justice to the
problem of designing STE systems. To aid the reader who is interested in further dis-
cussion of STE design, a selected list of references incorporating most of the recent
activity in STE has been compiled. All of these references should be available from the
National Technical Information Service (NTIS). In addition, some of the references from
the body of the chapter are repeated here for completeness.
Perhaps the most comprehensive published design efforts to date are incorporated
in the conceptual designs for the Shenandoah and Fort Hood STE installations listed
below. The NTIS report number refers to the number NTIS recognizes when responding
to requests for copies of the report. Some of the reports consist of several volumes.
202 Applications of Solar Energy
1. Solar Total Energy—Large Scale Experiment No. 2 Phase II, Conceptual Design
Final Report—General Electric Company, Valley Forge, Pennsylvania, for U.S.
Department of Energy, NTIS-ALO/3985-1.
2. Solar Total Energy—Large Scale Experiment Phase II, Conceptual Design Final
Report—Acurex Corporation, Mountain View, California, for U.S. Department
of Energy, NTIS-AL0/3986-1.
3. Solar Total Energy System—Large Scale Experiment, Shenandoah, Georgia,
Final Technical Progress Report—Stearns Roger, Denver, Colorado, for U.S.
Department of Energy, NTIS-ALO/3987-1.
4. Solar Total Energy—Large Scale Experiment No. 1, Westinghouse Corporation,
Pittsburg, Pennsylvania, for U.S. Department of Energy, NTIS-ALO/3988-1.
5. Solar Total Energy—Large Scale Experiment No. 1, Final Report, TRW Energy
Systems Group, Redondo Beach, California, for U.S. Department of Energy,
NTIS-ALO/3989-1.
Another group of reports that are very informative, especially for the baseline
designer, are the application analysis reports. Of the four contracts to execute STE
application analyses, two are currently underway and two have been completed, of which
only one final report has been actually completed. When the final reports for these
contracts are completed they will have the same NTIS number as that listed for the other
application analysis reports issued by that contractor except that the last digit will change
(this last digit of the NTIS number indicates the sequential number of the report pub-
lished under a given contract). The last digit reported below should be verified prior to
ordering any reports from NTIS.
Thus, although a final report number may not be available at this time, its number
can be predicted rather closely. An exception to this are the Atomics International
reports, which are published by NTIS under an AI number.
Another very useful source of specific design information is the series of progress
reports covering the Solar Total Energy Systems Test Facility operated by Sandia Labora-
tories, Albuquerque. These reports can be obtained through NTIS by referring to the
SAND numbers below. The reports currently come out semiannually. Rather than
repeat the titles, only the reporting period and report number are listed.
Sandia Laboratories, Division 4721, may be contacted for the status of additional
reports.
A complete annual listing of all Sandia Solar Energy reports is published around
June of each year. The last listing published by John Gardner of Sandia Laboratories in
June 1977 was, Sandia Solar Energy Titles, SAND 77-0929. This listing may be obtained
either directly from Sandia Laboratories, Division 3144, or through NTIS.
REFERENCES
EVERETT D. HOWE
Department of Mechanical Engineering
University of California
Berkeley, California
Solar energy can be used to produce fresh water from sea or saline water by three
different distillation processes. The first two involve the humidification and dehumidi-
fication of air and simulate the hydrologic cycle by which natural fresh water is extracted
from the oceans. The third process uses solar-heated water to energize vacuum distillers,
the latter being adapted from those designed for use with fuel heat.
In the hydrologic cycle, which produces fresh water as rain or snow, solar radiation
is absorbed by the oceans and causes evaporation of water from the surface. The vapor
so produced increases the humidity and temperature of the air above the ocean surface.
This humid air is moved by winds (also produced by solar energy) to locations where
cooling takes place and some of the humidity condenses as rain or snow. This process
of humidification followed by dehumidification is accomplished on a small scale by
basin-type solar stills. These consist of shallow pools of saline water covered by sloping
transparent enclosures. Solar radiation penetrating the transparent cover is absorbed by
the water in the basin and increases the humidity and temperature of the air immediately
above the water surface. This warmed and humidified air rises, comes into contact with
the relatively cool inner surface of the transparent cover, and some of the humidity
205
206 Applications of Solar Energy
condenses there. This condensate slides down the sloping cover surface, is collected in a
trough along the lower end of the cover and conducted out of the enclosure to form the
product fresh water. In this basin-type still solar heat is used only once, for it must be
dissipated from the outside of the cover to cause condensation of the humidity.
A more refined application of atmospheric humidification and dehumidification is
the use of solar-heated water in multiple-type distillers. Although these are analogous in
many respects to steam-operated multiple-effect and multiple-stage flash distillers, they
are very different from the latter both in construction and in operation. With solar heat,
the plants operate under atmospheric pressure, whereas in conventional steam distillers
it is necessary to operate under vacuum in order to evaporate water in the absence of air
at temperatures below the atmospheric boiling point. Variations of pressure through the
solar plants are accomplished by changes in the partial pressure of vapor that occurs with
changes in temperature. Although plants of this type should be superior to those of the
basin type because of their more efficient use of solar radiation, the added complications
cause increases in cost that have offset the savings in many of the experiments to date.
The third type of distillation possible with solar energy substitutes solar-heated
water for fuel-supplied heat in conventional distillers of the vacuum type. However,
because of the temperature limitations of solar collectors, equipment for these plants
must be especially adapted for this purpose. Vacuum-type plants are superior to
atmospheric-type plants because they do not experience heat loss from air convection.
An excellent summary of efforts to use solar distillation to produce fresh water from sea
water or other saline water sources has been prepared by Talbert, Eibling, and Lof [1]
and is recommended for further reading. A few of the highlights will furnish background
material for this chapter. Although the current activities in solar distillation began less
than 40 years ago, Nebbia and Menozzi [2] have pointed out that as early as 1742 there
was reference to the solar distillation of sea water. However, there seems to be little
written material available from that date until more than 100 years later.
The first large solar still of record was constructed in 1872 at Las Salinas in
northern Chile, to provide drinking water for men and animals engaged in the transport
of nitrates from the desert mines. As described by Harding [3] , this distillation plant
consisted of a number of long wooden water basins covered with glass panes sloping
downward to both sides of the basins. Wooden troughs along the lower edges of the
glass panes collected the condensate and conducted it to the end of the basin, from
where it was conveyed to a storage tank. The total area of all of the water basins was
nearly 5000 m2 , and the total output was said to have been nearly 19 m3 /day. It is
thought that this still was kept operating for some 30 years, until a pipeline was con-
structed to bring water from the Andes mountains to Antofogasta.
Very little use was made of the solar distillation principle until World War II, when
it was applied to the production of fresh water for use on the life rafts provided for
wartime emergency use by survivors of shipwrecks and downed airplanes. As described
by Telkes [4], the device consisted of an inflated transparent plastic envelope that would
Distillation of Sea Water 207
float in the sea, and which contained a black absorbent pad saturated with sea water.
A cross section of the device is shown in Fig. 33.1. The sun shining through the upper
part of the plastic envelope heated the absorbent pad and evaporated part of the sea
water from it. Part of the vapor so formed condensed on the inner top surface of the
envelope and was collected in the cup at the bottom. When the cup was full, the device
was lifted out of the sea and the fresh water drained out of it. The absorbent pad had
an area of about 0.2 m2 , and the device would produce nearly a liter of water on a clear
day. Work on these devices practically ceased when World War II came to an end in
1945.
In the early 1950s drought conditions in many parts of the world led to the serious
consideration of desalination of sea water and other saline water sources as a means of
augmenting dwindling fresh-water supplies. Many different distillation schemes were
tried, and one of these was solar distillation. Solar stills of small size were constructed in
Algeria, on some of the Caribbean islands, and on some of the Mediterranean islands. In
the United States the Department of the Interior and the State of California sponsored
work on solar distillers. At about the same time solar distillation efforts began in many
other countries, including Australia, Algeria, Italy, South Africa, Chile, Cyprus, Kenya,
the U.S.S.R., France, Senegal, India, Iran, Spain, Morocco, Egypt, Taiwan, Japan,
Tunisia, Greece, the Cape Verde Islands, Ethiopia, and Pakistan. Activities in Europe
were coordinated by the Organization for European Economic Cooperation. Research
and development efforts reached a peak in the middle 1960s, but decreased to a very low
level by 1970. It was generally concluded that solar distillation, using the devices
developed up to that time, produced fresh water at a cost greater than that of water
produced by desalination methods energized by fuels. However, the experimental work
done during the active period of the 1950s and 1960s did furnish an extensive body of
experimental data and an analytical basis for the design and prediction of performance
of simple solar stills.
TRANSPARENT
ENVELOPE
ABSORBENT
PAD
1.110111......111,100.011.11111.....1111.1111111111111.01111 1110.11111.111.0111111111111111111111/ .....110111111/10'
&-CONDENSATE
CUP
Figure 33.1 Floating solar still, designed for emergency use in the life rafts, schematic
only. After Ref. 1.
208 Applications of Solar Energy
Because the existing large solar stills are of the basin type, it is important to understand
their construction and the principle by which they operate. Figure 33.2 shows schemati-
cally one form of basin-type solar still. In this figure saline water stands at a shallow
depth, 5 cm or less, in an insulated basin lined with a waterproof black material, such as
polyethylene or butyl rubber. A sloping pane of glass, supported by an appropriate
frame, covers the basin and is sealed tightly to minimize vapor leakage. A distillate
trough runs along the lower edge of the glass pane to collect the distillate and carry it
out of the enclosure. As shown in Fig. 33.2, there is a tube for refilling the basin with
saline water, and a level-control pipe that also serves as a drain to carry away the brine
during flushing. Basin-type stills can be operated with saline water being fed continuous-
ly to the basin, or on a batch schedule, the amount of feed varying with the initial
salinity, and being about double the quantity of fresh water produced when sea water is
used. The ratio of the feed water to the water flushed away is controlled by the basin
water salinity, which is found to have a proportional relation to the rate of fresh water
produced. The whole assembly must be located in a shadowless area, and is oriented so
that the long axis of the still runs east and west and the sloping glass cover slants toward
the equator to ensure a maximum intake of solar radiation. For solar stills located in the
tropics, this orientation is less critical, because the sun passes back and forth across the
zenith during the seasons of the year.
TRANSPARENT
COVER FILLER
DISTILLATE BASIN
TUBE
LINER
TROUGH
7,
7 .v.V,
."-;'
INSULATION
LEVEL CONTROL
84 DRAIN
Figure 33.3 shows some of the geometrical configurations used for the various large
solar stills. The single-sloped design, used by McGill University at La Gonave, Haiti, is
shown in Fig. 33.3a. This still was constructed on a gentle hill sloping south, with its
several bays arranged like the steps of a stairway. The concrete curbs serve as glass
supports and also contain cast-in troughs for collecting both the distillate and rain water.
The butyl rubber basin liner, glued to the concrete side walls, was designed to terminate
a few centimeters above the water level. However, this led to some difficulties because of
base material subsidence, causing the liner to pull away from the sides and drop into the
basin water, thus reducing the useful evaporating area as well as causing some leakage of
basin water.
Figure 33.3b shows the double-sloped, symmetrical still used at Las Marinas, Spain.
This design was evolved from the deep basin type developed and tested at the Daytona
Beach (Florida) laboratory of the U.S. Office of Saline Water. This still has a continuous
basin 28 m wide and 31 m long, and is lined with heavy asphalt sheets hot-mopped
together to form a watertight basin. Precast concrete members are set on top of the liner
to support the glazing. The lower glass support member has a distillate collector consist-
ing of an aluminum or stainless steel trough set into a depression molded into the
concrete member. The saline water depth of this design was calculated to average about
10 cm over the entire basin area.
The double-sloped symmetrical design presently being used by the CSIRO in
Australia is shown in Fig. 33.3c. It consists of wide concrete curbs between relatively
narrow bays, with the glass cover resting on the edges of the curbs. Opposite glass panes
are hinged together along the ridge line with silicone sealant, thus eliminating the need
for a ridge support member. The butyl rubber basin liner extends across the bay and
distillate troughs and under the edges of the glass panes. The bottom is insulated, as is
the bottom of the distillate trough. These stills are designed for continuous flow of saline
water, and slope toward the discharge end. Cross dams at intervals of about 1 m divide
the length of the still into a series of pools, with the water depth in each pool averaging
a little over 3 cm. The first model of this still was designed with galvanized sheet-metal
side members and polyethylene basin liner, but both of these materials proved to be
less durable than originally expected and were replaced by the concrete curbs and butyl
rubber liner.
Figure 33.3d shows the double-sloped unsymmetrical still designed by Delyannis
for use on the Greek islands. The glass supports are formed from aluminum angles and
rest on concrete curbs. The distillate troughs are aluminum extrusions and the basin
liners are butyl rubber. The bays are arranged in pairs, with walkways between each pair.
The unsymmetrical design is intended to be more economical of glass than a symmetrical
unit would be, since the side away from the equator tends to pass far less of the direct
solar radiation than does the opposite side. No specific comparison of the two configura-
tions has been made, but the resuls of operational data do not seem to bear out
significant differences in performance.
A still further configuration, with plastic covers instead of glass, is shown in
Fig. 33.3e, a design used by McGill University at Petit St. Vincent Island in the Caribbean.
The cover of this still is Tedlar plastic film, 0.1 mm thick and mechanically deformed to
make it wettable. The edges of the cover are clamped against the concrete curbs on the
Distillation of Sea Water 211
0.66m
GLASS
BASIN
SAND
INSULATION
SAND
\ BASIN LINER
(b)
1.0m
LASS
INSULATION
SAND
( c)
I.5m
BASIN LINER ALUMINUM GLASS
SUPPORTS a
GLASS
AND
,c,-.08414,
• ..•-- LINER
I I Pao 044 47%/47.0.tkWele'.(W*9"47 Aftte.".
It)
Figure 33.3 Cross sections of large basin-type solar stills, schematic only. (a) Solar still
with single-sloped glass covers. (b) Solar still with symmetrical double-sloped glass covers
and continuous basin. (c) Solar still with symmetrical double-sloped glass covers and
basin divided into bays. (d) Solar still with unsymmetrical double-sloped glass covers
and divided basin. (e) Solar still with inflated plastic film cover and divided basin. (f)
Solar still with sawtooth glass cover.
sides, together with the butyl rubber basin liner, to form an airtight enclosure. Air in the
still enclosure is maintained at a pressure of about 6 mm of water to inflate the cover.
The distillate troughs are molded into the concrete curbs, as is also a trough to collect
rainwater. The brine depth is about 7.5 cm, and the basin is not insulated.
Although no large stills of the sawtooth design shown in Fig. 33.3f have as yet
been constructed, certain features of their design are worthy of consideration. Developed
at the University of California for use on the South Pacific Islands, this type of still has
212 Applications of Solar Energy
distillate troughs that run crosswise of the bays rather than lengthwise. Since the bays
are narrow and the distillate troughs are short, this design reduces the distance of distil-
late flow within the enclosure, actually averaging only half the width of the bay, and thus
exposes the distillate to reevaporation for considerably less time. Its modular construc-
tion lends itself to units of sizes varying from a "family size" of one or two modules to
much larger sizes of many modules.
In addition to these simple solar stills of the basin type, there are a few designs that,
although not having flat basins, are related to the basin-type still. Figure 33.4 shows two
of these designs, the first being the tilted tray type, which is sloped at an angle to receive
the direct solar radiation more nearly perpendicular to the absorbing surface than is
possible with the flat type. It consists of a series of steps having very narrow widths and
shallow depths of water, insulated in the rear and covered with a glass pane sloping with
the same angle as the tilted tray. In a sense, this is a series of basin-type stills arranged
GLASS COVER
BLACKENED TRAY
DISTBIR
INSULATION
(a)
WICK
SURFACE ----- , ..-
GLASS
COVER101111
11,
BRINE
DRAIN ALUMINUM PAN
DISTILL
OUTLET
(b)
Figure 33.4 Cross sections of tilted solar stills, schematic only. (a) Tilted-tray solar still.
(b) Tilted-wick solar still.
Distillation of Sea Water 213
in cascade. In operating, saline water is introduced at the top step and cascades down
the steps to the drain at the bottom. The distillate is collected in a trough at the bottom
of the glass. In spite of the good performance shown in tests, the added costs and compli-
cations seem to outweigh the slightly increased rate of production.
The second unit shown in Fig. 33.4, is the tilted wick still, introduced by Dr. Maria
Telkes during the 1960s. This unit substitutes a porous cloth in place of the trays of the
previous example, thereby reducing the amount of saline water in the still to a minimum.
Because of this reduction, the water in this type of still should heat up rapidly in the
morning and achieve a higher temperature than in either the flat basin or the tilted tray
type. On a laboratory scale this tilted-wick unit has been found quite effective, but
when tried on a larger scale it proved to have difficulties of construction and operation
that outweighed its advantages. Thus, because of complications, neither of these two
latter types of stills has as yet been constructed in a large-scale installation.
These several examples suggest many of the configurations that have been tried. It
should not be thought that these represent all of the configurations that have been con-
structed and tested. Reference to the extensive literature, summarized by Talbert et al.
[1], will furnish many more examples of solar still configurations and construction
features.
B. Performance Data
Much study and research has been done to determine the performance rating and efficien-
cy of the several types of solar stills. Although there are many ways of expressing the
effectiveness of performance, the preferred method seems to be to plot the amount of
fresh water produced per unit of basin area in one day versus the solar radiation intensity
over the same period. Such curves for several solar stills are shown in Fig. 33.5. Also
plotted on this figure are radial lines of constant efficiency, the latter being defined as
w Oh
n (33.1)
H
where
w = weight of distillate per square meter per day
Oh = enthalpy change from cold water to vapor
H = solar radiation intensity per square meter per day
It would seem that the above equation is exceedingly simple, but it becomes increasingly
obvious from the available literature (see Talbert et al. [1] ), that there is a wide range of
definitions for the value of w and H in this equation. For instance, in calculating either
the value for unit w or H per square meter, the area represented in square meters could
conceivably be (a) the water surface in the basin, (b) the area of the condensing glass
cover, or (c) the area of the base of the still. The one chosen for most performance data
is the area of the water surface.
Furthermore, the value for 61, which includes the latent heat of vaporization,
presents another ambiguity. Since the value of latent heat varies appreciably because of
214 Applications of Solar Energy
3
1!
—E
2
>-
i-
>
ir-
S
o
0 5 10 15 20 25 30 35
DAILY TOTAL IRRADIATION , mi•day X 106
(a)
200 I I I 1 1 I 1.
:1
••{ •
: ••1' ••••
• •: '•••• .
100
PRO DUCTIV IT Y ,
•. •• ... •• ;.—
•
50
900 • •
I •
0 II
10 30 50 70
SOLAR ALTITUDE AT NOON - DEGREES
(b)
Figure 33.5 Performance curves of basin-type glass-covered solar stills. (a) Clear-day
performance curves. (b) Effect of weather on solar still performance.
the range of water temperatures during the day's operation, the question is what value to
choose, an average value of water temperature or a series of instantaneous values related
to corresponding weights of water produced during the instantaneous short periods. As
of this date, the preference seems to be for average values, the most common encountered
being 594.5 kcal/kg (1070 Btu/lb).
In examining the curves of Fig. 33.5, it will be noted that all of the curves slope
upward to the right and, if extended downward, would cross the horizontal axis at the
various points to the right of the origin. This appears to indicate that there is some value
Distillation of Sea Water 215
of solar radiation below which there would be no production; that is, the radiation
absorbed would just equal the losses from the structure. Further, it is seen that the
efficiency of the operation increases with the radiation intensity. This is consistent with
the fact that as the partial pressure of the water vapor increases with the temperature, the
air pressure decreases, so that the heat loss by convection of air within the still enclosure
is reduced, and consequently the efficiency is increased.
It is also noted that where the curves of constant efficiency cross the performance
curves of all of the solar stills represented, the efficiencies are less than about 45 percent
of the upper values of intensity, and average less than 35 percent over the range of radia-
tion intensities likely to be encountered. Calculations will be presented later that will
explain these low values of efficiency, but it should be pointed out here that these curves
indicate the fact that instantaneous efficiencies increase from zero at sunrise to a
maximum around noon and decrease in the later afternoon. However, problems of
distillate collection are such as to prevent the accurate determination of instantaneous
efficiencies, so that the above qualitative conclusion cannot be numerically demonstrated
by test.
The curves shown in Fig. 33.5a give the averages of clear-day performance, but no
indication of the effect of weather. However, the effect of local weather is appreciable,
as is illustrated by the performance data shown in Fig. 33.5b. These data were accumu-
lated over the period of a year for a solar still with a basin area 37 m2 , located at
Richmond, California, and are plotted using the solar altitude at noon as the horizontal
coordinate in order to emphasize the weather factor. The use of this coordinate is
roughly equivalent to plotting the production rate versus the day of the year. For this
reason, the point representing the production on a cloudy day in summer will be far
separated from the point for the production measured on a clear day in winter, even
though the total daily radiation measured on the two days may have been the same.
Had the data been plotted using measured daily solar radiation as the horizontal coordi-
nate, as in Fig. 33.5a, many of the scatter points of Fig. 33.5b would lie much closer to
a clear-day curve, such as that in Fig. 33.5a. Thus, using measured solar radiation as the
horizontal coordinate tends to mask the effects of poor weather on production.
In a related experiment, Morse and Read [5] studied the actual output of a 418-m2
solar still in Western Australia for the period March through December. By plotting all of
the data, using the radiation as coordinate, they were able to draw separate regression
lines for the production rates of the various months. This, they concluded, seemed to
indicate that for a given solar radiation intensity there could be several different produc-
tion rates due to differences in the monthly weather conditions. Informative data of this
type are useful for the design of a particular still in a particular location, but are not
applicable to other locations having a different weather regime. One must be fully aware
that the effect of local weather conditions, in addition to radiation intensity, is indeed
appreciable and must be taken into account when sizing a solar still for a given location.
(l.-ay- r)
( I -aw )1-I
qr \qb\qe\
1...4111..
qb
iqb
(a)
a.
(I- a - .01
9 WI
clwr I
(I-d,„v )ri
q
b
lb)
Figure 33.6 Heat flow relationships in basin-type solar stills. (a) Vectors showing
direction of heat flow. (b) Sankey diagram showing heat flow quantities. Adapted
from Ref. 26.
heat quantities involved, the upper diagram indicating the direction of heat flows and the
lower diagram showing the interrelations between the heat flows at some instant in time.
Heat quantities associated with the inflow of makeup water and with the outflow of brine
and distillate have not been included because these are negligibly small.
Distillation of Sea Water 217
where
I = solar radiation intensity on a horizontal surface
ag = absorptivity of the transparent cover material
Table 33.2 Percent Energy Distributions in Basin-Type Solar Stills
Electrically
Heated Basins Solar-Heated Basins
Investigator Telkes Baum L6f Bloemer Grune Datta Morse Lawand Cooper
Reference 27 28 21 29 30 31 32, 33 34 35 5, 36 37 38 Ave.
Internal
Evaporation 68 69 61 34 47 40.5 32 41.2 29.6 32.5 35.6 26.9 40
Radiation, brine to glass 10 8 13 10 23 16.9 25 12.2 15.2 18.3 12.6 15
Convection, brine to glass 4 2 2 5 11 8.4 6 7.5 5.9 4.1 5
Brine preheating 3.2 1.6
Heating of glass 10 9.9 5
External
Reflected from glass 8 15 17 12 14.2 43.6 8.1 6.5 22.0 10
Radiated by glass 1 (60.8) 35.1 23.6
Convected by glass
Edge losses 6 6 3.5 2 6.3 2.4 22.4 11.8 6.3 5
Ground losses 7 5.1 5
Input energy, J/cm2 • d (h) (340) (340) (293) 2260 2143 2629 1508 2264 2348 2896 2527 2551 2268
where
qe = heat transferred to cover from water by evaporation per unit area of water
basin
qr = heat transferred to cover from water by radiation per unit area of water
basin
qc = heat transferred to cover from water by convection per unit area of water
basin
These three equations have been written for heat quantities per unit area and assume that
the basin area equals the glass area. The short-wave radiation reflected out of the still,
(1 - ag - T) + (1 - aw ) T, does not appear in the above equations because it does not have
any effect on either the evaporation or the condensation of vapor. An examination of
Fig. 33.6 will support the validity of this statement.
Heat quantities in the above equations can be calculated from heat transfer princi-
ples, for example, those introduced by Dunkle [6] and modified by Morse and Read [5].
Their selections of relationships are given in the following paragraphs.
where
Fe = configuration factor for the geometrical arrangement
a = Stefan-Boltzmann constant
Tg= temperature of the inner glass surface
In this equation the factor Fe includes the effect of geometry as well as of emissivity of
the cover and water surfaces. The idealized case of radiation between two parallel planes
can be used for the basin-type configurations insofar as geometrical effects are concerned.
220 Applications of Solar Energy
For these conditions the factor Fe can be assumed equal to the emissivity of the water
surface (Boelter et al. [7] , p. 684), which is approximately 0.9 for water in the tempera-
ture range typical of the solar still.
1/3
qc = 8.84 X 10-4 [(Tw - T ) + pw - Pwg Tw 1 (Tw - Tg) (33.6)
g 265pT - pw
= he (Tw - Tg)
where
pT = total pressure in still enclosure
pw = saturation pressure of water vapor at water temperature
pwg = saturation pressure of water vapor at cover temperature
he X 3 p2 g13 AT . auCpg] n
= constant [X (33.7)
P2 k
where
he = heat transfer coefficient
X = distance from water surface to transparent cover
k = thermal conductivity of fluid
p = density of fluid
g = acceleration of gravity
13 = thermal expansion coefficient of fluid
µ = viscosity of fluid
Cp = specific heat of fluid
For the conditions of the temperature and humidity obtaining in a solar still, and for
values of X larger than about 7.5 cm, the constants have been evaluated for the case of
parallel, horizontal plates, with the following result:
11/3
[X3p2 g13 AT PCp g
he = 0.075 (33.8)
X p2 k
It is noted that the term X cancels out of this equation, implying that the water-to-cover
distance does not affect the convective heat flow. The quantity i3 AT, buoyant force,
varies along the path of air circulation, because of the changes in humidity, and has been
evaluated in terms of partial pressures and molecular weights of the water vapor and dry
air by Boelter et al. [7] . The end result, when numerical values for properties are substi-
tuted in Eq. (33.8) is shown in Eq. (33.6).
Distillation of Sea Water 221
Pw Pwg
T
qe = (16.276 X 10 -3 ) qe Tg hfg (33.9)
Tµ
where
h fg = latent heat of vaporization
This equation is based on the analogy between heat transfer and mass transfer.
Heat loss through the base and periphery of the still, qb:
where
Ub = overall heat transfer coefficient
Ta = ambient temperature
This loss has been the subject of considerable differences of opinion. Formulated by
Morse and Read [5] , it has been criticized as giving values unreasonably large. In their
published calculations Morse and Read assumed the value of Ub to be 5.68 W/m2 • °C or
1 Btu/(ft2 • h • °F). For the conditions of their specific example, they calculated by
means of the above equation that the total loss through the base and the periphery was
6493 kJ/m2 • d or 572 Btu/(ft2 • d), which was nearly 79 percent as much as the heat
transferred by evaporation/condensation, that is, the heat used for the production of
fresh water. LA [8] has suggested that the total daily bottom and edge loss from solar
stills placed on the ground is not more than 567 kJ/m2 • d or 50 Btu/(ft2 • d). However,
careful experiments by Cooper [9] place the loss through the base of the still between
these two values. His experiments were conducted with heat meters under the still, and
showed the base loss to be approximately 8 percent of the incident solar radiation. This
would be roughly 30 to 50 percent of the values given by using the heat transfer coeffi-
cient recommended by Morse and Read. It is therefore suggested that the value of Ub be
reduced to 1.7 or 2.8 W/m2 • °C, the larger value being used for narrow and separated
bays and the smaller one for continuous basin plants, or those with no ground exposed
between bays.
The foregoing equations must be satisfied simultaneously, but can be solved only
by trial-and-error methods. Morse and Read [5] presented the type of chart shown in
Fig. 33.7 for reducing the labor involved in calculating the daily distillate production
rate for given solar and ambient conditions. The general scheme is to replace the differ-
ential in Eq. (33.2) (d Tw /d0), by A Tw ,, the hourly change in basin water temperature.
Hourly calculations of AT, are then made, beginning at sunrise with known values of
Ta , Tw ,, and the several constants. If daily values for solar radiation are known, but not
the hourly values, it will be necessary to estimate the hourly quantities. This can be
done by using either the scheme of Whillier [10] or by assuming the solar radiation to
vary sinusoidally from sunrise to sunset. In addition, it is necessary to assign values to
properties of the glazing material and the basin liner.
222
HE ATFLUXQUANTITIES , W/m2
Applications of Solar Energy
Having selected the values noted above, the first step is to determine the value of
Tg for the first hour of operation, a value of Tg for which the quantities of Eq. (33.3) are
in balance. This equation is rearranged in the form
Values of the quantities (qga - qe), (qr + qe) are read from the chart for the assumed Tg ,
and the value of a I determined. If these values satisfy Eq. (33.3a), then the value of
A Tw is computed from Eq. (33.2). This value is added to the initial value of Tw to
obtain the starting value for the next hour. If the equality of Eq. (33.3a) is not satisfied,
a second assumed value for Tgis used to repeat the process described. Successive trials
must be continued until the equality of Eq. (33.3a) is satisfied. The process is repeated
for each hour of the day, ending with the hour before sunrise. The value of Tw at the
end of this hour should be the same as that at the beginning of the calculation. If it is
not, the calculations could be modified by varying one or more of the constants assumed
in the beginning.
The above scheme of calculation involves the use of a slide rule or its equivalent for
calculation. If a programmable digital computer is available, the several equations can be
linearized and the accuracy of the calculation increased. Quantities such as the vapor
Distillation of Sea Water 223
pressure of water and the latent heat of water can be expressed as linear functions of the
water temperature throughout the relatively narrow ranges of temperature involved in
solar distillation. The procedure followed in this case is to use the same scheme noted
above for manual calculation. The program is set up to stop at the point where the
equality of the two sides of Eq. (33.3a) can be determined, repeating the first part of the
program until the equality is satisfied. The balance of the program is then run to give
the A Tw for the hour. The totals for the day are used to determine the quantity of
distillate produced.
Whereas the above notes apply to the prediction of solar still performance for a
single set of conditions, the digital computer can evaluate changes in various quantities.
For this purpose, Cooper [9] set up a mathematical model of a solar still and calculated
the daily performance under various assumed conditions of wind speed, ambient tempera-
ture, and other variables. The still was assumed to be an interior bay of a large installa-
tion, to have symmetrical glass covers, and to have a bay width of about 1 m, as in the
CSIRO design.
The equations used included heat balances on the glass cover and on the basin
water, together with heat flow equations similar to those already presented. Ambient
temperature was assumed to vary sinusoidally with time, and wind speed was assumed
to vary linearly from a maximum during the middle of the day to zero before midnight.
Solar radiation intensity was assumed to vary sinusoidally, the equation used being that
suggested by Close [11] , which is
irH irB
I— sin (33.11)
7.20s Os
where
I = hourly radiation intensity on a horizontal plane
H = total daily radiation
Os = time in hours from sunrise to sunset
0 = hours after sunrise
The heat loss through the base was computed using a finite difference scheme, with the
space beneath the still divided into 24 segments. The variables studied included wind
speed, basin water depth, cover slope, insulation of the base, use of double glass covers,
and the effect of weather type. Twenty-four computer runs were made and the results
presented in tabular and graphical forms.
The conclusions reached on the basis of the above calculations included the
following:
1. With no insulation beneath the still, the depth of water has little effect on the
daily productivity. With 7.5 cm of insulation, the productivity at 1.25-cm
basin water depth was found to be 30 percent greater than that at 30.5-cm
water depth.
2. Wind has little effect on productivity, the increase in cooling of the cover being
offset by the lowering of the basin water temperature. Low wind speeds could
224 Applications of Solar Energy
100
o 80
w
cr
ce
260
I
>-
o 40
cc
w
z
w
$ 20
0
o 2D 40 60 80
BASIN WATER TEMPERATURE, ' C
Figure 33.8 Effect of water temperature on internal heat flows in a basin-type solar
still. Adapted from Ref. 9.
Distillation of Sea Water 225
Tw •81•C
6 2
H•28.72X10 J/m .day
3 INSULATED BASIN
"0.6 UNINSULATED BASIN
59'
0.4 • 19.18 X106
48'
0
8 0.2— 37.
a.
a.
0 1
O0 5 10 15 20 10 20 30
WATER TO COVER TEMPERATURE DIFF.,*C DEPTH OF BASIN WATER , cm
(a) (b)
0.6
I.-
F 0.4
o 0.2
cc
a.
0 5 10 15 20 0
0 5 10 15
WATER TO COVER TEMPERATURE DIFF. , °C WATER 70 COVER TEMPERATURE , °C
(c) d
Figure 33.9 Basin-type solar still performance parameters. (a) Effect of basin water
temperature on productivity. (b) Effect of basin water depth on productivity. (c) Effect
of extreme cover slopes on productivity. (d) Effect on productivity of water-to-cover
distance. Redrawn from Ref. 1.
surface. Precautions were taken to minimize boundary effects. The still was operated
both on a steady-state basis, and also with the electrical heat input programmed to
simulate the diurnal variation of solar radiation.
The results of Bloemer's work has been presented in a group of charts, some of
which are reproduced in Fig. 33.9. Chart (a) shows the effect of brine temperature on
productivity, and emphasizes the importance of maximizing this variable. Chart (b)
shows the effect of basin water depth, and also that of bottom insulation. Chart (c)
shows data for two cover slopes, 10 and 45°. It is seen that there is no discernible dif-
ference in performance at these two extreme cover slopes. Chart (d) shows the effect
of brine-to-cover spacing for values of 15 and 40 cm. Again this showed no discernible
difference, thus indicating that the designer can be free to use whatever spacing is most
economical of materials. The experiments on floating wicks showed no gain in produc-
tivity as compared with free brine surface. Experiments with internal convection also
226 Applications of Solar Energy
showed no gain. Thus, the previous conclusions, based on the model calculations by
Cooper, are apparently borne out by experiment.
C. Materials
Equally important as the designing of an efficiently producing solar still is the need to
be aware of the effective materials that may be used in a still's construction. Since
amortization costs for materials account for more than 70 percent of the cost of the
water produced, it is essential to minimize these costs by using suitable and durable
materials.
Results of the experimental work of the past 20 years have provided much useful
information on materials for solar stills. The transparent covers used so far have been
predominantly glass, although a few units have been built with plastic film covers. Glass
has the primary advantages in that it is rigid and wettable by water, and it has high trans-
missivity for solar radiation and low transmissivilty for long-wave thermal radiation. Its
chief disadvantages are high cost, great weight, and vulnerability to breakage.
Although plastic films have been much improved over the past several years, none
as yet has been found to be entirely satisfactory for use in solar stills. Many of the film
materials have been subject to ultraviolet degradation, most of them transmit an appreci-
able fraction of long-wave thermal radiation, and most of them must be treated frequent-
ly to retain their wettable property. In order to determine their durability, Irwin and
Fischer [13] made a series of exposure tests of low-cost plastics, including cellulose
acetate, vinyl, polyvinyl chloride, and polyethylene, and concluded that the design
useful life period for these materials would not be more than 1 year. Tedlar PVF film,
with ultraviolet stabilizer, has been used in some of the inflated stills, but it, too, has a
durability limit of only about 1 year. Recently, semirigid plastic Fiberglass sheets have
been manufactured for use in solar heating and cooling installations and may be suitable
for use as solar still covers. These must be surface treated to render them wettable. The
unit area cost of these sheets is about one-third that of glass. Some of these are in the
process of being tested on small solar stills at the University of California, and are show-
ing productivities equal to that of glass-covered units.
The glass support structure and basin-forming structures have in the past usually
been built of wood, galvanized iron, and concrete in various designs. For permanent
installations the use of concrete, either in precast elements or on-site poured foundations,
is the most satisfactory. For short-term temporary service, plywood has been used with
some success. Corrosion-resistant aluminum alloy tee bars and angles have been used
extensively for upper glass supports, even though the cost is appreciable.
Distillate troughs must be highly resistant to corrosion because of the corrosive
properties of distilled water. Designs in the past have favored either stainless steel troughs
or troughs cast into the concrete base and lined with a plastic basin liner. Aluminum or
galvanized steel troughs exposed to distilled water are soon corroded, so should not be
used.
Basin liners of butyl rubber sheet are presently favored as compared with black
polyethylene. However, there has been some use of black polyethylene in the form of
double sheets with plastic mesh reinforcing between them. These liners cost much less
Distillation of Sea Water 227
than butyl rubber, but have a shorter useful life than the latter. Polyethylene tends to
break down under high temperatures, which can result from failure to keep the still basin
full, or from failure of the adhesive in scattered spots in the double sheet material, which
can cause the upper sheet to bulge above the water level. Asphalt sheet material, such as
used in stock tanks, has been used in large continuous basin solar stills, and is satisfactory
for that service. It has not been used in stills with separate narrow bays.
The limitations of basin typesolar stills have stimulated much experimentation directed
toward discovering possible schemes to reuse the solar heat. Both multiple-effect and
multiple-stage flash distillation have been the basis of much research. Several of the
schemes investigated have used the humidification/dehumidification process in order to
avoid including vacuum pumps in the system. In these systems, in which the total
pressure is maintained at atmospheric, the vapor pressure decreases along the path of
saline water flow, and simultaneously the partial pressure of the air component increases.
The vapor pressure at any point in the saline water stream is the saturation pressure
corresponding to the saline water temperature and salinity.
A. Multiple-Effect Distillation
In using the multiple-effect principle, the temperature interval between the maximum
and minimum values of the hot water and the cooling water supplies is divided into
steps, and evaporation and condensation are caused to occur at each step. The heat
energy released by condensation in each step furnishes the heat for evaporation in the
next lower temperature step. Thus the heat that causes evaporation in the maximum
temperature step is passed along by condensation to the next step, and so on to the
last step, from which the remaining heat is discharged to the condenser cooling water.
Although no unit of this type has become commercial, a review of the several experi-
mental developments may suggest ideas for further improvement.
Experimental Developments
From the literature on this subject it appears that all of the experimenters used a
series of parallel flat plates, spaced relatively close together, standing either vertical or
sloped to receive the maximum solar radiation on the first plate. One example, used by
Dunkle [6] , consists of five vertical plates, the first of which is heated by solar-heated
water flowing through tubes attached to the front side of the plate. This heat causes
evaporation of a thin film of saline water falling down the inner side of the plate. This
vapor is condensed on the front side of the second plate, and the heat from its conden-
sation causes evaporation from the thin film of saline water falling down its opposite side
(back of second plate). This operation is repeated in the five effects, the heat being
finally discharged into the condenser cooling water and the incoming cold saline water.
The condensate drips down the front side of each of the plates into the collection troughs
for the distillate.
228 Applications of Solar Energy
Experiments on multiple effect solar stills have been reported by Dunkle [6] and
Telkes [14], and patents in the field have been secured by Ginnings [15] and Petrek [16].
Performance ratings have been reported by some of the experimenters. Dunkle reports
that his five-effect solar still, when controlled electrically to maintain a first-plate
temperature of about 72°C, showed a productivity of 1.1 1/h (2.4 lb/h). Telkes reports
experiments on a three-effect solar still with reflector, in which the average solar radia-
tion intensity on the front plate reached 18.16 X 106 J/m2 • h (1600 Btu/ft2 • h), and
which produced distillate in the amount of 1.61/m2 • h. A five-effect still, constructed
by Petrek, is reported to have produced 14.251/m2 • d, with a solar intensity of 19.07 X
106 jim2 d • (1680 Btu/ft2 • d). When these rates of distillate production are compared
with an output of about 3.5 l/m2 • d from a simple basin-type solar still, it seems promis-
ing that the use of multiple-effect configurations could increase the output per unit area
considerably.
A major problem in all of these experiments has been that of maintaining a uniform
wetted surface on the saline water side of each plate. Most of the experimenters have
resorted to using a porous cloth covering the entire surface of the metal plate and extend-
ing into a feed trough along the upper edge of the plate, as shown in Fig. 33.10a. The
SALINE
WATER INLET
GLASS
COVERS
WICKS
BLACK
ABSORBER
PLATE
SALINE
WATER
GLASS COVERS INLETS
BLACK ABSORBER
PLATE
BRINE
DRAINS
DISTILLATE OUTLETS
( b)
Figure 33.10 Cross sections of typical atmospheric multiple-effect solar stills. (a) Still
using wick feed for saline water. (b) Still using multiple tray for saline water. Adapted
from Ref. 1.
Distillation of Sea Water 229
feed water in the trough is then drawn onto the plate surface by capillarity. Excess
water moving down the porous surface drips into a collection basin along the lower edge
of the plate and is conducted out of the still. The condenser of the last effect can be
cooled by blowing air across it, or by circulating cold saline water through tubes attached
to the back of the plate.
Another scheme used by one experimenter is shown in Fig. 33.10b. Saline water
supplied to narrow trays on the back of the metal absorber plate is heated by conduction
along and through the plate. The figure shows a still with two such effects. The back
plate could be cooled by free convection, or by air blown across it by a fan.
Prediction of Performance
For the purposes of this chapter, the arguments presented by Dunkle [6] furnish
an adequate basis for the following comments concerning performance predictions of
the output of a multiple-effect still. One important aspect of performance is to deter-
mine values for the heat transfer between the parallel plates, which obviously can be
accomplished by radiation, convection, conduction, and diffusion of vapor. The effect
of air convection can be suppressed or eliminated by appropriately spacing the plates
closely together. The criterion by which this distance is determined depends on relation-
ships expressed in Grashof's number. Since suppression does not occur when that
number is larger than 2 X 103 , it is important to design the still so as to ensure a low
Grashof number. The Grashof number for humid air at 65.5°C can be expressed as
where
X = plate spacing in meters
AT = differential temperature between plates
By solving the above equation for X at a AT of 11.1°C (20°F), it is found that convection
will be suppressed for spacing less than 1.33 cm, leaving only radiation, conduction, and
particularly vapor diffusion as the means for transmitting heat from one plate to the next.
One way to evaluate the amount of heat carried by diffusion is based on the follow-
ing diffusion equation:
DMw PT X dpw
(33.13)
ww - I RT x PT - Pw dx
where
Ww = rate of diffusion
The value of the diffusion coefficient, D, can be computed from Sherwood's equation:
T3'1 1 )1'2
D = 0.0043 + — (33.14)
PT (\Cl/3 + VB113 )2 (Mw MB
230 Applications of Solar Energy
where
V = molecular volume
M = molecular weight
T"2 M w 1+ 1 )x 1 dp
Ww = -0.0043 w (33.15)
R (Vw "3 + VB "3 )2 ( Mw MB PT - pw dx
From this equation it can be deduced that the rate of diffusion will be largest when the
molecular volume of air, VB , and the molecular weight of air, MB , are small. Thus, if air
could be replaced by a lighter gas, such as hydrogen, the rate of diffusion would be
increased appreciably. Furthermore, since the equation includes the pressure gradient
dpw /dx any decreasing in the plate spacing (Ax), would increase the pressure gradient and
thereby increase the rate of diffusion. It is also seen from the equation that by appreci-
ably reducing total pressure, PT (in this case atmospheric pressure), the rate of diffusion
is increased. Normally, since the operation of the still is based on atmospheric pressure,
to reduce pi, below atmospheric pressure would be difficult from a practical point of
view. However, when one compares atmospheric pressures at sea level and the relatively
lower pressures in mountain areas, diffusion rates will understandably be found to be
greater in the latter areas.
In an effort to explore further the partial pressure relationships by which multiple-
effect stills operate, constants appropriate to an air-water vapor mixture are inserted in
the equation:
Ppw Ps - Pwrn
Ww = (I)w X (33.16)
Ax X PT - Pw m
where
Ps is standard atmospheric pressure.
This diffusion constant, Ow , for the air-water vapor mixture is a function of temperature,
but only for the given air-water vapor combination.
The heat transferred by evaporation, qe , can be written as
qe = Ww hfg (33.17)
where
hfg = latent heat
k (T i - T2)
qcond - Ax (33.18)
where subscripts 1 and 2 relate to the surfaces of the two plates, and Ax is the plate
spacing.
The heat transferred by radiation is determined by the usual radiation equation:
Distillation of Sea Water 231
qr = Fc u (Ti 4 - T2 4 ) (33.19)
In estimating the heat flow through a group of plates in this type of plant, it is necessary
to allow for the temperature drop through each plate, and the water films on both its
faces. This can be estimated using a unit thermal conductance for each plate equal to
1533 W/m2 • °C, which assumes the plates to be made of thin copper. Using the above
scheme, Dunkle [6] calculated the output of his five-effect solar still to be 14.0 kg/m2 • h
when the temperature range was 48.9 to 37.8°C.
Most of the large modern steam-operated distillation plants are of the multiple-stage flash
type. It was therefore to be expected that an atmospheric solar unit would be con-
structed using this principle. In the early 1960s work at the University of Arizona led to
the installation of a 5000-gallon/d unit at Puerto Penasco in Mexico on the shore of the
Gulf of California. This plant had the general characteristics of a multiple-stage flash
distiller, although its designers usually referred to it as a "multiple-effect humidification
plant," because it had the capability of multiplying the effect of the energy used.
Plant Description
As described by Hodges et al. [17] and shown in Fig. 33.11, this plant consisted
of three sections: the heat collector section, the water storage section, and the evaporator-
condenser section. The solar heat collectors were shallow, insulated water channels lined
with black plastic and covered with two layers of transparent film. The lower plastic
film floated on the surface of the water to minimize evaporation, and a small air pressure
between the two films maintained a separation between them so that the upper film
could serve its intended purpose of reducing heat loss. As indicated in Fig. 33.11 two
SOLAR
COLLECTO
BRINE
DRAIN
DISTILLATE
OUTLET
SALINE WATER INLET
insulated underground water tanks were used, one to store the preheated sea water
coming from the condenser and the other to store the sea water after further heating in
the solar collectors. The evaporator-condenser section consisted of two vertical columns
connected by cross-tubes at several elevations. Hot sea water from the storage tank was
pumped into the top of the evaporator column and cascaded by gravity over the packing
in the column, before it was pumped out of the bottom to waste. Air was pumped up
through this column countercurrentiy to the cascading water, and was humidified by
contact with the water surfaces on the packing. This rising humid air, increasing in
temperature as it came into contact with the hotter, higher water, passed through the
several cross-connections into the condenser column. The elevations of the cross-
connections represent the stages of the process, the temperature being higher at each
higher level. As the humidified air expanded over the finned-tube condenser coils,
through which was flowing the cold sea water, part of the humidity was condensed on the
tubing coils thereby heating the incoming sea water. Condensate was withdrawn from the
bottom of the column, and the down-flowing air was circulated back to the evaporator
column by a fan in the bottom cross-connection.
Power Requirement
One characteristic of multiple-stage flash distillers is that they require very large
flow rates of saline water. For example, steam-operated plants require the circulation
through the stages of from 10 to 20 times the product flow rate. This, in turn, requires
large-capacity water pumps and hence appreciable power for pumping. The atmospheric
solar plant described above also requires large flow rates of saline water, but in addition,
requires considerable air circulation to carry the vapor from evaporator to condenser. It
is noted that the Puerto Penasco installation had three water pumps driven by motors
rated at a total of 15.5 hp, and an air blower with a motor rated at 40 hp. For compari-
son, a motor-operated vapor-compression distillation plant, designed for the same output,
would require a motor rated at about 20 hp. Thus, the multiple-stage flash distiller using
Distillation of Sea Water 233
Limitation of the atmospheric-type solar distillers has led to several proposals for
applying solar energy to conventional distillers. Since the distillers that use fuels to
furnish heat for the evaporation process also use appreciable amounts of electrical energy
for pumping and for vacuum production, it should not be considered a detriment that the
substitution of solar energy for fuel energy must also be accompanied by the use of
electrical energy for the same purposes. When solar energy is considered in this manner,
the plant in effect would be a solar-assisted distiller, with the solar energy supplying the
heat required but also depending on appreciable amounts of electrical energy for other
purposes. This is in contrast to the more simple concept of using solar energy alone.
There have been several proposals advanced for such solar-assisted distillers. Brice
[18] and Weihe [19] have proposed adaptations of the multiple-stage flash (MSF)
process, which has been the popular conventional form of distiller. The proposals of
others, Eibling et al. [20] , Baum [21], and Tleimat and Howe [22, 23] , have been based
on various configurations of multiple-effect distillers. Because of the very large quantities
of water required in MSF plants and the pumps needed to control the pressures in the
successive stages, the electrical power needed in the multiple-stage flash process is much
greater than that used in the multiple-effect plant. The major differences in the multiple-
effect proposals relate to the type of solar collector required. Eibling and Baum pro-
posed the use of concentrating collectors, so that steam could be generated at a pressure
slightly above atmospheric and used in existing types of distillers. Tleimat and Howe [22]
proposed the use of flat-plate collectors designed to produce heated water at temperatures
around 60°C. The effective use of this temperature for distillation requires relatively
efficient distillers, and recent development of plants of high efficiency is the basis for
the proposals.
The efficiency of a distiller is often expressed as the ratio by weight of the distillate
produced to the steam used, and this ratio is generally referred to as the gained output
ratio (GOR). The GOR for a multiple-effect distiller plant is proportional to the number
of effects in series, and, for a first approximation, is usually taken as 0.8 times the
number of effects. The number of effects possible between the inlet steam temperature
and the cooling water temperature is determined by the temperature interval that must
be reserved for heat transfer in each effect. In a solar plant the extreme temperatures of
steam and cooling water are of the order of 65 and 20°C, respectively, and thus provide a
differential of about 45°C. If this differential is divided into equal increments, one for
each effect, the influence on the GOR of the temperature interval per effect is relatively
clear. For instance, the use of three effects would give a GOR of 2.4 and reserve a 15°C
temperature interval per effect, whereas the use of nine effects, with a GOR of 7.2, would
reserve only 5°C for each effect. If a plant could be designed to operate with as little as
2°C per effect, then 22 effects could be used and the GOR would increase to about 17.6.
234 Applications of Solar Energy
SCOOP
(STATION STEAM IN
CONDENSATE
OUT
ROTATING
DISKS
VAPOR OUT
( TO NEXT EFFECT)
As has been noted, distillers capable of operating with small temperature intervals are
now available, at least in development stages. For a more complete discussion of the
GOR, see Ref. 26.
Researchers at the University of California have made studies on two types of
distillers that seem to be appropriate for solar plants, the Vertical Tube Evaporator (VTE)
type using foamy flow techniques, developed by Sephton [24], and the rotating wiped
film type, developed by Tleimat [25] . In the past, VTE distiller plants have been
designed using heat transfer coefficients of the order of 5 kW/m2 • °C. The recent
developments of foam flow techniques in special vertical tubes increased heat transfer
coefficients to about 20 kW/m2 • °C, and were the basis for the first solar distiller plant
described by Howe and Tleimat [23] .
The second high-efficiency distiller plant, the rotating wiped film type, is capable
of transferring heat energy at even higher rates than noted above, with experiments
showing values between 20 and 50 kW/m2 • °C. Tleimat and Howe [22] proposed a
plant of this type, having 15 effects and capable of producing 37.9 m3 /d (10,000 gal/d)
of distilled water. A plant of this effectiveness should require solar collectors having an
area of only about 1/15 that of basin-type stills to provide the same amount of water.
Figure 33.12 shows a schematic cross section of one effect of the rotating wiped film
distiller. This device consists of a flat rotating cylinder, with the saline water evaporating
on the outside of the two ends, and steam condensing on the inside. Butyl rubber wipers,
held against the end plates on the outside of the cylinder, distribute the saline water in a
thin film over these plates. Steam, condensing on the inside of the cylinder, is thrown to
the outer periphery and collected by scoops, which discharge the condensate through the
hollow shaft. In the proposed distiller plant, 15 of these are placed in series, with the first
one receiving steam produced by the solar-heated water expanding through an orifice.
Solar collectors for a plant of this type could be of the same construction as those
developed for the heating/cooling of buildings, since these can heat water to about 65°C.
Distillation of Sea Water 235
The solar-heated water would be expanded through a small temperature range and pro-
duce steam by flashing, this steam being used to energize the first effect of the distiller.
The plant can be designed with hot-water storage capacity sufficiently great to render
operation of the distiller unit independent of that of the collectors.
The economics and other practical features of this type of plant can only be
evaluated fully when a prototype plant is constructed. Although the operation of both
the rotating evaporator and the solar collectors have each been studied experimentally,
their operation together could give rise to problems not predictable from separate tests.
NOMENCLATURE
General Subscripts
w water, water surface, water vapor
m mean
g glazing, glazing surface
T total
REFERENCES
2. G. Nebbia and G. Menozzi, "A Short History of Water Desalination," Acque Dolce
Dal Mare, IIa Inchiesta Internazionale (Proceedings International Symposium,
Milano, April 1966), pp. 129-172, 1967.
3. Josiah Harding, "Apparatus for Solar Distillation," Proceedings of Institution of
Civil Engineers 73, 284-288 (1883).
4. Maria Telkes, Solar Distiller for Life Rafts, U.S. Office of Science Report No. 5225,
P.B. 21120, June 19, 1945.
5. R. N. Morse and W. R. W. Read, "A Rational Basis for the Engineering Develop-
ment of a Solar Still," Solar Energy 12 (1), 5-17 (1968).
6. R. V. Dunkle, "Solar Water Distillation: The Roof Type Still and a Multiple Effect
Diffusion Still," International Developments in Heat Transfer, ASME, 895-902,
Part 5, August 28-Sept. 1, 1961 (also CSIRO, Victoria, Australia, publ. 108, 1961).
7. L. M. K. Boelter, V. H. Cherry, H. A. Johnson, and R. C. Martinelli, Heat Transfer
Notes, McGraw-Hill, New York, 1965.
8. George 0. G. Lof, "Letter to Editor (Regarding earlier paper by Morse and Read,
Solar Energy 12, 1968)," Solar Energy 12 (4) (1969).
9. P. I. Cooper, "The Maximum Efficiency of Single Effect Solar Stills," Solar Energy
15, 205-217 (1973).
10. A. F. Whillier, "Solar Radiation Graphs," Solar Energy 9 (3), 164 (1965).
11. D. J. Close, "A Design Approach for Solar Processes," Solar Energy 7, 112-122
(1967).
12. J. W. Bloemer, "Factors Affecting Solar-Still Performance," ASME Paper 65-WA/
SOL-1, November 1965.
13. J. R. Irwin and R. D. Fischer, The Use of Low-cost Plastic Films in Solar Stills,
Battelle Memorial Institute, Columbus Laboratories, Research Report, 1967.
14. Maria Telkes, Solar Still Construction, Office of Saline Water, U.S. Dept. of
Interior, Res. and Dev. Rep. No. 33, 1959.
15. D. C. Ginnings, "Multiple Effect Solar Still," U.S. Patent 2,445,350 (1948).
16. J. B. Petrek, "Solar Desalting Process Breakthrough," Solar Energy Digest 6, (6)
(1976).
17. C. N. Hodges, T. L. Thompson, J. E. Groh, and D. H. Frieling, Solar Distillation
Utilizing Multiple Effect Humidification, Office of Saline Water, U.S. Dept. of
Interior, Res. and Dev. Rep. No. 194, 1966.
18. D. B. Brice, "Saline Water Conversion by Flash Evaporation Utilizing Solar
Energy," Adv. Chem. Ser. 38, 99-116 (1963).
19. H. Weihe, "Fresh Water from Sea Water, Distilling by Solar Energy," Solar Energy
13, 439-444 (1972).
20. J. A. Eibling, R. E. Thomas, and B. A. Landry, An Investigation of Multiple-Effect
Evaporation of Saline Waters by Steam from Solar Radiation, Office of Saline
Water, U.S. Dept. of Interior, Res. and Dev. Rep. No. 2, 1953.
21. V. A. Baum, "Solar Distillers," UN Conference New Sources of Energy, Paper
35/S/119, Rome, August 1961.
22. B. W. Tleimat and E. D. Howe, "Solar Assisted Distillation of Sea Water," in
M. A. Kettani and J. E. Soussou, eds. Heliotechnique and Development, DAA Press,
Cambridge Press, Cambridge, Mass., 1976.
23. E. D. Howe and B. W. Tleimat, "Twenty Years of Work on Solar Distillation at the
University of California," Solar Energy 15, 97-105 (1974).
24. H. H. Sephton, "Interface Enhancement for Vertical Tube Evaporators: A Novel
Way of Substantially Augmenting Heat and Mass Transfer," ASME No. 72-HT-38,
presented at the ASME-AICHE Heat Transfer Conference, Tulsa, Oklahoma, 1971.
Distillation of Sea Water 237
25. B. W. Tleimat, "Rotating Disk Still with a Hydrodynamically Applied Thin Film,"
U.S. Patent No. 3,764,483, Filed Oct. 5, 1970.
26. E. D. Howe, Fundamentals of Water Desalination, Science and Technology Series,
Vol. 1, Marcel Dekker, New York, 1974.
27. Maria Telkes, "Distillation with Solar Energy," in F. Daniels and J. A. Duffie, eds.,
Solar Energy Research, University of Wisconsin Press, Madison, 1955.
28. Maria Telkes, "Solar Still Theory and New Research," Proceedings, Symposium on
Saline Water Conversion, Nat'l. Acad. of Sciences, Nat'l. Res. Council, Washington,
D.C., Publication No. 568, pp. 137-149, 1957.
29. G. 0. G. L6f, Solar Distillation of Sea Water in the Virgin Islands, Office of Saline
Water, U.S. Dept. of Interior, Report No. 5, PB 161380, 1955.
30. G. 0. G. Lof, "Design and Cost Factors of Large Basin-type Solar Stills," Proceed-
ings, Symposium on Saline Water Conversion, Nat'l. Acad. of Sciences, Nat'l. Res.
Council, Washington, D.C., Publication No. 568, pp. 157-174, 1957.
31. G. 0. G. L8f, J. A. Eibling, and J. W. Bloemer, "Energy Balances in Solar Distillers,"
AlChE Journal 7 (4), 641-649 (1961).
32. J. W. Bloemer, R. A. Collins, and J. A. Eibling, Study and Field Evaluation of Solar
Sea Water Stills (June 1960), OSW Report No. 50, PB 171934, September 1961.
33. J. W. Bloemer, R. A. Collins, and J. A. Eibling, "Field Evaluation of Solar Sea
Water Stills, in Saline Water Conversion, Advances in Chemistry, Series No. 28,
Amer. Chem. Soc., Washington, D.C., pp. 166-177, 1960.
34. W. N. Grune, R. A. Collins, R. B. Hughes, and T. L. Thompson, Development of
an Improved Solar Still, OSW Report No. 60, PB 181144, 1962.
35. R. I. Datta, S. D. Gomkale, S. Y. Almed, and D. S. Datar, "Evaporation of Sea
Water in Solar Stills and Its Development for Desalination," Proceedings of 1st
Internat'l. Symposium on Water Desalination, Vol. 2, pp. 193-199, Washington,
D.C., October 1965 (1967).
36. R. N Morse and W. R. W. Read, "The Development of a Solar Still for Australian
Conditions," Trans. Inst. Eng., Mech. and Chem. Eng., Australia MC 3 (1), 71-80
(May 1967).
37. T. A. Lawand, Engineering and Economic Evaluation of Solar Distillation for Small
Communities, Technical Report No. MT-6, Brace Research Institute, McGill
University, August 1968.
38. P. I. Cooper, "Digital Simulation of Transient Solar Still Processes," Solar Energy
12 (3), 313-331 (May 1969).
chapter 34
IRRIGATION PUMPING
34.1 INTRODUCTION
Agricultural production accounts for about 3 percent of total U.S. energy consumption.
About 1.3 quads of energy* were used directly in agricultural production in 1974, and an
additional 0.7 quad was used in the manufacture of fertilizers and pesticides, as shown in
Table 34.1 [1]. This energy breakdown covers only those functions directly contributing
to actual production and does not include energy consumed by farmers for residential or
nonbusiness transportation.
239
240 Applications of Solar Energy
We see from Table 34.1 that irrigation pumping accounts for about 20 percent
of overall direct agricultural energy use. To illustrate further the economic importance
of irrigation, the estimated energy cost per product value of irrigated crops is in the range
of 20 to 40 percent. Furthermore, the total domestic value of irrigated crops is a sub-
stantial $7 billion per year [2].
Solar energy offers several features that make its utilization for irrigation pumping
quite attractive. First, the greatest need for pumping occurs during the summer months
when solar radiation is greatest. Second, pumping can be intermittent to an extent.
During periods of low solar radiation when pumping decreases, evaporation losses from
crops are also low. Finally, relatively inexpensive pumped storage can be provided in the
form of ponds.
In the United States, about 42 percent of irrigation water comes from ground
sources, another 42 percent from irrigation organizations, and the remainder from surface
sources [3]. In 1974, irrigated crops in the 17 Western states (Arizona, California, Color-
ado, Idaho, Kansas, Montana, Nebraska, Nevada, New Mexico, North Dakota, Oklahoma,
Oregon, South Dakota, Texas, Utah, Washington, and Wyoming) accounted for more than
88 percent of the nation's irrigated croplands and for 94 percent of the irrigation energy
used. The disproportionate energy usage in the West arises from greater pumping depths
and higher water application rates.
The amount of irrigated cropland in the United States has grown steadily during
this century. Between 1935 and 1970, irrigated land more than tripled to 15.4 million
hectares (38 million acres). By 1976, irrigated lands in the United States were estimated
to cover 23 million hectares (57 million acres). Of this total, 20 million hectares (50
million acres) were in the 17 Western states [4] . Table 34.2 summarizes cultivated and
irrigated land by country and world region [5] .
United States croplands are irrigated by an estimated 300,000 to 500,000 irriga-
tion power units with power ratings up to about 225 kW (- 300 hp). Pumping depths
are steadily increasing. For the 17 Western states, the average pumping depth is now
58.5 m. The deepest pumping (yearly average) occurs in New Mexico, where irrigation
water is pumped from a depth of 109 m [6] .
As energy costs have risen and conventional fuels have become less available, the
search for alternative energy sources, such as solar energy, has received new impetus. A
number of recently constructed solar irrigation pump installations are now operational,
as discussed in Sec. 34.6. The major obstacle to increased use of solar irrigation systems
at this time is their relatively high capital cost.
If the costs of solar pump systems can be substantially reduced-and assuming
that conventional fuel costs continue to rise-solar irrigation could become economical,
and increased use of such systems might be anticipated in the future.
242 Applications of Solar Energy
Although the use of solar energy for various purposes dates back to many centuries
B.C., credit for the first use of solar energy for raising water should probably go to the
French engineer, Solomon de Caux (1576-1626). A. S. E. Ackermann [7] discusses
Caux's first solar work in 1615 and describes the subsequent achievements of several
engineers and inventors. Another Frenchman, August Mouchot, built several solar
machines. A solar-driven steam engine was used in 1875 to pump water. An American,
John Ericson (1803-1889) spent a good deal of his own money on solar engines, con-
cluding that solar steam is many times more costly than steam produced by burning
coal. He employed a modified solar hot air engine to be used as a small pump. An
Englishman, William Adams, started his work in 1876 and demonstrated a solar-driven
pump in Bombay, India. Abel Pifre, of France, used a parabolic reflector to power his
rotary pump to raise 991 of water 3 m in 14 min (1878). An American, A. G. Eneas,
built several solar plants using a truncated, cone-shaped reflector. One of his plants,
erected at Pasadena, California, in 1901, was used for pumping. It pumped at the rate
of 5.3 m3 /min (1400 gpm) for a 3.7-m lift, developing about 3.2 kW (— 4 hp), and was
the largest pump ever built to that time.
H.E. Willsie and John Boyle, Jr., were probably the first to use a two-fluid system.
In 1904, a complete 4.5-kW (6-hp) sun-power plant was built at St. Louis by the "Willsie
Figure 34.1 A view of the Shuman-Boys solar collector system constructed at Meadi,
Egypt, in 1913. Courtesy of Solar Energy Library, Arizona State University.
Irrigation Pumping 243
Sun Power Company" using an ammonia engine. Several plants were built subsequently.
A 15-kW (20-hp) plant, using sulfur dioxide, was built near Needles, California. The
engine operated a centrifugal pump to lift water from a 13.1-m well.
By far the best known of the early solar pump pioneers is the American, Frank
Shuman, who built several solar plants. By 1907 he had a 2.6-kW (3.5-hp) plant running,
and in 1911 developed a unit that generated about 24 kW (32 hp). His best-known plant
was the 1913 Shuman-Boys solar pump installation at Meadi, a suburb of Cairo, Egypt
(Fig. 34.1). As reported by Shuman [8] , the Meadi plant utilized parabolic collectors
with "five concentrations of sunshine." Five parabolic troughs were utilized, 4.06 m wide
at the top and 62.2 m long, that is, a total collector area of 1263 m2 . The boilers were
located at the focal line. The heat absorbers were placed about 7.6 m apart to avoid
mutual shading. The collectors were placed in a north-south orientation with an east-to-
west rotation. The mirrors were set in a light steel frame. The Meadi plant had a reported
average capability of 37.3 kW (50 hp). Shuman noted that the same plant located "1000
miles or so farther south should give about 65 hp." A. S. E. Ackermann, who was a con-
sulting engineer to Shuman, reported that the "best overall thermal efficiency" of the
Shuman plant in Egypt was only 4.32 percent [7] , largely because of the low steam
pressure developed. The project was abandoned in 1915.
With the advent of cheap fossil fuel energy, interest in solar water pumping de-
clined. In the 1950s, the SOMOR Corporation of Lecco, Italy, marketed a number of
small solar pumps with power ratings from fractions to a few kilowatts. Renewed interest
has now been stimulated, and several solar pump projects are under way (see Sec. 34.6).
Design requirements for current solar pump systems are heavily influenced by the fact
that these are usually demonstration units. For example, units may be designed on the
basis of peak power output per dollar invested rather than on the more rational basis of
maximum seasonal water delivery per dollar invested. Current solar pump systems are
often backed 100 percent by auxiliary power—either by a directly coupled electric
motor or by a separate motor-driven pump. Nevertheless, many general design con-
siderations for such demonstration units are believed to be relevant to future units of
more mature design. A discussion of the more significant design requirements is pre-
sented in this section.
A. Insolation Patterns
many U.S. cities [11] . The average daily amounts of solar radiation received in the
United States for each month of the year can be obtained from various publications
[12, 13] . These values can then be used to calculate the monthly and seasonal perfor-
mance of solar pumps. They should not be used to calculate the instantaneous, or design
peak, performance levels. For design peak calculations, the expected peak solar radiation
should be used.
In general, a farmer can choose from three sources of irrigation water: (a) on-farm under-
ground water, (b) on-farm surface water, and (c) irrigation organizations.
Table 34.3 shows the sources of irrigation water for farms in the 17 Western states
and Louisiana. The relative source distribution differs somewhat from the overall U.S.
distribution.
Water delivery is dependent, of course, on the type of farm crop being irrigated.
A representative seasonal water-delivery schedule for a cotton farm in Arizona is shown
in Fig. 34.2.
As noted earlier, the average lift for ground-water pumps is increasing yearly in
most heavily irrigated regions. This is a result of the lowering of the water table. Conven-
tional irrigation pumps are usually sized to provide the required water delivery by opera-
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Months of Year
ting 24 hours a day. However, unassisted solar-powered pumps, which operate for less
than 10 h/d for most of the season, must be oversized to provide the same daily total
water delivery.
C. Pump Curves
Three general pump configurations are used to supply irrigation water. For low-lift appli-
cations (< 6 m), single-stage propeller pumps are used. Single-stage mixed-flow pumps are
used for intermediate lifts (6 to 15 m), and multistage mixed-flow or turbine pumps are
used for greater lifts.
Figure 34.3 shows typical characteristics for a single-stage propeller pump for a
low-lift application. Performance characteristics for any one pump design can be
extended to match a range of system requirements by trimming or selecting the proper
impeller diameter. Pump suppliers typically provide performance data at selected speeds
for a range of several impeller diameters.
Head and water delivery for a typical low-lift application with a 3.7-m static lift
will fall along the system curve plotted on the pump performance data of Fig. 34.3.
Required pump power input at design flow can easily be obtained from such a plot.
Figures 34.4 and 34.5 show dimensionless performance characteristics of low-lift
and deep-well pumps, respectively, in which speed-dependent data are relaxed into two
plots, one for pump efficiency and another for dimensionless head. Whenever possible,
actual manufacturers' product literature should be used to select a pump to match a
particular pumping requirement. However, pump performance at speeds not covered by
246 Applications of Solar Energy
75 % efficiency
24 . 80 %
• • 83 %
20 --___ 35 hp
4rit 16
a)
12
I
Propeller pump
0
,N., °O Sao..
..o
'>2 J
1 .47i 1 I
1000 2000 3000 4000 5000 6000 7000 8000 9000 10,000
Capacity, gpm
Figure 34.3 Typical irrigation pump performance curves for a single-stage propeller
pump.
the product literature can be obtained from the dimensionless plots. In addition, the
dimensionless plots are useful in a computer simulation of a solar-powered irrigation
pump.
If values of shutoff head pressure and flow rate at the best efficiency are known at
one speed, performance values at any other speed can be obtained. For example, per-
1.0 •
•
•
0.9 — •
Efficiency
0.8
0.7
0.6
0.5
0.4
Head /Head s.o.
0.3 Value at 1170 rpm
Head sn. = 63.7 ft
Q2
Flow BE = 2630 gpm
HPso =165 hp
0.1
o
o 0.2 0.4 0.6 0.8 1.0 1.2
Flow/Flow B. E.
10
Head / Head so
Head/ Head so , Eff iciency (770, fraction
0.9
0.8
0.7
06
05
I
02 04 06 0.8 10 12 14 16 18
Flow/ Flow B E
formance at any speed (rpm) and water delivery (flow) can be calculated for a four-stage,
deep-well pump using Fig. 34.5 as follows:
Then, the dimensionless flow parameter, flow/flowB.E., can be computed and used to get
values for dimensionless head, head/heads0 , and pump efficiency, rip , from Fig. 34.5.
Pumping lift or head is
( head )
Head — X heads .0 .
heads .0
The minimum power input at the selected speed (rpm) for water to begin to flow is given
by HPs.o.
The dimensionless plots also provide a realistic basis for scaling pump performance
to other power levels by altering the design flow at best efficiency and/or the shutoff
head at design speed.
248 Applications of Solar Energy
H = lift, ft
For an electrically driven pump, the electrical power required in kilowatts to provide a
specified water flow rate and lift is
H X gpm (34.2)
kW —
5331 np 77mechnmot
where
Electrical power requirements are plotted in Fig. 34.6 from this equation, for typical
efficiency values.
A rule of thumb for electrical power consumption for irrigation water pumping is
1.6 kWh per acre-ft of water delivered per foot of lift from which the water was pumped.
The collector area required by a solar irrigation pump for a specified maximum
power output at noon on a design day can be obtained from relatively simple relation-
ships. The thermal efficiency of power output from a Rankine-cycle power system is
Tcond
771-c = 1 1 — )flerir (34.3)
Thigh
where
1 1
Assumed Efficiencies, percent
1000 Pump 82
Mechanical 94
900 Electric motor 80
800
Electrical Power Input, k W
700
600
500
400
300
200
100
o 8000
o 2000 4000 6000 10,000
Water Flow Rate, gpm
The ratio of collector area to maximum power output, Ac /HP (ft2 /hp), is given by
Ac 2545 (34.4)
HP nre Qsn,
where
These relationships are plotted in Figure 34.7 for ready reference. The use of this
plot is illustrated by the following example.
Thigh = 350°F
260
Engine efficiency =70 percent
240 Solar flux =300 Btu/hr-ft2 600
Collector efficiency =50 percent
220 Condenser temperature= 90 F
Rankine -cycle efficiency = 75
percent of Carnot efficiency
Co llector Area/ Power Output, ft / hp
200 500
2
180
HighTemperature, F
160 400
140
120 300
100
80 200
60
40 100
20
0 1 1 1 1 1 I I I I 1 I 1 0
0 2 4 6 8 10 12 14 16 18 20 22 24 26
Rankine-Cycle Efficiency, percent
rirc = 16.8%
A,
= 101.6 ft2 /hp
HP
Therefore, the collector area required is
The design of a solar power irrigation system involves a large number of design options.
Some of the more important of these are discussed below.
Irrigation Pumping 251
A. Power Systems
Heat Engines
Generally, heat engines are used to convert solar energy to mechanical energy to
power the solar pumps. Several basic types of power systems, or working cycles, can be
selected to convert the thermal energy from the solar collectors into useful shaft power.
Figure 34.8 shows schematics of three such power systems, together with their thermo-
dynamic cycles.
The Rankine cycle is the only one of these systems to use both the liquid and
vapor phases of a working fluid. The other cycles use only the vapor (or gas) phase. At
the temperature levels generally used in solar power systems, for example, 25 to 300°C
(80 to 600°F), the Rankine cycle is generally superior to the other cycles in terms of
overall cycle efficiency and component sizes. Therefore, it is presently the most widely
used in solar power systems.
As depicted in Fig. 34.8a, the working fluid in a Rankine cycle may be either a
"wetting-type" fluid or a "drying-type" fluid. This simply means that as vapor expan-
sion occurs through a turbine; for example, some fluids will expand into the "wet"
region under the saturated vapor line, whereas others will expand into the superheat,
or "dry" region. Expansion into the wet region should be avoided to prevent liquid
droplets from damaging the expansion device. Therefore, a wetting-type fluid should
be superheated before expansion, whereas a drying-type fluid may be expanded directly
from the saturated vapor line. However, as a drying-type fluid expands farther into the
superheat region, it often becomes necessary (and advantageous) to add a regenerator
between the turbine and the condenser, which recovers this superheat to preheat the
liquid going to the boiler. A more detailed description and sample calculations for a
typical solar-powered, Rankine-cycle system are given later.
The Brayton cycle (Fig. 34.8b) is generally used when higher working tempera-
tures can be achieved. This power system can be either a closed or open system. A closed
system can use working fluids other than air when desirable. Dual turbines can also be
used to improve the overall cycle efficiency, with the first turbine driving the compres-
sor, and the second (separate) turbine providing the output shaft power.
The Stirling cycle (Fig. 34.8c) can achieve high overall cycle efficiencies, but has
been limited to rather small power outputs in solar applications [141. Relatively large
heat exchanger areas are required, together with high gas temperatures and pressures.
Figure 34.9 is a schematic of the major components and fluid circuits of the 37-kW
(50-hp) solar-powered, Rankine-cycle irrigation pump designed and built by Battelle
(BMI) in a cooperative venture with the Northwestern Mutual Life Insurance Company
(NML) of Milwaukee, Wisconsin (see Sec. 34.6). This layout is fairly typical of Rankine-
cycle systems.
Referring to Fig. 34.9, the operation of a typical solar pump is as follows. In the
morning, whenever the solar flux reaches a high enough intensity to heat the water in the
collectors to a usable temperature, the concentrating collectors automatically begin
tracking the sun.
Simultaneously, a small electric motor operates two pumps: one to circulate
water through the collectors, boiler, and preheater, and another to pump the liquid
working fluid through the various heat exchangers. As soon as the turbine develops
252 Applications of Solar Energy
Temp Temp
Wout
4
Win
3
Qou
Wetting fluid Drying fluid
Qout Entropy Entropy
a Rankine Cycles
Qout Regenerator
b Brayton Cycles
Expansion space
Heater
Qin Wout
Regenerator a
Displacer
piston
Cooler Temp
Qout
Working Compression space
piston
Win Qout
Power
output
shaft
Entropy
c. Stirling Cycles
Figure 34.8 Selected power systems used for converting thermal energy to shaft power.
sufficient speed to maintain operation, the starting motor is disconnected and the unit
continues to operate on solar energy throughout the day.
At design conditions, the Rankine-cycle power unit operates as follows. The
working fluid, in a liquid condition, is pumped up to a relatively high pressure by the
feed pump, which, in turn, is driven by the engine. The liquid is then partially heated
Irrigation Pumping 253
Legend
- --Hot water system
-Freon® system
----Cooling water
Solar collector
(concentrator type)
I
8
n Boiler ID
Preheater
0
Pump
Regenerator
6 e--1 Turbine
J L-
c< Condenser 17.E--
4
5 Pump
Feed pump
Irrigation
canal
in the regenerator and preheater before it passes to the boiler. In the boiler, the high-
temperature water from the collectors causes the working fluid to boil and vaporize.
This high-pressure vapor is then expanded through the turbine, where it develops the
required shaft power. The output shaft in this example drives an irrigation pump, but
it could just as easily drive an electric generator or an air conditioning system. The
vapor exits the turbine at a low pressure and passes through the regenerator to give up
some of its remaining heat before going to the condenser, where it once again becomes
a liquid. A small portion of the water being pumped is used to remove heat from the
condenser.
This process is repeated continuously unless the sun goes down, there is insuffi-
cient water to be pumped, or the wind speed exceeds a design point. Any of these events
will cause the collector array to return automatically to the stowed position.
Figure 34.10 is a pressure-enthalpy diagram for Refrigerant 113, a typical working
fluid which can be used in Rankine-cycle power systems as described above. The turbine
derives work from the system as the vapor expands from point 1 to point 2. The regen-
erator transfers heat from the vapor (point 2 to point 3) to the liquid (point 6 to point
7). The water-cooled condenser removes heat as the vapor condenses (point 3 to point 5).
254 Applications of Solar Energy
1000
100
Absolute Pressure, psi
10
10
'0 20 40 60 80 100 120 140
Enthalpy, Btu/lb above saturated liquid at -40 F
The pump raises the pressure of the liquid from point 5 to point 6. The preheater raises
the temperature of the liquid as it moves from point 7 to point 8. Finally, the boiler
converts the liquid into a gas when heating it from point 8 to point 1.
The fluid flow rates, heat transfer requirements, and cycle efficiencies for typical
Rankine-cycle power systems described above, and depicted in Fig. 34.10, can be accu-
rately calculated using an available computer analysis program, such as CYCLE [15] .
If necessary, however, the parameters can also be calculated by hand as follows.
First, the maximum and minimum temperature (and pressure) levels should be
selected for the boiler and condenser, respectively. The maximum working fluid tem-
perature (T1 ) will generally be 15 to 20°F lower than maximum collector-fluid tem-
perature, because of temperature differences in the heat exchangers. Likewise, the
condensing temperature will be 15 to 20°F above the cooling-medium temperature.
Enthalpy (H1 ) and entropy (Si ) of the vapor at entry to the expander can be found
from saturation property tables at the temperature of T1 (e.g., Ref. 9, Chap. 31).
Next, values for working-fluid properties at the outlet of the expander (point 2
in Fig. 34.10) are calculated from an assumed value for expansion efficiency and
assumed pressure drop in the superheated vapor flowing from the expander outlet
through a regenerator (if used) to the condenser.
where
P = pressure, lb/in.2
Property values for the various state points can be obtained from Ref. 9, Chap. 31, or
similar tables and charts of thermodynamic properties. Finally, other cycle parameters
of interest can be calculated.
2545 HPnet
w= , lb/h (34.8)
(H1 - H2) 77ni -(H6 -H5)
where
Then,
Q
—re (34.10)
H7 + H6
w
256 Applications of Solar Energy
Hpfp w H6 — H5 ) (34.14)
2545
Also, the overall Rankine-cycle efficiency is
100(2545)HPnet
arc = %
Qboil Qpreheat
For a numerical example, the following assumptions are made:
Then,
P3 = 8.54 psia
P1 = 283.3 psia
H1 = 124.80 Btu/lb
S1 = 0.1905 Btu/lb-°R
H4 = 91.98 Btu/lb
P4 = 96.63 lb/ft3
H5 = 26.80 Btu/lb
144 283.3 - 8.54 = 27.55 Btu/lb
H6 = 26.80 +
778 (0.7) (96.63)
2545(100) = 16,806 lb/h
(124.80 - 108.07)(0.95) - (27.55 - 26.80)
Qreg = 16,806(0.162)(0.7)(190.7 — 90) = 191,914 Btu/h
For an assumed solar flux, Qs = 300 Btu/h-ft2 , and a collector efficiency, fie = 0.50,
the collector area can now be determined from Eq. (34.4):
Ac /Hp
= 2545 = 96.18 ft2 /hp
(0.1764)(300)(.50)
For 100 hp,
B. Storage Schemes
When extraction rates from irrigation wells are limited, or when the daily water delivery
period must exceed the 6- to 10-hour period of solar pumping systems, storage is needed.
Water storage is most easily accomplished. It involves the pumping of excess water during
the day for storage in an elevated pond or tank and later delivery to the field by gravity
flow. Such systems may be designed for short-term (1 to 3 days) or long-term (seasonal)
storage.
Tanks are generally preferred for short-term storage where ponds cannot be econo-
mically constructed without substantial loss of water by leakage or evaporation. Ponds
are most appropriate for long-term pumped-water storage because of the large contain-
ment volume needed. When ponds are used, ground loss can be minimized by compacting
the bottom surface of the pond and then covering it with Bentonite clay, black poly-
ethylene, or rubber. Evaporation from ponds can be minimized by floating a black cover
of UV-resistant material on the surface of the pond.
Energy storage systems are used when water extraction rates from irrigation wells
are limited. They extend the daily pumping duration of the solar power system by
storing energy between the energy collection unit and the pump drive unit. Systems with
photovoltaic collectors or those employing an electric generator/electric motor drive
between the solar power unit and the pump drive unit may utilize batteries for storage.
However, the efficiency of battery storage systems is relatively low. As much as 40 per-
cent of the energy delivered by the solar power system may be lost at the generator,
battery, and electric motor.
Another means of extending daily operation of the power system is some form of
thermal energy storage between the collection and power units. Thermal energy may be
stored as sensible heat, as latent heat with a phase-change material, or as a combination
of the two. In sensible-heat storage systems, the storage medium heats up or cools down
in storage and retrieval of energy; in latent-heat storage systems, the majority of energy
is stored at a constant temperature corresponding to the phase-change temperature.
Most solar power systems will operate more efficiently at a constant tempera-
ture corresponding to design-point conditions; hence, there is considerable effort under-
way in the United States to develop phase-change storage units, particularly for solar
space heating and cooling and solar power generation [17] . Such systems are experi-
Irrigation Pumping
Storage Capacity, kca l/gm 259
mental in nature and, at present, are costly, requiring large areas for transfer of heat into
and out of the solid-phase material.
Sensible-heat storage systems include those in which heat is stored in liquids,
solids, or a mixture of the two. Table 34.4 gives the physical characteristics of some
liquid sensible-heat storage materials [18] . Figures 34.11 and 34.12 show the specific
0.20
Mg0
Storage Capac ity, kca l/gm
0.15
A1203
Si02
0.10
Granite
0.05
500
H2O
400
Storage Capacity / Cost, kcal/I
300
,Caloria HT43
200
1
Therminal 55
„A.-Hitec
NaK
100
Therminal 66
0 I 1 1
0 200 400 600
Temperature Swing, C
Figure 34.13 Capacity-cost comparison for liquid storage media. From Ref. 18, used
with permission.
heat storage capacity of these fluids and of some selected solid materials. The data
show that, on the basis of specific storage capacity, the liquids are more useful than
the solids. However, in the irrigation pumping environment, weight is less important
than cost.
Figures 34.13 and 34.14 show that inexpensive granite and sand have very favorable
storage/cost values. A liquid heat transfer medium is often used for transport of energy
into and out of packed beds of the solids. A recently designed storage unit for a 10-MW
solar power system utilizes Caloria HT43 fluid and crushed granite [19] .
C. Collector Types
Generally, collector field configurations can be selected from three possible options:
flat-plate collectors, concentrating-tracking collectors, and power tower configurations.
500
Granite
Si02
400
Storage Capacity/Cost, kca l/1
300
200
100
0
0 200 400 600
Temperature Swing, C
Figure 34.14 Capacity-cost comparison for solid storage media. From Ref. 18, used with
permission.
The angle or tilt of the flat-plate collectors toward the south should be adjusted
to maximize the amount of solar energy collected during the irrigation season, unless
the energy is to be used for some other purpose during the off-season.
To maximize the amount of energy collected between March and September in
the United States, for example, flat-plate collectors should be tilted at an angle of 5 to
.
- June
Mar Aug
A
Sept June
8
70.3°1.5°
58.0°
Dec
34.6°
June
Solar altitude at
noon on 21st day
of month at 32°
north latitude
fo
Go
45°
15° less than the local latitude. (For example, at 40°N latitude, the collectors should
be tilted southward at about 30° above the horizontal.) To maximize year-round energy
use, the collectors should be tilted at an angle about equal to the latitude. To favor
winter use the angle should be 5 to 15° higher than the latitude.
500
Based on noon values for Phoenix,
Arizona (10-yr monthly averages)
400
Solar Flux at Noon, Btu/hr -ft 2
300
200 —
Legend
o
Total on 45° col lector
100
with 35° reflector
o Total on 32° collector
+ Total on horizontal surface
3 Direct on horizontal surface
I I I I I I 1
J FM A M J J A S 0 N D
Month
Figure 34.16 Incident solar radiation on collectors and ground.
264 Applications of Solar Energy
Concentrating/Tracking Collectors
The second collector option provides arrays of concentrating/tracking collectors.
These arrays have two main advantages over flat-plate collectors when used with solar-
powered irrigation pumps. First, they can reach higher operating temperatures because
of the concentration of sunlight, which in turn produces higher Rankine-cycle efficien-
cies. Second, they collect more usable solar energy each day during the irrigation season
because they follow, or track, the sun all day long. Other advantages of these collectors
include fewer plumbing connections (which is an important cost consideration) and a
relatively small absorber area, which minimizes heat losses.
There are two major disadvantages of concentrating/tracking collectors. First, they
do not collect the diffuse component of sunlight (which can be appreciable in some areas
of the country). Second, they require accurate focusing and relatively sophisticated
tracking mechanisms, which use auxiliary electrical power.
There are three ways to orient concentrating/tracking collectors, depending on the
daily and seasonal pumping requirements. The collector rows can be mounted horizon-
tally, with their axes either north-south or east-west, or the axes can be tilted at the
latitude angle in a polar mount. The north-south arrangement favors the summer season.
In this orientation, the collectors gather the most daily energy because they track the sun
all day from east to west while maintaining a high efficiency. However, performance
declines considerably in the winter months as the angle of the sun above the horizon
decreases.
An east-west axis orientation maintains the peak noontime design values through-
out the year, because the collectors track the altitude angle of the sun. However, daily
performance is very similar to that of a fixed flat-plate collector in that it falls off sharply
as the sun's azimuth angle departs from solar noon.
A polar-mounted tracking concentrator collects the most energy on a year-round
basis, but its summer output is not as high as the north-south horizontal-mounted collec-
tor, nor is its winter peak noon output as high as with the east-west orientation. Its peak
output occurs during the spring and fall when the sun is perpendicular to the axes. Dis-
advantages of polar mounting are that it normally requires a substantial framework for
mounting the collectors and, if very large arrays are required, shading of successive rows
Irrigation Pumping 265
must be considered. In addition, wind forces are a more important design consideration
with polar mounting.
Figures 34.17, 34.18, and 34.19 show some water-pumping comparisons between
flat-plate and tracking collectors when operated with a 50-hp low-lift solar pump. Figure
34.17 shows the daily variation in pumping characteristics for a typical average day in
May in Phoenix, Arizona. The focusing collectors with north-south axes pump about
twice as much water each day as a flat-plate collector system designed for the same peak
power output at solar noon. Figure 34.18 compares the monthly variation in daily pump-
ing time for both focusing and flat-plate collectors. With the focusing collectors oriented
north-south, performance reaches its maximum during the summer and then falls off
as the sun moves lower in the sky.
Because the system requires some minimum power before water can be pumped
to a higher elevation, neither collector system will pump any water on a typical
December day. However, as mentioned earlier, if an east-west axis orientation is used,
the design power output can be reached at noon each day, but naturally the daily pump-
ing duration during winter will be shorter. Figure 34.19 compares the total daily water
delivery of the focusing and flat-plate collector systems for each month of the year. Once
again the superior performance of the tracking collectors can be seen. As would be
expected, the higher pumping performance during the summer months coincides nicely
with the usual requirements for irrigation water.
10
/
9
E 8
o.
0 7
0 1
cc
1
5 I Flat plate system 1
E
er (collector 50°/
reflector 35°)
3' 3 1
1
Tracking collector
(N-S axis)
1
1 1
1
1 1 1 1 1 1 1 1 1 1
r 8 9 10 II 12 13 14 15 16 17 18
Time of Day, hours
Figure 34.1 7 Daily variation in water delivery for flat-plate and tracking collectors
(50 hp, low-lift pump).
266 Applications of Solar Energy
10
8
Daily Pumping -Time Durat ion, hr s
Jan Feb Mar Apr May Jun Jul Aug Sept Oct Nov Dec
Month
Figure 34.18 Predicted monthly variation in daily pumping time duration for flat-plate
and tracking collectors.
Power Towers
Power tower systems consist of a large number of tracking mirrors (heliostats)
that direct the reflected solar rays to a receiver mounted on top of a tower. The receiver
absorbs the reflected radiant solar energy and transfers it to the working fluid of a power
system.
Heliostats may be rectangular or circular. Typically, they are slightly concave to
enhance focusing on the absorber. Each heliostat must be "steered" on two axes to
direct the reflected rays onto the absorber as the sun moves across the sky.
Energy flux levels at the absorber surfaces of power tower systems are equivalent
to those on boiler tubes within fossil-fueled power plants. This is possible because of the
high concentration ratio, which may exceed 1500 in comparison to 30 to 50 for distri-
buted parabolic trough collectors. Energy heat losses during transport from the collector
to the power system are small because the power system boiler is usually an integral
part of the absorber surface and heat losses from the boiler outlet to the power system
at the base of the tower can be minimized with insulation. Reflective surfaces on ab-
Irrigation Pumping 267
5
Daily Water Delivery, 10 ga llons
3 —
/ •••••_ /sr — \
It '--e, \\
/ \ ,//
_ `vi• \ \\
2 /
/
Flat plate system
Flat
/-4,- (collector 50 0/
1 — / reflector 35°)
/
/
d
0
I I 1 1
Jan Feb Mar Apr May Jun Jul Aug Sept Oct Nov Dec
Month
Figure 34.19 Predicted monthly variation in daily water delivery for flat-plate and
tracking collectors (50 hp, low-lift pump).
sorber cavity walls and selective surface coatings on the boiler tubes will minimize
radiant energy losses from the absorber.
Power tower systems do not utilize diffuse solar radiation and therefore should be
located in regions with a high proportion of clear weather. These systems require con-
siderably more land area than conventional systems with distributed single-axis tracking
collectors because individual reflectors must be spaced apart to avoid shadowing. The
design of optimum reflector fields for power tower systems is quite complex [20] .
Power tower solar systems become cost competitive with distributed concentrating
collector systems only in relatively large solar power systems. This higher cost is due to
the need for two-axis tracking mechanisms on hundreds of heliostats and a relatively
high-cost absorber. On large tower systems, computers may be used to steer the helio-
stats, whereas clock-driven mechanisms may be used on smaller systems.
A. Startup
solar flux must supply enough heat to the system to maintain operation. Then (a) the
collector fluid must be circulated to the boiler, (b) the working cycle liquid must be
pumped to the boiler, and (c) cooling of the condenser must begin. Usually these func-
tions are accomplished by means of startup motors and pumps. Once the system has
reached an operating condition that will sustain itself, the starting motors can be turned
off. This startup condition must be determined for each system, and sensors can be used
to detect when predetermined levels of speed, pressure, temperature, and so on, are
reached.
B. Maintenance
Several components in the solar power system require regular maintenance, including
cleaning the collectors, purging the condenser of noncondensible gases, maintaining the
proper charge of working fluid in the system, lubricating where necessary, adding corro-
sion inhibitors and/or antifreeze to the collector fluid, and cleaning the strainers and
filters.
A gradual buildup of dust on the collectors can lower the operating efficiency by
10 percent or more. In areas where rainfall occasionally washes off the collectors, addi-
tional cleaning may not be necessary. However, in dry, dusty environments, periodic
cleaning of the collectors will improve efficiency. This can be acomplished quite easily
by using a water spray apparatus with detergents added, being careful to use detergents
that will not damage the collector materials.
Noncondensibles can be purged from the condenser either by continuously ex-
hausting a small portion of vapor from the location where they will collect, or by
periodically venting the condenser from the proper location. If the condenser normally
operates at pressures below atmospheric, then either the vapor must be pumped out of
the condenser, or the condenser temperature must be raised until a positive pressure is
available for purging.
Periodically, the system must be charged with working fluid, unless, or course,
a hermetically sealed, leak-tight system can be achieved. Leak tightness is very difficult
to obtain because of rotating seals, periodic purgings, and leaking joints. The system can
be charged by pumping or bleeding liquid or vapor into it through a charging valve.
Periodic lubrication is usually necessary when rotating seals or bearing surfaces are
present. Therefore, a lubrication schedule should be determined for each system design.
Corrosion inhibitors and antifreeze solutions are usually required if water is used
in the collector loop. These chemicals should be checked once a year to assure a safe
operating condition.
In addition to corrosion inhibitors, sacrificial corrosion materials, such as zinc
plugs, may also be desirable to protect heat exchanger and pipe materials,
If strainers or filters are included in the system to protect pumps, heat exchangers,
gearboxes, and so on, they should be cleaned or changed occasionally.
C. Safety Aspects
Because of high pressures and temperatures in solar power systems, the equipment must
be safety tested in accordance with ASME specifications. For example, all piping should
Irrigation Pumping 269
be proof tested at pressures 1.5 times higher than the normal operating pressures, with
proper consideration given to strength-at-temperature for the various materials. Pressure-
relief valves and burst diaphragms should also be tested to ensure that the maximum
pressures attainable are below the test pressures. Emergency shutdown provisions that
sense overspeed, overpressure, and overtemperature conditions should be included. Such
conditions might occur, for example, if a driveshaft breaks or if the collector flow is
interrupted. If the collector array could be damaged by high winds or hail, then provi-
sions must be made to return the collectors to a safe position during such storms, or
during loss of sunlight.
270
Irrigation Pumping 271
The Nebraska pump is powered by silicon photovoltaic cells [24] . The system has
a peak power rating of 25 kW and drives a 7.5-kW (10-hp) pump. The collector consists
of 120,000 individual silicon solar cells mounted on 28 flat panels. The panels are wired
to produce 6.2 A at 150 V. The solar array consists of two rows 99 in long facing due
south. One row is set about 1.5 m lower than the other to prevent shadows from falling
on the cells (see Fig. 34.22). Power from the collectors is fed to a bank of large lead acid
batteries, capable of storing 85 kWh of energy for off-peak power use, and to a set of
inverters. The inverters convert the DC electricity produced by the cells into AC at 220
V (three-phase) used to power the irrigation pump motor and other loads. The 7.5-kW
pump is interconnected with a gated pipe irrigation system. This pump can operate 12
hours a day and pump 3.8 m3 /min (1000 gpm) of water out of an irrigation reservoir.
Current plans call for switching the power from the irrigation system to a nearby corn-
drying bin during the nonirrigation portions of the year.
In recent years, the French Company SOFRETES, through RENAULT Moteurs Develop-
ment as general sales agent, has marketed a line of solar pumps worldwide.
The largest of the SOFRETES pumps was installed in San Luis de la Paz, Mexico,
in 1975 (see Fig. 34.23). This is a 25-kW (33.5-hp) installation that supplies irrigation
and drinking water to the citizens of San Luis de la Paz. Low-temperature water, about
71°C (160°F), circulates from 1500 m2 of flat-plate collectors to a heat exchanger. The
exchanger generates superheated Freon for an expansion motor, which drives an
alternator-generator. Power from the generator is used to drive electric pumps at each of
two well sites. The output of the installation is about 2.5 m3 /min (660 gpm) of water.
272 Applications of Solar Energy
Figure 34.23 Solar pump installation, San Luis de la Paz, Mexico. Courtesy of Societe
Francaise d'Etudes Thermiques et d'Energies Solaire.
The dynamic pumping levels are 54 m and 30 m. Daily average operating time is from
5 to 6 h.
About 40 SOFRETES "one-kilowatt" stations are now operational in 12 countries:
The Arab Emirates, Brazil, Cameroon, Upper Volta, Cap Verde Islands, India, Mali,
Mexico, Mauritania, Niger, Senegal, and Chad [25] . Additional units are under construc-
tion. A 70-kW unit became operational in the Republic of Mali in 1978.
Present SOFRETES 1-kW units utilize 100 to 120 m2 of flat-plate collectors with
fluid circulation via thermosiphon or circulators. A two-cylinder expansion motor
employing butane or Freon is used. The pumps operate about 5 to 7 h/d with a daily
output of about 30 m3 for a discharge height of 20 m.
There is a large concentration of solar pumps in Mexico [26]. Sixteen are opera-
tional, three are under construction, and one is on exhibition. Most of these pumps are
small, 1-kW units.
system. The objective of these design studies was to develop a system applicable to irri-
gation requirements in the Southwest, specifically for a demonstration site in Coolidge,
Arizona.
Acurex [27] proposed to use a distributed collector field, employing parabolic
trough tracking collectors, with an outlet temperature of >315°C (600°F). A thermo-
cline fluid-storage system was to be used to accommodate insolation variations and
operation during the evening. An organic Rankine turbogenerator was to generate the
required 150 kWe, three-phase 440 V to drive the pump motor.
Honeywell's proposed system [28] consisted of a paraboloid, two-axis, tracking
(radar dish) solar-heat concentrator with a cavity receiver, a heat engine, and an electri-
cal generator mounted at the focal point. Storage and backup systems were to be
provided.
The Black and Veatch concept [29] incorporated a central receiver, solar-thermal-
electric energy conversion system. The system was to utilize a 36.6-m-high power tower
surrounded by heliostats. Power was to be furnished to the pump via a six-stage turbine
and 875-kV generator set. Sixteen megawatt-hours of thermal storage were to be pro-
vided. The pump system was to be located on 1.54 hectares (3.8 acres) of land and was
designed to supply 171 hectares (423 acres) of crop land from a 90-m-deep well.
The designs were compared and evaluated and the Acurex design was selected for
construction.
REFERENCES
1. Energy and U.S. Agriculture: 1974 Data Base, Vol. I, FEA, USDA, September 1976.
2. Solar Energy Applications in Agriculture, Potential, Research Needs, and Adoption
Strategies, NSF/RA-760021, Agricultural Experiment Station, University of Mary-
274 Applications of Solar Energy
land, College Park, Maryland, January 1976. (Also, Summary and Recommendations
report, same title.)
3. 1969 Census of Agriculture-Irrigation, Vol. IV, U.S. Department of Commerce,
Bureau of the Census.
4. "1976 Irrigation Survey," Irrigation Journal 26, no. 6 (November-December 1976).
5. A Hungry World: The Challenge to Agriculture, General Report of the University of
California Food Task Force, July 1974.
6. D. Dvoskin and E.O. Heady, Energy Requirements of Irrigated Crops in the Western
United States, Iowa State University, Ames, Iowa, November 1976.
7. A.S.E. Ackermann, "The Utilisation of Solar Energy," Journal of the Royal Society
of Arts (April 20, 1915).
8. Frank Shuman, "Sun Power Plant," Proceedings, Manchester Association of En-
gineers 1913-1914 (Read Saturday, March 14, 1914, supplemented by lantern
illustrations).
9. Handbook of Fundamentals, Chap. 22, Solar Radiation Section, The American
Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc., New York,
1972.
10. TRNSYS, A Transient Simulation Program, University of Wisconsin, Solar Energy
Laboratory report and computer program, December 1976.
11. SOLMET weather data for 26 cities in the United States, available from National
Climatic Center, Asheville, North Carolina.
12. George 0. G. Lof, John A. Duffie, and Clayton 0. Smith, "World Distribution of
Solar Radiation," Solar Energy, 10, no. 1, 27-37 (1966).
13. Eldon C. Boes, Irving J. Hall, Richard R. Praire, Robert P. Stromberg, and Herbert
E. Anderson, Distribution of Direct and Total Solar Radiation Availabilities for the
U.S., Sandia Laboratories Energy Report No. SAND 76-0411, August 1976.
14. A.M. Zarem and D.D. Erway, eds., Introduction to the Utilization of Solar Energy,
Chap. VII, McGraw-Hill, New York, 1963.
15. J.P. Albin and W.R. Lenenberger, Program Cycle: A Rankine Cycle Analysis Routine,
Sandia Laboratories Energy Report No. SAND 74-0099, October 1974.
16. R.W. Matlin (MIT Lincoln Laboratory) and M.T. Katzman (Harvard University),
The Economics of Adapting Solar Photovoltaic Energy Systems in Irrigation, June
22, 1977.
17. T.T. Bramlette, R.M. Green, J.J. Bartel, D.K. Ottesen, C.T. Schafer, and T.D.
Brumleve, Survey of High Temperature Thermal Energy Storage, SAND 75-8063,
Sandia Laboratories, Livermore, Calif., March 1976.
18. R.M. Green, D.K. Ottesen, J.J. Bartel, and T.T. Bramlette, "High Temperature
Thermal Energy Storage," Proceedings from ISES Conference at Winnipeg, Canada,
Vol. 8, August 1976, pp. 5-47.
19. Central Receiver Solar Thermal Power System Phase 1: Thermal Storage Subsystem
Research Experimental Detail Design Report, McDonnell Douglas Report No. MDC
6360, Huntington Beach, California, April 1976.
20. C.S. Hoyle, "Use of Calculated Displaced Shapes to Define the Reflected Light
Pattern from a Focused Collector, Proceedings of ISES Conference at Winnipeg,
Canada, August 1976, p. 208.
21. S.G. Talbert, R.D. Fischer, G. Alexander, D.H. Frieling, and J.A. Eibling, Battelle-
Columbus Laboratories, "The Development of a 50-Horsepower Solar-Powered
Irrigation Pump," presented at the International Solar Energy Society Meeting,
New Delhi, India, January 1978.
Irrigation Pumping 275
FOOD DEHYDRATION
35.1 INTRODUCTION
People have used naturally dryed foods for thousands of years. Artificisl dehydration
of foods, however, dates back only about two centuries. In artificial dehydration, an
artificial source of heat is provided to take the place of direct exposure of the food to
the sun in the field.
The food industry in the United States currently utilizes mainly natural gas and
some propane and oil for the dehydration of fruits, nuts, vegetables, and milk. The
largest part of the industry is located in California, which processes almost all commer-
cially dried fruit and vegetables with the exception of apples (Washington, 50 percent)
and potatoes (Washington and Idaho, 16 percent).
With the recent announcement by the California State Public Utilities Commission
that the demand for natural gas by the food dehydration industry cannot be satisfied
beyond 1979, the industry is currently evaluating a number of prototype solar dehydra-
tion installations.
Although solar food dehydration plants are new to the United States, a small
number of prototype units have been constructed in other parts of the world during
the past decade. A solar dehydration facility was installed in Barbados, West Indies,
in 1967 to dry corn. Another system was constructed in Puerto Rico in 1962 to dry
coffee beans, and a third facility was installed in Turkey in 1971 for the dehydration
of fruit. Although all of these solar installations have reported varying degrees of success,
277
278 Applications of Solar Energy
it was not until the emergence of the current energy crisis that solar dehydration received
the widespread attention it deserves.
Although the food dehydration industry is by no means the largest section of the food
processing and services industry, it is nevertheless one of the most essential elements of
that industry. At the present time, dehydration offers the most efficient method of
preserving food. With the world reserve capacity of food at a dangerously low level, food
preservation by means of dehydration provides the only meaningful way to meet tem-
porary food shortages.
Current and projected food dehydration production volumes for the United States
are shown in the second column of Table 35.1. These estimates are based, conservatively,
on present and expected future normal demands and do not take into account emergency
conditions that might arise at any time as a result of droughts, wars, or international
food shortages. It is essential that this nation should continue to prepare for such likely
eventualities by maintaining, at all times, an economically sound and highly productive
food dehydration industry.
In addition to fulfilling this essential national need, food dehydration offers the
following production advantages:
Table 35.1 Projected Production Volumes, Retail Values, and Energy Demands for
Selected Dried Foods
Equivalent Fuel Demand
Retail Heat Gas Propane
Productiona Value Required (Millions (Millions Oil
(Millions of (Millions (Millions of Cubic of Cubic (Millions
Year of Pounds) of Dollars) of Btu's) Feet) Feet) of Barrels)
1975 13,740 $7,031 107,029,450 107,029 42,812 17.8
1980 14,840 $7,282 110,812,390 110,812 44,325 18.5
1985 15,890 $7,712 113,308,620 113,305 45,322 18.9
2000 19,110 $8,882 115,072,550 115,073 46,029 19.2
a
Includes milk, rice, potatoes, onions, garlic, prunes, raisins (machine), apples, walnuts, and almonds;
excludes eggs, coffee, tea, corn, and minor dried food products.
Food Dehydration 279
Based on the production volumes listed in Table 35.1 and dehydration heat demands
ranging from 1.93 MJ/kg (830 Btu/lb) for rice to 87.9 MJ/kg (37,800 Btu/lb) for milk,
the U.S. food dehydration industry accounted for a heat demand of more than 1.13 X
10'4 kJ (1.07 X 1014 Btu) (i.e., Table 35.1, column 4) in 1975). This represented a not
insignificant 0.14 percent of the total U.S. energy consumption in that year.
How can this energy need be met from 1979 onward, without the availability of
natural gas? For various reasons, electricity is unlikely to provide even a short-term
energy solution for the food dehydration industry. With the curtailment of natural gas
to all commercial and industrial customers in California from 1979 onward, some in-
dustries, such as the glass industry, will have to convert to electricity as the only work-
able energy alternative for manufacturing. It is highly questionable whether existing
public and private utility companies will be able to accept this influx of new electricity
users without the imposition of mandatory quotas. Second, at least 60 percent of the
electricity generated in California and more than 80 percent of the national electric
power supply is produced by the combustion of rapidly dwindling supplies of fossil
fuel. Finally, the cost of electricity today is more than five times greater than the cost
of natural gas, per energy unit, in many areas of the United States.
Under these circumstances, it is only natural that the food dehydration industry
should turn to solar energy as a source of heat. Apart from the natural affinity that
exists between the sun and plant growth, the favorable climatic conditions prevalent in
California where the largest section of the industry is concentrated, and the availability
of low-cost land for large solar collector arrays in proximity to commercial dehydration
facilities, recent research studies funded by ERDA* have indicated the following:
1. The peak energy needs of the food dehydration industry generally coincide
with periods of most intensive solar radiation, from July to November.
2. Solar food dehydration systems costing $100/m2 of collector area, amortized
over 10 years and operating for 200 days per year, can be cost competitive
with oil conversion systems for oil at $30.00 per barrel.
3. Solar food dehydration system components can be manufactured by existing
carpentry and sheet metal tradespeople utilizing readily available material and
unsophisticated technology.
The ability to preserve food by dehydration is based on the fact that bacterial action and
the growth of fungi and molds are largely eliminated if the moisture content of food is
reduced to about 15 percent on a dry weight basis.
Moisture can be held within a food material in a variety of ways, purely mechan-
ically by means of surface tension, absorbed moisture held in the material by weak inter-
*Energy Research and Development Administration, now replaced by the U.S. Department of Energy.
280 Applications of Solar Energy
2.0
1.0
0 2 4 6 8 10
TIME (HOURS)
Figure 351 Relationship between moisture content and drying time for a typical food
product.
28
o 24
20
16
z
12
w
cr
co
—
3!,74
13:1 Figure 35.2 Equilibrium moisture
00 20 40 60 80 100 curves of a number of materials at room
EQUILIBRIUM RELATIVE HUMIDITY , temperature, approximately 25°C (77°F).
Food Dehydration 281
molecular forces, and by chemical absorption in which water is held in the material by
chemical bonds over a wide range of strengths. The dehydration process consequently
proceeds in two phases, the constant-drying-rate and the falling-drying-rate phases.
There is a sharp dividing line between these two drying phases.
During the constant-rate period, free moisture evaporates from the surface. After
all the free moisture has been removed, the rate of drying decreases in the falling-rate
phase. As shown in Fig. 35.1, for most agricultural products much of the drying time is
spent in this phase because the moisture must first move within the material to the
surface before it can evaporate.
When a food product is brought into contact with air at a given temperature and
relative humidity, it will eventually come to a fixed moisture content and will not change
with further exposure to the same air. This is known as the equilibrium moisture content
of the material. If a food product is above the equilibrium moisture level, it will lose
water, and if it is below, it will gain moisture. Some typical equilibrium moisture curves
are shown in Fig. 35.2.
Five principal types of dehydrators are in common use in the food dehydration industry
today: tunnel, belt, bin, column, and rotary dehydrators.
As shown in Fig. 35.3a, a tunnel dehydrator consists of an elongated enclosure
through which trucks, loaded with trays filled with drying food, are moved in contact
with or against a current of air heated from 49 to 88°C (120 to 190°F) for moisture
evaporation purposes. The trucks are normally pulled continuously through the tunnel
by a moving chain. Loaded trucks are introduced periodically at one end of the dehy-
drator, each displacing a dried truck at the other end of the tunnel.
In operation, the prepared food is loaded in a thin uniform layer on the drying
trays, and these are stacked one above the other to a height of 1.5 to 2.1m. A single
tunnel may contain 5 to 15 trucks. A 12-truck tunnel, with 0.9m X 1.8m trays stacked
24 high, can hold the equivalent of 4540 kg (5 tons) of fresh produce at one time.
A schematic diagram of a typical belt dehydrator is shown in Fig. 35.3b. Each
stage is 9 to 18m long and 2 to 3m wide. Only stainless steel comes in contact with the
food. Sectionalizing the dehydrator into multiple stages allows for efficient operation.
In the first stage, the food is spread thinly (approximately 10 cm thick), and relatively
high-temperature, high-flow air at 71 to 104°C (160 to 220°F) is used to remove 90
percent of the original moisture. In the final stages, the partly dried product is redis-
tributed to a deeper layer approximately 30 cm thick and drying continues at a lower
temperature: 38 to 71°C (100 to 160°F).
It is common practice to complete the dehydration process of fruit and vegetables,
predried in tunnel or belt dehydrators, in bin dehydrators (Fig. 35.3c). A typical bin
dehydrator consists essentially of a metal or wooden box equipped with an air inlet at
the bottom, allowing warm air to pass up through food piled on a wire mesh deck. Bin
sizes generally vary from crop to crop, from small boxes lm on each side to large con-
tainers 20m or more on a side.
282 Applications of Solar Energy
HEATER FAN
FRESH
AIR
SE3E1 IF)
EXHAUST AIR STACK
(a)
PRODUCE
SLATTED BOTTOM
HEATED AIR
(c)
t=_I
SLOTTED SIDES
HEATED AIR
(d)
FAN
Figure 35.3 Five principal types of
E3U7;.ER i
llDRYING
DRUM
commercial dehydrators. (a) Tunnel
ka dehydrator. (b) Belt dehydrator.
(c) Bin dehydrator. (d) Column
(e) dehydrator. (e) Rotary dehydrator.
The continuous-feed column dehydrator (Fig. 35.3d) is very well suited for granu-
lar free-flowing materials such as grain. The wet material is loaded in a hopper at the top
of a long, vertical, perforated column. The dried material is removed from the bottom
at a continuous rate proportional to the size of the drying column. Heated air, from 3 to
55°C (5 to 100°F) above ambient temperature is blown across the column at right
angles to the direction of movement.
Rotary dehydration (Fig. 35.3e) is best suited for material that is not free flowing
and can withstand a lot of handling, such as chopper forage and fruit and vegetable
residue. The drum is either horizontal or at a small angle. The inside of the drum may
have ribs to help keep the material in the air stream. Finely chopped wet material, such as
chopped green alfalfa, may be exposed to temperatures as high as 815°C (1500°F).
Food Dehydration 283
A typical industrial dehydration unit is capable of processing some 1.78 kN (400 lb)
of wet food per hour and, operating at an efficiency of 40 percent, requires about 2.6
MJ (2500 Btu) of heat and 56.6m3 (200 ft3 ) of air to evaporate 0.5 kg of water. Since
almost 90 percent of the weight of wet food is typically attributable to the presence of
water, the attendant heat energy requirements of a commercial food dehydration facility
are quite large. A typical dehydration unit will utilize nearly 1 GJ (1.0 million Btu) of
energy per hour with a corresponding air flow requirement of 560m3 /min (20,000 ft3 /
min).
As the heated air passes over the wet food particles (e.g., diced carrots), it absorbs
moisture and is finally exhausted at a higher humidity and lower temperature. The con-
stant enthalpy relationship between the temperature and humidity of the air entering
and leaving the dehydrator may be represented in the form of constant wet-bulb tempera-
ture lines, as shown in Fig. 35.4. While the food particles are wet, they keep relatively
cool due to the evaporation of water from the surface (i.e., by latent heat of evaporation),
until they assume a temperature approximately equal to the wet-bulb temperature of the
air. Thereafter, as the surface dries, the rate of evaporation and the cooling effect
decrease, until the temperature of the food reaches the dry-bulb temperature of the air.
To prevent scorching of the food particles, the temperature of the air is normally
adjusted to between 70 and 93°C (160 and 200° F) during the initial or "constant-rate"
drying phase and 50 to 70°C (120 to 160°F) during the final "falling-rate" drying phase.
Most commercial dehydration facilities employ some form of heat recirculation to
improve the efficiency of their operation. As much as 50 percent of the warm, moist
exhaust air is mixed with the fresh incoming air, then reheated and passed over the food
particles again. The increase in efficiency that may be gained by recirculation is readily
appreciated by reference to the psychometric chart as shown in Fig. 35.5. If point A is
w
3.1.1110S8V
rn
co
0 I
-n c
o
o
>
MOISTURE
EVAPORATED
WET BULB
LINE LEAVING DEHYDRATOR AIR
WITH RECIRCULATION
H2(humidity)
,.., „.., _ t.
LEAVING DEHYDRATOR -1; r a
AIR WITHOUT 0 w E cn cil
RECIRCULATION z -0 — 0
C 3, 73 71 C
AIR ENTERING
DEHYDRATOR
HI (humidity)
Figure 35.5 The dehydration process with and without heat recirculation.
plotted as the temperature (T1) and humidity (H1) of the fresh air entering the dehy-
drator, then point B represents the temperature of the air after heating. The quantity of
heat required to heat the air from T1 to T2 is proportional to the rise in enthalpy, repre-
sented in Fig. 35.5 by the length of line AB. In the case of a dehydrator operating with-
out recirculation, the air in condition B will be cooled adiabatically (i.e., without heat
entering or leaving the system) along line BC, while passing over the food particles, and
then exhausted in condition C. On the other hand, if recirculation is employed, only a
portion of the air is exhausted and the remainder together with fresh air is heated to
temperature (T2 ) and humidity (H2 ) shown at point D. Again, as the air in condition D
is passed over the food particles, it cools adiabatically along line DE to condition E. A
portion of this air is exhausted and the remainder mixed with fresh air to produce a new
condition F, and so on.
As shown in Fig. 35.5, line DE is a little shorter than line BC, thus indicating that
whereas recirculation can increase the efficiency of a dehydrator by some 30 percent,
it tends to decrease the drying rate by about 10 percent.
Unfortunately, many commercial dehydration facilities operate only during a
few months of each year. After a crop has been harvested, it must be processed and dried
within a few days to avoid spoilage. Even today there are still a considerable number of
dehydration facilities that rely on a single crop type and close down after the harvesting
or dehydration season. This season may range from a brief period of 40 days in the case
of prunes to 200 days for the dehydration of potatoes. During this relatively short drying
season, commercial dehydration facilities are normally in operation for 24 hours each
day, 7 days per week.
One of the largest solar crop dehydration projects ever undertaken in the United States
is presently under construction in Fresno, California. Sponsored by the U.S. Energy
Research and Development Administration as a cooperative study among The California
Food Dehydration 285
) )
solocita
Figure 35.6 Lamanuzzi and Pantaleo (L&P) solar crop dehydration system under con-
struction in Fresno, California.
Polytechnic State University, San Luis Obispo, TRW Systems and Energy, Lamanuzzi
and Pantaleo (L&P), and the Pacific Gas and Electric Company, the project comprises
a solar system (Fig. 35.6) capable of delivering 70 percent of the total heat require-
ments of a single tunnel dehydrator during a 200-day operating season.
286 Applications of Solar Energy
Present indications are that the total construction costs of the solar system
including heat-recovery facilities will not exceed $350,000, or $180/m2 of collector
area. The solar facility comprises 1885m2 of single-glazed flat-plate air collectors,
350m3 (12,500 ft3 ) of rock storage, a heat-wheel type heat-recovery unit, and an auto-
mated control system. All of the materials and components of the facility are readily
available and were assembled by local contractors and sheet metal workers.
As shown schematically in Fig. 35.7, the ambient air is first preheated by heat
obtained from the moist dryer exhaust air using a rotary wheel-type heat-recovery unit.
The air is further heated in one of the following modes: by passing through the solar-
heated rock-bed storage unit; by standard gas-fired burners in the dehydration tunnel;
or by a combination of these methods, determined by the availability of heat from the
collector and the rock storage unit.
When there is insufficient solar radiation, the air will pass through the storage unit
as long as heat is available. Any shortage is automatically made up by the thermostatically
controlled burners. The control system will also divide the flow between the collector
and the storage in a ratio to ensure that the two air flows are at equal temperatures, thus
allowing the collectors to operate whenever solar energy is available.
During the day, the collector output temperature is maintained at the dehydrator
temperature by increasing the flow through the storage fan and thus "loading" the
storage for night operation. If overheating occurs, the air to the dehydrator is diluted
with air from the heat-recovery unit in order to maintain the desired temperature.
During all of the operation modes, both the fan in the dehydration tunnel and the
inlet fan operate continuously at a constant flow rate. These fans turn off only when the
tunnel doors are opened for purposes of loading, unloading, or checking the product. In
view of the varying system pressure drops that occur during different modes of operation,
66•C
(I50°F) DEHYDRATOR
("6
--0.1X3 1
BURNER
EXHAUST r/ //////////
AIR — PRODUCE
II //////// 4
66°C ( ItYrn 49°C (120°F)
FRESH HEAT RECOVERY UNIT
ROCK STORAGE AIR
43°C 4...A - 43°C (110°F)
(
2I°C
43°C (I10°F) STORAGE FAN (70°F)
a- INLET FAN
300
200
100
0
14
c)
DOWN
MASSFLOW( Ibihr -ft
10 —) FLOW UP
(DRYER
6
2
0
—2
—6
160 22. Storage at Drying Temperature =150 °F
top Temperature Deficit Made Up By Burner
150
7 22
140 7 2,22
ST , Average Storage
130
TEMPE RATURE( OF )
ST
120
110 141nlet
to Inlet Fan
100 Collector
2
/Outlet
90 Temperature Ambient
Boost by Heat
80 Recovery
70
60
0 4 8 12 16 20
TIME OF DAY
Figure 35.8 Hourly system operating conditions, on September 21, predicted for L&P
solar facility at Fresno, California.
the inlet fan flow is kept at a constant flow rate by means of automatically controlled
inlet guide vanes.
The temperatures indicated in Fig. 35.7 represent average design values corres-
ponding to September operation and weather conditions. The system is designed to
supply a fixed air flow rate of 9.4 m3 /s (20,000 cfm) to dehydrator 24 hours a day.
The drying temperature varies through the season beginning in June at 60°C (140°F) and
increasing to a maximum of 66°C (150°F) in August and September. Operations from
November onward are dedicated almost exclusively to the final dehydration of sun-dried
raisins at about 50°C (120°F).
Figure 35.8, based on a computer stimulation study, indicates the predicted "design
day" (i.e., September 21) temperatures and air flow rates for the Fresno solar facility. On
the day under consideration, the system operation encompasses four distinct operating
modes. Except for electrical energy used by the fans, the curves show that the system
uses solar energy exclusively for 17.5 h of the day. The solar energy is supplied directly
by the collectors from 9:00 a.m. through 4:00 p.m., and by the rock-bed storage from
288 Applications of Solar Energy
TRANSITION
GLAZING (LEXAN)
GLAZING SUPPORT
11111
COLLECTOR ABSORBERS
COLLECTOR FRAME
atio eq..-
EXPANSION BOOT
EXTERIOR SIDE
441.
°
COLLECTOR
SUPPORTS
Figure 35.9 Cross section of collector design used in L&P solar installation.
4:00 p.m. to 2:30 a.m. Thereafter, until 8:00 a.m., the stored heat is augmented by the
use of auxiliary fuel. Starting at 8:00 a.m., the collectors begin to supply heat. By 9:00
a.m., the collectors are handling all of the dehydrator needs, and the additional heat
collected begins to be stored in the rock. The rock storage can be seen to be about
"half-full" by 12:00 noon. Prior to this time, as a consequence of the storage identifica-
tion, the air from the bottom of the storage unit is returned to the collectors at the mini-
mum design temperature (equal to that delivered by the heat-recovery unit). After 12:00
noon, the temperature at the bottom of the storage unit begins to rise, as can be seen by
the increase in the temperature at the collector inlets. By the end of the collection storage
period at 4:00 p.m., the average rock temperature is greater than the dehydrator inlet
temperature, indicating an "overfull" storage unit. The collection and storage of heat at
temperatures above 66°C (150°F) and the overfilling of the storage are both detrimental
to collector performance. These are consequences of component size decisions to mini-
mize life-cycle costs for the intended percent of solar utilization. Thus, although costs
are minimized, maximum component efficiencies are not always attained. The daily
average and peak collector efficiencies for the design day are 35 and 48 percent,
respectively.
The solar collector unit (Fig. 35.9) utilizes flat oval air ducts 7.6 cm X 38 cm X
54.8m long, enclosed in an insulated metal-timber frame. SL-1139 LEXAN plastic sheet-
ing, 0.5 mm thick, is used to provide single glazing. At a cost of just under $64/m2 , this
collector design should prove to be attractive for full-scale implementation in the food
dehydration industry.
The future economic viability of solar food dehydration systems is strongly linked to
government policy and the implementation of solar incentives. Using the prototype
L&P solar dehydration facility in Fresno, California, as a basis, Table 35.2 presents an
Food Dehydration 289
analysis of the relationship between solar system cost and solar energy delivered. The
graphical representation of these data in Figure 35.10 clearly indicates a steep rise in sys-
tem costs for higher percentages of solar energy. Figure 35.11 assumes that no form of
government financing is available and is based on standard economic equations and
assumptions, such as 10 percent interest on 20-year loans and no tax credit (see Chap.
46). Under these circumstances, the base price of oil will always coincide with the "zero"
solar percentage position on the graph, indicating a singular lack of economic incentive
for the utilization of solar energy.
The noticeably disporportionate increase in solar system costs that occurs at the
20 percent solar energy contribution mark in Fig. 35.10 is due to the fact that a solar
dehydration system can supply approximately one-fifth of the total heat requirements
of a food dehydration tunnel operating for 24 hours each day without the need for a
heat storage facility.
As shown in Fig. 35.12, the implementation of low-interest loans at 3 percent and
the availability of 20 percent tax credits are sufficient to bring solar dehydration systems
to a competitive level with oil. At a cost of $25/bbl of oil, the total life-cycle system cost
of a solar dehydration system does not rise significantly above that of a comparative oil-
fired system until the 70 percent "solar" mark is reached. If the cost of oil increases to
$50/bbl, then a solar dehydration system contributing an optimum solar percentage of
between 70 and 80 percent would definitely become an economic proposition.
On the other hand, if the government were to allow a direct tax deduction for each
barrel of oil replaced by solar energy, then an economic solar dehydration system would
be possible today. Based on this proposition, Figs. 35.12 and 35.13 clearly indicate the
favorable relationships that could result between the economic feasibility of a solar
600,000 —
500,000-
-44
E 400,000 —
a,
>•
Cl)
,_
Lo 300,000,
-
o
tn
0
g 200,000-
0
100,000 —
20 40 60 80 100
Percentage Solar
Figure 35.10 Projected system cost versus percentage solar to deliver 3358 MBtu per
year to a dehydration tunnel.
40-
30-
Energy Cost in $/ M Btu
oil at $50/bbl
10—
oil at $25/bbl
oil at $15/bbl
I — I
20 40 60 70 80 100
Percentage Solar
Figure 35.11 Cost per MBtu versus percentage solar for a system as shown in Fig. 35.7
and 10 percent interest, 20-year life, no tax credit, and 51 percent tax rate.
290
40 -
30—
oil at $50/bbl
oil at $15/bbl
oil at $25/bbl
20 40 60 80 100
Percentage Solar 5 )
Figure 35.12 Cost per MBtu versus percentage solar for the same system but with a
government loan, 3 percent interest, and a 20 percent tax credit.
20 —
oil at $50/bbl
oil at $25/bbl
oil at $15/bbl
1 I i 1
1 I I I
1 1 1 i
i i
20 4U do -go 100
5 Percentage Solar
Figure 35.13 Effect of low-interest loans of 3 percent, 20 percent tax credit, and a
50 percent rebate for the oil not used.
291
292 Applications of Solar Energy
dehydration system and the price of oil. Already at $25/bbl of oil, a solar system supply-
ing more than 60 percent of the total energy requirement could result in a negative
energy cost; that is, the tax return would exceed the entire cost of the energy used (Fig.
35.13).
In answering the question, "What is the most economical way to use solar energy
in food dehydration today?," we must turn back to Fig. 35.10. A system utilizing no
more than 20 percent of solar energy requires no heat storage facility, and must there-
fore be considered as having the greatest potential for economic feasibility under present
conditions. This can be further illustrated by integrating the area under a cosine curve for
a time period of 12 hours (i.e., representing 12 hours of sun availability per day) and
dividing this area by the area representative of a constant, 24 hours per day, energy
demand (Fig. 35.14).
In reference to Fig. 35.14, the demand for fossil fuel (i.e., 80 percent) could be
reduced to 40 percent by the use of a heat-recovery unit, which should be capable of
supplying energy at around $0.90/MBtu. Under these conditions, the solar system would
function as a preheater reaching maximum output at 12 noon each day. Naturally, such
a solar dehydration system would result in greatly reduced construction costs. Based on
the costs listed in Table 35.2 for the prototype L&P facility in Fresno, California, system
cost of $86/m2 of solar collector area could be reasonably estimated, that is, less site
preparation required, 50 percent reduction in concrete and steel work, elimination of
storage facility, 33 percent reduction in ducting and thermal insulation, 80 percent
reduction in electrical work as a result of elimination of dampers, and minimal controls.
In this solar dehydration system, the 20 percent solar energy contribution would cost
$3/MBtu, whereas the 40 percent energy contributions from heat-recovery and fossil
fuel would cost $0.90/MBtu and $3.90/MBtu, respectively. At a total cost of $2.50/
MBtu, such a solar dehydration system should become competitive, even with natural
gas, in the near future.
ENERGYDEMAND
0 12 24
TIME(HOURS)
Figure 35.14 Solar energy supplying 20 percent of the heat requirements of a demand
system.
Food Dehydration 293
Since this chapter has focused on the drying of foods utilizing mechanical dehy-
drators, grain, which is still dried largely by nonmechanical means, has been omitted
from consideration. Grain drying is admirably suited to the utilization of solar energy,
because it involves large volumes of low-temperature heat. In view of the short drying
season, a low-technology solar system without heat storage facilities is recommended.
The reader may wish to pursue an interest in grain drying by reference to the supple-
mentary publications listed in the Bibliography.
BIBLIOGRAPHY
M. Akyurt and M.K. Selcuk, "A Solar Dryer Supplemented with Auxiliary Heating
Systems for Continuous Operation," Solar Energy 14, 313-320 (1973).
K. Bird, "The Food Dehydration Industry's Prospective Changes," USDA Marketing
Econ. Div., 1967.
H.R. Bolin and F.P. Boyle, "Use of Potassium Sorbate, Diethylpyrocarbonate and Heat
for the Preservation of Prunes of High Moisture Levels," Journal of Science and
Food Agriculture 18, 289 (1967).
F.P. Boyle, "Dehydrated Mashed Potatoes-Potato Granules," in W.F. Talburt and 0.
Smith, eds., Potato Processing, AVI, Westport, Conn., 1967.
C.W. Browning, D.B. Brooker, R.M. George, and C.E. Browning, 'Batch in Bin Drying
by Alternating Heated and Unheated Air." Transactions of the American Society
of Agricultural Engineers 14, 193-194 (1971).
F. H. Buelow, "Drying Crops with Solar Heated Air," UN Conference on New Sources of
Energy, E/Conf., 1961 35/S/17.
F.H. Buelow, "Solar Energy Collector Design," Transactions of the American Society
of Agricultural Engineers 5, 1-5 (1962).
F.H. Buelow and J.S. Boyd, "Heating Air by Solar Energy," Agricultural Engineering
Journal (January 1957).
California Polytechnic State University, Application of Solar Energy to Industrial
Dehydration, Final Phase I Report, ERDA Contract No. E(40-1)-5123, 1977.
V. Cervinka, W.F. Chacellor, R.J. Coffelt, R.J. Curley, and J.B. Dobie, Energy Require-
ments for Agriculture in California, U.C. Davis, 1974.
D.J. Close, "Solar Air Heaters for Low and Moderate Temperature Applications, Solar
Energy 7, (3) (1963).
C. Davis, and Lipper, "Solar Energy Utilization for Crop Drying," U.N. Conference on
New Sources of Energy. E/Conf. 35/S/53, August 1961.
W.C. Dickinson, and H.J. Freeman, An Economic Methodology for Solar Assisted Indus-
trial Process Heat Systems, UCRL-52254, 1977.
A.J. Ede, and K.C. Hales, The Physics of Drying in Heated Air, with Special Reference
to Fruits and Vegetables, G. Brit. Dept. Sci. Ind. Res., Food Invest. Spec. Rept.
53, 1948.
B. Feinburg, "Vegetables," in Food Dehydration, 2nd ed., vol. 2, W.B. Van Arsdel,
M.J. Copley, and A.I. Morgan, eds., AVI, Westport, Conn., 1973.
J.P. Gentry, M.W. Miller, and L.L. Claypool, "Engineering and Fruit Quality Aspects of
Prune Dehydration in Parallel-/and Counter-Flow Tunnels," Food Technology
19(9), 121-125 (1965).
E.G.B. Gooding and C.G. Tucker, "Dehydration of Vegetables in Multistage Cross-Flow
294 Applications of Solar Energy
B.S. Bauman, M.F. Firmer, and G.C. Shove, "Low Temperature Grain Drying with
Supplemental Solar Heat From an Adjacent Metal Building," American Society
of Agricultural Engineers 75-3514,1975.
F.H. Buelow, "Drying Grain with Solar Heated Air," Quarterly Bulletin Michigan Agri-
cultural Experiment Station, 1961.
F.H. Buelow, Corrugated Solar Heat Collectors for Crop Drying, Sun at Work: No. 4,
1962.
H.H. Converse, A.H. Graves, and D.S. Chung, "Transient Heat Transfer Within Wheat
Stored in a Cylindrical Bin," American Society of Agricultural Engineers 16(1),
129-133 (1973).
M.T. Danziger, M.P. Steinberg, and A.I. Nelson, "Drying Field Corn with Silica Gel,"
Transactions of the American Society of Agricultural Engineers 15(6), 1071-1074
(1972).
G.H. Foster, "Drying Cereal Grains and Other Field Crops," W.B. Van Arsdel, M.J.
Copley, and A.I. Morgan, eds., in Food Dehydration, 2nd ed., Vol. 2, AVI, West-
port, Conn., 1973, pp. 465-488.
G.H. Foster, "Energy Requirements and Techniques for Drying Corn," Proceedings of
the Regional Research Conference, University of Illinois, AE-4228, pp. 1-3.
Food Dehydration 295
C.C. Huxsall and C.W. Hall, "Effects of Sonic Irradiation on Drying Rates of Wheat and
Shelled Corn," Transactions of the American Society of Agricultural Engineers
13(1), 21-24 (1970).
W.H. Johnson and B.J. Lamp, Principles, Equipment and Systems for Corn Harvesting,
Agricultural Consulting Associates, Wooster, Ohio, 1966, p. 370.
G.E. Meyer, H.M. Keener, and W.L. Roller, Solar Heated Air Drying of Soybean Seed
and Shelled Corn, 1975 American Society of Agricultural Engineers, 75-3002,
1975.
R.V. Morey, H.A. Cloud, and W.W. Nelson, Field Evaluation of a Solar Energy Grain
Drying System, American Society of Agricultural Engineers, 75-3515.
D.W. Morrison and G.C. Shove, Bare Plate Solar Collector Grain Drying Bin, American
Society of Agricultural Engineers, 75-3513.
R.M. Peart and G.H. Foster, "Grain Drying with Solar Energy," ISES, Los Angeles, 1975.
W.H. Peterson, "Solar Heat for Drying Shelled Corn," American Society of Agricultural
Engineers, Paper No. NC 73-302, 1973, p. 14.
I.J. Ross and G.M. White, "Thin Layer Drying Characteristics of White Corn," Transac-
tions of the American Society of Agricultural Engineers, 15(1), 175-176 (1972).
T.L. Thompson, G.H. Foster, and R.M. Peart, Companson of Concurrent Flow, Crossflow
Grain Drying Methods, USDA Agricultural Marketing Research Report 841, 1969.
USDA, "Drying Shelled Corn and Small Grains," Farmer's Bulletin, 2214 (1965).
G.M. White, I.L. Ross, and P.W. Westerman, "Drying Rate and Quality of White Shelled
Corn as Influenced by Dew Point Temperature," Transactions of the American
Society of Agricultural Engineers, 16(1), 118-120 (1973).
chapter 36
36.1 INTRODUCTION
Industrial process heat represents a large potential market for solar thermal collector
systems, yet it has not received a great deal of attention from either the solar industry
or the government. As shown in the U.S. energy flow diagram of Fig. 36.1, about 40
percent of total U.S. energy consumption is accounted for by the industrial sector.
About 60 percent of industrial energy use goes to producing process heat, either in the
form of steam or direct heat.
Industrial process heat is used over a wide spectrum of temperatures, from heating
water to 50°C for food and can washing to providing 1000°C direct heat for processing
metals, glass, and ceramics. As seen in Fig. 36.1, nearly one-half of process heat is used
at temperatures below 175°C (350°F). It is in this low-to-medium temperature range
that flat-plate collectors, nontracking concentrators, and single-axis parabolic trough
concentrators, or perhaps combinations of these collector types, should be ideally suited
to deliver solar heated air, water, or steam.
There are some significant differences between a residential solar space heating
system and an industrial solar process heat system. For example, consider the problem of
system maintenance. The typical homeowner installs a conventional water heater or gas
furnace with the expectation that it will be essentially maintenance free for at least 5
to 10 years. The industrial plant manager, on the other hand, has a full-time maintenance
staff whose job it is to keep the plant running at peak efficiency. These people are fam-
iliar with structural and plumbing systems and are able to make adjustments and change
components whenever necessary. Daily maintenance attention could be given a solar
system, and this might make simpler and less expensive systems more attractive.
297
298 Applications of Solar Energy
12 Other
Process >175°C
/0' Industry
steam -s
<175°C I Near-term
solar
Process 7 Direct <175°C applications
heat heat J
->175°C
->75Q
Other
45 (residential and
commercial,
transportation)
Figure 36.1 Distribution of total 1975 use of primary energy; 1 Q = 1015 Btu.
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Figure 36.2 Comparison of monthly utilization of solar heat from a typical residential
system and a typical industrial process system. Energy delivered per unit area of collector
at different times of the year.
Industrial Process Heat 299
greater return on investment by building a solar system or, for example, by increasing
its production capacity? The intangibles, such as public relations and civic duty, usually
take a back seat to maximizing return on investment. A company's chief concern is
to remain competitive in selling its products, and it must choose carefully among al-
ternative uses of its investment capital. Rumors of large increases in the price of fuel
may encourage industrial planners to recommend marginal systems, but solar systems
that show payback periods of greater than 10 years will probably not be seriously
considered.
One reason that solar collectors for industrial applications have not received much
attention in the national solar program is that large collector areas are required to pro-
vide meaningful amounts of heat to any individual plant. Typical domestic heating sys-
tems'generally consist of 10 to 50 m2 of collector area and a storage system of a few
cubic meters. A typical industrial solar system might run from 1 to 100 hectares (2.5 to
250 acres). Of course, these large-size solar systems translate into very large markets for
solar collectors and very substantial fossil fuel savings. A 1 hectare (-100,000 ft2 ) solar
system, collecting heat on an annual basis, would provide an energy savings roughly
equal to that of 500 solar heated homes. Only 5000 such systems would provide a
national energy savings approximately equal to that of 2.5 million solar heated homes,
President Carter's stated goal for 1985. Significant economies of scale are expected for
large industrial solar systems, leading to a reduction in installed system costs.
Another difference between residential and industrial solar systems is in the re-
quired fluid temperatures. Space heating typically requires peak temperatures of 40°C
104°F) for efficient operation, whereas domestic hot water needs to be provided at 55
to 60°C. Industrial processes, on the other hand, require a wide variety of temperatures,
some of which must be closely controlled to avoid damaging the equipment or the
product. This means that a wider variety of collector types will be needed.
Finally, the fact that most industrial heat loads are constant throughout the year
(or peak in the summer and fall for some food processing plants) gives a decided ad-
vantage to industrial solar systems over space heating systems. The solar heat can be
fully utilized the year around for the industrial system. Domestic solar heating systems,
on the other hand, are sized to provide substantial output during the winter months,
which means that most of the heat collected during the summer must be thrown away.
In Fig. 36.2 we show typical values for monthly energy contributed by a solar space
heating system and by an industrial solar system, both normalized to unit collector
area. The winter space heating curve lies above the process heat curve because a col-
lector space heating system is tilted for maximum collection in winter whereas an in-
dustrial collector system is tilted to maximize annual collection of heat. The area under
the industrial curve is typically a factor of 1.5 to 2.5 greater than under the space heating
curve.
Present industrial practice is often to provide higher-quality process heat than is actually
required in the process. For example, steam is often generated in a boiler at 175 to
300 Applications of Solar Energy
250°C (350 to 500°F) and then piped to the process area where some fraction may be
used at the top temperature while the remainder is used to heat unpressurized water.
The efficiency of fuel-fired boilers is insensitive to working temperature, and hence
this is a reasonable procedure. A solar process heat system should be designed to satisfy
the needs of a particular process. Delivery of solar heat at higher temperatures requires
more sophisticated collector systems so that a matching of the process with the simplest
and most cost-effective solar system is of utmost importance.
There have been several recent studies of the end-use temperature distribution of
process heat in U.S. industry [1-3] . Table 36.1 shows the distribution of process steam
temperatures according to one of these studies [3] . Note that, of the total 11.3 Q
(12 X 10" kJ) of process steam, about 7 Q is required at temperatures below 175°C.
In addition, Fig. 36.1 shows that about 1 Q of direct process heat is used below 175°C.
This total of 8 Q represents about 10 percent of total U.S. energy useage or about 1.3
billion barrels of oil equivalent per year.
Table 36.1 shows that there is little potential for solar energy in the primary metal
industry but a great potential in the chemical, petroleum, food processing, and paper
industries, where major amounts of process steam are used at temperatures below 175°C.
The other two studies [1,2] confirm this finding.
Each industrial process and each plant has unique features, and a careful design
study must be made to adapt solar energy to the process requirements. Because of the
daily and seasonal cycles of solar radiation and its intermittent character, solar systems
must be combined appropriately with fossil-fuel backup systems so that the process
line is never interrupted because of lack of solar heat.
In general, there are three possible modes for the application of solar energy to an
industrial process [1] :
The first mode is quite attractive because the cost of a storage system can be elimi-
nated. The maximum heat output of the solar system is utilized during the daytime when
the sun is shining. A conventional backup system provides heat at other times. In this
mode, a maximum of 30 percent of the total process heat load could be supplied by the
solar system.
The second mode, including energy storage, would be required if a greater fraction
of the total process heat load were to be supplied by the solar system. Also, if the plant
operates only 5 or 6 days per week, storage would be desirable so as not to waste one-
seventh to two-sevenths of the available solar heat. Even in a solar system without energy
storage, a small amount off buffer storage would still be desirable so as to make the inte-
grated fossil fuel/storage system less perturbed by rapid fluctuations in solar radiation.
In the third mode, the solar system would, for example, preheat boiler feed water.
The advantage of this operating mode is that there are preheat possibilities in almost all
process applications. Also, solar collectors operate at higher efficiencies at lower tem-
peratures. Another possibility is to produce low-pressure solar-generated steam, which
could then be compressed and superheated by an engine-driven compressor.
Industrial Process Heat 301
Table 36.1 Estimated 1975 Uses of Process Steam (at Various Temperatures) in
Industrial Sectors
Total Steam Use
Temperature Range Total Steam Use of Sector
Industry (°C) (°F) (% of Sector) (1012 Btu)
Primary metals >250 >500 —100 390
The U.S. Department of Energy (DOE) is funding the design, construction, and testing
of a number of prototype systems supplying solar heat to industrial processes in various
regions of the United States [4] . The solar collectors range from 250 to 2000 m2 in area
and will provide solar-heated hot water, hot air, or process steam to selected industrial
plants. These prototype systems, in most cases, contribute only a small percentage of
the total process heat required in the plant. The purpose in building these first-generation
industrial solar systems is to discover what problems exist in installing and operating
large-area solar systems at industrial plants and find solutions to these problems that will
lead to the demonstration of engineering and, it is hoped, economic viability.
The essential characteristics of the 11 projects now underway are summarized in
Table 36.2. A brief description of each project is provided below.
The AAI Corporation has designed and constructed an 856-m2 solar system at a
concrete block curing plant in Harrisburg, Pennsylvania. The X24 concentrating collector
is of the "Venetian blind" type with fixed absorbers, designed by AAI. Hot water is
provided in the range of 60 to 80°C for curing concrete blocks. A toroidal underground
curing tank, "rotoclave," also serves as the storage tank for the solar system. Figure 36.3
is a view of the completed system mounted on the roof of a concrete block storage shed.
This project has the advantage of built-in storage and requires only moderate tempera-
tures for the hot water.
The Acurex Corporation has designed and constructed a roof-mounted solar system
to provide hot water for washing empty soup cans at a large plant in Sacramento,
Figure 36.3 The AAI solar system, which provides 80°C hot water for the curing of
concrete blocks in a Pennsylvania plant.
Industrial Process Heat 303
Figure 36.4 View of the Acurex solar system, consisting of both flat-plate and parabolic
trough collectors, mounted on the roof of a soup company in California. This system
supplies 90°C water for the washing of soup cans.
California. This system consists of 360 m2 of flat-plate collectors and 250 m2 of para-
bolic trough concentrators of Acurex design. Water at ambient temperature is first heated
by the flat-plate collectors and then by the concentrators, typically to a temperature of
90°C. Storage is provided by a 64-m3 (17,000-gal) water tank. Figure 36.4 is an overall
view of the system. This project demonstrates several features: two types of collectors
in series, storage, roof-top mounting, and a heat exchanger.
The General Electric Company has designed and constructed a 620-m2 solar system
to provide water at 80 to 120°C for textile dyeing at a plant in LaFrance, South Carolina.
The evacuated-tube collectors are of General Electric design and are mounted on a ground
site adjacent to the plant. Storage is provided by a 30-m3 (8000-gal) water tank. An
important aspect of this project is the opportunity to test a new collector design.
Members of the engineering faculty at the California Polytechnic State University
have designed and supervised the construction of an 1800-m2 , site-built solar air heating
system for the drying of raisins and prunes at a facility in Fresno, California. The system
delivers heated air at 65°C to one of 12 tunnel dehydrators. There is a 340-m3 (12,000-
ft3 ) rock-bed storage unit. A heat exchanger wheel that conserves exhaust heat is an
important aspect of this project.
The Lockheed Missiles and Space Company has designed and constructed a 230-m2
flat-plate solar system for the drying of lumber at a lumber kiln in Canton, Mississippi.
Water is heated in the flat-plate collectors and is circulated through a water/air heat
exchanger to provide hot air. Storage is provided by an 18-m3 (4800-gal) water tank.
UJ
O
41.
Table 36.2 Essential Characteristics of Department of Energy Solar Process Heat Projects Under Construction or in Operation
Estimated Values
Installed Annual
Maximum Collector System Heat Heat per
Temperature Collector Area Cost Supplied Unit Area
Contractor Project Fluid °C (°F) Type (m2 ) ($/m2) (GJ)a (GJ/m2 -yr)b
AAI Corporation Concrete block Water 80 (180) "Venetian 860 520 1600 1.9
curing blind"
concentrator
Acurex Washing food Water 90 (195) Flat-plate 360 930 2300 3.8
Corporation cans parabolic 250
trough
General Electric Textile dyeing Ethylene 120 (250) Evacuated 620 930 1500 2.4
Co. glycol/ tube
water
California Poly- Prune and Air 65 (150) Site-built 1800 300 2400 1.3
technic State raisin drying flat-plate
University
A2Jau] m os Jo suoppon ci dv
Lockheed Kiln drying of Water 70 (160) Double- 230 1350 1000 4.3
Missiles and lumber glazed
Space Co. flat-plate
Teledyne-Brown Soybean Air 70 (160) Single-glazed 1220 600 3900 3.2
Engineering drying flat-plate
00.1d I Pp 1S rIp Ul
Honeywell, Inc. Fabric drying Water 195 (380) Parabolic 770 830 1300 1.7
(230 psig) trough
2POH SS 3
Trident Onion drying Water 100 (210) Evacuated 550 910 2400 4.4
Engineering tube
Assoc., Inc.
Jacobs Engin- Commercial Water 215 (420) Parabolic 600 860 1260 2.1
eering Co. laundry trough
Acurex Gauze Water 215 (420) Parabolic 1070 900 1400 1.3
Corporation bleaching trough
General Electric Orange juice Water 230 (450) Evacuated 930 850 3400 3.7
Co. processing tube and
reflectors
Calculations showed that it was more cost effective to heat water and use a heat ex-
changer than to heat air directly.
The Teledyne-Brown Engineering Company has designed and constructed a 1220-
m2 flat-plate solar air heater system for drying soybeans at a plant in Decatur, Alabama.
No storage is provided; the solar heated air is fed directly into the inlet of the corn-
bustors of the existing oil-fired dryers. A serious problem with this project is that soy-
bean dust falls on the collector glazings and is difficult to remove.
Honeywell, Incorporated, has designed and constructed a 770-m 2 solar system to
provide 160 to 195°C steam for drying textiles at a plant in Fairfax, Alabama. The dry-
ing process involves passing wet yarn over the outer surface of a steam heated can dryer.
The collectors are parabolic troughs of a Honeywell design. No storage is provided.
The Trident Engineering Company has completed the design and construction of
a 550-m2 evacuated-tube collector system to be used for onion drying at a plant in
Gilroy, California. The collectors heat water to 100°C which is then put through a
water/air heat exchanger to provide hot air to the dryer burner inlets. No storage is
provided.
The Jacobs Engineering Company has designed a 600-m2 solar system using the Del
parabolic trough concentrators to produce 0.76 MPa (110 psig) steam for processing
laundry and drycleaning in Pasadena, California. The pressurized water in the closed
collector loop will be circulated to a 1.1-m3 (300-gal) high-temperature water storage
tank, the steam generator, or the hot water heat exchanger. The collectors are located
directly above the laundry on an independent structural steel framework.
The Acurex Corporation has designed a 1070-m2 solar system using the Acurex
3001 parabolic trough concentrators. The collectors are mounted on the ground adjacent
to a Sherman, Texas, pharmaceutical plant. The heated water from the collectors is fed
to a 19-m3 (5000-gal) flash boiler/storage tank. The output from the flash boiler is
175°C (350° F) steam, which is directed to the plant's main steam supply line and is
utilized 365 days a year.
The General Electric Company has designed a 930-m2 ground-mounted solar
system for an orange juice processing plant in Bradenton, Florida, using the G.E. TC-
300 evacuated-tube cusp collector. This is a 2.9X nontracking concentrator that requires
four seasonal tilt adjustments per year and is made of materials that can withstand
stagnation temperatures. A closed-loop water system supplies energy to heat the fluid
in a steam generator, which in turn supplies 180°C (360° F) steam to a pasteurizer unit
during the week and 125°C (260°F) steam to a glycol evaporator on the weekend. The
system has no separate thermal storage tank.
As of September 1978, the U.S. Department of Energy has placed seven additional
contracts for the design of intermediate temperature solar steam prototype systems. The
overall objective of the program is to design a system with approximately 1000 m2 of
collector area to provide industrial process steam within a temperature range of 150 to
300°C. Because these activities are just underway, only limited information on them is
available at this time. The seven projects are as follows:
The large proportion of projects at 175 to 180°C is reasonable, since a large per-
centage of the steam systems in this country use 1 MPa (150 psi) steam. Industry has
settled upon 1 MPa steam as a standard, regardless of the lower end-use temperature,
because there is very little increase in efficiency at lower temperatures when the heat
source is a 2000°C flame. It is easier to interface the solar system with the rest of the
plant by simply connecting directly into the main system header, so the majority of these
systems are naturally designed to provide —1 MPa steam.
As noted in the introduction to this chapter, there are a number of significant differences
between domestic heating systems and industrial process heat solar systems. The differ-
ences can impose limits that must be observed by planners and designers if solar systems
are to be marketed successfully. These limits fall into two broad categories: economics
and land use. The first of these is generally applicable to all industrial process heat
systems. The second is site specific and must be evaluated for each individual situation.
The economic requirements for any capital equipment in a commercial environ-
ment represent the biggest factor affecting the application of solar energy to industrial
process heat today. Almost no industry has an allowable payback period longer than 7
years and most require 5 years or less. This is generally regarded as an absolute require-
ment. Little allowance is made for public relations benefits or the possibility that fuel
costs 10 or more years in the future might rise faster than present rates. If a solar system
cannot show the proper payback, it will not be installed. Competition in the market
place forces such decisions.
Assuming that fuel costs = $5 .00/GJ ($30/bbl of oil or 50¢/therm of natural gas)
and assuming that a typical solar system in the southwestern United States can provide,
at most, 4.5 GJ/m2 (0.4 MBtu/ft2 ) per year, we can see that the value of such a system
is only about $20/m2 ($2.00/ft2 ) per year. Applying the 5-year payback limitation to a
system with this return, the company has to be able to install the entire system (including
308 Applications of Solar Energy
pumps, piping, valves, control systems, mounting structures, and interfacing connec-
tions) for $100/m2 ($10/ft2 ). This is admittedly a simple analysis and income taxes
and investment tax credits do affect these numbers, but the numbers one derives from a
detailed economic analysis of the individual plants come reasonably close to this esti-
mate. (See Chap. 46 for a full discussion of industrial process heat economic method-
ology.) It is apparent that costs of existing solar systems that now sell for $300 to
$1000/m2 (see Table 36.2) are going to have to be drastically reduced if they are to
compete with fossil fuels.
Most plants will profit far more by installing energy conservation equipment than
they will by converting part of their heat load to solar. Such conservation measures can
range from very simple procedural changes, such as restricting open warehouse doors to
the downwind side of the building, to very extensive modifications such as the installa-
tion of heat exchangers to recover waste heat. Many companies are installing additional
insulation and have instituted a regular maintenance program. (One company cut its
steam consumption 5 percent simply by checking its pipes for leaks.) Cogeneration is
one conservation technique that is especially promising, even though it detracts from
possible solar applications.
Cogeneration is applicable at industrial plants with both a sizable electrical power
demand and a large process heat load at low temperatures <260°C (<500°F). High-
pressure steam is produced in the plant's boilers and sent to a turbine which drives a
generator. Rather than expanding the steam down to a low pressure where it is con-
densed and the excess heat thrown away (typically two-thirds to three-fourths of the
original thermal energy supplied by the boiler), the steam is extracted from the turbine
at the pressure required by the process. Thus, the "waste heat" is put to use and the
overall plant energy efficiency goes up. Where the electrical load is very high, additional
power must be obtained from outside sources. On the other hand, a plant with a rela-
tively small electrical load must bum some fuel just for process heat and usually will
not bother with cogeneration.
Until high-pressure steam, 4 MPa (600 psi) and up, can be produced economi-
cally from solar energy, or photovoltaic power sources become much cheaper, solar
energy cannot compete with cogeneration when a plant's heat and electrical loads are
large. Cogeneration, like solar energy, represents a large technological step for many
companies. To overcome this problem, some companies are colocating their plants
with local utility-owned electriclal power plants and purchasing both electricity and
steam from the utilities.
After economic factors, the most obvious limitation on the installation of
industrial solar systems is a lack of suitable land area to locate a collector array. Since
cost is a major factor, it is obvious that a collector array must be located very close to
the process to cut down the cost of the piping. The problem is that many plants do not
have access to adequate land adjacent to their operation.
To provide a clear picture of this problem, consider the land area required for a
plant that uses 60 MW of heat and is located in the western United States where the daily
beam radiation is typically 28 MJ/m 2 (2500 Btu/ft2 ) per day during the summer. We
assume, optimistically, that a tracking solar system can delivery 50 percent of this energy
to the user after going through heat exchangers, distribution piping, and storage systems,
and we assume a 50 percent packing factor of the collectors on the land, allowing for
Industrial Process Heat 309
roads, pipeways, and shadow effects from adjacent collectors. Then we collect about
7 MJ/m2 (600 Btu/ft2 ) of land area per day. We further assume that this 7 MJ/m2 is all
delivered during a 10-hour period and find that our solar collector array can deliver about
2 MW/hectare (2.5 MBtu/h per acre) of land area. This equates to about 2.8 tonnes of
steam per hour per hectare (2500 lb/h-acre) of land. The actual systems will probably
provide less than half as much energy per unit land area.
Now refer back to our 60-MW example load. Even with the optimistic estimate,
this installation will require 30 hectares (75 acres) of land for solar collectors, just to
meet its heat load when the sun is shining in the summertime. If storage is contemplated
to extend the solar energy supply to a full 24-hour day, these land area requirements
will more than triple.
The obvious conclusion is that solar energy does have the potential to supply sig-
nificant quantities of process heat to industries located in areas such as mining and some
chemical and food processing plants, where relatively low-value land is available. In
urban areas, land is generally too valuable to permit its dedication to a solar collector
system that will provide an annual income equivalent of only $20/m2
Urban plants that have large areas of flat roofs may be able to get double duty from
their land by mounting the collectors on the roofs. This solution can be expensive, how-
ever, as most industrial roofs are not designed to accommodate any extra load.
Rebuilding or reinforcing roofs is an expensive proposition and presents a severe handicap
to roof-mounted systems.
A third category of factors affecting the potential for solar industrial process heat
is the design and operational features of the specific operation. This category may contain
both assets and limitations.
Limits on land area and the process temperatures have several implications on the
nature of the solar system. Most of the DOE industrial process heat prototype systems
are being set up to operate in parallel with the rest of the plant. For example, they will
heat 1 dye beck out of 10, 1 onion dryer out of 6, or 1 raisin dryer out of 12. This may
be appropriate for these projects in order to show the performance to be expected from a
solar system. Parallel operation may not be the best approach for plants with limited land
area and high-temperature processes. If the size of the collector field must be limited, a
higher thermal efficiency and a better return on investment may be obtained by operating
the solar collector as a preheater in series with the existing process heat system. Operating
the collector field at a lower overall temperature will increase its efficiency, regardless of
the type of collector, and will reduce heat losses from the distribution piping. (It may
also permit the use of less expensive collectors.) An exception is provided by plants with
boilers having large economizer sections sized to operate with cool feed water. Raising
the temperature of the feed water will lower the boiler's efficiency and end up displacing
a minimal amount of energy. For these plants, parallel operation may be best, regardless
of available land area.
As noted above, steam is the most popular working fluid to move heat from boiler
to process when fossil fuel is the energy source. In the case of solar systems, however,
there are three alternatives that should be considered before establishing the system
working fluid.
1. Make steam at the same pressure and temperature as the existing system; or, if
land is limited as discussed above, preheat the feed water going to the boiler. This has the
310 Applications of Solar Energy
distinct advantage of being able to connect to the existing system with very little addi-
tional modification. It has a disadvantage in that there is a limited storage capability. A
system connected in series will have trouble responding to large load changes; a series-
parallel arrangement may be preferable.
2. Heat some other, nonvaporizing fluid such as oil or pressurized water. The ad-
vantage here is that energy is easier to store as sensible heat in a liquid than as latent heat
in a vapor because of the smaller volume required. Larger heat exchangers will probably
be required with this option, and it is likely that other plant modifications will also be
required. Also, the working fluid must be carefully chosen to assure compatibility with
the process fluid.
3. Heat the process fluid directly. This option allows the total system to operate
efficiently because there are no intermediate heat exchangers. Further, if the object is
to provide only sensible heat to the process fluid rather than latent heat (as in evapora-
tion cycles), then the collector will operate at a lower average temperature if it heats the
process fluid directly rather than if it makes steam. One additional benefit is available
if the process fluid is used. Energy can be stored directly in the process loop in the inter-
mediate holding tanks that are associated with most operations. A difficulty with this
approach arises if the process fluid cannot tolerate either the random heating rates pro-
vided in solar collectors or the possible stagnation temperatures that can occur if circu-
lation equipment fails. Fluids that are especially corrosive or have a high fouling rate
would not be circulated directly through a collector field.
The choice of process fluid is also constrained by its acceptability to the plant's
operating personnel. Unfamiliar systems are not likely to be accepted and will not re-
ceive the wholehearted maintenance support that old, known systems do. The solar
system's impact on the plant maintenance program is a major consideration. Any sys-
tem should be designed to operate with a minimum of maintenance, but large plants
with maintenance staffs are in a better position to absorb the additional load imposed
by a solar system than smaller ones. Solar systems can even benefit a continuing main-
tenance program if they are installed in parallel with other heating equipment. With
sufficient solar capacity, some of the conventional heating equipment can be scheduled
for preventive maintenance during the day shift, usually at a lower labor rate. The solar
system, of course, is always shut down at night and thus is available for maintenance
without cutting into its productivity.
The conclusion to be drawn is that it will be difficult to sell a package solar system
to a given industrial plant the way package boilers are sold today. The economic and
thermal efficiencies of solar systems are much more sensitive to the specific conditions at
individual plants than conventional boilers; therefore, the designer must give careful con-
sideration to the integration of the solar system into the plant operation. Any attempt
to design a solar industrial process heat system without reference to specific plant para-
meters is likely to end in failure.
REFERENCES
37.1 INTRODUCTION
In recent years, photovoltaic (PV) power generation has been receiving considerable
attention as one of the more promising energy alternatives. The reasons for this rising
interest lie in PV's direct conversion of sunlight to electricity, the nonpolluting nature
of the PV conversion process, and PV's nondependence on fossil and nuclear fuels.
To date, the widespread use of PV generation has been hampered by economic
factors. Heretofore, the low cost of conventional energy sources has obviated the develop-
ment of a broad-based PV technology. At the present time, PV generation can be justified
only for special situations, mostly for remote sites where utility lines or other conven-
tional means of furnishing energy may be prohibitively expensive. The relatively un-
mechanized production processes employed to date exacerbate the poor competitive
313
314 Applications of Solar Energy
PHOENIX,
0 / 1953
_1
0
In
z
0.5 _J
< 5
z
NJ
BOSTON,
C—C 1960
0
0
I
12 12 12 0
MIDNITE NOON MIDNITE J F MAMJ J ASOND
•E
h
0
z
_J
Z 5
0
N/
I
I-
>—
0
7c J F M A M J J ASOND
0
0
J F MA M J J A S 0 N D
)
basis, the daily mean available total horizontal solar energy will range from 5.0 kWh/m2
in a clear, lower-latitude location such as Phoenix to 3.5 kWh/m2 in a stormy, higher-
latitude location such as Boston. Direct normal insolation, important for tracking and
concentrator systems, will exceed total horizontal insolation by about 20 percent in
midsummer to 50 percent in midwinter. This is, of course, greatly dependent on the
site location.
Figures 37.1c and 37.1d show data for daily total horizontal insolation for the
Phoenix and Boston locations based on analysis of hourly weather data of representative
years. These plots roughly show how many days out of each month can be expected to
have insolation of some value or greater. Thus, in Boston during May, the plot shows
a probability of 50 percent (or 15 out of 30 days) that the daily insolation will be
55 kWh/m2 or greater. Knowing the relationship between daily insolation and PV
generation for given system types (see Sec. 37.5), the probability that a particular array
size can furnish a particular energy demand can be established. A comparison of curves
(c) and (d) shows that the solar availability spread for Phoenix is narrow, indicating
relative constancy in weather conditions, whereas the broader spread for Boston indi-
cates larger weather variations. Plots of this type for other climatic regions are available
from the authors of this chapter.
The annual total energy availability based on recent 20-year periods shows one
sigma variations from a mean insolation year ranging from less than 2 percent for
Phoenix to less than 5 percent for Boston. Maximum clear-day daily energy availability
with two-axis tracking systems following the sun can exceed 12 kWh/m 2 . The solar
availability time profile as tempered by meteorological conditions, location, collector
orientation, and tracking determines the photovoltaic power-generation profile.
The curves of Fig. 37.2 indicate the power-generation characteristics for a single solar
cell or, by suitable scaling, for a series-parallel matrix of solar cells. Curve (a) shows the
typical variation of voltage with current for a solar cell under constant illumination at
constant temperature. The curve is developed by varying the load impedance across the
cell and measuring the corresponding voltage and current as shown on the sketch near
curve (a). The maximum power point (Pmax s-
represents the conditions for which the
voltage-current product is maximum. The curve designates Voc (open circuit voltage);
Vmp (max. power voltage); Isc (short-circuit current); and Imp (max. power current).
The fill factor for a solar cell is defined by:
Pmax
Fill factor — Imp X Vmp
ISc X VOC 1sc X Voc
Solar cell designers strive to increase fill factor values, to minimize internal losses. Typical
fill factors fall in the range of 0.70 to 0.78. Within the insolation and temperature ranges
of interest, curves (b) and (c) illustrate the following important relationships. 'Sc is
directly proportional to insolation; with slight deviation IMP is also proportional to
insolation; both Isc and IMP are relatively insensitive to temperature. Voc and VM P
Electric Power Generation: Photovoltaics 317
INSOLATION
VOLTAGE, V -WAN..
iVV aw
CURRENT, I i
SC
(a)
HI INSOLATION
HI TEMP
I LO IN
SC
LO TEMP
CURRENT VOLTAGE - '1
IMP -.,..
\.....s• VOC
7/ -.,... V
MP
(curve c) are inversely proportional to temperature; both are relatively insensitive to the
insolation level. For a given level of insolation, curve (c) shows that reduced operating
temperatures are significant for achieving higher performance.
Some typical numbers illustrate these characteristics. For a single crystal silicon
cell of 12 percent efficiency, the DC maximum power output at a peak insolation of
100 mW/cm2 and a 28°C cell temperature would be 12 mW/cm2 of cell area. An open-
circuit cell voltage of 0.60 VDC would drop to 0.46 VDC at the maximum power point,
and short-circuit current of 27.2 mA/cm2 would be 26.1 mA/cm2 at the maximum
power point. The efficiency would fall by 0.05 percent per °C cell temperature rise,
that is, 11 percent efficiency at a cell temperature of 48°C. A free-standing air-cooled
flat-plate assembly of such cells with a passive thermal design would experience a tem-
perature rise of about 30°C above ambient in still air. Wind velocity of 5 m/s would
limit the rise to about 10°C above ambient.
Solar cell panels usually consist of series-parallel interconnected solar cells bonded to
supporting substrates and encapsulated with transparent materials to provide environ-
318 Applications of Solar Energy
BYPASS DIODE
ELEMENTARY SERIES/PARALLEL
SEGMENT - E.G., CELL, PANEL,
CONCENTRATOR
(-) RETURN
mental protection. Solar arrays consist of multiple panels (or tracking concentrators)
appropriately deployed, supported, and interconnected.
Solar arrays are electrically defined in terms of "strings" or "circuits", each of
which contributes a portion of the total current output at some nominally specified
array voltage. Each string may consist of series-connected panels or of a series-parallel
matrix of panels. As shown in Fig. 37.3, each string is usually terminated with a single
or paralleled group of protective diodes inserted between the string positive termina-
tion and the solar array bus. The diodes protect the string by preventing the possible
backflow of current that might arise under certain shading or failure conditions.
Bypass diodes, also illustrated in Fig. 37.3, are another protective feature common-
ly used on solar arrays. These diodes are wired in parallel around segments of a string and
serve to limit the level of back bias voltage that could be developed across segments as
a result of inadvertent shading or cracked solar cells. They are-also useful in permitting
continued string operation in the event of open-circuit failures within the protected string
segment. Such protective devices are essential for line-focusing concentrators that
employ single-cell strings.
Solar array voltages are established on the basis of the application and the electrical
insulation capabilities of the basic panel or concentrator designs. Small remote applica-
tions commonly employ voltages in the range of 6 to 30 V, often determined by the
characteristics of batteries connected in parallel with the solar array. Larger systems strive
to operate at higher voltage to minimize wiring losses. Residential solar arrays of perhaps
50 to 100 m2 in size (approximately 5 to 10 kW peak output) might be designed for
200 VDC. Larger installations, possibly for commercial or institutional use of about 200
to 500 kW peak might operate in the range of 400 to 600 VDC. Central power plant
solar arrays would be designed for much higher voltages. Conceptual designs for such
purposes have considered array voltages of 10 kV. Such high voltages require that panels
Electric Power Generation: Photovoltaics 319
A wide variety of methods are available for matching the output of PV solar arrays to
specific load needs. These range from solar arrays connected directly to loads which can
accept the variability in output associated with changing insolation, to more demanding
loads which require extensive processing of the PV power, including the use of storage,
voltage or current regulation, and conversion to other required forms of power. Largely
as a result of the space program, considerable experience has been accumulated for
handling the variety of situations likely to be encountered in utilizing PV power. Numer-
ous forms of power conditioning have been developed, though these are not generally
available in commercial, inexpensive configurations.
Power conditioning technology per se is not considered to be a barrier to the wider
usage of PV systems—the circuit techniques and devices are readily available. The problem
is more the lack of an existing market justifying the use of cheaper mass production tech-
niques. There is also the broad technology of non-PV power conditioning, which could
be easily adapted to PV needs once the market conditions are favorably established.
Most of the present-day PV installations for remote terrestrial applications employ
lead-acid battery energy storage. As shown in Fig. 37.4, the battery is usually connected
CHARGE LOAD -)
CURRENT CURRENT
SOLAR ARRAY
STR NG, I2-VOLT
30 SERIES SOLAR LOAD
BATTERY
CELLS EACH
LOAD LOAD
20 IMPEDANCE 20
IMPEDANCE
ARRAY OPERATING
POINT LOAD DISCHARGE
or CHARGE CURRENT
VOLTAGE,
-.43.2 VOLTS 10 DISCHARGE
I0 VOLTAGE
VOLTS VOLTS VOLTS
LOAD
CHARGE POINT
CURRENT
CURRENT CURRENT
(a) HI ILLUMINATION, (b) LO ILLUMINATION,
BATTERY CHARGE BATTERY DISCHARGE
in parallel with the solar array whose output furnishes both load and battery charging
power. At a nominal solar cell voltage of 0.40 V, a 12-V battery would operate in con-
junction with a solar array designed with strings having about 30 solar cells in series.
Charge-to-discharge voltage limits for lead-acid batteries range from about 13.2 V (full
charge) to about 10.8 V (full discharge) for a six-cell battery. These levels are overlaid
on the I-V plots in Fig. 37.4 and show how they affect the solar array operating point.
As long as the number of series solar cells is reasonably selected to match the nominal
battery voltage, only a small penalty in array output is associated with the charge-to-dis-
charge variations. The average array temperature affects the match to some degree, and
may be taken into account in determining the number of series solar cells. Higher tem-
peratures would require a larger number of series solar cells, and vice versa. Adjustments
would be made according to a sensitivity of about 0.4 percent increase in the number of
series solar cells for each °C increase in average array temperature. Section 37.3 provides
estimates of the nominal operating cell temperature (NOCT) for representative flat-panel
solar arrays.
The simple battery tie-in approach described above operates in a straightforward
manner. The battery is charged during periods of excess PV generation and is discharged
during nightime or periods of high demand. Continuity of service is maintained as long as
the accumulated PV-generated energy is equal to or in excess of demand and as long as
the battery is not fully discharged. In this latter regard, several practical considerations
must be taken into account. First, excessive discharge can damage batteries, and therefore
steps must be taken to prevent this possibility. Various low-voltage cutouts have been
used for this purpose. Their activation, of course, results in the interruption of service,
a factor that may lead to the need for other backup power sources, depending on load
criticality. Of equal importance is the prevention of overcharge, a condition that could
severely damage or destroy batteries. Here too, cutouts and power-shunting devices are
employed to divert the flow of PV current, once appropriate sensors detect that a full-
charge state has been achieved.
More elaborate regulation techniques may be employed to improve PV system
performance over that possible with batteries alone. Some of these methods are described
briefly below.
1. Voltage and current regulation of PV systems is simply achieved by introducing
regulation devices at appropriate points within the system. For the example shown in
Fig. 37.5, a current regulator controls the rate of battery charge while the voltage regu-
lator receives the unregulated array/battery input, with voltage variations of about 30
percent, and regulates it to the limits required by the particular load, typically of the
order of several percent. Several basic regulator types serve a wide range of operating
conditions. Series dissipative regulators are simply variable-resistance devices that absorb
the power associated with the input/output voltage difference and the load current
(identical to the input current). The efficiency of such regulators is defined simply by
the ratio of the output to input voltage, ignoring a small amount of power required for
sensing and control. Such regulators are cheap, uncomplicated devices, but they do have
the constraint that the unregulated input must always be larger than the regulated output.
Pulse-width-modulated (PWM) regulators employ high-frequency switching techniques
and essentially serve as DC-to-DC transformers. Typical devices include input filter,
switching, transformation, and output filter stages. The dominant feature of these devices
Electric Power Generation: Photovoltaics 321
SOLAR VOLTAGE
ARRAY REGULATOR "'LOAD
CURRENT
REGULATOR
PWM
SOLAR POWER VOLTAGE —.LOAD
ARRAY Di REGULATOR
TRANSFER
REGULATED BUS
SOLAR
ARRAY -N LOAD
associated charge and discharge regulators. High system efficiency is achieved by fur-
nishing solar array power directly to the load without the use of in-line regulating ele-
ments. Voltage regulation is maintained by either the shunt or discharge regulator
according to system logic, which considers the relative availability of solar array power
and load demand. Such approaches with several variations have been integrated suc-
cessfully into spacecraft systems.
4. DC/AC inverters are an important class of power conversion devices, particular-
ly as PV systems are applied to terrestrial-type loads involving utility tie-ins. Self-
commutated inverters are designed with internal circuitry for generating the frequencies
associated with AC operation. Line-commutated inverters depend on utility-line AC
characteristics to provide driving signals required for inverter operation. Figure 37.8
illustrates several alternative arrangements for integrating PV power with utility power.
The system shown in (a) is designed to displace as much utility power as possible in
UTILITY
BACKUP
ONE- WAY
POWER FLOW
SOLAR
ARRAY --N DC /AC
INVERTER
LOAD ( a ) SYSTEM WITH STORAGE,
NO FEEDBACK
UTILITY
BACKUP
1 :TWO - WAY
POWER FLOW
SOLAR DC /AC ( b ) SYSTEM WITHOUT STORAGE,
ARRAY INVERTER LOAD FEEDBACK TO UTILITY
meeting the needs of the specific load. The system uses batteries to store excess PV
generation. The stored energy is dispensed to the greatest extent possible during night-
time periods. The inverter is line-commutated to be in synchronism with the utility,
and meets all load demands up to the limit of its power rating or up to the limit of
battery energy availability. Should these limits be exceeded, the hard utility tie-in is
always available as backup. This latter feature is particularly important for transient
startup conditions (motors, etc.). Providing the PV inverter with such startup
capability might be prohibitively expensive. The arrangement shown in Fig. 37.8b is
designed to permit the feedback of PV power to the utility grid. With this feature
battery storage is unnecessary, since the PV feedback power is absorbed by loads else-
where along the utility grid. In essence, the PV system becomes another generator con-
tributing to the bulk energy of the utility system. Maximum power tracking techniques
are easily applied to this concept by varying the impedance imposed by the DC/AC in-
verter. The elimination of batteries and their associated inefficiency makes this feed-
back approach attractive, although the economics must be evaluated carefully in terms
of the business arrangement between the PV system owner and the utility.
E. Other Considerations
Economics
Numerous system studies have been performed during the last few years to examine
the economic viability of photovoltaic power generation applications. These have con-
sidered both the projected goals for solar cell costs and the breakeven costs with compe-
titive energy forms in specific applications. Based on the reasonable assumptions necessary
for such studies and the use of life-cycle costing, it appears that for many current applica-
tions remote from conventional power sources, PV power generation is presently viable.
PV generation is potentially competitive for widespread applications if long-term cost
goals for the PV systems are met.
During the 1970s costs for flat panels have been reduced by one to two orders of
magnitude to a range of $10 to $20 per peak watt.* Goals for the early 1980s of $2 per
peak watt seem very likely to be met. Goals for the late 1980s of 50cr per peak watt pose
greater difficulty, as do subsequent reduction goals. To reach these goals technical im-
provements, cost reductions, and creation of a market are required.
Concentrator systems are also targeting a set of goals similar to those for flat
panels for the 1980s. The substitution of reflective or refractive optical systems for cells
shifts the emphasis from improvements of cell costs to improvements of cell efficiency.
Although both types of improvements are required for all system types, cell efficiency
improvements directly reduce all area-related costs (the bulk of the costs for systems of
several kilowatts and larger) for both system types. Cell cost improvements tend to im-
prove the position of flat-plate systems vis-a-vis concentrator systems, since cell costs
constitute a much larger part of flat-plate system costs.
Remarkable strides have been made in the reduction of cell costs and improve-
ment of cell efficiency in recent years. However, it is becoming apparent that, in addition
*1975 dollars are used as the economic measure of worth throughout this chapter.
324 Applications of Solar Energy
to cell improvements, significant strides in cost reduction will also be required for the
other mechanical and electrical elements in a photovoltaic system.
Storage
Remote and free-standing systems not connected to backup sources of power
require storage to provide power for load demands when direct solar power is not
available. Storage capacity increases the system costs, and creates its own energy round-
trip efficiency losses, but is a necessary element in isolated systems. Many studies of
applications involving utility backup power, including applications of the photovoltaic
plant as a generating unit in the utility, have drawn conclusions that dedicated storage
for the photovoltaic generation plant is uneconomical and not technically required. For
large utility systems that depend on a variety of generation sources, the question of
storage viability should be examined in the context of total utility needs rather than
those of the individual generation sources.
Utility Interface
Atmospheric Pollutants
Dirt and grime deposited on installed arrays reduce the power output through
partial obscuration of the solar radiation input. Recent observations of installed flat-
plate arrays in many locations of the United States show that this power loss shortly
after installation ranges from 10 to 30 percent. The lower figure is representative of
rural, relatively clean environments, and the higher loss is representative of major metro-
politan areas. Although cleaning the arrays will restore the full power output, in the
long run the frequency of cleaning versus power loss economic trade-off adds an addi-
tional economic hurdle for photovoltaic systems to overcome.
Institutional Issues
Photovoltaic systems pose many of the same issues as other solar energy systems.
These include "sun rights," property taxes, system financing, insurance, building and
construction codes, system safety, and so on (see Chaps. 47 and 48).
A. Solar Cells
Silicon solar cells for application on flat-panel arrays are generally circular in shape
because of the economy associated with slicing the wafers directly from Czochralski-
grown crystals. Wafer diameters commonly range from 50 to 100 mm; the geometry
of the contact metalization pattern varies with manufacturer, but is usually character-
ized by relatively wide grid line spacing and redundant main collector bars. The use of
silver-printed contact metalization is also common for cells of this type.
Recent advances in laser scribing and the growth of sheet single-crystal silicon may
make it economically possible to utilize rectangular or hexagon-shaped cells in flat-panel
solar arrays. The efficiency of these cells generally ranges from 11 to 14 percent at an
insolation of 100 mW/cm2 when measured under the following atmospheric conditions
700-
600 —
500 —
400—
Zz.
E
— 300 —
z
w
cc
cr
200 —
c.)
100
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
VOLTAGE (VOLTS)
Figure 37.9 I-V characteristics for a typical 53-mm -diameter terrestrial solar cell.
326 Applications of Solar Energy
A reference cell that has been calibrated as described in Ref. 1 is generally used to
establish the irradiation level for cell or array testing. Figure 37.9 gives a set of typical
I-V characteristics for a 53-mm-diameter solar cell under this illumination but at various
operating temperatures.
Thin-film Cu2 -CdS solar cells have also been applied to flat-panel terrestrial solar
arrays. Peak efficiencies of 7 to 8.5 percent have been demonstrated at room temperature
[3] , but problems of environmental stability lead to significant degradation if adequate
protection from atmospheric constituents is not provided.
The majority of the flat-panel solar cell modules produced for terrestrial applications
have been sponsored by the U.S. Department of Energy through Jet Propulsion Labora-
tory's Low-Cost Solar Array (LSA) Project. The first phase of this program, called
Block I, provided about 50 kW of modules of the types normally supplied by manufac-
turers as a commercial product. The second, or Block II phase, which represents 130 kW
of peak output, procured modules to a common specification and mechanical interface
requirements that included the following:
Figure 37.10 is a photograph of each of the four types of rigid flat-panel solar cell modules
procured under the Block II phase.
SOLARLX 18 6 WATTS
Electric Power Generation: Photovoltaics
12 INCHES
09000110011
0000110000
(1),
`kk
seeeepooto
In 1., ....".4410e SENSOR TECHNOLOGY 9 7 WATTS
• '---t46-16afiij
Figure 37.10 LSA Block II solar cell modules. Photo courtesy of Jet Propulsion Laboratory.
328 Applications of Solar Energy
Table 37.1 Pertinent Design Characteristics of LSA Block II Solar Cell Modules
Module Supplier
Sensor Solar
Design Characteristic Technology Solarex Power Spectrolab
1. Overall dimension (mm)
Length 582 581 1168 1168
Width 289 581 389 388
2. Weight (kg) 1.50 4.08 7.57 6.12
3. Number of solar cells 44 42 40 120
4. Solar cell diameter (mm) 55.9 76.2 101.6 50.8
5. Total solar cell area (m2 ) 0.1080 0.1915 0.3243 0.2432
6. Module area (m2 ) 0.1682 0.3376 0.4544 0.4532
7. Overall module packing
factor (Item 5 ± Item 6) 0.64 0.57 0.71 0.54
8. Nominal operating cell
temperature, NOCT
(°C) 42.9 47.1 46.0 41.1
9. Electrical performance at
the standard operating
conditionsa
(a) Maximum power (W) 10.4 18.7 31.3 28.5
(b) Voltage at max.
power (V) 18.7 16.3 16.6 17.3
(c) Open-circuit
voltage (V) 23.4 22.4 22.0 21.9
(d) Short-circuit
current (A) 0.59 1.44 1.98 1.88
10. Areal specific power
output (W/m2 module
area) (item 9a - item 6) 61.8 55.4 68.9 62.9
Source: Based on data from Ref. 4.
aStandard operating conditions are module irradiation of 100 mW/cm2 and cell temperature equal
to the NOCT.
Each of these module designs is intended for mounting to a frame structure for
general application in an array of free-standing east-west rows. Table 37.1 lists the per-
tinent characteristics of each of these module types.
The Nominal Operating Cell Temperature (NOCT) is the module cell temperature
under open-circuit conditions when operating in the Nominal Terrestrial Environment
(NTE), which is defined as
30
25
T
CELL - TAIR. NTE
• 20
rc
1-▪ 15
10
f-
50 60 70 80 90 100
INTENSITY , mW/cm2
The NOCT is determined from actual measurements of cell temperature for a range of
environmental conditions similar to the NTE. These data are then plotted as shown in
Fig. 37.11 to obtain the cell temperature rise above ambient at 80 mW/cm2 intensity.
The NOCT value is then determined by adding Tait = 20°C to the value of (T„ll - Tair)
thus obtained, that is,
The value of module electrical power output at 100 mW/cm2 insolation and at the
NOCT, as listed in Table 37.1, is thus a realistic measure of module performance since it
includes both the effects of solar cell efficiency and operating temperature. When this
output power is divided by the overall module area, the resulting areal specific output is
a measure of the relative performance of these various module types. As indicated in the
last row of Table 37.1, these values range from 55.4 to 68.9 W/m2 of module area for the
Block II procurement.
Figure 37.12 shows another terrestrial module type that is currently being de-
veloped as part of the LSA Project. This module is intended as a direct replacement
for shingles on the sloping roofs of residential or commercial buildings. The hexagonal
configuration of this module permits nineteen 53-mm-diameter solar cells to be closely
packed within 507 cm2 of exposed module area for an overall packing factor of 0.827.
The calculated NOCT for this module is 61°C assuming mounting against the insulated
roof of a residential building.
At a total AM1 irradiance of 100 mW/cm2 and at the NOCT, this module is ex-
pected to have a maximum power output of 5.14 W at 7.8 V. This represents an areal
specific power output of 101.4 W/m2 of module area.
330
Applications of Solar Energy
A. General Characteristics
Photovoltaic concentrators are designed to replace solar cells with less costly reflective
and refractive optical elements. The basic principle of concentrators is to intensify the
incoming light incident on a relatively small solar cell target. The geometric concentra-
tion ratio of a PV concentrator is defined as the aperture area (projected concentrator
opening in the direction of the incoming light) divided by the solar cell area. The optical
efficiency is defined as the percentage of the energy passing the aperture that reaches the
solar cells. The optical inefficiency is due to the reflection, transmission, and divergence
losses along the optical path of the concentrator. The optical concentration ratio is the
product of the geometrical concentration ratio and the optical efficiency. It represents
the ratio of the average flux density incident on the solar cells to the flux density at
the aperture plane.
Basic optical characteristics only allow concentrators to use incoming light rays
effectively within narrow angular limits, depending principally on the geometrical con-
centration ratio. Two important limitations arise from this fact: (a) sunpointing of the
concentrator is generally required; (b) except for rays within the acceptance angle limits,
other incoming rays will not be incident on the solar cell target. Thus, diffuse light rays
from the skydome are practically negligible in contributing to useful energy collection.
Concentrators may be grouped into one-axis and two-axis tracking categories.
One-axis systems require motion of the concentrator about a single axis in order to reach
the orientation at which the collected energy is incident upon the solar cell target. One-
axis systems almost exclusively use cylindrical parabolic troughs. Line focusing is
achieved when the sun's rays are parallel to the plane bisecting the parabolic cylinder.
Single-axis systems have been built (though very few with solar cells) that track about
east-west, north-south, polar, and vertical axes. Several limitations of one-axis systems
are the cosinusoidal drop off in output for nonnormal sunlight, and the associated para-
bolic trough end losses. Considering that solar cells would be located along the focal line
and connected in series, the end losses reduce the flux reaching the solar cells in the end
regions, causing a disproportionate loss in overall output. Bypass diodes connected across
groups of cells in the end regions can be used to minimize the nonuniform light distri-
bution effects. The end effects can also be circumvented by paralleling the cells within
each optical segment, thereby avoiding any series choking effect. This approach is des-
cribed later for a periodically tilt-adjusted compound parabolic concentrator (CPC).
Polar and north-south axis systems tend to minimize end losses, or at least the daily
variations in end losses. However, higher physical profiles and greater self-shading are
characteristics of these approaches.
Two-axis systems are those in which the aperture plane is positioned normal to the
sun's direct rays. Combined motion in azimuth and elevation are commonly employed.
Motion about a polar axis with an additional axis for periodic updating of solar declina-
tion is also commonly used. A variety of optical approaches can be adapted to two-axis
systems including Fresnel lenses (both circular and linear types), solid lenses, parabolic
cylinders, and paraboloidal dishes. The problems of nonuniform cell-to-cell illumination,
peculiar to end effects in one-axis systems, generally do not exist for two-axis systems.
332 Applications of Solar Energy
Two-axis systems yield the highest integrated value of collected energy per unit of aper-
ture area. On a very approximate basis, subject to correction for specific tracking modes
and site latitudes, two-axis systems collect about 50 to 60 percent more energy than
one-axis systems because of the cosine falloff of the latter. A more quantitative compari-
son of performance is presented in Sec. 37.5.
24
23
22
21
AR G a As/GaAs CELL
20
19
SOLARCELLEFFICIENCY, ck
18
17
16
15 SILICON CELL
14
13
11
10
Figure 37.13 Typical efficiency versus concentration ratio curves for silicon and Al-Ga-
As/Ga-As solar cells.
Electric Power Generation: Photovoltaics 333
Concentrator solar cells are usually designed to fit the requirements of specific applica-
tions regarding such factors as concentration ratio (CR), anticipated operating tempera-
tures, the line- or point-focusing nature of the concentrator optical system, and the uni-
formity of the concentrated light. Silicon and gallium-arsenide solar cells are the principal
types presently being considered for concentrator applications. Silicon cells are typically
used for concentration ratios up to 50:1 or so, although successful designs have been built
with CR's of 500:1. Efficiencies are in the range of 12 to 15 percent at cell temperatures
of 29°C. As established for nonconcentrating solar cells, output decreases about 0.4
percent for each 1°C temperature increase. Gallium-arsenide cells can operate at CR's in
excess of 1000:1 and are reported to show efficiencies greater than 20 percent. Figures
37.13 and 37.14 portray typical performance characteristics of both cell types.
24
A/GaAs/GaAs CELL
23
22
21
20
SOLARCELLEFF ICIENCY, %
19
18
17
16
15
14 SILICON CELL
13
12
11
10
Figure 3734 Typical efficiency versus temperature curve for silicon and Al-Ga-As/
Ga-As solar cells operating at design concentration.
334 Applications of Solar Energy
Concentrator solar cells are usually distinguished by a high density of front surface
metallized gridding which overlays the electricity-producing surface of the cells. The
gridding reduces ohmic losses, although it also shades a certain portion of the cell. The
amount of gridding is optimized for particular concentration ratios. Because of the high
flux incident on concentrator solar cells, they are fabricated using materials and coatings
that are more temperature resistant than those required for nonconcentrating systems.
C. Concentrator Cooling
D. Hardware Examples
A summary of concentrator systems that have been built or are under development is
given in Table 37.2. Brief descriptions of each are presented below:
The Spectrolab design shown in Figs. 37.15 and 37.16 consists of a sun azimuth
tracking turntable about 24 m in diameter on which are mounted spaced rows of
trunnion-mounted sun elevation tracking semiparabolic collectors. The collectors are
ganged together so that a small number of sun-sensing devices and drive motors are
required. A distinguishing feature of the Spectrolab design is the use of a compound
elliptical concentrator (CEC), which serves as a secondary stage after primary concen-
tration by the parabolic collector. This two-stage arrangement relieves contour accuracy
requirements on the primary stage and also tends to distribute the light reaching the
solar cells more uniformly. The CEC surrounds the solar cells and also functions as the
passive cooling system by means of fins built into the exterior surface. The design uses
series-connected 5- X 5-cm silicon solar cells mounted on insulating mica wafers. An
enclosing window protects the cells against environmental effects.
The proposed General Electric design shown in Fig. 37.17 employs similar track-
ing and optical concepts to those of the Spectrolab design. Distinguishing features are
the use of an active cooling system by means of coolant circulation through a pipe lo-
cated along the collector focal line. Silicon solar cells are vee-mounted along the pipe,
with each cell surface receiving light energy from each half-parabola. The currents
developed in the solar cell strips of each vee surface are used to sense and control the
elevation pointing of the collector.
Table 37.2 Summary of photovoltaic concentrator systems
System
Characteristic Spectrolab General Electric Martin-Marietta RCA MIT Argonne
Optical system Parabolic Parabolic trough Circular Fresnel Circular Paraboloidal Compound
trough Fresnel dish parabolic
collector
Tracking system Two-axis, Two-axis, Two-axis, Two-axis, Two-axis, One-axis,
azimuth azimuth azimuth polar- polar- seasonal
elevation elevation elevation declination declination adjust
Geometric 25.2 34 39 423 >200 9.2
concentraion ratio
Optical efficiency < 0.75 to 0.85 >
Concentrator size 3.17 m2 22.3 m 2 0.37 m2 0.37 m2 0.29 m2 5.83 cm 2
1.28 X 2.48 m 2.44 X 9.14 m 0.3 X 1.22 m 0.61 X 0.61 m 0.61 m diam. 1.91 X 3.05 cm
Cell type and size Silicon Silicon Silicon Silicon Silicon Silicon
3.2 X 5.0 cm 3.5 X 4.0 cm 5.5 cm diam. 0.56 cm 4.3 cm diam. 0.25 X 2.54 cm
diam.
Cooling system Passive Active Passive Passive Active Passive
mos Jo suo!lpon d dv
Figure 37.19a Fresnel housing and heat sink. Photo courtesy of Martin-Marietta.
E l ectri c PowerGeneration : Photovol tai cs
Figure 37.19b Fresnel housing and heat sink. Photo courtesy of Martin-Marietta.
342 Applications of Solar Energy
approach. The tracking approach employs an azimuth rotating hub with two extended-
elevation tracking tubes onto which are mounted a series of four-element Fresnel lens
enclosures. Circular solar cells 5.5 cm in diameter each receive the concentrated light of
30 cm X 30 cm Fresnel lenses. The cells are mounted on alumina wafers. Heat is trans-
ferred to finned aluminum extrusions and convectively rejected to the surrounding air.
The RCA polar-mounted system with declination angle updating provides another
example of a refractive two-axis pointing system. The RCA system, illustrated in Fig.
37.20, employs rather small 0.56-cm-diameter silicon solar cells which operate in con-
junction with 10.2-cm by 10.2-cm Fresnel lenses. The system yields a geometric concen-
tration ratio of 423:1. The solar cells are hardmounted through copper plugs to finned
aluminum extrustions, which serve both as a passive heat rejection system and as con-
ductors for the generated current.
CR = 1
sin 0
2
where 0 represents the angular limits of the incoming light rays projected onto the plane
perpendicular to the CPC trough axis. A fuller explanation of the theory of these collec-
tors is presented in Chap. 10. The sun's rays emanate from different sectors of the sky
over the period of a year. For this reason, it is necessary to adjust the tilt of the CPC's
periodically in order to include the incoming rays within the acceptance angle limits.
This adjustment represents the one-axis tracking aspect of this concept, although, strictly
speaking, continuous tracking as such is not involved. An interesting version of the CPC
shown in Fig. 37.24 uses a solid dielectric material with parabolically shaped sides instead
of reflective surfaces. Light entering through the aperture plane is internally reflected off
the sides and ends and ultimately reaches the cells mounted along the bottom. The cells
of each CPC segment are parallel connected, permitting nonuniform light distribution.
By this means, the end effects present in other types of trough systems are avoided. Since
Electric Power Generation: Photovoltaics 343
the conditions of illumination are identical in each CPC segment, it is possible to connect
the segments in series to achieve voltage buildup. The refractive index of the dielectric
material has the effect of increasing the acceptance angle for a given concentration ratio.
Thus, fewer seasonal tilt-angle adjustments are required.
344 Applications of Solar Energy
Secondary
Focused mirror
light
Primary
mirror
Cell---''
r-"Li..--Water-cooled Inner Outer
Main frame block gimbal
Figure 37.23 Compound parabolic concentrator. Photo courtesy of Argonne National Laboratories.
346 Applications of Solar Energy
' II
0101'
4s1 111111
lio EXTENDED-SURFACE
INTEGRAL SOLAR CELL HEAT SINK
HEAT-SINK ASSEMBLY
(a)
LIGHT RAYS
REFRACTION
TOTAL
INTERNAL REFLECTION----....„..
Several unique PV concentrator concepts have been recently proposed and are
undergoing development. One example being developed by AAI uses an array of two-
axis tracking heliostats mounted on a tilted framework that reflects incoming sunlight
onto a solar cell target deployed above the array.
Electric Power Generation: Photovoltaics 347
A. Flat-Panel Systems
The performance of the flat-panel solar cell modules described in Sec. 373 has been
analytically simulated using hourly weather data at 12 specific locations within the
continental United States. These locations represent a wide range of both climatologic
and insolation conditions and are felt to be a representative sample of the various com-
binations of conditions that influence photovoltaic module output. Each of these loca-
tions represents a weather station for which hourly SOLMET data [6] are available.
Table 37.3 summarizes the results of this performance simulation for six basic
types of flat-panel orientation and/or tracking modes. The first of these is the east-
west row of fixed sloping panels. For this most commonly used mounting arrangement,
two slope angles have been selected: 30° and 40° (measured from the horizontal),
depending on the site latitude. Previous studies [7, 8] have indicated that optimum
photovoltaic system performance, in terms of annual collected energy, is obtained for
roof slope angles that are approximately 10° less than the site latitude, but the devia-
tion from the maximum possible system output is not significant for 5° variations from
this optimum slope. The measure of system performance used in this tabulation is the
effective annual operating time at the standard operating conditions, which are defined
as a total insolation on the module of 100 mW/cm2 and a cell temperature equal to the
NOCT. This approach to array performance assessment allows the determination of the
annual energy output of any solar cell module by simply multiplying the areal specific
power output for the module (Table 37.1) by the effective annual operating time as
given in Table 37.3. Thus, for the Spectrolab Block II module mounted in east-west fixed
tilted rows in Albuquerque, New Mexico, the annual energy output is given by
If the 12 data points for the east-west fixed tilted panels are plotted as a function of
annual total horizontal insolation, the results are as shown in Fig. 37.25. The scatter of
these data points from a best-fit linear relationship is influenced primarily by cell
operating temperature, which is, in turn, strongly related to ambient temperature and
wind speed. For locations other than those listed in Table 37.3, it is suggested that es-
timates of flat-panel system annual energy collection can best be made by using judg-
ment to rank the location in terms of total horizontal insolation and ambient tempera-
ture relative to the 12 data points tabulated. These values for a number of locations
throughout the United States can be obtained from Ref. 9.
The four single-axis tracking flat-panel orientations listed in Table 37.3 each yield
a higher annual energy collection at the added expense of associated tracking mechanisms
and controls. Either polar axis tracking or latitude tilt vertical axis tracking, with nearly
identical performance, would seem the best choice for a single-axis tracking flat-panel
installation. In either case annual energy collection will be about 20 percent greater than
the fixed E-W tilted panels. However, tracking installations of this type bear further
detailed investigation, since the land area required to avoid significant mutual shadowing
Table 37.3 Performance of flat-panel solar cell modules
Effective Annual Operating Time at Standard Operting Conditionsa (h)
Annual Annual
Total Direct E-W N-S
Horizontal Normal E-W Horizontal Horizontal Latitude Tilt
Latitude Longitude Insolation Insolation Fixed Axis Axis Polar Axis Vertical Axis Two-Axis
Station (°N) (°W) (kWh/m2) (kWh/m2) Tiltb Tracking Tracking Tracking Tracking Tracking
Albuquerque, N.M. 35.0 106.6 2108. 2601. 2521 2701 3011 3243 3131 3335
Bismarck, N.D. 46.8 100.7 1433. 1624. 1873 2018 2194 2390 2395 2473
Boston, Mass. 42.3 71.0 1278. 1195. 1614 1715 1852 1965 1970 2025
Charleston, S.C. 32.8 80.0 1551. 1369. 1756 1842 1947 2071 2036 2110
Fort Worth, Tex. 32.6 97.2 1697. 1706. 1935 2042 2232 2389 2308 2449
Madison, Wisc. 43.0 89.5 1378. 1369. 1648 1764 1894 2053 2036 2112
Miami, Fla. 25.8 80.3 1706. 1387. 1871 1964 2138 2220 2146 2256
Nashville, Tenn. 36.2 86.8 1460. 1323. 1699 1778 1903 2027 1982 2071
Omaha, Neb. 41.3 96.0 1515. 1633. 1808 1951 2149 2308 2274 2360
Phoenix, Ariz. 33.5 112.0 2154. 2528. 2425 2584 2874 3057 2954 3159
Santa Maria, Calif. 35.0 120.0 1843. 1998. 2202 2331 2520 2702 2657 2787
Washington, D.C. 39.0 77.0 1387. 1259. 1523 1640 1664 1856 1822 1910
aStandard operating conditions are module irradiation of 100 mW/cm2 and cell temperature equal to NOCT.
b .
30 slope angle for latitude 537°, 40° slope angle for latitude >37°.
Electric Power Generation: Photovoltaics 349
2600—
ALBUOUEROUE,N.M.°
/
/0
co 2400 / PHOENIX, AZ
cr
I— 0 /
< I
Lai v; /
X z 2200 — SANTA MARIA , CA 0 /
1-0
0 i-7 /
z—az /
ao
I-- o 2000 —
cc /
co FORT WORTH , TX
Cl-—Z
o /o
BISMARK, ND ,
—I1-< 0 7 0 M I A M I , FL
a cc
0
,w
:i. 1800 —
Z
a
0
OMAHA, 9/
NE / ()CHARLESTON , SC
w cr BOSTON, /0 NASHVILLE , TN
<
o 1600MA 0
_ 0 / MADISON , WI
z
Lai
(f) / 0 WASHINGTON , DC
1400 I I I I I 1
1200 1400 1600 1800 2000 2200 2400
ANNUAL TOTAL HORIZONTAL INSOLATION , kWII / m2
Figure 37.25 Performance of low fixed tilted flat-panel arrays at various locations.
losses may be an important factor to consider. Also, the vertical height of such tracking
installations may involve a significant structural cost penalty implicit in the requirement
to survive design wind loads.
A two-axis tracking flat-panel installation affords the highest possible annual
energy collection with the cost implications associated with tracking controls and
mechanisms.
B. Concentrator Systems
The photovoltaic concentrator system most commonly used is the two-axis tracking
system which positions the aperture plane normal to the direct rays from the sun. This
is generally achieved by combined azimuth and elevation tracking, but polar axis rota-
tion with periodic adjustment for solar declination has also been employed. In either
case the small acceptance angle for such a system limits the available insolation to the
direct normal component. Annual performance results based on hourly simulation of
a typical passively cooled two-axis tracking system are given in Table 37.4 for 12 selected
geographic locations. This example system has an optical concentration ratio of 24 but
the results, in terms of annual maximum power output energy per unit of aperture
area, should be relatively independent of this ratio over the range from 15 to 35. An
arrangement of fins on the solar cell mounting heat sink provides the cell cooling,
which is a strong function of the ambient temperature and wind speed. The details
350 Applications of Solar Energy
Annual Direct
Normal Insolation Two-Axis N-S Horizontal
Site Location (kWh/m2 ) Tracker Axis Tracker
of this heat exchanger design will influence the annual energy output, but the example
calculation is typical of an existing design in this area. An overall optical efficiency of
72 percent has been used in this calculation together with a peak solar cell efficiency of
14.7 percent at 28°C. Again, these parameters may vary with the specific details of a
design, but the performance values given can be directly proportioned for small changes
about these assumed values.
A review of these simulation results for the two-axis tracking concentrator reveals
that the performance in Miami and Boston are nearly identical, whereas the total hori-
zontal insolation for Miami is 33 percent greater than for Boston. The large fraction of
diffuse radiation and higher ambient temperatures in Miami account for this difference in
performance.
A N-S horizontal axis tracking approach is also included in Table 37.4 as an
example of a single-axis tracking array. The performance of this system, as well as other
single-axis tracking concepts, is very dependent on the details of the particular design, so
care should be exercised in applying these results.
REFERENCES
LORIN L. VANT-HULL
Energy Laboratory
University of Houston
Houston, Texas
38.1 INTRODUCTION
353
354 Applications of Solar Energy
DEGREES RANKINE
CARNOT CYCLE
0.8
>-
0.6
z
w
U
u- 0.4
0.2
Figure 38.1 Contributions to the overall efficiency of a solar thermal conversion system.
The practical thermodynamic cycle efficiency is taken as 70 percent of the Carnot
efficiency. We assume that 360 W are absorbed by the receiver from each square meter
of collector surface. The family of curves diverging from an efficiency of 1.0 show the
effects of radiation loss on collector efficiency for optical concentrations ranging from
1 to 3000. The cusps at zero efficiency represent the stagnation temperature and the
associated numbers (3, 10, ... , 3000) give the effective concentration uAcoll/eArec.
The products of the practical cycle and the collector efficiencies are also shown, and give
the efficiency for thermodynamic use. The maxima of these curves indicate the desirable
temperature range for each concentration.
Electric Power Generation: Thermal Conversion 355
An important consideration for any system that is to produce mechanical energy from
thermal energy is the overall system efficiency. This is determined by the product of
the thermodynamic cycle efficiency (in a typical case, about 70 percent of the Carnot
cycle efficiency), and the collection efficiency of the solar system. If the latter is pri-
356 Applications of Solar Energy
manly limited by radiation losses, the losses will be negligible at low temperatures,
rising to 100 percent at the temperature of radiative equilibrium for the system. A
series of such curves for various concentrations are also shown on Fig. 38.1. As the
thermodynamic cycle efficiency increases with temperature, it is clear that there will be
a maximum overall efficiency for any specified concentration factor, as shown by the
domelike curves. For our purposes we define the concentration factor to be a product
of the geometric concentration and the value of ale for the surface in question. Thus,
with an effective concentration of unity, temperatures much above ambient cannot be
reached. For the example of Fig. 38.1, where we have assumed an average of 360 W is
absorbed for each square meter of collector surface, to account for reflectivity, angle
of incidence, and so on, a concentration of 10 provides a stagnation temperature of
approximately 500 K, giving a maximum overall efficiency at 400 K of about 8 percent.
In contrast with a concentration of 300, the radiative equilibrium temperature is 1200 K
and a maximum system efficiency of about 32 percent is achieved at a temperature
between 700 and 800 K. Provided that the cost per unit area of the surface intercepting
the sunlight is comparable, the higher efficiencies available with higher concentration
systems strongly suggest that these systems will be more economical if mechanical energy,
for example, to drive an electric generator, is required since the size of the solar collector
system to produce a given amount of output is inversely proportional to the system
efficiency. Efficient use of the concentrator surface area, high-reflectivity mirrors, and a
highly absorbing receiver also contributes to good system efficiency.
This consideration of system efficiency has the important result that collectors
for solar thermal electrical power production must produce relatively high temperatures
with good collector efficiency, or they must be extraordinarily cheap per unit area. Under
certain conditions, nonconvective solar pounds can be extraordinarily cheap collectors of
solar energy at 100 to 150°C, so they remain viable contenders even though the net
system efficiency may be quite low. The ocean is an even cheaper (free) collector, so at
1 to 2 percent efficiency the ocean thermal gradient system remains a viable contender
which must be carefully examined. For both of these systems, the fluid-handling com-
ponents and the turbines tend to become large and expensive, and great care is required
in design so the parasitic losses do not use all the energy collected.
It can easily be shown that flat-plate collectors are not practical for thermal elec-
trical generation. Simple flat-plate collectors have such large thermal losses that their
collector efficiency is poor even at 80 to 90°C, the collection of the low-temperature
absorbed heat to the turbine is wasteful and expensive, and finally, the Carnot efficiency
is very low. Improved flat-plate collectors can work at temperatures of 100 to 120°C,
but such collectors are very sophisticated, using selective surfaces, multiple glazings,
evacuated chambers, and so on. Consequently, their price is much higher, wiping out the
advantage. However, for small systems below a few tens of kilowatts or so, the simplicity
of a nontracking receiver may lead to its use in special applications—Sofretes of France
has sold a number of flat-plate systems for isolated village electric plants and isolated
well pumping. The economics are not good, but then, diesel fuel imported by mule is
not cheap either.
For the remainder of this chapter we shall consider concentrating systems only.
These systems collect only direct beam sunlight. The diffuse component (typically no
more than 20 percent of the direct beam in the U.S. Southwest) is wasted.
Electric Power Generation: Thermal Conversion 357
Figure 38.2(a) Artist's conception of the 10-MWe pilot solar tower plant, courtesy of
McDonnell Douglas Astronautics Co.
•
COOLING TOWER
illuminated at other than normal incidence (i = 0), the focus degenerates into a circle of
least confusion as the focal length for rays in the optical plane is multiplied by cosine i
while the focal length for rays perpendicular to the optical plane is divided by cosine i,
where i is the angle of incidence on the mirror. The resulting circle of least confusion
grows as the angle of incidence increases until, for i >60°, the diameter of the image is
greater than the minor axis of the foreshortened image from a flat mirror of equal size.
In the 5-MWth test facility (Fig. 38.3) [7] that recently began operation at
Albuquerque, New Mexico, focusing heliostats with an area of about 40 m2 are being
used. The 25 segments are canted to superimpose their beams, and each segment is stress
curved to provide for 4 to 6 h per day the high beam accuracy and high concentration
required of this test apparatus. The high beam accuracy is required so that the flux can
be tailored to meet the requirements of a variety of tests. The first tests will be of proto-
360 Applications of Solar Energy
a.
-,,s111P
r, I i
"Milisamajzza Si ,4Fi / Si
# ^••••••• i
*4111(0,10,, , 41, ISr
Figure 38.3 Aerial view of 5-MW solar thermal test facility with 72 of 222 heliostats
installed. Operation with 222 heliostats was initiated in February of 1979. Photo
courtesy of C. E. Robertson of Sandia Laboratories.
type receivers of the 10-MW, pilot plant to be constructed at Barstow, California. These
are a scale-model 5-MWth cavity with a 2.67-m square aperture, a single segment of an
external receiver m wide X 17 m long, and a downward-looking annular aperture
cavity receiver (0.D. 4.9 m, annular width 1.5 m). A multitude of other experiments and
tests are envisioned. The more specific requirements of the 10-MW, (40-MWth) pilot
plant will place less exotic restrictions on the heliostats and thus probably unfocused,
canted segment heliostats will be adequate for each receiver.
Let us consider the various subsystems of the solar tower system and discuss the primary
requirements on each one. For clarity, many aspects of this section will be illustrated
with figures related to the McDonnell Douglas design study [8] recently completed for
the Department of Energy (DOE) and selected as the basis for the final pilot plant design
[10] .
The imposed requirements in this case were to demonstrate a feasible and poten-
tially economic central receiver system to generate 100 MW of electricity at 2:00 p.m. on
December 21, while charging a system capable of providing 6 h of deferred operation.
The design condition was A,35° N latitude in the sun-rich, but seismically active U.S.
Electric Power Generation: Thermal Conversion 361
Southwest. A 30-year life and 90 percent availability (exclusive of weather factors) were
also specified, implying survival in 40 m/s (90 mph) winds at 10 m height, seismic
accelerations of 1/4 to 1/3 g, and temperatures from -30 to +50°C.
Design of the actual system that will best meet these requirements is an iterative
process. That is, the designer is required to first specify a set of additional requirements
and ground rules or assumptions leading to a reasonable baseline design. From this a
final design evolves as a result of a series of cost and performance trade-off studies.
Prior analysis [9] had shown that large, single-tower systems were generally more
effective than smaller multiple systems, that the cost of the heliostats would predominate,
and that ground coverage by the heliostats, 0, should be nonuniform, but be in the range
of 0.2 to 0.5 (averaging 1/4 to 1/3) to be most cost effective. Rather than choose a
receiver and design a heliostat field to best illuminate it, we chose to design the more
expensive heliostat field first, choosing a reasonable configuration that would deliver
the required power at minimum cost. A receiver was then designed that would effectively
intercept the resulting convergent beams and produce steam at conditions compatible
with a versatile but effective dual-admission 100-MWe turbogenerator set.
A first-order calculation gives
where I is the direct insolation (W/m2 ), the optical system is taken to have an efficiency
n of 2/3, the turbogenerator to have a net efficiency nt of 1/3, and the solar multiple
(SM = peak power/rated power) of 1.7 is required to use effectively the 6 h of storage
at 35° latitude. The resulting field diameter is approximately 1.8 km: at 35°N latitude
the receiver should be located about one-third of a diameter from the south border for
best annual performance due to excessive cosine effects in the south field. The corre-
sponding visual range to the receiver will be 1.2 km or perhaps larger, requiring a tower
of order 300 m high and a receiver aperture at least 12 m in diameter to accept the entire
solar disk, assuming perfect optics.
With these constraints in mind, four heliostats were designed, under DOE funding,
for adequate performance and minimum installed cost per square meter (Fig. 38.4). The
heliostat selected as offering the greatest potential for adequate performance coupled
with low cost under mass production is labeled McDonnell Douglas in Fig. 38.4. This
heliostat is square and composed of 6 to 12 segments in which 6-mm-thick float glass
is supported by a steel structure. A cost analysis shows that the foundation, pedestal,
and tracking mechanism tend to drive to larger heliostat sizes, while the mirror support
structure and handling costs begin to increase when the dimensions exceed 5 m. In
addition, excessive mirror diameters will produce enlarged beams at the receiver. Con-
sequently, a mirror diameter of 6 to 7 m was chosen. When sufficient strength was
designed into this heliostat to survive the maximum wind, mirror surface deflection
under operating winds was essentially negligible. No significant cost savings were found
for aiming and steering errors larger than a = 3 mrad. Consequently, the field analysis
was carried out with a square mirror 6.5 m in diameter and with 3-mrad beam accuracy
from all causes.
HONEYWELL
MARTIN Mc DONNELL
MARIETTA DOUGLAS
44,
anu] Rios jo suo!ivoHdde
Figure 38.4 Sketches of four heliostat concepts involved in the DOE conceptual design studies.
Electric Power Generation: Thermal Conversion 363
Neglecting off-axis aberrations, one may represent a partially focused or flat mirror
as a point mirror having a uniform angular dispersion of r/Q, where r is the actual image
size produced at a range of Q by parallel rays reflected from a perfect mirror. One could
add to this the solar limb half-angle a, and a beam error due to steering, surface irregu-
larities, and so on, taken as twice the standard deviation of the beam, that is, as 2u, to
encompass 95 percent of the area under the normal error curve. The sum of these dis-
persions, r/2 + a + 2u, is an exceedingsly conservative estimate; essentially no photons
will be found with this large a dispersion.
A more reasonable approach is to compute the first moment of each distribution
and add these moments in quadrature. Adequate approximations to these first moments
are r/212, a/2 and a. Identifying the square root of the sum of the squares of the resulting
moments, as the standard deviation of the beam, essentially all the energy will be
found in an image 2µ in angular radius, or 20 in radius. Thus (see Fig. 38.5), if
= R/cos 0 is the slant range to the receiver, we have
2 2
2112 2 F — + 021
r + a
2Q 2
as an estimate of the beam radius encompassing nearly all the beam energy. For a flat
heliostat 7 m in diameter, r = 3.5 m, a = 4.66 mrad, a = 3 mrad, and we may take the
field radius R to be 1200 in. If 0 is moderately small, Q = R/cos B "=-• 1200 m also. Then
Note that if perfect focus is assumed, we simply delete the term in r/k and the 10
changes to a 9. Thus, the use of flat heliostats is not a substantial disadvantage at this
4/LR
2a
cose
SPHERE
a/COS 8
4 CYLINDER
APERTURE
•
SING
2µR 2A0
Cylinder 2rraL 27r 2/1R Ao = fr(2µR)2
cos 0 cost cos3 0
fra 2 2µR 2 A0
Aperture 7
sin2 COS 0 sin 0 sin2 0 cost
Plots of these equations appear in Fig. 38.6. Note that in this formulation Ao scales
as the square of the field radius, not the tower height. This is a reasonable scaling relation-
A0 =7r( 2/1.1R)2
15A0 APERTURE
SURFACE A REA
I0A0
CYLINDER
5A0 -
~SPHERE
0 I 2 3 4 5
R/h (= cot 0)
Figure 38.6 Receiver area versus normalized field radius for various central receiver
configurations. The effective beam radius is a = 2µR/cos 0.
Electric Power Generation: Thermal Conversion 365
ship, as the total solar energy available is proportional to the area of the collector field
and the average efficiency of its use varies by only 10 percent or so between different
well-designed fields. For the sphere and the cylinder, the minimum occurs for the
smallest 0, that is, the lowest tower. Hence, to produce 500 MWth , the tower height is
set at '\,250 m simply by the optical requirements of the field, that is, to minimize
blocking. The aperture (or cavity), in contrast, has a minimum area, when 0 = 45°, or
4 X Ao . However, this requires a tower height of 1200 m, or '\,900 m if the aperture
is tipped toward the north. If a suboptimum is accepted, a tower half this high can be
used, but the aperture area is doubled. The compensating advantage of a cavity is that
the reflection loss of 5 percent is overcome and convection losses are minimized. Fur-
thermore, the absorbing surface can be many times larger than the aperture, resulting
in reduced flux density and lessened thermal fatigue problems.
The receiver chosen for the commercial system design [8] , 500 MWth , was a
cylinder 17 m in diameter and 25.5 in tall. The outer surface of the cylinder consists
of 24 panels, each composed of 170 vertical tubes, each 1.3 cm in diameter of Inconel
800 welded together to form a membrane. The maximum design flux will be 0.85 MW/
m2 , and feed water will be heated from 205°C in a single pass to superheated steam at
516°C and 11.1 MPa (108 atm). The pilot plant will essentially be designed as a scale
model of the commercial system design.
There are three levels at which one may design a heliostat field. At the simplest
level, one simply distributes heliostats in a rational fashion, estimates the resulting losses,
and readjusts the field for improved performance. However, realistic calculation of per-
formance requires the capability of computing the losses. The primary losses encoun-
tered are atmospheric attenuation, shading, cosine foreshortening, reflection losses,
blocking, attenuation in the path to the receiver, interception losses at the receiver,
reflection by the receiver, and finally, the thermal losses of convection and radiation
at the receiver and in the heat transport system.
To deal with each of these effects adequately, one is soon led to develop extensive
computer models [11] . The existence of such models leads to a second level of sophisti-
cation. Here one evaluates many of the effects at a given location in advance. Either for a
given instant, averaged over a day, or averaged over a year, tables of atmospheric attenua-
tion, reflection loss, cos i, attenuation to the receiver, reflection, and interception by the
receiver can be tabulated in advance. The remaining losses due to shading and blocking
can be evaluated for several geometries and "footprints" can be generated (Fig. 38.7).
These "footprints" can be very informative, showing shading "wings" to the east and
west, and a blocking "nose" in the direction of the tower. One can then hold shading
and blocking losses constant over the field by properly choosing the spacings from the
"footprints." Alternatively, and more logically, one can hold the total of losses from all
effects constant over the field by varying the shading and blocking loss to compensate
for the other losses. Where other losses are high, the shading and blocking is held low by
separating the heliostats more widely. The resulting field is a kind of equal performance
optimum.
Finally, a true optimization procedure can be implemented. Based on a consistent
analysis of heliostat performance, an iterative technique may be used to generate a field
of heliostats having economically balanced performance and the best figure of merit,
366 Applications of Solar Energy
VECTOR TO
RECEIVER
LOCATION OF NEIGHBORS LOCATION OF NEIGHBORS
o RADIAL STAGGER o N - S STAGGER
x RADIAL ORTHOGONAL x NORTH-SOUTH
EAST-WEST
Figure 38.7 Contours of annual percentage loss due to shading and blocking from a single
neighbor. The contours (e.g., the line marked 2%) show a line along which the single
neighbor should be moved to give an annual shading and blocking energy loss of the
specified amount (e.g., 2%). The contours are normalized to the nominal heliostat dia-
meter, Dmir. Cell 1, 1 is 2-1/2 tower height west and 3-1/2 north of the receiver, whereas
cell 5, 5 is 1/2 tower height west and 1-1/2 north.
that is, the lowest system cost per megawatt hour delivered annually. All the effects
above must be taken into account, as well as accounting for the energy loss experienced
by adjacent heliostats when an additional heliostat is crowded into a specific area. The
radial staggered arrays show substantially better performance than any other over most
of the field. Amazingly, the optimized results show a great deal of regularity. To first
order, the field can be represented by a set of circles of radial separation given by a simple
function such as
The radial field produces a varying AAz, so from time to time it is necessary to
break the pattern and relax the azimuthal separation by introducing slippage planes. A
final field design for a 10-MWe pilot plant [81 (Fig. 38.8) consists of 32 circles, 13 of
which are trimmed away on the south in response to severe cosine losses there. 1970
heliostats, each '\40 m2 in area, will be adequate to produce 10 MWe at 35°N latitude
at 2 p.m. on December 21 with a direct beam solar intensity of 950 W/m2.
Electric Power Generation: Thermal Conversion 367
A commercial-sized version of this system (100 MWe) will generate steam at 510°C
at a peak rate of 430 MWth . A portion of this thermal energy will be stored and the rest
will drive a turbine to produce 100 MWe of net energy for delivery to a utility grid. On
most clear days the excess energy collected will be sufficient to provide stored heat for
6 h of deferred operation (after sunset) at 70 percent of capacity. The receiver is ele-
vated 250 to 275 m above grade, and about 0.8 km2 of reflective surface are used, dis-
tributed over 3 to 4 km2 of desert.
In this design a cylindrical external receiver is used because operating receiver
losses are only about 12 percent, 5 percent due to reflection and 7 percent due to
radiation and convection. These percentage losses are small because the heliostat field
provides high concentration. Although 2000 suns are possible, the aim points are dis-
tributed over the receiver so the peak flux is 900 suns.
The water-cooled receiver was chosen in 1974 for early development because
water/steam systems are familiar to the utilities and present no new problems. With
funds committed to build a 10-MWe pilot plant (Fig. 38.9) [12] , attention was turned
to more advanced systems. DOE is currently funding four studies of advanced central
receiver systems (Fig. 38.10) [12] , which show promise of significantly better perfor-
mance than the water/steam system. Three of these systems use a molten salt or metal
as the receiver coolant, heat transfer fluid, and thermal storage medium. The excellent
heat transfer properties of these fluids and the maintenance of single-phase flow at
nominal pressure allow the safe use of higher fluxes, reducing the receiver size, weight,
losses, and potential cost. Use of the same material for thermal storage allows operation
from thermal storage with no reduction in system efficiency (cf. '\,70 percent for the
water/steam system with hot oil thermal storage). Furthermore, existence of the hot
stored energy allows one to operate a reheat turbine, with a gain of about 15 percent in
Rankine cycle efficiency. Overall, the advanced systems will produce about 10 percent
more electricity with a 15 percent smaller collector field. The fourth advanced system
under study is a gas-cooled receiver operating at about 900°C.
Studies and tests of both open-cycle air and closed-cycle helium systems have
been funded by the Electric Power Research Institute. Operating at 800 to 1100°C,
these systems project a significantly higher turbine cycle efficiency and somewhat
higher receiver losses, even with a cavity receiver. Thus, the overall performance is
roughly comparable to the external receiver system. However, the cavity places some-
what more severe optical requirements on the heliostat field. In order that beams
reflected by the heliostats effectively enter the aperture, the field is typically restricted
to a 63° cone half-angle. In contrast, with the exception of a small central exclusion
area, the external receiver does not limit the acceptance angle. Consequently, helio-
stats can be placed where they produce the best economic performance.
At the 100-MWe level, it appears the 500 to 600°C external receiver with a steam
Rankine cycle and the 1000°C cavity receiver with a Brayton cycle will provide com-
parable efficiencies, although materials problems are obviously more severe at the higher
temperature. If one considers smaller systems operating at the same temperature, cavity
receivers are obviously favored, although requirements for focusing heliostats to reduce
the beam size become bothersome. Alternatively, by lowering the operating temperature
to be compatible with conventional steam turbines at each size, for example, 510°C at
100 MWe, 470°C at 10 MWe, and 300°C at 1 MWe, the nominal performance trade be-
368 Applications of Solar Energy
e s:0 0 0: 00
esss se00 000 • ••
eee G GG s ° sG0 0 000
eeGG e0 e 0 0 0 00 0
e
ee e s e
ee 69s
s : s6:0®0v°0
69,,,0 00 _° '0000 0000
0 e0-eet
e (Deeeeee ee ® 669 69
e eee Ge ee 69'e 0e69000 000°
0 @ GO e e
e ee e e ab0 e0 e aaeeee
ee
eee e eG G Ge® G-0® ® 000 69 0 0 0 aeeeee
eeee eee e e eeeees 00 00
eeee e eeee eee 69e4E9G 6VG0 000 0 0eee
e e eeee
e eeee eeee eeeE9eeED ss(00 00 edD e ee eee
a eeee eeeeeeeeeseeegge asTeee
e0ee0-eeeGGG
mee eeGl)
aeaaaeeeeee epees,
eeeeeeeeeeeeeeereee ®EDGED®
eeee e eaeee „,„,onvo e iEDGEDEDEDG
e _ e ee ,se eeei„ ee SEDEDGGs
®® ;,,eee
0 " ee'
0 0 000 0 eee° eaeaeelDee
eae ee
0 :0 aleTa e a
0 0 0 0®
'Th 0 0 0 00 0 000 0°00
0 00 00 00
0 0 0 0°0° °0
0 e a 0 0000
0 0 ° 0-0 00 - 00°0
00 00 0 0 ®0 0
0 GO 0 OG0 G 00® ® ®:: ®00:: s ov-,,00
e 0 0€9G00
0 G®
0 0 0 0® 0) el ji9 0 0 pith g f 69 6969
0 0;1%0 0 e eer Geg,Deese ED0e
®e:Oe'31 ee-
on EDEDEDeeee
® ® G 0 0690 G 0 06)1421884a a, Geaeeee
ee
O 069 GG 0 G0000€9es@wEDRe- wee eee
0 €90000630690® , e@GEDT, $0:: ®®
O 0 0 G G G ® ED G G2Xe EDEDe
,?g,,. (Deft e
ee® ,,e
sseee eeee we-eee'47-a se7,0,G
e s e ee e e EDGe eeeeS a
0ei se?:
®G GED,, e 0sEDeeeeee e 0100
00 0 0 eew
GEDGEDGwlee eZED:Te:06: Dee 000 00 0®
fp ED e e a e e 0 00 000 0 69
eED e e eeeeaae
e e e 0 0 000:000 ®® ®1 ® ®®®0
0)- -e e e 000 0 0 00,00 0
eeee
-eeeee eaeeeae ee 000 e
eee ae °000i00 0
e e ae e ae00 000 II
e ae 0 000 0
e a a 0- 62, 00
e a
a ea
Electric Power Generation: Thermal Conversion 369
tween cavities and external receivers remain even, and the less constraining requirements
of the external receiver lead to greater versatility and economics in the heliostat field.
According to the Sandia evaluation of the preliminary design studies as shown in
Fig. 38.11 [10] , the first commercial 100-MWe plant of the chosen design (including
6 h of thermal storage) will cost about 332 M$, whereas a mass-produced version, the
"Nth" plant of the same technology, would be expected to cost 210 M$ in 1977 dollars.
The cost goal for an advanced-technology plant is $1250/kWe . This goal is being ap-
proached along three avenues, the first two of which are already active (Fig. 38.10)
[12] : (a) Advanced central receiver systems studies to develop more efficient and less
costly overall systems; (b) low-cost heliostat development studies designed to develop
lower-cost heliostats through conceptual and prototype studies and through integra-
tion of mass manufacturing techniques; and (c) production of a sizable number of helio-
stats. A working schedule for several application projects that would require about
20,000 heliostats by 1985 is shown in Fig. 38.9 [12] . The larger modules scheduled
for 1985 and 1986 operation would require 10,000 to 60,000 heliostats each and would
justify construction of an efficient dedicated heliostat facility in 1982 or 1983. The
R&D proposed to support such development is shown in Fig. 38.12 [12] . Here
"repowering" refers to supplying a heliostat field to an existing generating plant
to operate in a "fuel-saver" mode. This requires the solar collector and receiver to
be designed to match the operating condition of the existing turbine. As this turbine
will be some years old, it will probably have mediocre efficiency, therefore excessive
collector size and cost may result. However, if alternative fuels are not available, or
their use is not allowed by governmental regulation, the utility may still consider this
a viable option and its implementation will aid in fostering the mass production of helio-
stats. In contrast, the hybrid system will be designed to take best advantage of the
solar collector, which will be backed up with fossil fuels rather than a large storage
unit. Full capacity credit will accrue to the unit, and fossil fuel use will be minimal
in an intermediate-load plant and perhaps halved in a base-load plant.
Solar thermal electric systems much smaller than 100 MWe are not likely to be considered
by utilities because the efficiency of both the solar collection system and of the turbine
system is lower. Consequently, both capital and operational costs per kilowatt-hour will
be higher. In addition, manning and dispatching a multitude of small plants becomes an
intolerable burden.
Figure 38.8 Heliostat distribution for a 10-MWe pilot plant representative of a cost-
optimized early commercial 100-MWe intermediate-load plant. Each small circle
represents the exclusion area of one heliostat. The tower is centered in the central circle
with the power house to the south, the thermal storage unit to the east, the water treat-
ment facilities to the southeast, and the assembly and warehouse area to the west. The
cooling tower is "downwind" of the field, presumably in a southerly direction. Helio-
stats range from 65 to 365 m from the tower, which supports the receiver with its center
line 80 m above the ground. A heliostat placed in "bare" areas would be severely blocked
by the heliostat in front.
CENTRAL POWER APPLICATIONS PROJECTS
APPROVED
SELECT START
CONCEPTUAL DETAILED
CAPACITY (MW) DESIGN DESIGN OPERATIONAL
TENTATIVELY PROPOSED
Figure 38.9 Approved and proposed central power applications projects. Requests for proposals with June 1978
due dates were issued by DOE for the Barstow detailed design and the Hybrid Critical Module conceptual design.
In April of 1980 over 50% of the heliostat foundations had been installed at the Barstow site, the tower steel had
been ordered, and the receiver and heliostat fabrication process was under way. Six utility repowering and six
industrial process heat retrofit conceptual design studies involving specific sites funded by DOE are nearing com-
pletion, and cogeneration proposals have been requested by DOE.
A2Jau3 m os Jo suo!iPoH ddv
Electric Power Generation: Thermal Conversion 371
- JANUARY, 1978
HELIOSTAT DEVELOPMENT
- JANUARY, 1978
Figure 38.10 DOE activities designed to produce more efficient and lower-cost central
receiver systems.
FIRST COMMERCIAL PLANT
TOTAL COST
332.2 MS
MASTER CONTROL
2.2 M
LAND/BUI LDINGS
5.5 M
DISTRIBUTABLES/INDIRECTS
21.2 M
TOTAL COST
210 MS
COLLECTOR
107 M THERMAL STORAGE
24 M
MASTER CONTROL
Ilbo EPGS 2M
14M
LAND/BUILDINGS
5M
DISTRIBUTABLES/INDIRECTS
15 M
CONTINGENCY
10 M
*Represents projected costs for the Nth central receiver plant, which
continues first generation technology. (It should be noted that an Nth
commercial plant probably would not use first generation technology.)
372
Electric Power Generation: Thermal Conversion 373
PROTOTYPE HELIOSTAT
PHASE 1 PHASE 2
RFPA A ' & A
ADVANCED RECEIVER
SYSTEM PHASE 1 PHASE 2
RFP • A A A
PHASE IA
REPOWERING A A
RECOMMENDATION
HYBRID
RFP PHASE I • PHASE 2
• • • •
Figure 38.12 Sandia proposed schedule for central power systems research and
development. Phase 1 of the repowering element has been delayed by about 18 months,
but will be completed in June 1980.
There are applications of smaller systems, however. There are requirements for
electricity (or shaft power) in isolated areas for agricultural or industrial applications.
Many farms, buildings, and companies have stand-alone power systems that require
expensive and/or scarce fuel supplies such as natural gas, propane, low-sulfur oil, or
diesel fuel. Interruptable power can readily be supplied for any of these requirements
by solar thermal power systems in the 10-KW, to 10-MW, range. At the lower end of
this range a small array of dishes or troughs is all that is required, and collecting the heat
from 10 or 20 units is no real problem. In the range of 1 MWe there are many com-
peting systems. One must compare arrays of collectors, each driving its own small heat
engine, with distributed arrays driving a central heat engine through an array of plumb-
ing lines, with small central receivers using partially focusing heliostats in a north field
array. The costs and operating complexities of each of these systems are still not well
Figure 38.11 Plant cost estimates for a 100-MWe intermediate-load commercial plant.
Sandia revised cost estimates for the subsystem designs recommended for the Barstow
pilot plant; ratio of peak receiver power to turbine rating of 2; storage capacity of 7 h
at 70 MWe.
374 Applications of Solar Energy
defined. The author feels that central receivers will eventually become preeminent for
most electrical generating applications above 1 MWe . The trade between cavity and ex-
ternal receivers will depend on the power level and the operating temperatures. This
area is currently the subject of a study monitored by the Jet Propulsion Laboratory for
DOE. In the 0.5- to 5-MWe range, central receivers, an array of parabolic dish thermal
collectors powering a central turbine, and an array of parabolic dishes with individual
heat engines will be compared by three industrial contractors. After an initial 10-month
conceptual design ending March 30, 1978, this program is scheduled to proceed to an
18-month design and construction phase followed by several years of test operations.
Although construction, operation, and maintenance costs of these smaller systems will
be difficult to keep tolerable, they offer promise as stand-alone interruptable systems
for isolated areas, as fuel savers in many applications, and as a possible means of reducing
distribution costs and reliance on the utility grid.
Statements are frequently heard that solar thermal electric technology can have no
impact until the year 2020 or later. This will undoubtedly be true, in spite of utility
and industry interest in central receiver systems, if sufficient government support is not
forthcoming to achieve a dedicated heliostat production plant capable of producing at
least 20,000 heliostats per year, and to keep it in operation for 5 or more years. At this
point sufficient experience coupled with increasing demand for heliostats at costs of 60
to 80 $/m2 could lead to construction of a truly large-scale integrated heliostat produc-
tion facility producing 100,000 or more heliostats per year. Further cost reduction to
the range of 55 to 60 $/m2 could lead to construction of one such production facility
each year. This growth pattern, if aggressively pursued [13] , Fig. 38.13 [14] , would
lead to the displacement of about 106 bbl of oil (6 X 1015 J) per day in the year 2000
and 20 times this amount is 2121. This would correspond to meeting the entire inter-
mediate load-generating capacity in the southwestern portion of the United States,
amounting to about 8000 plants of 100-MWe capacity. Adequate suitable land is
available. More rapid growth than this would require penetration into other markets
such as base-load electrical, process heat, or high-voltage transmission out of the South-
west [13] . Extension of the region of installation beyond the high-insolation "sun bowl"
of the Southwest is also possible, but a substantial penalty will result because of the
lower available sunlight. Whether this is competitive will depend on the availability and
cost of fossil fuel used in a nonpolluting manner. It should be noted that a very large
capital investment is involved in such a program. However, more than half of the invest-
ment would be required to build the required fossil fuel generating plants if the solar
option were not implemented. A substantial portion of the remainder would represent
required investment in mining and transportation. Thus, the overall capital expenditure
is only increased by a few tens of percent, although certain economic displacements
would occur, such as transfer of some demand from coal mining to iron mining, and
so on.
• PLANT TYPE:
• CENTRAL RECEIVER
2
• HELIOSTAT AREA/STORAGE — 1 km /6 hr
1000 8000 • INTERMEDIATE MODE
• SITING CONSTRAINT
10 A INTERMEDIATE / _. 80 • SUITABLE AREA (SW US) — 21,500 sq mi
CAPACITY / IN.
INSTALLED • COOLING WATER
(including /
1
A INTERMEDIATE • ENVIRONMENTAL IMPACT
replacement I SOLAR CAPACITY --- 40
i INSTALLED — 32 • LAND USE (1 sq mi/100 MWe)
\/ 24 • NO POLLUTANTS
i INTERMEDIATE • AESTHETIC
CAPAC ITY-1000
/ SOLAR CAPACITY — 16
,1985 = 100 MWe I (growth — 50%/yr) • INSTITUTIONAL
1 / I 8 • HIGH CAPITAL INVESTMENT COST
1960 1970 1980 1990 2000 2010 2020 2030 • SOLAR $1360/KWe
YEAR • CONVENTIONAL $343/KWe
Figure 38.13 Aerospace projection of the intermediate-generation capacity in the Southwest and the possible market
capture potential for solar thermal conversion. The indicated solar capacity growth curve can be reasonably approximated
by bringing on line, each year starting in 1988, one additional factory with an output of ten 100-MW, solar plants.
376 Applications of Solar Energy
REFERENCES
5. G. Francia, "Pilot Plants of Solar Generating Stations," Solar Energy 12, 51 (1968).
6. F. Trombe, "Solar Furnaces and Their Applications," Solar Energy 12, 9 (1957).
7. D. M. Darsey et al., Solar Thermal Test Facility Experiment Manual, Report SAND
77-1173 prepared by Sandia Laboratories for the DOE under contract AT(29-1)-
789, Albuquerque, New Mexico, October 1977.
8. R. W. Hallet, Jr., and R. L. Gervais, Central Receiver Solar Thermal Power System,
Phase 1, McDonnell Douglas Astronautics Company, Final Report on DOE contract
EY-76-C-03-1108, NTIS report number SAN/1108,vols. 1 through 7, Huntington
Beach, California, October 1977.
9. L. L. Vant-Hull and C. R. Easton, Solar Thermal Power Systems Based on Optical
Transmission, Reports NSF/RANN/SE/GI-39456/PR/73/4, PR/74/2, and FR/75/3,
University of Houston, Houston, Texas, February 1974, December 1974, and
October 1975.
10. Anonymous, Recommendation for the Conceptual Design of the Barstow, California,
Solar Central Receiver Pilot Plant—Executive Summary, Report SAND77-8035,
Livermore, California, October 1977.
11. Proceedings of The ERDA Solar Workshop on Methods for Optical Analysis of
Central Receiver Systems, Organized by the Solar Energy Laboratory of the Uni-
versity of Houston for Sandia Laboratories Under ERDA Contract AT(29-1)-789,
Houston, Texas, August 10-11, 1977.
12. A. C. Skinrood, "Technical Overview—Solar Thermal Central Power Systems Pro-
gram," presented at the Department of Energy Central Power Systems Semi-Annual
Review, San Diego, California, March 9, 1978.
13. L. L. Vant-Hull, "Solar Energy Resources—Electric Power Generation," in Proceed-
ings of the Conference on Magnitude and Deployment Schedule of Energy Resources,
Portland, Oregon, July 21-23, 1975, edited by Loveland, Spinrad, and Wang, Oregon
State University, Corvalis, Oregon, September 1975.
14. Anonymous, Solar Thermal Conversion Mission Analysis—Southwestern United
States, NSF/RANN Report No. ATR-74(7417-16)-2, vol. 1, under contract NSF-
C797, The Aerospace Corporation, El Segundo, California, January 1975.
15. Anonymous, Solar Porgram Assessment: Environmental Factors—Solar Thermal
Electric, Report No. ERDA 77-474, Environmental and Resource Assessments
Branch, Division of Solar Energy, ERDA, Washington, D.C., March 1977.
16. M. Davidson and D. Grether, The Central Receiver Power Plant: An Environmental,
Ecological, and Socioeconomic Analysis, prepared for ERDA under contract No.
W-7405-ENG-48, LBL, University of California, Berkeley, California, June 1977.
17. A. C. Meyers, and L. L. Vant-Hull, section 2, of the Final Report, part II, Liquid
Metal Cooled Solar Central Receiver Feasibility Study and Heliostat Field Analysis,
prepared by the Solar Energy Laboratory for ERDA under grant No. EG-76-G-05-
5178, University of Houston, Houston, Texas, May 1978. Summarized in section
III-B of the Proceedings of the 1978 annual meeting, American Section of the
International Solar Energy Society.
chapter 39
ROBERT COHEN*
Division of Ocean Energy Systems
Office of Solar Power Applications
U.S. Department of Energy
Washington, D.C.
39.1 INTRODUCTION
The earth's oceans intercept solar radiation, much of which they convert to and store as
thermal energy. On the other hand, cold waters near the freezing point circulate from
polar regions toward the equator at depths less than 1000 m. Thus, a vertical tempera-
ture difference amounting to 21°C (38°F) or more exists throughout the year at many
tropical and subtropical locations. This temperature gradient can be exploited as an
electrical energy source, as was first pointed out by d'Arsonval in 1881 [1] .
*The views expressed herein are those of the author, and are not necessarily, in all or in part, those of
the U.S. Department of Energy.
379
380 Applications of Solar Energy
The process to accomplish this, now being referred to as ocean thermal energy
conversion (OTEC), requires that the warm surface waters and cold water from depth be
brought into proximity so they can act as the heat source and heat sink, respectively, for
a heat engine. In other words, solar energy—collected and stored as heat by the world's
major oceans—can be converted into electricity through a generation process similar to
that of conventional power plants, except that in the case of OTEC, no depletable fuel
is required. Furthermore, although there is some seasonal variation in the ocean thermal
resource at a given OTEC power plant location, there is little diurnal variation. Accord-
ingly, OTEC power plants are analogous to solar hydropower plants in that they smooth
out the diurnal intermittence of the solar radiation, in contrast to other solar-electric
energy options. Thus, OTEC power plants provide a potentially substantial renewable
source of base-load electricity, albeit located mainly at sea.
Although it is possible to find good land sites where OTEC power plants can be
located, by bringing the warm and cold water onto shore via aqueducts, it is clear that
such opportunities will be much more limited on a global basis than the ample opportun-
ities for generating substantial amounts of OTEC electricity aboard floating OTEC
platforms. This is both because of the special technical requirements for onshore OTEC
plants, and because of the limited market potential (at least in the near term) for OTEC
electricity at such sites. Onshore OTEC power plants will be viable mainly at locations
where three requirements are all simultaneously satisfied with satisfactory economics:
(a) coastal-zone land must be available; (b) the sea floor must descend sufficiently
rapidly from the shore-based plant location; and (c) the seasonal availability of warm
and cold water—without undue degradation by the warm and cold water effluents from
the OTEC plant—must meet certain criteria. In any event, it is probable that available
and attractive onshore and near-shore OTEC power plant locations will be populated
early in the development and implementation of the OTEC concept, both as convenient
locations for pilot and demonstration plants, and because they will constitute attractive
intermediate markets for OTEC electricity and by-products. In fact, likely island
locations for OTEC are typically in short supply of fresh water and fish protein, both
of which might be provided as adjuncts to OTEC operation, as pointed out by Othmer
and Roels [2].
This chapter constitutes a description of the background, status, and future pros-
pects of OTEC technology. This is a controversial subject, since there are widely differing
perceptions as to the appropriate rate and manner of OTEC development. Also, some
have reservations as to the viability of OTEC power generation, which relies on unfamil-
iarly small temperature differences. For example, with warm water at 26°C (79°F) and
cold water at 5°C (41°F), the temperature difference of 21°C (38°F) corresponds to a
theoretical Carrot efficiency of about 7 percent. However, because of the OTEC require-
ment for parasitic power (such as for pumping up the cold water supply) and other losses,
the achievable net conversion efficiency is only about 2.5 percent. This compares to net
efficiencies of 30 to 40 percent associated with conventional power plants.
Some engineers question whether such an extremely low net efficiency will ever
allow OTEC to become economically viable. However, it is important to consider the
matter in more sophisticated terms than net efficiency, since in the case of OTEC there is
no fuel cost, only the requirement to pay for circulating much more warm and cold water
Electric Power Generation: OTEC 381
than is normally associated with power generation. This means that extensive areas of
heat exchangers will be required for "closed-cycle" OTEC plants (which would employ a
working fluid such as ammonia), or that degasifiers (to remove gases dissolved in the sea
water) and tremendous turbines would be required for "open-cycle" OTEC plants that
would operate by the flash evaporation of sea water. Thus, although the net efficiency
of the OTEC plant must certainly be positive and as high as is readily attainable, the key
economic question is the resulting cost of OTEC electrical energy, not the actual value of
the net efficiency. Also, OTEC economics are frequently perceived as being linked and
relevant to the economics of "bottoming cycles," which would use the rejected heat from
conventional power plants to operate a Rankine power cycle much like OTEC. However,
the viability of bottoming cycles and the viability of OTEC cycles are really separate
technical and economic questions with relatively independent answers.
At this point a simple comparison may lend some insight into the energy supply
potentially available from OTEC. A temperature difference is analogous to a hydraulic
head; each centrigrade degree corresponds to 427 m of head (each Fahrenheit degree
corresponds to 778 ft of head). Thus, a temperature difference of 21°C would corre-
spond to having the water at a height of 8967 m if 100 percent conversion efficiency
were attainable. However, although this hydraulic head can be converted into a 21°C
temperature increase, the process is not reversible; that is, a temperature difference of
21°C cannot be converted into the equivalent hydraulic head, but only to the percentage
of the theoretical head corresponding to the net thermodynamic efficiency. Thus, a net
2.5 percent conversion efficiency would enable a height of 224 m to be attained. The
analogy of OTEC to solar hydropower is again evident. Thus, it is as if much of the
world's ocean water were captured behind invisible "thermal dams" of signficant heights.
The question then remains whether this inconspicuous thermal potential head can be
harnessed usefully and economically.
Siting of OTEC power plants involves both the ocean thermal resource require-
ments of plant operation and the market potential for OTEC products. The marketing
of OTEC products from sea-based power plants is possible via two basic avenues: (a)
submarine umbilicals (such as electrical cables, or pipelines for gases such as hydrogen,
ammonia, and compressed air) and (b) marine transportation of energy-intensive products
(such as ammonia, hydrogen, chemicals, and metals) derived from OTEC electricity. In
OTEC applications where submarine umbilicals are required, the additional capital invest-
ment for such umbilicals will obviously have to be provided, along with the cost of
station keeping. Where OTEC products are transported in other ways, there are much
less stringent station-keeping requirements, and in fact the OTEC plant can "graze" to
ensure an optimum thermal resource input. Let us now consider OTEC siting and the
ocean thermal resource in more detail.
Siting of OTEC power plants must take into account the availability of an adequate
thermal resource, station keeping, possible impacts on the environment, possible impacts
of the environment on plant design and operation, ocean logistics, and—where submarine
382 Applications of Solar Energy
The results of the computer and laboratory modeling experiments are also being
used to estimate the horizontal and vertical distributions of temperature, salinity,
nutrients, species, and substances (such as working fluid and biocides) that might be
discharged by OTEC power plants. These distributions can be considered in the immedi-
ate vicinity (or "near field") of an OTEC plant, or in the "far field," that is, at distances
where the effluent from a single OTEC plant is controlled by ambient conditions. Also,
the combined large-scale effects of a number of OTEC power plants operating at appro-
priate spacings can be analyzed. Computer or "numerical" modeling seems better
adapted than experimental or "physical" modeling for handling large-scale, far-field
effects, whereas physical modeling is preferred—but numerical modeling is also
amenable—for simulating near-field regions (up to a few kilometers away from the OTEC
plant). It is premature to reach firm conclusions regarding all possible OTEC environ-
mental effects. However, from the results to date, it appears that local surface tempera-
tures will not be depressed more than a fraction of a degree centigrade (if recirculation is
avoided), hence no significant local or regional climatic effects would be expected.
The parameters of the ocean environment that are likely to affect the design and
operation of OTEC plants have been cataloged for relevant ocean areas by Bretschneider
[7] . He analyzed the sea state and design waves for potential OTEC sites on a worldwide
basis, and included the effects of winds, waves, and surface currents. Also, the ocean
thermal resource available to OTEC plants at certain locations may be disrupted by
coastal upwelling and other phenomena, as discussed by Thompson, Hurlburt, and
Lin [8] , and is reduced for periods of a few days at some locations following hurricanes,
which also extract ocean thermal energy.
The global ocean thermal resource can be mapped, as shown in Fig. 39.1, in terms
of annual average monthly temperature differences between the surface waters and the
water at a depth of 1000 m. The most valuable ocean thermal resources should offer
average temperature differences of about 20°C or greater. The contours of greatest
interest in Fig. 39.1 are therefore those for 20, 21, 22, 23, and the nominal 24°C contour
(which corresponds to temperature differences of 24°C or larger). A considerable region
of the globe thus appears advantageous for OTEC exploitation, including many locations
accessible to land via submarine cable and extensive ocean areas where OTEC plant ships
could produce energy-intensive products at sea. This region happens to coincide geo-
graphically with the locations of numerous developing nations and, if commercially
viable, could provide a substantial addition to world energy resources. For applications
where mooring of the plant is required, there are practical upper limits of about 2000 m
for tolerable mooring depths. In such cases, sites of interest would be limited to regions
where ocean depths ranged from about 1000 to 2000 m.
Although the ocean thermal resource at a given location is quite stable from day to
day, there is a seasonal variation of the resource. The amplitude of variation increases
with distance of departure north and south of the equator. There are substantial ocean
thermal resources in the Gulf of Mexico available to the continental United States via
submarine electrical cable. Also, U.S. islands such as Puerto Rico, the Virgin Islands,
Guam, and the Hawaiian Islands have excellent ocean thermal resources very close to
their shores.
An example of the monthly temperature differentials available at a location 230 km
west of Tampa, Florida, is shown in Fig. 39.2. The annual variation of AT between the
surface waters and a depth of 1000 mat that site is shown in Fig. 39.3, together with
160W 150W 140W 130W 120W 110W 100W 90W 80W 70W 60W 50W 40W 30W 20W 10W OW 10E 20E
40N
40t1 .....
.•:.*** prkx;
:P:. 1
•• • , •16 • Np5C1i5C5% ••*. ..: •
K., '-.453i.eft
..:•:..•:4:,.*. - • 58' • •.• ..* . • ......
•18•••• atioiAktiu
30N .....17 30N
• ::$:::::...... 146wo oh., 1111110::k:;:i!,,:t.:•... -iqt/11:41&
•:tie • ••••:* ... %az" 1 , 11 20'
21' . - iiilphigitk4i*Wn
' oe. . . ;.% • ils,X0i 1 1 il 1 . P----'N: 11111M101111Iligagkat:* r
20N 6.,,a6481 .39', •' 20N
d 1•1111111 • - -;!il::;i:111111111
I. 'Iniiiirilli -20' • ' % QRSIRRIP
IIII I III 111 I :
...„
:t imiiiiiiii0111 II, 21' • 0
'11:22' lib, 0 ItLA.)::$
10N ....iii11111111;111(11 10N
11111111 1 111111111in b.,...............i.,
1111 11111111
EO II IIII1111 1 ill 1 11 1 11111 111 1111111 11110111111111011.12gri EO
20N 20N
109 109
EO EO
10S 10S
20S 20S
30S 30S
40S 40S
30E 40E 50E 60E 70E 80E 90E 100E 110E 120E 130E 140E 150E 160E 170E 180E 170E 160W 150W
50
100 100
150 150
200 200
250 250
300 300
350 350
14
400 400
450 450
500 500
550 550
600 600
D EPT H INMETERS
650 650
700 700
750 750
23
800 800
19
850 850
900 900
950 950
1000 1000
1050 1050
1100 1100
1150 1150
1200 1200
1250 1250
1300 1300
1350 1350
1400 1400
1450 1450
1500 1500
Figure 39.2 Contours of most probable monthly temperature differentials (in °C)
between surface and depths up to 1500 m in Section 29G west of Tampa, Florida.
similar variations for sites off New Orleans, Key West, and Brownsville, Texas. Potential
sites in the Gulf of Mexico for providing electricity to U.S. shores are included in the
shaded area of Fig. 39.4, which indicates the oceanic region having annual average AT
20.6°C at ocean depths between 1000 and 2000 m. The annual average OT's at three
Gulf of Mexico locations are compared in Fig. 39.5 with values for a site east of Brazil
and for sites near Puerto Rico and Hawaii. (It would probably be preferable to use as a
Electric Power Generation: OTEC 387
27
48
26
46
25
44
24
4 42
23
40
22
\
a 21 ‘ 38
16
28
15
26
14 Legend
-*-111- New Orleans/Mobile - 24
13 • f West Coast Of Florida
-41-111r Key West - 22
12
AO. ---e- Brownsville. Texas
11
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan
Months
statistic the root-mean-square (RMS) AT rather than the average AT, since the net power
obtainable from a given OTEC plant varies with (AT)2 in the neighborhood of the plant's
design temperature, as will be discussed in Sec. 39.5. However, in practice these statistics
differ only sightly; that is, the RMS AT exceeds the average AT by about 0.2°C.)
The seasonal variation of the ocean thermal resource, hence of the output power
from OTEC power plants, is probably quite advantageous, fortuitously, at least for pro-
viding electricity to the seasonally varying electrical load in the southern United States.
There, because of the significant winter-to-summer load variation, present utilization of
fixed-baseload power plants is inefficient from the standpoint of matching to the load.
However, the combination of appropriate mixes of seasonally varying OTEC power with
95 90° 85° 80°
30' 30°
BROWNSVILLE
.?.."
25 25° N
42 1,1111,1,1 1P
23
40
22
38
21 z
36 20 <,
34 19 r"
17
30
5500
Average Monthly Power Load I MWel
5000
4500
4000
I 1 1 1 I I 1 I 1
D J F M A M J J A S 0 ND J
Figure 39.6 A comparison of the 1976 seasonal variation of electrical load served by the
Florida Power and Light Company with the combination of the projected seasonal varia-
tion of 1500 MWe of OTEC capacity, obtained from OTEC power plants located 230 km
west of Tampa, combined with 2744 MWe of fixed-baseload capacity.
fixed-baseload power (such as from coal and nuclear power plants) would match the
seasonal load variations quite nicely, as shown in Fig. 39.6.
Two basic power cycles have been advocated for conversion of ocean thermal energy:
the so-called open cycle, employing sea water as the working fluid, and the so-called
closed cycle, utilizing other working fluids (such as ammonia, hydrocarbons, or halo-
carbons). The first published work on OTEC by d'Arsonval in 1881 [1] suggested a
closed cycle, and that article proposed sulfur dioxide as the working fluid. However,
the first OTEC experiments by Claude in the 1920s [9] utilized an open cycle, where
sea water was evaporated under a partial vacuum. In this section, these two system
approaches are treated separately, and the subsection on the open-cycle system includes
a discussion of several variations of that approach.
A. Closed Cycle
A schematic diagram of a closed-cycle OTEC power plant is shown in Fig. 39.7. Heat
exchangers known as evaporators and condensers are a key ingredient, since extensive
areas of material are needed to transfer significant amounts of low-quality heat at the
low temperature differences being exploited. In other words, large volumes of water
Electric Power Generation: OTEC 391
WARM
WATER
INTAKE /".
COLD
WATER
OUTLET t
PUMP
r"
3. ire
CD
EVAPORATOR CONDENSER z
37.
73 PUMP
PUMP
WARM
COLD
WATER
WATER
OUTLET
INTAKE
must be circulated through the OTEC power plant, requiring commensurately large heat
exchangers. The actual components employed in an OTEC closed-cycle system would
appear more like the hardware illustrated in Fig. 39.8, another closed-cycle schematic.
A fundamental requirement in closed-cycle systems is to transfer heat efficiently
across the heat exchange surfaces constituting the evaporators and condensers, so as to
achieve a high value of overall heat transfer coefficient, U, measured in watts per Kelvin
per square meter, or W/K • m2 (more commonly, in Btu/h • °F • ft2 ). For the evapora-
tor, this overall heat transfer coefficient is a measure of how efficiently heat is transferred
sequentially from sea water through the heat exchanger material (a metallic alloy) and
thence to the working fluid (e.g., ammonia). For the condenser, an overall U character-
izes the reverse heat transfer process.
In an ocean environment, it is likely that a layer of slime known as "biofouling"
will eventually accumulate on the water side of the heat exchangers. Such slime is first
comprised of microorganisms, at which stage the biofouling is called "microfouling."
Subsequently, if the slime is not removed, additional biofouling in the form of macro-
organisms will become attached, augmenting the slime layer. The occurrence of micro-
fouling seems to be a prerequisite for the attachment of macroorganisms. A film of
corrosion and possibly of calcareous (i.e., mineral) deposits can also accumulate on the
water side (and conceivably—through leakage—even on the working-fluid side) of the heat
transfer surfaces. The total formation of biofouling, corrosion, and so on, is referred to
as "fouling" (or "scaling") and will tend to inhibit heat transfer through it. The "fouling
392 Applications of Solar Energy
EVAPORATOR
SURFACE WATER WARM
700-85° WARM WATER OUT
AMMONIA LINES
GENERATOR
PUMP
AMMONIA TURBINE
TURBINE EXHAUST
PUMP
COLD WATER OUT
CONDENSER
DEEP COLD
WATER PUMP
40°
factor" is a measure of the thermal resistance, Rf , of a fouling film. This thermal resist-
ance is the reciprocal of the corresponding heat transfer coefficient, hf, of the fouling
film. To maintain viable OTEC heat exchangers, provision must be made to inhibit the
formation of fouling layers and to remove any significant fouling that forms. Removal
can be accomplished by periodically cleaning the heat exchanger surfaces through
mechanical, chemical (within EPA-acceptable limits), or other means.
The overall heat transfer coefficient, U, is the resultant of combining the heat
transfer coefficients involved in each step of the process. The reciprocal heat transfer
coefficients (i.e., thermal resistances) are combined additively to calculate the overall
thermal resistance, R = 1/U. Thus, the case of an aluminum heat exchanger employing
ammonia as a working fluid, the heat transfer coefficients on the ammonia and sea-water
sides combine with those of the fouling layer (assumed here to be on the sea-water side
only) and of the aluminum to give an overall heat transfer coefficient calculated as
follows:
(39.1)
U hNH 3 hH20 hAl hf
We note from this equation that relatively high values of h contribute toward
achieving a high U, whereas relatively low values of h are detrimental. In other words, a
high fouling factor (low hf) should be avoided, and the h's for the ammonia and sea-water
sides should be as high as feasible. Augmentation of heat transfer is being studied for
both the working-fluid sides and sea-water sides, by using special (e.g., fluted) metallic
Electric Power Generation: OTEC 393
surfaces. However, such surfaces result in additional cost, so that a trade-off must be
considered between factors such as the cost effectiveness of heat transfer enchancements,
the producibility and joinability of special surfaces, the ability to control the biofouling
thereof, and other power system parameters. For the OTEC application, a likely possi-
bility is to employ heat transfer enhancement only on the working-fluid side of the heat
exchangers.
Heat exchangers for conventional process heat and power system applications
operate with U's having typical values of about 1700 to 2300 W/K • m2 (300 to 400
Btu/h • °F • ft2 ). However, because of the special requirements of OTEC power systems,
it is desirable to achieve U values of at least twice these rates. This means that Rf (the
reciprocal of hf) should be kept as low as possible, preferably no greater than 2 to 5 X
10 -5 m2 • °K/W, and that augmentation of heat transfer should be considered, at least
on the working-fluid side.
It is anticipated that biofouling on heat exchangers located in open ocean waters
will not develop as rapidly as at most coastal locations, where nutrients are more abund-
ant. Measurements of the thermal resistance, Rf, of a biofouling layer forming on the sea-
water side of aluminum and titanium tubes located in water masses characteristic of the
open ocean have been conducted in Hawaii and Saint Croix, V.I. The rates of accretion
(cf. Fig. 39.9) were such that the thermal resistance of the biofouling layer built up at the
BIOFOULING TESTS
KEAHOLE POINT, HAWAII
ST. CROIX ISLAND, VI
1976.1977
6 — a Al (6061-76): 6 Ft/Sec
A Ti (B337-76): 6 Ft/Sec zm PENALTY GOAL.
Al 16061.761: 3 Ft/Sec 25% 1977
5
d
• 4
3
15% '978
Tr. O
2
MAN BRUSH
5%
0 a
2 4 6 8 10 12 14 16 18
TIME (WEEKS)
Figure 39.9 Time variation of fouling factor measured in Hawaii and the Virgin Islands,
indicating the consequences of wiping heat exchanger tubes after 16 weeks with M.A.N.
(trade name) brushes. Titanium and aluminum tubes were tested at the water flow
velocities indicated. The fouling factor, Rf, measured in ft2 • h • °F/Btu is converted to
m2 • K/W by multiplying by the factor 0.176. A given fouling factor is associated with a
heat transfer penalty shown in percentages on the right-hand axis of ordinates.
394 Applications of Solar Energy
rate of about 2 X 10-s K • m2 /W) per week, and preliminary results indicate that the
layer was amenable to cleaning with a commercial brush produced by the M.A.N. (trade
name) organization. The apparatus used to measure the thermal resistance of a biofouling
layer was developed by John Fetkovich [10] of Carnegie-Mellon University, for deploy-
ment from ocean buoys and from vessels.
Even though biofouling can be inhibited by the use of biocides such as chlorine
(through continuous or intermittent dosing), provision for mechanical and/or chemical
cleaning is regarded (from the standpoint of having a fail-safe method) as an important
adjunct or substitute. Two mechanical devices presently in use to clean heat exchanger
tubes are the M.A.N. brush and the Amertap (trade name) sponge-rubber ball. Other
techniques include abrasive slurries and water jets. In order to test biofouling counter-
measures and control systems under the accelerated rates of slime accretion available in
coastal waters, the U.S. Department of Energy has established a heat exchanger cleaning
test facility at the Naval Coastal Systems Laboratory, Panama City, Florida. A schematic
drawing of that installation is shown in Fig. 39.10.
Although corrosion of OTEC heat exchangers would probably not be a problem if
they were made of titanium, that metal is usually regarded as being a somewhat costlier
alternative than use of some other candidate metals. In particular, aluminum is regarded
as a strong candidate from the standpoint of cost, if it can be qualified technically. The
key technical problem is the ability of a material to withstand erosion and corrosion in
conjunction with mechanical cleaning methods in the presence of sea water and ammonia.
Indications from previous experience are that aluminum surfaces maintained free of
biofouling will not corrode as readily as those where biofouling is allowed to accumulate.
Alloys under consideration for OTEC heat exchangers include aluminum alloy 5052 (con-
taining 2.5 percent magnesium), stainless steel (AL-6X), and copper alloy 706 (10 percent
nickel). With ammonia as a working fluid, there is some concern as to whether the
resulting corrosion rate of copper-nickel would be tolerable [11] . Plastics have also been
mentioned as a possible candidate heat exchanger material [12] .
Calcareous deposits can form if there is leakage of ammonia into sea water. This
process is associated with the resulting increase of pH. Similar deposits and or corrosion
could result on the working-fluid side if sea water were allowed to leak into ammonia.
Although efforts can be made to design OTEC heat exchangers that are leakproof, a small
percentage of heat exchanger tubes that leak can be tolerated by detecting such leaks and
plugging the offending tubes.
Other components of the closed cycle include turbines and demisters. Ammonia
turbines will require special seals to contain the ammonia, and some additional develop-
ment will probably be required to attain optimum sizes at maximum efficiency.
Demisters for ammonia service pose no special problems. The ammonia working fluid
needs to be kept free of water concentrations above 0.1 percent to prevent degradation
of system performance [13] and to avoid corrosion. Accordingly, the ammonia closed-
cycle system will need to include provision for removing water from the ammonia.
"Open cycle" refers to the utilization of sea water as the working fluid, wherein sea water
is flash evaporated under a partial vacuum. The low-pressure steam is passed through a
HEAT EXCHANGER
CLEANING TEST FACILITY
1 SEAWATER INTAKE 7 RECIRCULATING SPONGE
BALL CLEANING SYSTEMS 121
2 SUCTION PUMPS
3 FEED TANK B FLOW DRIVEN BRUSH
GLEAMING SYSTEMS (21
4 FEED PUMPS
9 INSTRUMENTATION BUILDING
5 SEAWATER MANIFOLD !CONDITIONING EOUIPMENT/
5 CONTROL HEAT TRANSFER 10 INSTRUMENTATION TRAILER
UNITS 141
11 DATA ACQUISITION SYSTEM
Figure 39.10 The OTEC heat exchanger cleaning test facility located at the Naval Coastal Systems Laboratory, Panama City, Florida.
396 Applications of Solar Energy
turbine, which extracts energy from it, and then the spent vapor is cooled in a condenser.
This cycle derives the name "open" from the fact that the condensate need not be
returned to the evaporator, as in the case of the "closed" cycle. Instead, the condensate
can be utilized as desalinated water if a surface condenser is used, or—if a spray (direct-
contact) condenser is used—the condensate is mixed with the cooling water and the
mixture is discharged back into the ocean. A schematic diagram of the open-cycle system
is shown in Fig. 39.11. Since the early OTEC experiments performed by Claude [9]
utilized an open cycle, the open-cycle system is sometimes referred to as a "Claude
cycle."
Because of the need in the open cycle to harness the energy in low-pressure steam,
extremely large turbines (comparable to wind turbines) must be utilized. Furthermore,
degasifiers (deaerators) must be used to remove gases dissolved in the sea water unless one
is willing to accept large losses in efficiency. On the other hand, since there are no heat
transfer problems in the evaporator, the problem of biofouling control is minimized.
The cost of an open-cycle system for producing substantial numbers of megawatts
is presently regarded by most OTEC workers as being significantly greater than for a
closed-cycle system. An evaluation of costs for an open-cycle system was recently com-
pleted by Watt, Mathews, and Hathaway [14]. The turbine cost constituted almost half
the cost of the power system, but may be amenable to reductions that could result from
design innovations.
There are several variations on the standard OTEC open-cycle system. One varia-
tion is the "hybrid cycle," which is an attempt to combine the best features and avoid the
worst features of the open and closed cycles. First, as shown in Fig. 39.12, sea water is
EXHAUST COMPRESSOR
ro->---0NONCONDENSABLE GASES
STEAM TURBINE/GENERATOR
DEAERATOR
FLASH
EVAPORATOR
L.
BAROMETRIC
CONDENSER
xi
.6
I / /
WARM WATER WATER WATER
t
Figure 39.11 Schematic of the
COLD WATER OTEC open cycle.
Electric Power Generation: OTEC 397
DEAERATOR r n NH3
TEAM EVAPORATOR
STEAM NH 3
CO NDENSATEaCONDENSER
~
FLASH
EVAPORATOR
-
NH3
t
WARM WATER WATER
WATER
COLD WATER
Figure 39.12 Schematic of the OTEC hybrid cycle, combining features of an open cycle
with features of a closed ammonia cycle.
flash evaporated to steam, as in the open cycle. The heat in the resulting steam is then
transferred to ammonia in an otherwise conventional closed Rankine-cycle system. A
comparative study of the closed (ammonia), open (steam), and hybrid cycles showed the
closed-cycle system to be most economical in cost and to require the least parasitic
power [15].
Several other approaches to ocean thermal energy conversion systems have also
been suggested and are being investigated. They may all be regarded as variations on the
open cycle. An idea proposed by Beck [16] and patented by him [17] was to utilize the
heat in sea water to create a column of water through producing cavitation bubbles, as in
an airlift pump. Studies of such a steam lift pump have subsequently been reported by
Beck [18]. An advantage of this approach is that a hydraulic head is produced, thus
allowing a hydraulic turbine to be used, rather than requiring a gas turbine. Zener and
Fetkovich [19] suggested that the two-phase mixture of Beck have a foam structure.
Ridgway [20] proposed that warm sea water be introduced as a mist that is then lifted
against gravity by the flow of steam from a higher-pressure region to a lower-pressure
region.
Figure 39.13 is a schematic diagram of these "lift-cycle" approaches. They are
analogous to the naturally occurring hydrological cycle that leads to the production of
solar hydropower. In the case of OTEC lift cycles, an artificial hydrological cycle is
created within a large, ocean-going vessel. A 10-MW, version of a mist-flow OTEC
power plant concept by Ridgway [20] is shown in Fig. 39.14. The bubble, foam, and
398 Applications of Solar Energy
BUBBLES,
SEPARATOR
MIST, VAPOR
OR
FOAM
ADDITIVES NON-
LIQUID
(IF USED) CONDENSIBLES
A
V
/
NUCLEATION HYDRAULIC
CHAMBER CONDENSER
TURB NE
Figure 39.13 Schematic of the OTEC steam lift-cycle concept, wherein the ocean
thermal gradient results in the lifting of water bubbles, mist, or foam. The potential
energy of the elevated liquid water is used to propel a hydraulic turbine.
mist approaches thus convert a "temperature head" into an hydraulic head. They are
advanced concepts that offer certain attractive features and are being investigated. How-
ever, they present a number of practical problems such as the potential stability and/or
instability of the bubbles, foam, mist, and of the associated ocean platform. Some of
these questions are discussed in an exchange between Henrie [21], Beck [22], and Zener
and Fetkovich [23].
OTEC power plants will usually be floating structures consisting of a power system
contained by an ocean platform. The platform can be connected to shore via a sub-
marine umbilical to convey its products in the form of AC or DC electricity, or via
compressed air, liquid ammonia, and gaseous hydrogen. If an umbilical is utilized, then
there will be a need for platform station keeping. This can be accomplished by anchoring
and mooring, and/or dynamic positioning. On the other hand, if OTEC products are
transported to market via barge or other vessel, the requirement for station keeping is less
stringent.
Electric Power Generation: OTEC 399
WARM WATER
INLET
50 METERS
Figure 39.14 Schematic of a 10-MWe pilot plant that would utilize a mist-flow concept
proposed by Ridgway of R&D Associates.
Cold water pipes up to about 40 m in diameter have been considered for OTEC
plants, but it will probably be possible to reduce the required diameter somewhat if water
flow velocities are increased. Methods for deploying the cold water pipe and interfacing
it with the OTEC platform need to be developed. Also, the design of the cold water pipe
will require some advancement in the state of the art. Two CWP deployment approaches
have been suggested. One is the "float-and-flip" technique, where the pipe is towed to
the OTEC station and flipped into position. The other technique, reminiscent of drilling
procedures, is to deploy the pipe vertically, section by section.
Calculations of bending moments for representative cold water pipes indicate that
significant stresses can exist at various depths, as shown in Fig. 39.15, for a concrete pipe,
unless the sections of pipe are jointed or flexible. For jointed pipes, these bending
moments are reduced to uniformly low and tolerable values at all depths. Materials being
considered for the cold water pipe include low-density concrete, plastic-reinforced fiber-
glass, steel, aluminum, or rubber. The largest precast circular concrete pipe yet produced
was manufactured for the Central Arizona Project. A 6.3-m section of this irrigation pipe
is shown in Fig. 39.16. It is 0.9 m thick, and has an outer diameter of 7.35 m.
90
Dynamic Moment
Distribution
Concrete CWP
Tuned Sphere
(100MWe )
1000
Non jointed
2000 —
Figure 39.15 Bending moments versus depth [32] for a 100-m concrete cold water
pipe (CWP) of 17.7 m diameter located beneath a tuned-sphere-design OTEC platform.
The two curves correspond to reinforced concrete cold water pipes of two different
designs: One design was for a continuous concrete CWP, the other consisting of 46-m
sections connected by flexible members. The pipe in each case has a wall thickness of
0.6 m. The modulus of elasticity, E, was estimated as 2.1 X 1010 Nim2 (3 X 106 lb/in.2
or psi) for the continuous pipe, and 10 percent of that for the linked pipe. H113 refers to
the significant wave height assumed in this calculation, 19 m.
Electric Power Generation: OTEC 401
Figure 39.16 A section of precast concrete irrigation pipe manufactured for the Central
Arizona Project.
It has recently been suggested that, analogous to some of the technology presently
being used in the offshore oil industry, the OTEC cold water pipe could be utilized for
mooring the OTEC platform to the sea floor. This concept is known as "tension-leg
mooring," one concept of which is diagramed in Fig. 39.17.
402 Applications of Solar Energy
Riser Tensioner
Cables
Dynamic Positioned
Hull
II
Riser
Section
200 ft
Reentry Funnel
Acoustic Beacons
12
Syntactic Foam
Subsea
Buoyancy Joint Connector-Hydraulic
//1
1
Latch
Free-Standing CWP
Cold
Water Intake
Deadweight Anchor
Pile Optional
Figure 39.17 A "tension-leg mooring" concept for station keeping of an OTEC plat-
form utilizing the cold water pipe for connecting to the sea floor. The cold water pipe
penetrates the OTEC platform through a moon pool occupying the central portion of
the platform, which is dynamically positioned to equalize tension on the riser cables
shown. The platform is maneuvered over the 840-m moored pipe section shown at
the lower left, and the 60-m riser section is connected via the subsea connector/hydraulic
latch to the moored section.
Intakes for OTEC warm and cold water need to be screened to prevent the entry
of fish and other marine biota that would otherwise damage themselves and/or plant
operation if allowed to enter. Screens for OTEC plants have been studied by Nath,
Ambler, and Hansen [26] and by Thomas and Bason [27]. Similar screens to those
required for OTEC are already in use for cooling-water intakes at coastal power stations.
Electric Power Generation: OTEC 403
Large volumes of sea water—about 4 m3 /s per net MWe (6 X 106 gal/min per net
100 MWe)—need to be circulated via the cold water pipe, necessitating heavy-duty sea-
water pumps. The cost of the warm and cold water pumps needed to circulate a total of
about 8 m3 /s per net MWe (12 X 106 gal/min per net 100 MWe) probably constitutes
about 10 percent of the total cost of the power plant and platform [28] . The parasitic
power requirement for circulating the cold water is associated mainly with the need to
overcome the pressure head resulting from the variable compressibility and variable
density of water as a function of depth. This power requirement consumes about 10 to
20 percent of the gross power output of the plant.
For commercial size (ca. 400 MWe) OTEC systems relying on a submarine electrical cable
to transmit power to shore, certain advancements will be necessary in the state of the art
for design and deployment of electrical cables. Studies of OTEC bottom cables and
OTEC riser cables are presently being conducted by the Pirelli Cable Systems and Simplex
Wire and Cable companies, as reported by Morello [29] and by Pieroni et al. [30] ,
respectively.
A key engineering distinction between present-day submarine cable technology and
that required for OTEC applications is the need to connect the cable to a moving plat-
form in a way that can withstand these motions for extended periods. Also, OTEC
installations require traversal of depths up to about 2000 m, whereas bottom cables have
so far been laid in depths up to only 500 to 600 m [31] . OTEC bottom cables at depths
less than 100 m may need to be embedded to depths up to about 2.5 m in order to
prevent mechanical damage from anchors and trawlers.
These requirements give rise to the need for technology development in the area of
submarine cable design, deployment (laying and embedment), and interfacing between
bottom cables, riser cables, and OTEC platforms. Some of the problems that must be
surmounted include cable armoring to resist twisting, minimization of the number of
splices, and the ability to recover and repair. Antitwist armoring can probably be
achieved by using a double layer of steel wires. Engineering of a quick-disconnect
capability may also be desirable, but would increase system complexity.
For OTEC plants located out to about 30 km from shore, AC transmission can be
utilized. At greater distances, the marginal costs for power-factor compensation of long
submarine AC cables for inductive and capacitive effects become excessive compared to
the inversion and conversion costs incurred when resorting to high-voltage DC submarine
transmission cables. In fact, at distances approaching 100 km, an AC cable could not
transmit any real power, owing to the large capacitive charging currents, hence there is
also a technical reason for choosing DC transmission, beyond economic considerations.
For most island applications, and for some mainland applications, the ocean thermal
resource is within the 30-km trade-off distance, so that AC submarine cables will prove
more economical. However, much of the ocean thermal resources is located at distances
from shore exceeding 30 km, hence DC transmission will be required. Present projections
indicate that the costs of such transmission will probably limit DC submarine cable
transmission to distances of about 300 km from shore.
404 Applications of Solar Energy
C. OTEC Platforms
Various concepts are being examined as to the configuration of ocean platforms for
400 MWe (net) commercial OTEC plants. A conceptual analysis of six hull options,
shown in Fig. 39.19, has been conducted by teams headed by Lockheed [32], M. Rosen-
blatt and Son [33] , and Gibbs and Cox [34]. These studies were for potential sites off
Tampa, New Orleans, Puerto Rico, and Hawaii. Similar studies have been conducted by
the Applied Physics Laboratory (APL) of the Johns Hopkins University for a 325-
MWe ammonia plant ship located in the Atlantic Ocean about 400 km east of Brazil [35].
Figure 39.20 shows an APL concept for a 100-MWe ammonia plant ship producing about
Electric Power Generation: OTEC 405
OTEC PLANT
D TO 1.6D
Figure 39.18 A concept of Simplex Wire and Cable Company for interconnecting the
bottom cable to an OTEC plant via a riser cable suspended from a subsurface buoy.
Relative dimensions are expressed in terms of D, the depth of the bottom cable below
the subsurface buoy, which is installed at a depth of 60 to 100 m.
270 metric tons of ammonia per day. This plant ship consists of 20 power modules, each
of 5-MWe net power output.
Baseline conceptual designs for complete OTEC power plants were developed and
costed in previous studies by teams headed by Lockheed and TRW. The TRW concept
for a 100-MWe OTEC power plant consists of four power modules, each of 25-MWe net
power output. A cutaway diagram of one such module is shown in Fig. 39.21. Detailed
analyses and cost studies for this system have been developed [36] . The initial Lockheed
conceptual design [25] was later reoptimized so as to produce a net output of about 260
MWe from four 65-MWe power modules. A cutaway diagram of this concept is shown in
Fig. 3922.
A concept of a 100-MWe OTEC power plant [37] developed by Sea Solar Power,
Inc., employs a minimal amount of hull to envelop the working components, which are
submerged at depths where their operating internal pressures match the hydrostatic pres-
sure of the ocean. A cutaway diagram of this concept (Fig. 39.23) shows the condensers
(at marking 2) located at a level above the evaporators (at 1). Besides balancing internal
and external pressures, this configuration avoids the need for a feed pump to supply
406 Applications of Solar Energy
SEMISUBMERSIBLE
TANKER
SPAR
SUBMERSIBLE
DISK
TUNED SPHERE
Figure 39.19 Six concepts for candidate OTEC commercial hulls for producing 400
MWe net power. These concepts were studied by Gibbs and Cox, Lockheed, and M.
Rosenblatt and Son.
working fluid to the evaporator, since gravity performs that function. Other items
indicated in Fig. 39.23 are as follows: 3 = warm water inlet; 4 = warm water inlet
screens; 5 = cold water pump; 6 = warm water pump; 7 = exhausts for warm and cold
water effluents; 8 = adjustable louvers; 9 = "stockade" cold water pipe, with cross
section shown at A-A; 10 = cold water pipe connection; 11 = location of deaerators;
12 = location of auxiliary power system (for plant startup and station keeping); 13
buoyance chambers; 14 = equipment deck; 15 = living quarters; 16 = turbine-generator.
Electric Power Generation: OTEC 407
TRAVELING GANTRY
CRANE
AMMONIA ----
SYNTHE IS
C R Eksi
QUARTERS----
OLD
ATER
,—DEMISTERS
EQUIPMENT HOUSES
ELECTROLYSIS CELLS
AMMONIA PUMPS
!I
GDIERATORS
COLD WATER PI PE
Figure 39.20 An APL conceptual design of a 100-MWe plant ship for producing
ammonia, containing twenty 5-MWe modules and the ammonia fabrication equipment.
D. Station Keeping
Station keeping of OTEC platforms will be required, both for platforms connected to
shore via submarine umbilicals (such as electrical cables) and for platforms producing
energy-intensive products on board (i.e., OTEC plant ships). However, the station-
keeping requirements in the former case will be more stringent. Station keeping can be
achieved through dynamic positioning and/or by mooring. Dynamic positioning refers
to the utilization of effluent thrust, thrusters, or a combination of both [38]. However,
use of solely sea-water effluents to achieve this thrust would probably require additional
pumping power during some sea conditions, compared to the power required purely to
408 Applications of Solar Energy
VENTILATION
DUCTS
44s •
r
WARM •!:oV I ,
WATER INLET rwl.n2ailtior
N110—
$1
ill RN i
COLD WATER
PLENUM CHAMFER
ads
— . 961,
_ Fairte—
OTEC - BASELINE SYSTEM CONFIGURATION
Figure 39.21 The TRW concept of a 100-MWe OTEC baseline power plant. One of the
four 25-MWe power modules is shown in the cutaway diagram. The cold water pipe has
a diameter of 15 m, and is made of fiberglass-reinforced plastic.
discharge effluents for plant operation. Also, this extra pumping power could constitute
a prohibitive power/cost factor for overall plant sizes in the tens of megawatts (but not
hundreds of megawatts) if sea conditions at the plant site are severe. Mooring costs in a
representative severe plant environment also exhibit an economy of scale, decreasing by
approximately an order of magnitude for plant net power outputs in the range of 20 to
1000 MWe [38] .
A study [39] by Hydronautics, Inc., intercompared the station-keeping perform-
ance of five generic candidate platforms for OTEC power plants. These hull forms con-
sisted of a ship, a submarine, a vertical column stabilized semisubmersible, a disk, and an
axisymmetric spar. Each platform had connected to it an OTEC cold water pipe. The
motions and accelerations of the platforms and the bending moments of the pipe were
BALLAST
TANK WARM SURFACE
EQUIPMENT
HANDLING AREA WATER (80° F)
INLET
CREW
EVAPORATOR LIVING QUARTERS POWER
MODULE
................
WARM AMMONIA
- WATER STORAGE
INLET
Electric PowerGeneration : OT EC
411
ELECTRICAL
EQUIPMENT AREA
GENERATOR SWIVEL-TRAPEZE
MOORING SYSTEM
TURBINE PREVENTS PIPE
AMMONIA FROM BECOMING
VAPOR ENTANGLED AS THE
POWER PLANT ADJUSTS
COLD 70 CHANGING
CONDENSER CURRENTS
A. WATER
INLET
LIQUID BUOYANCY
AMMONIA TANKS
Figure 39.22 The Lockheed concept of a 265-MWe OTEC baseline power plant, showing details of one of the four 65-MWe external
power modules. The cold water pipe is 38 m in diameter and composed of reinforced concrete.
410 Applications of Solar Energy
Figure 39.23 A 100-MW, (OTEC) plant concept, as envisioned by Sea Solar Power, Inc.
utilizing a "stockade" design of cold water pipe, 9 m in diameter, made of aluminum or
steel.
compared for several wave conditions. Based on these intercomparisons, some of the
conclusions of the study were as follows. From the standpoint of station-keeping per-
formance independent of platform costs, the semisubmersible platform appeared most
attractive. Motions and accelerations do not appear to represent serious problems for
Electric Power Generation: OTEC 411
platform or plant operation, but ship and disk heave must be reckoned with. The
platforms appear to be in little danger of capsizing or being subject to catastrophic
motions or accelerations even in the most severe sea conditions (sea state 9, with 25-m
significant wave height). Cold water pipe design problems can be more serious if rigid
pipes are used, but can be alleviated by using flexible or articulated pipes. Platform
motions tend to decrease, but cold water pipe bending moments tend to increase with
increasing plant size [39].
A key factor in projecting the economic viability of commercial OTEC power plants is
the cost of the energy produced. The energy cost can be estimated from the capital cost
of the power plant based on a complex set of assumptions [40]. Since OTEC, like other
solar energy options, requires no fuel, the major cost is the amortization of the capital
investment. Some of the key factors that enter the calculation of energy cost are the
plant's capacity factor, the financial life of the OTEC plant, its tax life, the cost of
capital before and after taxes, the portion of the power system that is taxable, the
insurance and property tax rates, the amount and applicability of any investment tax
credit, the tax rate on gross receipts, the plant construction time, the rate of inflation,
and the cost of operation and maintenance of the power plant. Because of their modular-
ity and standardization, it is likely [25, 36] that baseload OTEC power plants will attain
a capacity factor of about 80 percent and will be capable of being constructed in about
3 years. Their annual operation and maintenance (O&M) costs are variously estimated at
about 1.5 percent of the capital investment.
Although OTEC power plants will provide steady outputs throughout the day and
from day to day, plant output will vary seasonally, as shown for example in Fig. 39.6.
This variation for a given plant will depend [41] sensitively on the temperature differ-
ence, AT, of the input warm and cold water supplies. The plant capital cost, C, will de-
pend inversely on the design temperature difference, AT*, through the relationship [42]
C a (AT*)-k (39.2)
where k 2.5. The seasonal variation of the net power output, Po , of a given plant
designed at AT* will depend on AT through the relation
Po = ( i Lo) ( AT
AT */
— 110 (39.3)
where Lo is the fractional pumping loss at the design temperature difference. This calcu-
lation assumes that the pumping rate is maintained constant, independent of AT. Lo is
about 0.25. Equation (39.3) infers that an OTEC power plant can be designed for an
intermediate temperature difference AT* somewhere between the seasonal maximum
AT (ATmax ) and the seasonal minimum AT (ATm in). Of course, to take advantage of
temperature differences in excess of AT*, certain system components such as turbines,
generators, and the power conversion/inversion/transmission equipment will need to be
412 Applications of Solar Energy
oversized, that is, designed for ATmax rather than for AT*. The preceding conclusions
assume that the turbine efficiency is not appreciably degraded by operating the plant
above or below the design condition. Further verification is required as to how far off
design a relationship such as Eq. (39.3) will be applicable.
If the seasonal modulation of AT is given by a simple sinusoidal relationship with
amplitude e, then the temperature difference at time t (measured in months) can be
expressed [41] as
where AT0 is the annual average temperature difference. Then the annual average plant
output, Pe , is given by
[ 0 )2+€2/2 ]
- Lo
Po (1 Lo) (AT (39.5)
(AT*)2
Then, for each year of operation, at a fixed charge rate, R, and a capacity factor, Cf, the
energy cost, E, will depend on these parameters through the relationship [41]
C•R
E- (39.6)
Cf • 8760 h • -Po
where C is the capital cost from proportionality, Eq. (39.2), E is usually stated in mills
per kilowatt-hour. For a 30-year lifetime, allowing a 7 percent investment tax credit,
and 1.5 percent for O&M, a value of R of about 15.7 percent is derived [40]. For a
capital cost of $1500/kWe, at a Cf of 0.85, Eq. (39.6) would estimate E at about 32
mills/kWh. As discussed below, OTEC capital cost targets for 400-MW, power plants
range from about $1500 to $1900 in 1978 dollars for electrical energy delivered to Gulf
Coast ports such as Tampa and New Orleans, and about $1200 to $1700 for U.S. islands
such as Hawaii and Puerto Rico. (These amounts would be about 25 percent less in
1975 dollars, which have frequently been used for energy cost intercomparisons.). The
corresponding energy costs in 1978 dollars range from about 32 to 41 mills/kWh and 26
to 36 mills/kWh, respectively.
The conceptual baseline studies by Lockheed [25] and TRW [36] in 1975 led to
conservative, state-of-the-art estimates of OTEC system costs for the baseline commercial
power plants considered by those organizations. These cost estimates differed consider-
ably from those obtained previously by OTEC proponents [42]. The explanation for
this discrepancy was that the proponents, in formulating their own cost estimates, had
assumed in their cost projections that certain engineering improvements and innovations
had already been achieved. Lockheed and TRW did not do this, but pointed to signifi-
cant cost reductions from their baseline estimates that could be realized through an
OTEC engineering development program, which they both independently recommended
should be pursued by the U.S. government.
Since 1975, additional system cost estimates can be projected as the result of the
more detailed succeeding studies. In particular, power system costs are being considered
Electric Power Generation: OTEC 413
through three concurrent contracts with Lockheed, TRW, and Westinghouse, who exam-
ined candidate OTEC power systems utilizing competitive varieties of conceptual shell-
and-tube heat exchanger designs [43-45]. Meanwhile, the three platform studies [32-
34] previously referred to obtained cost estimates for several candidate OTEC platforms.
Also, system costs and platform costs for their ammonia plant ship concept are being
obtained by APL [46] .
Cost estimates are based on extrapolating to production units the projected econo-
mies expected to be achieved through "learning curve" or "experience curve" cost
reductions, and are typically the result of estimating the costs resulting from building
eight or more plants. Critical integration of the results of early cost studies [42, 47]
and of more recent studies [48, 49] yields a composite of costs for providing OTEC
electricity to shore. The more recent cost estimates include the results to date for cost
data projected for OTEC submarine bottom [29] and riser [30] cables. The costs
projected for each OTEC subsystem and component can be represented as a range of
values extending from optimistic to pessimistic, and cost projections for the total OTEC
system can be obtained by aggregating the extremes of these constituent costs, as was
done in Fig. 39.24.
Note in Fig. 39.24 that the initial cost of a system with aluminum heat exchangers
might well be several hundred dollars per kilowatt smaller than the system cost of a
system with titanium heat exchangers. However, it remains to be established whether
an aluminum system can be qualified from the standpoint of corrosion resistance to
provide a longevity comparable to that of titanium. If so, then the relatively smaller
initial acquisition cost would represent a considerable advantage for aluminum versus
titanium. If not, then the aluminum exchangers would have to be replaced one or more
times during a 30-year period, possibly leading to higher life-cycle costs relative to
titanium. The cost projections for heat exchangers are also complicated by the need
to trade off optimum augmentation of heat transfer with the associated cost effectiveness
of achieving such enhancements.
The system costed in Fig. 39.24 includes about $250/kW for electrical transmission
to points in the Southern United States from the Gulf of Mexico. For U.S. islands, the
cable transmission costs will range from about $50 to $100/kWe .
A breakdown [48] of likely materials from which an OTEC power plant (based on
the Lockheed spar buoy concept [25]) might be constructed is shown in Table 39.1.
This breakdown indicates a weight of about 1350 kg/kWe . At a cost of $1,500/kWe,
this would require the fabricated cost of the plant to be $1.10/kg. To achieve such a
target, the use of reinforced concrete as a key hull ingredient seems well advised as com-
pared to steel. [Automobiles, it should be noted, are currently mass produced at a cost
of about $3.30/kg ($1.50/1b).] In contrast to the spar buoy, the recent Lockheed ship
design [32] is much less materials intensive, requiring only about 435 kg/kWe .
The range of targeted OTEC energy costs for baseload power is comparable to
projected costs for other baseload power sources (such as coal and nuclear) in the 1990
to 2000 Gulf Coast market for electricity. However, in U.S. island markets (e.g., Hawaii
and Puerto Rico), OTEC power plants could even sooner achieve costs that are less than
the projected costs of oil-derived electricity. This is the case even for small OTEC power
plants. These qualitative statements are illustrated in Fig. 39.25.
414 Applications of Solar Energy
COMPOSITE COST
AT = 40° F
1975$ 2000 - TOTAL
800
700
TITANIUM a
1000
600
ALUM.
SPECI FICCOST (S/K W)
500
400
300
200
100
EJ
Figure 39.24 Ranges of costs projected for production units of 400-MWe OTEC power
plants, measured in 1975 dollars. Costs shown are for the subsystems stated and for the
total plant cost. The submarine cable costs are predicated on cable links of about 130 to
230 km.
100 -1
OIL
OTEC
80 -
COST (m il ls/kWh)
COAL
60 - 40 MW
40 -
20 - 400 MW
Figure 39.25 Qualitative estimates for ranges of projected OTEC energy costs estimated
for year 2000 in constant 1978 dollars compared to ranges of projected energy costs for
baseload electricity derived from coal, uranium, and oil for Gulf Coast and U.S. island
markets. Note the variation in energy cost estimated at two different OTEC plant sizes
for the island markets. OTEC plant sizes for the Gulf Coast market were assumed to be
400 MWe .
An example of cost breakdown estimates in 1978 dollars per kilowatt for pro-
duction units of 400-MWe OTEC power plants is provided for Gulf Coast points in
Table 39.2, utilizing Lockheed estimates [32] for a ship platform and a composite
estimate from results [43-45], of the power system development contractors. These
costs range from $1400 to $1550 and $1250 to $1400, respectively, when measured in
1975 dollars. The power plant subsystems in Table 39.2 include turbines and generators,
each at about $50/kWe , and sea-water pumps at about $100/kWe .
Table 39.2 Cost Estimates in 1978 Dollars per Kilowatt for Production Units of
400-MWe OTEC Power Plants [AT = 22°C (40°F)]
Titanium Aluminum
Heat Exchangers Heat Exchangers
1750-1950 1550-1750
416 Applications of Solar Energy
In recent years, the U.S. OTEC program has moved rapidly from paper and laboratory
studies into testing of hardware of significant sizes in both land and sea environments.
Note that funding in each fiscal year somewhat exceeds the cumulative funding for the
preceding fiscal years, with the result that the total available information and the com-
plexion of the program are changing rapdily. The U.S. OTEC program began at the
National Science Foundation (NSF), and was subsequently transferred to the Energy
Research and Development Administration. ERDA became part of the U.S. Department
of Energy (DOE) on October 1,1977.
The DOE OTEC technology development program is aimed at developing and
testing viable OTEC components, subsystems, and complete systems. Activities in that
program are divided into three complementary facets, as shown in Fig. 39.26. In par-
Figure 39.26 Three complementary facets of the U.S. national OTEC development
program.
Electric Power Generation: OTEC 417
ticular, candidate heat exchanger designs for closed-cycle ammonia systems are being
produced and tested in laboratory and core test (1-MWth) units. Heat exchanger con-
cepts under consideration fall into shell-and-tube categories and plate configurations.
Sketches of some of the options are shown in Fig. 39.27. Ammonia has been selected
as the most likely working fluid, although the possibility of utilizing propane or a halo-
carbon is still preserved. Testing is so far being conducted only with ammonia, however.
The dimensions of typical heat exchanger test elements are shown in Fig. 39.28.
Presently, preparations are underway to provide a floating engineering test facility, to
be known as OTEC-1, in order to develop the capability to test heat exchangers up to
1 MWe (40 MWth) in size. Beyond that capability, several modular system experiments
at about 10 MWe are contemplated to test complete prototype OTEC systems. A sche-
matic diagram of the 1-MWth OTEC core test facility at Argonne National Laboratory
is shown in Fig. 39.29, and Fig. 39.30 is a photograph of that installation. Test results
for both single tube and core tests have exceeded program goals, with values of U for
enhanced surfaces of 4000 to 7000 A//m' • K being achieved, more than twice as great
as for conventional evaporators and condensers.
Although key emphasis is being placed on developing hardware for closed-cycle
ammonia systems, system studies of the open and hybrid cycles are being conducted.
WATER WATER
IN
41,...............,—, I
4
WATER IN
OUT
NH3 OUT NH3 IN
3 (V) OUT
LIQUID NH3 (L) NH
LIQUID
IN
NH3 IN NH3 IN
I (( I
I
I f CONCRETE
WATER SHELL
NH3
FLOW
VAPOR OUT PUMP
LIQUID
RECIRCULATION
SEA
WATER TROMBONE EVAPORATOR
VERTICAL FALLING FILM
Figure 39.27 Some candidate OTEC shell-and-tube and plate heat exchanger designs
now being studied in the U.S. national OTEC development program.
418 Applications of Solar Energy
Lab Testing
4.25' Core-Test Unit
75 t t i ol e
Single (t .Mu 1 MWt
Tubes to 2.5")
0.;t4. AIN
Pilot Plant Test Module
etr-AN , %
,
12.5 MWe
52'
Figure 39.28 Typical dimensions of present shell-and-tube heat exchanger bundles for
testing at laboratory scale, for core tests at 1 MWth , and for ocean tests at 1 MWe ,
together with projected size estimates for tests at about 10 MWe .
Also, the program is supporting analytical and laboratory studies of the foam and mist
approaches to the open cycle that were discussed in Sec. 39.3. Testing of OTEC sub-
systems and components aboard the OTEC-1 engineering test facility is expected to
begin in early 1980. The function of this testing will be to screen various candidate
heat exchangers and other components under actual ocean conditions and thereby lead
to improved performance and reduced costs for subsequent commercial OTEC plants.
OTEC-1 will employ a cold water pipe about 3 m in diameter at a depth of about 1000 m.
Deployment and engineering tests of a shorter cold water pipe of 1.5-m diameter are
planned for the fall of 1978.
To provide performance evaluation of operating pilot plants of significant size, one
or more modular system experiments at about 10 MWe are being considered for operation
in 1982 and thereafter. Options being discussed for the modular experiments include a
sea-based or land-based platform providing electricity to shore, and a grazing plant-ship
platform to demonstrate the operation of a system that could manufacture ammonia at
sea. The platforms and cold water pipes for these modular experiments would probably
be oversize; that is, they could accommodate more than the initial 10-MWe complement
of power modules. For example, Fig. 39.31 shows a conceptual design [46] by the
Applied Physics Laboratory of the Johns Hopkins University for a 20-MWe pilot plant
ship that would contain two 5-MW, power modules but with provision for two additional
power modules. A modular experiment of nominally 10 MWe might conceivably accom-
modate up to about 40 MWe capacity, via power modules of 5 or 10 MWe , and thus have
STEAM
3.2 x 10° Btu/h
HEATER CHILLER
• TO COOLING TOWER
SPRAY
D110 :uoRenua9 nmod oploal i
DESUPERHEATER
80°F 98°F 1di a. 40°F
LIOUID
50°F 50°F
r• —C
80°F j 40°F
( ) 3.41--,
C4 ,...__.."
C>4 D. Ni 1.-
TEST TEST
,1i
0, t EVAPORATOR *-1Q,,t J CONDENSER 1 f
WARM WATER LOOP AMMONIA LOOP COLD WATER LOOP
3200gpm 20.5gpm 3200gpm
Figure 39.29 Schematic diagram of the 1-MWth OTEC heat exchanger core test facility at Argonne National Laboratory.
Figure 39.30 Photograph of the Argonne core test facility diagramed in Fig. 39.29.
20 MWe (NET) CAPACITY
TWO 5 MWe (NET) POWER MODULES
INSTALLED; SPACE ALLOWED FOR
TWO MORE
WARM WATER
INLETS (4)
Figure 39.31 Conceptual diagram of a 20-MW, OTEC pilot plant ship designed by the Applied Physics Laboratory (APL) of The
Johns Hopkins University.
422 Applications of Solar Energy
the flexibility to commence operation with initial equipment of early design, followed by
substitution and/or addition of power systems of more advanced design. Optimum sizes
for OTEC power modules are being considered as part of the power system development
studies [43-45] by Lockheed, TRW, and Westinghouse. Earlier, Lockheed [55] per-
formed a producibility study for a 25-MWe shell-and-tube OTEC heat exchanger.
Besides the development and testing of system hardware, concurrent studies are
underway at various ocean and coastal locations to measure biofouling and corrosion,
and to test cleaning techniques and countermeasures, as discussed in Sec. 39.3. Corrosion
and the formation of deposits of calcareous scale are being studied at the Dow Chemical
facility, Freeport, Texas [56] , and at the University of Delaware laboratory at Lewes,
Delaware [57]. Cleaning tests are being conducted at the Naval Coastal Systems Labora-
tory, Panama City, Florida [58] (Fig. 39.10). Measurements of biofouling using ocean
buoys have been conducted for several years off Ke-Ahole Point, Hawaii [59] , and more
recently in the Gulf of Mexico using an ocean buoy. Also, biofouling and corrosion
measurements were conducted [60-62] off the Virgin Islands on the U.S. Navy barge
YFN-1126.
Proposals for an OTEC Seacoast Test Facility are being considered. Such a coastal
facility would utilize an intake pipe of about 30 cm diameter extending to depths of
about 1000 m to study the biofouling characteristics of OTEC condensers, together with
warm water intakes to simulate OTEC evaporators. Biofouling results up to this time
have simulated conditions in OTEC evaporators only. Development of OTEC bottom
cables [29] and riser cables [30] is being pursued as described in Sec. 39.4. However,
the major developmental challenge will be the riser cable, because of the combined
mechanical and electrical stresses therein.
The development of key OTEC hardware requires the design, construction, and
testing of experimental components, subsystems, and systems. This can be accomplished
on land and at sea, utilizing a whole spectrum of possible sizes of equipment undergoing
testing. Some clarification of possible test program and test facility requirements for the
OTEC development program was provided through two concurrent studies completed in
1976 by TRW and Lockheed [63, 64].
An important adjunct to the OTEC development program is the assessment of
environmental and resource questions. Oceanographic data relevant to thermal resource
and siting questions are being obtained from archival sources and through ocean measure-
ments. Measurements are being conducted from both ships and moored ocean buoys.
Some aspects of this subject were discussed in Sec. 39.2, together with the environmental
studies presently underway. A specialized workshop on OTEC Resource and Environ-
mental Assessment was held in June 1977 in Florida [6]. An OTEC Environmental
Development Plan (EDP) has been formulated by the U.S. Department of Energy [65] ,
and an interagency OTEC environmental working group has been established. This
working group includes representatives from the U.S. Environmental Protection Agency
and the U.S. Department of Energy. The OTEC environmental and resource assessments
are concerned with analytical, research, and experimental activities in support of the
development program and are to ensure minimization of possible OTEC environmental
impacts and compliance with existing environmental regulations. Besides studying
possible effects of OTEC on the environment, these assessments provide engineering
Electric Power Generation: OTEC 423
inputs so that possible impacts of the environment (such as sea condition) on OTEC
design, siting, and operations can also be considered.
Besides the research and development activities described above, the OTEC Program
Office is also concerned with programatic questions associated with the marketability of
OTEC electricity and energy-intensive products. Those questions revolve about the
projection of costs for commercial OTEC systems, and relate to potential markets,
market penetration, and legal, institutional, and financial matters. These subjects and
somt mechanisms for introducing OTEC technology into the marketplace will be
discussed below.
The introduction of OTEC technology into the marketplace will not necessarily occur
simply as a consequence of its successful demonstration and achieving competitive cost
goals. There are numerous factors that will affect OTEC marketability and commerciali-
zation. In particular, this technology is confronted with a situation more complex
institutionally than most other solar energy technologies, in that most of the products
derived will be manufactured beyond state boundaries. Besides creating questions as to
who will be the owner/operators of OTEC plants and plant ships, and how OTEC capital
formation will occur, there is the key problem associated with at-sea operation concern-
ing the prevailing legal regime. It will clearly be essential to resolve this problem in a
fashion conducive to the attractiveness and legal stability of OTEC commercial opera-
tion. However, the fact that the offshore oil industry and the offshore nuclear industry
have made considerable progress in solving comparable problems is encouraging.
Before exploring the legal, institutional, and financial questions confronting OTEC,
let us examine some of the intrinsic market questions, such as resource availability,
potential markets, market penetration, resource requirements, and net energy considera-
tions. From a global standpoint, OTEC represents a substantial potential increment of
world energy supply. In particular, it could provide (cf. Fig. 39.1) large amounts of
electricity via submarine cable to many nations in tropical and subtropical regions up to
about 25° of latitude on either side of the equator. The prime ocean thermal resource
is roughly bounded by the 20°C contour of Fig. 39.1, and would of course provide
considerable potential for the manufacture of energy-intensive products beyond the
regions readily accessible to shore via cable.
From a more parochial viewpoint, it is clear that considerable ocean thermal
resources are available to the United States. Besides the prime area in the Gulf of Mexico
shown in Fig. 39.4, there are excellent ocean thermal resources located a few kilometers
off U.S. islands such as Hawaii, Puerto Rico, Guam, the Virgin Islands, American Samoa,
and Micronesia. The total amount of OTEC power that could be supplied via cable to
Gulf Coast locations such as New Orleans and Tampa from the shaded areas of Fig. 39.4
is conservatively estimated [66] at upwards of 200 GWe , and perhaps as high as, or
higher than, 600 GWe. This estimate assumes that the constraint is a significant degrada-
tion of the ocean thermal resources flowing into those regions. As will be noted in the
discussion below, this power potential is comparable to or greater than the electrical
424 Applications of Solar Energy
market expansion projected for the southern United States for at least the next 20 years.
Also, the ocean thermal resource in the vicinity of U.S. islands is measured in the tens of
gigawatts, and far exceeds the power required to replace existing production and to
satisfy the potential incremental electrical markets in those islands during the next
century.
Thus, the available ocean thermal resources in the Gulf of Mexico can serve a
potentially growing electrical market in the southern United States commencing between
1985 and 2000. This assumes that cost reductions can be realized associated with travers-
ing the learning (or experience) curve through manufacturing a group of initial OTEC
plants and then entering a "production" phase. It also assumes that the cost of energy
from these production units will be comparable to the cost of energy from competitive
sources of baseload electricity, that is, coal and nuclear. Making the assumption that
OTEC electrical energy would become competitive with other baseload options in the
Gulf Coast market in the 1985-2000 time frame, a General Electric Tempo study [67,
68] intercompared growth projections for the United States and southern regional
baseload electrical markets, and then projected plausible penetrations of OTEC into
the southern regional market.
This set of projections is shown in Fig. 39.32. GE Tempo utilized two basic
scenarios for electrical load growth for the United States as a whole and for the three
southern United States regional reliability councils (known as SERC, SWPP, and
ERCOT). One is a fairly rapid rise, projected by GE Tempo, and the other is a slower
growth rate projected by Dr. Alvin Weinberg of the Institute of Energy Analysis.
Based on these two growth rates, GE Tempo arrived at three possible OTEC implemen-
tation rates, shown by the set of curves in the lower part of Fig. 39.32 for high, inter-
mediate, and low market penetrations. These projects lead to a total OTEC baseload
contribution ranging from 6 to 35 GWe by the year 2000. The most optimistic projec-
tion would amount to about 30 percent of the incremental market in the year 2000.
These scenarios assume that penetration of southern United States markets would
be achieved by DC cables into locations such as Tampa, New Orleans, and Brownsville.
Additional market penetrations would be realized by short AC cables into U.S. island
locations such as Puerto Rico, Hawaii, and Guam. From southern U.S. locations, OTEC
energy could be wheeled inland via high-voltage transmission lines. Cable runs of about
150 km will be required to reach locations such as New Orleans, and about 250 km to
reach locations such as Tampa and Brownsville.
According to the GE Tempo study [68] , OTEC power wheeled inland from Gulf
Coast submarine cable terminals could penetrate various distances, depending on the
initial cost of OTEC energy at shore, the cost of high-voltage DC overland transmission,
and the costs of competitive baseload electricity. Figure 39.33 represents [68] a possible
geographical penetration, based on an assumed busbar OTEC energy cost of 35 mills/kWh
in the year 2000 (1976 dollars), utilizing a transmission cost of 2 to 3 mills/kWh per
160 km. Good ocean thermal resources are also situated within several miles of the
west coast of Mexico, and just south of Baja California. Thus, OTEC power could readily
become economic for Mexican utilization and potentially for export to U.S. locations
such as southern California and Texas.
U.S. islands, such as Puerto Rico and Hawaii, will probably constitute excellent
early markets for OTEC electricity. These markets will likely be competitive in cost
Electric Power Generation: OTEC 425
10,000
o Scenario
C + SW PP +ERGOT
Baseload
SE R Capacity In Temp
'C 9
‘.e ro
C"?'"
ti Scen
,`moocs et 9
<23'
‘.e \14
cF P:a"
FQ
eF
ce
IO.
<b (C`Q.
$ <ei° 6 0 To 35 GWe
C.) k In Year 2000
4., 0
A. ,t,S
Q?
0
Figure 39.32 Some scenarios for baseload electricity utilization in the United States and
in southern United States, as derived by General Electric Tempo, together with projec-
tions of possible OTEC market penetrations in the Southern states. The abbreviations
SERC, SWPP, and ERCOT refer to the southern utility regions defined by the regional
electric utility reliability councils.
immediately, even for the first commercial OTEC plants. The key reason for this state-
ment is that the existing electricity supply on U.S. islands is derived largely from
imported oil. Also, short AC cables will suffice. A comparison of post-1985 new plant
costs for oil-derived and OTEC-derived electrical energy is shown for Puerto Rico in
Fig. 39.34. The cost of the first 250-MW, OTEC power plant is estimated at $2800/kWe
in deriving the first OTEC point plotted, followed by learning-curve cost reductions
estimated at 10 percent for each doubling of the cumulative number of OTEC power
plants.
In principle, much of the present utilization of oil to produce electricity could be
replaced by OTEC power production at U.S. islands, especially as existing gas turbines
are amortized and phased out. Each megawatt of oil-derived electricity is equivalent to
about 40 bbl/day of oil displacement, hence the eventual utilization of 2000 MW, of
426 Applications of Solar Energy
*
New Conventional Generating Equipment (Yr 2000)
Projected BBEC Range (MOPPS)
New England
(33-56)
Middle
(33-49) Atlantic
(33-53)
Pacific •
West North C -ntral,East North
Central
(32-50)•
• (32-45)
• Mountain
(32-50)
Figure 39.33 A possible geographical penetration for baseload OTEC electricity in the
year 2000, estimated by General Electric Tempo for OTEC power plants in the Gulf of
Mexico off New Orleans and Tampa. OTEC busbar energy costs were projected at 35
mills/kWh, plus about 2 to 3 mills/kWh per 160 km of transmission, versus costs of
competing baseload electricity projected at the values shown in parentheses.
OTEC electricity in Puerto Rico, for example, would save about 80,000 bbl of imported
oil (costing about $1 million) per day. The achievement of this "island strategy" of
market penetration could result in a penetration of 5000 MWe or greater in U.S. island
markets. The associated cost reductions achieved could then enable penetration of the
Gulf Coast market. Figure 39.35 is a qualitative representation of the sort of cumulative
market penetration that might be achievable in the U.S. island, Gulf Coast, and inter-
national markets in the 1985-2010 time frame.
A scenario for commercial implementation and deployment of OTEC power plants
has been projected by Curto [40] , based on specific assumptions for the timing of OTEC
plant orders and the completion of additional OTEC hull facilities. However, considera-
tion must be given in such scenarios to production constraints, including requirements for
material and energy resources. In particular, if aluminum alloys were to be qualified for
use in OTEC heat exchangers, then the supply of aluminum would probably be quite
adequate even if 5 or 10 GWe of OTEC power capacity were added annually. On the
other hand, if titanium were employed for OTEC heat exchangers, then a considerable
expansion of the present U.S. production capacity for titanium would have to be
provided. Fortunately, there is an abundant supply of titanium ores. Similarly, signifi-
cant numbers of OTEC power plants requiring electrical cables would seriously tax
world submarine electrical cable production capabilities, which would have to be greatly
augmented.
Electric Power Generation: OTEC 427
100 --
90 --
80 I--
70 --
FIRED PLANTS
COMBINED CYCLE OIL
COST (MILL S/ K WH1
60
50
40 00/0
30 -
OTEC PLANTS
20 - 20%
10
1 I I 1 I I I I I I I I l I I
1985 1990 1995 2000
YEAR
Figure 39.34 Comparison (in constant 1976 dollars of baseload electrical costs projected
for Puerto Rico for oil-derived electricity with energy costs for electricity derived from
250-MWe OTEC plants a few kilometers offshore. Possible cost projections (solid lines)
are shown within bands of estimated uncertainty. Numbers near points refer to cumula-
tive number of OTEC plants.
100
CUMULATIVECAPACITY ( GWe)
75
50
25
From a net energy standpoint, the energy payback time (i.e., "breeding time") for
an OTEC plant seems favorable. Even though considerable aluminum would be required,
for example, it is easy to calculate that at about 18 kWh/kg of aluminum, it would take
about 1 month of energy output to provide the electrical energy invested in the heat
exchangers, assuming the 22 tons of aluminum per megawatt stated in Table 39.1. If
energy inputs for other components of the OTEC plant are included, the entire breeding
time is about a year, even if the heat exchangers are made of titanium [25, 64, 69] .
The market potential for OTEC products other than electricity has been examined
by GE Tempo [67, 68] , DSS Engineers [70] , and by the Applied Physics Laboratory
(APL) of the Johns Hopkins University [35, 71] . The products examined included
aluminum, ammonia, hydrogen, chlorine, and other sea chemicals. Also, the Institute
of Gas Technology (IGT) has studied the production costs of chemical energy carriers
such as ammonia, hydrogen, carbonaceous fuels (such as methanol), of refining lithium
for lithium-air batteries, and of thermal energy carriers such as molten salts [72, 73] .
Floating OTEC plant ships would achieve some economies not experienced by
OTEC plants attached to submarine electrical cables—namely, no cable costs and lower
station-keeping costs. If the stability of OTEC platforms is compatible with the onboard
refining of alumina or the production of other electrically intensive metals, then OTEC-
derived metals may well become a viable product. Indeed, they could be produced
onshore by utilizing OTEC electricity supplied via cable, especially at U.S. island
locations such as Puerto Rico and Hawaii that are convenient to sources of bauxite/
alumina. Hawaii is also convenient to locations where manganese nodules will be
removed from the sea floor.
On the other hand, ammonia synthesized aboard floating OTEC plants from
nitrogen removed from the air and hydrogen electrolyzed from water could well become
a viable product for the large and growing fertilizer market, presently being supplied by
ammonia derived from natural gas. However, manufacture of products such as ammonia
and hydrogen represents a degradation of high-quality OTEC electricity. This results in
a requirement that the cost of producing the ammonia or hydrogen will have to be fairly
low before such OTEC-derived products will be cost competitive, whereas OTEC electric-
ity costs, even on early commercial plants, can be relatively high (cf. Fig. 39.36) and still
be competitive in certain markets. In this connection, note the intercomparison of costs
in Fig. 39.36 for ammonia derived from coal, natural gas, and OTEC electricity, as esti-
mated by the Institute of Gas Technology [72]. The range of costs for coal-derived
ammonia results from a spread associated with high-Btu-content bituminous coal to low-
Btu-content lignite coal. In the 1980 period, costs for coal and natural gas lead to
ammonia costs of $175 to $220/ton and $140 to $160/ton, respectively. This assumes
projected cost ranges of the feedstocks of $13 to $15/ton for low-Btu coal, $20 to $25/
ton for high-Btu coal, and $2.40 to $2.60/million Btu of natural gas. In the year 2000
time frame, however, these ranges of feedstock costs are projected by IGT at $35 to $45/
ton, $55 to $65/ton, and $5.80 to $6.70/million Btu, respectively. This leads to pro-
jected costs ranging from $250 to $300/ton for coal-derived ammonia, and $240 to
$260/ton for natural gas-derived ammonia.
These cost projections for the production of ammonia from depletable feedstocks
mean (cf. Fig. 39.36) that in order for OTEC-derived ammonia to be cost competitive in
the early time frame, OTEC electricity would have to cost about 10 mills/kWh. Similarly,
Electric Power Generation: OTEC 429
500
400
Ammonia Cost, $/ S hort Ton
300
200
F.1Coal Gasification
100
0
0 10 20 30 40 50 60 70 80 90
Coal Cost, $/Short Ton
(Mississippi Valley Ammonia Production)
1 I I I I I I I I J
0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0
Natural Gas Cost, $/106 Btu
(Texas Gulf Ammonia Production)
I I I I I I I I I
0 10 20 30 40 50 60 70 80 90
OTEC Electricity Cost, Mills/ kWhr
(Puerto Rico OTEC Site, New Orleans Delivery)
it would have to cost no more than about 20 mills/kWh to be competitive from a cost
standpoint in the year 2000. Thus, the probability of OTEC-derived ammonia becoming
cost competitive seems more likely after initial OTEC production units are manufactured
to serve electricity markets, thereby achieving cost reductions, and as the cost of ammonia
derived from depletable fuels increases. However, cost is not the only consideration; the
possibility of providing an additional domestic source of ammonia from a renewable
energy resource with an acceptable environmental impact may well outweigh other
factors.
430 Applications of Solar Energy
The preceding discussion summarized the two basic OTEC market options, viz.,
electricity-to-shore via submarine electrical cable, and energy-intensive products. How-
ever, there is another key option for delivering electricty to shore, via an "electrical
bridge" consisting of a chemical energy carrier such as hydrogen, ammonia, or a primary
battery such as lithium/air. In the case of hydrogen or ammonia, electricity can be
regenerated through combustion or by using a fuel cell. It is likely that the ammonia
would first be cracked and the resulting hydrogen would then be utilized in the fuel cell.
Another possibility is to convert OTEC electricity to high-temperature heat contained in
molten salts, then to convey the salts to shore for reconversion to electricty. The
"thermal bridge" was examined by IGT [73], and it is evident that the electricity
produced in this way would be costly. However, the thermal application might be
economic if the heat were used for industrial processing on or near the OTEC platform.
As a subcontractor to General Electric Tempo, IGT performed [68] a cost inter-
comparison for four electrical bridges, namely, the hydrogen bridge, the lithium bridge,
the ammonia bridge, and the thermal bridge. The lithium bridge turned out to provide
the least expensive electrical energy. However, lithium/air batteries are as yet not com-
mercially available and require some technological development. Although it is likely
that aluminum/air primary batteries are closer to commercial readiness, the relative cost
of an aluminum bridge would probably exceed that of the lithium bridge. In the case of
primary batteries, the procedure envisioned is to ship material such as lithium in bulk
form from the OTEC platform to shore, then to insert it into batteries for generation of
electricity, followed by bulk shipment of the resulting lithium hydroxide back to the
OTEC platform for reconversion to lithium. This process would enable the transmission
of electricity from ocean thermal sources in the Gulf of Mexico, for example, to points
in the northeastern United States at costs of about 80 to 100 mills/kWh for busbar
electricity costing 20 mills/kWh. Such a procedure would enable this electrical energy to
be marketed as peaking power, intermediate power, or even baseload power. Estimates
by GE Tempo [68] indicate that OTEC electricity conveyed in this manner can be
competitive for peaking applications by the year 2000.
A related process, with additional market potential for OTEC electricity, is the
reprocessing of chemicals resulting from batteries utilized in electric vehicles. For exam-
ples, if lithium/air batteries were utilized in the large fleet of electric automobiles that is
projected for the year 2000 by many observers, then the lithium hydroxide could be
recycled to lithium on OTEC plant ships.
There are two possible OTEC by-products that may be marketable; these by-
products are (a) fresh water and (b) shellfish, kelp, or other food/energy crops resulting
from open-ocean or onshore mariculture utilizing the nutrients upwelled in the cold
water circulated through OTEC condensers. The intrinsic economics of OTEC energy
production may well be benefitted in certain instances from such by-product manufac-
ture. However, the market possibilities are sensitive to geography and to many uncertain-
ties, especially in the case of mariculture.
The cost of transporting fresh water produced as an adjunct to OTEC power
production would be substantial, except for onshore or near-shore applications.
Accordingly, it is probable that the marginal cost effectiveness of manufacturing fresh
water as an OTEC by-product will not be favorable for floating OTEC power plants at
significant distances from shore. Another option that has been mentioned for the
Electric Power Generation: OTEC 431
production of fresh water is the utilization of icebergs. An OTEC engine has been
suggested [74] for propelling icebergs from arctic to temperate latitudes. In this cycle,
the cold water would be obtained from the iceberg rather than via a cold water pipe.
A similar cycle has been proposed [75] for producing energy and fresh water from the
iceberg once it reaches its destination.
The technology for open-ocean mariculture of protein crops, such as shellfish, will
probably differ from the technology of open-ocean mariculture of energy crops, such as
kelp, according to a study by Laurence and Roels [76]. This is because both horizontal
and vertical containment of the artificially upwelled cold water is necessary in the former
case, and only horizontal containment is required in the latter case. Although the utiliza-
tion of OTEC electricity and upwelled nutrients for associated mariculture activities
would be a potential market for available OTEC power and cold water, there may be
certain incompatibilities between the technologies. In particular, the retention of signifi-
cant volumes of cold water effluents in the vicinity of the OTEC warm water intakes
might result in recirculation problems (whereby the efficiency of power production could
be reduced because of degradation of the warm water entrance temperature) and/or
might cause biofouling problems. On the other hand, plants such as kelp typically
thrive in a cold water environment, so that any protracted loss of cold water environment
through OTEC plant shutdown might result in crop damage.
The resource value of upwelled nutrients, as pointed out by Laurence and Roels
[76] , is very great if they are converted into protein. They could become very important
to a hungry and malnourished world. However, it would require surprisingly few OTEC
power plants to pump amounts of cold water containing sufficient nutrients to produce
enough shellfish protein to saturate world markets, assuming a viable mariculture technol-
ogy. Acordingly, even in the most optimistic case, the marginal economics of protein
production using OTEC power plants as sources of pumping power would appear attrac-
tive for only a modest number of OTEC plants. On the other hand, a fraction of the
cold water effluent from a greater number of OTEC plants might be marketed for this
application.
In contrast, if the open-ocean mariculture of energy crops such as kelp proves
viable and compatible with OTEC operation, and vice versa, then the synergism of such
combined operations may indeed prove economically advantageous someday to both
energy sources. The remarkable fact that OTEC power production intrinsically provides
an artificial upwelling of nutrients and a source of electricity for distributing those nutri-
ents within open-ocean mariculture farms still does not justify making an exception to
the author's maxim that "one should not prematurely combine immature technologies."
Another OTEC application has been proposed [77] for producing energy at Arctic
locations. This application would utilize the temperature gradient that exists between
cold arctic air and near-freezing ocean surface waters. During much of the year, air
temperatures below -30°C are experienced, suggesting that an air-cooled condenser
could be operated in conjunction with a surface water-heated evaporator, probably for
land-based electrical production.
Finally, we consider the legal, institutional, and financial questions confronting the
commercialization of OTEC technology. These are probably much more severe in some
respects than related problems associated with the introduction of new terrestrial energy
technologies. In particular, it is by no means clear who will be the owner/operators of
432 Applications of Solar Energy
OTEC power plants and plant ships. Commercial OTEC ventures and the manufacturing
of OTEC system components will necessitate considerable capital formation and a favor-
able investment climate. On the one hand, it will probably require [78] significant
federal incentives and assumption of risks to encourage private investors to participate in
the commercial introduction of OTEC, but on the other hand, such federal promotion of
OTEC will probably need to avoid regulatory features that could make such participation
unattractive.
The investment in and operation of OTEC power plants and plant ships in territori-
al seas, economic zones of coastal states, and in international waters will require a predict-
able and stable applicable legal regime. Some of the relevant legal, institutional and
financial aspects of the OTEC plant operation were examined in a study [79] conducted
by the American Society of International Law (ASIL). There is a renewed ASIL study
presently underway. One of the key considerations concerning applicable law and
economics is whether OTEC platforms can be regarded as "vessels" from a legal stand-
point. This classification may differ if the platform is moored, as in the case of providing
electricity to shore via submarine cable, or if it operates as an unmoored plant ship,
manufacturing energy-intensive products. On the other hand, in neither case would the
platform be a vessel in the sense of plying between ports. The safety, insurance coverage,
and physical protection of OTEC platforms will need to be ensured.
The facilitation of OTEC commercialization and ownership will require an intricate
institutional structure, generically described by Ezra [80] as a "technology delivery
system." Some of the scenarios for OTEC commercialization have been analyzed and
discussed by Naef [81] , who also identifies possible OTEC owner/operators. Entrepre-
neurial arrangements that may evolve include completely private ownership of OTEC
facilities, public ownership, or mixed public and private ownership. A study of possible
domestic entrepreneurial alternatives and the legal aspects and pros and cons of each is
being conducted by the law firm of Tefft, Kelly and Motley, One such arrangement that
they examined in this connection was recently reported [82] . The relative attractiveness
of a capital-intensive option such as OTEC as an investment opportunity in an era where
the demand for capital will strongly exceed its supply will probably be a strong factor in
determining its rate of market penetration, perhaps outweighing questions of how cost
competitive are its products.
The prospects for OTEC technology thus depend both on economic factors and on
institutional factors. Both sets of factors will need to be satisfactorily resolved before
OTEC commercialization can occur [83] . If and when this is achievable, the ocean
thermal resource could provide the world a new source of renewable energy having
substantial potential to help meet growing worldwide demands for additional energy
supplies. In a global climate where aspirations for energy are beginning to exceed the
plateau in the supply of depletable energy reserves, OTEC-derived electricty and energy-
intensive products, by increasing world energy supply, could help reduce forseeable
polarizations between nations over energy resources.
This experiment, a 50-kWe (gross) power plant, is known as Mini-OTEC. It was complete-
ly funded by a consortium headed by Lockheed Missiles and Space Company, Inc., in-
cluding the State of Hawaii, Dillingham Corporation, Alfa-Laval Energy Systems Division,
Worthington Pump Corporation, and the Rotoflow Corporation. The total cost of the in-
stallation, including its operation from August through November 1979, was about $3
million. Predicted net powers ranging from 10 to 15 kWe were produced from this sys-
tem, whose components were not optimally matched. (In commercial OTEC systems,
net power of about 75% of gross power is expected to be achieved.) Results from Mini-
OTEC were presented in April 1980 to the Seventh Energy Technology Conference in
Washington, D.C. by Frederick E. Naef and Delbert N. Burwell of Lockheed. Their paper
is available in the published proceedings of that conference.
REFERENCES
Conference, Miami, Florida, U.S. Govt. Printing Office, stock no. 061-000-00181-6,
1978.
28. Thomas E. Little, OTEC Seawater Pump Study, Westinghouse Electric Corporation,
Oceanic Division Report COO-4071-2, 1978.
29. A. Morello, "Bottom Power Cables Connecting Floating Power Stations to Shore,
in Abrahim Lavi and T. Nejat Veziroglu, eds., Proceedings, Fifth OTEC Conference,
Vol. 1, Miami, Florida, February 20-22, 1978, U.S. Govt. Printing Office, stock
no. 061-000-00178-6, 1978.
30. C. A. Pieroni, R. T. Traut, B. S. Clark, and W. G. Smith, "OTEC Riser Cable:
Problems and Costs," in Abrahim Lavi and T. Nejat Veziroglu, eds., Proceedings,
Fifth OTEC Conference, Vol. 1, Miami, Florida, February 20-22, 1978, U.S. Govt.
Printing Office, Stock no. 061-000-00178-6, 1978.
31. 0. Hauge, A. Berg, J. Normann Johnsen, G. Wettre, and K. Bj4r1ciw-Larsen, "The
Skaggerrak HVDC Cables," paper 21-05 in Proceedings, International Conference
on Large High Voltage Electric Systems, Paris, France, August 30 to September 7,
1978
32. OTEC Platform Configuration and Integration, Lockheed Report LMSC-D566214,
1977.
33. OTEC Platform Configuration and Integration, M. Rosenblatt and Son, Incorpo-
rated, Report 5142-6, 1978.
34. OTEC Platform Configuration and Integration, Gibbs and Cox, Incorporated,
Report 18351-10 (W-10,000), 1978.
35. W. H. Avery, R. W. Blevins, G. L. Dugger, and E. J. Francis, Maritime and Construc-
tion Aspects of Ocean Thermal Energy Conversion (OTEC) Plant Ships, Applied
Physics Laboratory, The Johns Hopkins University, Report SR 76-1B, 1976.
36. Research on an Engineering Evaluation and Test Program, TRW Systems Group,
Vols. 1 through 5, Reports SAN-1089-TI-P1, 2, 3, 4, 5, 1975.
37. Power from the Sea, Sea Solar Power, Incorporated, York, Pa., 1976.
38. Hunter Davidson, Jr., and Thomas E. Little, OTEC Platform Station Keeping
Study, Westinghouse Electric Corporation, Oceanic Division, Report COO-4071-1,
1977.
39. Roderick A. Barr and John F. O'Dea, "Theoretical Evaluation of the Seakeeping
Performance of Five Candidate OTEC Platforms," in Proceedings, Second Ship
Technology and Research (STAR) Symposium, San Francisco, California, May
25-27, 1977, The Society of Naval Architects and Marine Engineers (SNAME),
New York, 1977.
40. P. A. Curto, "An Update of OTEC Baseline Design Costs," in Abrahim Lavi and
T. Nejat Veziroglu, eds., Proceedings, Fifth OTEC Conference, Vol. 1, Miami,
Florida, February 20-22, 1978, U.S. Gov. Printing Office, Stock no. 061-000-
00178-6, 1978.
41. Gay Heit Lavi, "The Commercialization of OTEC: Problems and Opportunities,"
in T. N. Veziroglu, ed., Proceedings of Miami Internattoal Conference on Alterna-
tive Energy Sources, Miami, Florida, December 5-7, 1977, published by Clean
Energy Research Institute, University of Miami, 1978.
42. Abrahim Lavi, "Solar Sea Power Plants, Cost and Economics," in Gordon L.
Dugger, ed., Proceedings, Third OTEC Workshop, Houston, Texas, May 8-10, 1975,
Applied Physics Laboratory, The Johns Hopkins University, Report SR-75-2, 1975.
43. OTEC Power Systems Development, Phase 1, Conceptual Design, Lockheed Report
LMSC-D-566744, 1978.
436 Applications of Solar Energy
44. OTEC Power System Development, Phase 1, Conceptual Design, TRW Report
SAN/1570-1, 1978.
45. OTEC Power System Development, Phase 1, Conceptual Design, Westinghouse
Electric Corporation Report on Contract EG-77-C-03-1569, 1978.
46. J. F. George, Private communication of data from Abam Engineering regarding
scale-up of results on the preliminary engineering design studies of a 20-MWe OTEC
plant ship relevant to a 325-MWe commercial OTEC plant ship. Applied Physics
Laboratory, The Johns Hopkins University, 1978.
47. R. Eldridge and R. Roberts, OTEC Systems Descriptions and Engineering Costs,
MITRE Corporation, Report MTR-7485, Vol. VII, 1977.
48. P. A. Curto, An Update of OTEC Baseline Design, MITRE Corporation, Report
MP 79W00007, 1979.
49. P. A. Curto, Private communication regarding OTEC cost estimates, 1978.
50. Takuya Homma and Hiroshi Kamogawa, "Conceptual Design and Economic Evalua-
tion on OTEC Power Plants in Japan, in Abrahim Lavi and T. Nejat Veziroglu, eds.,
Proceedings, Fifth OTEC Conference, Vol. 2, Miami, Florida, February 20-22,
1978, U.S. Govt. Printing Office, Stock no. 061-000-00179-4, 1978.
51. Association Europeene Oceanique (EUROCEAN), Ocean Thermal Energy Conver-
sion—Final Report, 1977.
52. B. A. P. L. Lachmann, "EUROCEAN OTEC Project," in Abrahim Lavi and T. Nejat
Veziroglu, eds., Proceedings, Fifth OTEC Conference, Vol. 1, Miami, Florida,
February 20-22, 1978, U.S. Govt. Printing Office, Stock no. 061-000-0178-6, 1978.
53. U.S. Energy Research and Development Administration, OTEC Program Summary,
Report ERDA 76-142, 1976.
54. U.S. Department of Energy, OTEC Program Summary, Report DOE/ET-0021/1,
1978.
55. Lloyd C. Trimble, OTEC Heat Exchanger Design and Producibility Study,
Lockheed Report LMSC-D507632, 1976.
56. C. F. Schrieber, W. D. Grimes, and W. F. Mcllhenny, Interleakage of Ammonia
and Seawater in OTEC Heat Exchangers—Effects on Corrosion and Scale Forma-
tion, Argonne National Laboratory Report on OTEC Workshop on Biofouling,
Corrosion and Materials, Rosslyn, Virginia, January 8-10, 1979.
57. S. C. Dexter, "On the Localized Corrosion of Aluminum Alloys for OTEC Heat
Exchangers," in Proceedings, OTEC Biofouling and Corrosion Symposium, Seattle,
Washington, October 10-12, 1977, Battelle Pacific Northwest Laboratories,
Richland, Washington, 1978.
58. V. J. Castelli, A. B. Fritsch, and W. L. Adamson, "An Evaluation of Some Mechani-
cal Cleaning Methods for the Control of Microbiofouling in Heat Exchangers," in
Abrahim Lavi and T. Nejat Veziroglu, eds., Proceedings, Fifth OTEC Conference,
Vol. 4, Miami, Florida, February 2-22, 1978, U.S. Govt. Printing Office, Stock no.
061-000-00181-6, 1978.
59. J. G. Fetkovich, G. N. Grannemann, L. M. Mahalingam, and D. L. Meier, "Measure-
ment and Mechanisms of Biofouling in OTEC Heat Exchangers," in Abrahim Lavi
and T. Nejat Veziroglu, eds., Proceedings, Fifth OTEC Conference, Vol. 4, Miami,
Florida, February 20-22, 1978, U.S. Govt. Printing Office, Stock no. 061-000-
00181-6, 1978.
60. R. P. Aftring, D. G. Capone, L. Duguay, J. W. Fell, I. M. Master, and B. F. Taylor,
Biofouling and Site Characterization Studies, Part III of Tracor Marine Report
PNL-2739, 1978.
61. H. L. Craig, Jr., S. C. Munier, and J. Morse, "Corrosion Results of Saint Croix
Electric Power Generation: OTEC 437
MARSHAL F. MERRIAM
Dept. of Materials Science and Mineral Engineering
University of California
Berkeley, California
40.1 INTRODUCTION
Winds, waves, and tides are of interest today as possible future energy sources primarily
because of their nondepletable character and their favorable environmental aspect, in
comparison with energy technologies presently in use. The environmental advantage,
though important, is not decisive; were it not widely believed that oil and gas supplies
will be substantially depleted within the present century, interest in winds, waves, and
tides would be much less than it presently is.
Studies comparing the cost of energy that could (perhaps) be obtained from future
wind, wave, or tidal installations with the cost of energy from today's fossil fuel and
*Adapted and reprinted, with permission, from Annual Review of Energy, Vol. 3. Copyright CD 1978
by Annual Reviews Inc. All rights reserved.
439
440 Applications of Solar Energy
uranium sources show wide variation, and often have difficulty making a case for immedi-
ate economic viability of the renewable sources. If, however, oil and gas resources of the
world do in fact become progressively and substantially depleted over the next 25 years,
price increases and shortages will ensue and implementation of presently uneconomic
alternatives will occur on a large scale. This chapter is a review of the present state of
knowledge and experience with wind, wave, and tidal energy, and presents an estimate
of the magnitude, in energy terms, of possible future use.
One characteristic that the three sources being considered have in common is their
inability to provide firm power on demand. (When the power is demanded, the wind may
not be available.)
Energy delivered according to a schedule controlled by the source is certainly less
valuable, per kilowatt-hour, than energy supplied as required by demand. The question
of how much less valuable is not so simple. This question is discussed later in this
chapter.
A. Wind
If the total quantity of energy delivered into a nation's commercial economy is taken as
the measure of importance, wind energy is nowhere important today. It has, however,
been important in times past. Wind energy was used on a large scale for commercial and
agricultural purposes in England, Germany, France, Holland, and elsewhere until dis-
placed by the steam engine.
Disruption of petroleum supplies, either actual or threatened, has at various times
in the recent past, and now again, motivated national wind energy programs in Europe
and North America. In addition to the national programs, important work has been done
under sponsorship of private firms or other institutions. The Smith Putnamwind genera-
tor, which operated in the United States in the early 1940s, is an example of an effective
private-sector program [1].
Efforts to harvest the energy of the winds on an appreciable scale have, in recent
years, been oriented toward electricity generation, specifically toward generation for
distribution through existing electrical networks. Small units for pumping water, pro-
ducing electricity, or generating heat for decentralized, independent applications are at
present important in several locations. Recent decades have seen a drastic decline in the
number of decentralized wind units in use but this trend appears to be reversing.
The aerodynamics of wind machines [2] is reviewed in Chap. 22; however, a few
salient facts will be repeated here. The power in the wind itself is given by
P = .1 pAV3 (40.1)
In this formula A is the area of wind intercepted (A = 7rR2 for a horizontal-axis propeller
whose blades sweep out a circle of radius R), V is the wind speed, and p is the air density
—about 1250 g/m3 . Variations in p (altitude, humidity, etc.) are of minor importance in
affecting the power. Variations in V are extremely important.
Electric Power Generation: Wind, Waves, Tides 441
available from the wind than the generator can handle. To accommodate this situation,
the wind machine is designed to operate in such a manner that above rated wind speed
a progressively smaller fraction of the power in the wind is extracted as wind speed
increases. Blade pitch control is one way in which this may be accomplished. Thus,
above rated wind speed, output is constant. In some machines, output actually decreases
moderately with increasing wind speed above the rated wind speed.
The value of rated wind speed (the wind speed at which the machine is rated) is
chosen on the basis of a trade-off between value of energy and cost of equipment. If
orated is chosen to be low, say, equal to the average wind speed at the site, then actual
wind speed will frequently exceed rated wind speed and much of the wind energy avail-
able will not be harvested. If on the other hand, the rated wind speed is made very high
and a large electrical generator is installed, then the oversized equipment works to full
capacity only occasionally and full value is not obtained for money spent on equipment.
Usually the rated wind speed is chosen to be somewhere in the range 1.5 to 2.0 times the
average wind speed.
The dimensions and electrical ratings of a number of large wind generators are listed
in Chap. 22, Table 22.1.
The technology of wind machines, and also their history, is exhaustively discussed
in the literature [5-18]. The physics and engineering of the machines is reasonably well
understood, although the engineering is not yet optimized. A considerable amount of
testing and modification remains to be done before mass production can be undertaken
on today's designs. All large wind generators to date have been prototypes. All have
operated for limited periods, and in each case the performance has been more or less as
calculated. Failure or shutdown occurred either for economic reasons or for mechanical
reasons that were not fundamental, and that could have been corrected with more
engineering development work. There are no substantial technical problems that would
limit large-scale deployment of wind machines in the near future.
The dominant technology in all the national programs today, with one exception,
is the two- or three-blade propeller turning about a horizontal axis. The vast majority of
large wind-electric generators constructed in the past have been two-blade or three-blade
propellers. Other technologies, such as the vertical-axis Darrieus rotor, vortex concen-
trators, single-blade horizontal-axis rotors, and others may offer advantages, but the
advantages are not likely to be crucial for the development of wind energy. The crucial
factors that will determine whether or not wind energy contributes in a substantial way
to human energy needs over the next half-century are not technological in nature.
B. Waves
The power in ocean waves has been part of the human experience for thousands of years.
However, the history of attempts to exploit this power for human purposes has not been
extensive. Patents have been issued on a variety of devices, and many schemes have been
described conceptually. Some small-scale prototype devices have been tested. Up to
now no major development program has been carried through in any country. Small
devices are available, however, and are in limited use as power supplies for buoys and
navigational aids. From the engineering development point of view, wave energy
development is not nearly as far along as wind and tidal energy.
Electric Power Generation: Wind, Waves, Tides 443
Wave energy derives from wind energy, which derives in turn from solar energy.
Wave energy has the advantage over wind that the energy has been naturally concentrated
by accumulation over time and space and transported from the point at which it was
originally present in the winds. Thus, waves striking the west coast of the United States
in summer often originate in the Antarctic winter winds, whereas waves coming in winter
bring energy from the North Pacific storms.
The degree of power concentration effected by waves is quite substantial. Whereas
the power density in the wind may, at a good site, average some hundreds of watts per
square meter, wave power being transported across a plane perpendicular to the wave
propagation direction at a good site is from 10 to 100 times as large.
The major disadvantage of wave energy, as compared to wind, is that the energy
is available only on the ocean. The extraction equipment must operate in a marine
environment—with all that implies in terms of maintenance, construction cost, lifetime,
and reliability. The energy may have to be transported a great distance to shore. Also,
the energy concentration means that wave energy converters must be capable of with-
standing very severe peak stresses in storms.
The mechanical energy in waves takes different forms. There is the energy of
forward motion of the wave—the highly noticeable energy that slams into ships and
cliffs. Some of the proposed schemes are oriented toward this forward-motion kinetic
energy. Any geometric arrangement that absorbs energy by converting the forward
momentum of the wave into motions of its internal parts, without reemitting as much
energy as it absorbs, can extract this forward-motion energy.
Also very promising from the energy-extraction point of view is the potential
energy of the raised water at the wave crest. The gravity head between crest and trough
is not large enough to permit practical power generation by driving a hydro turbine
directly, but the potential energy in the wave is very considerable. The rate at which
work is done on a large ship by the ocean as the ship is lifted up by the swell is typically
several times the power being delivered by the ship's engines.
It is not difficult to invent a variety of ratchet, valve, or intermittent-pump mech-
anisms that can convert wave energy to hydraulic, pneumatic, or electrical form. The
engineering challenge is to find a cost-effective way to do it on a large scale. Since waves
come in a wide range of wavelengths and amplitudes, any effective device will either have
to be broadband—that is, nonresonant—or it will have to have its resonance frequency
continuously adjusted [19]. Many mechanisms have been proposed, and a number are
being investigated. For example, waves can be made to compress air in the top of a float-
ing tank, using one-way air valves. Electricity is generated as the air is bled out through a
pneumatic turbine [20].
In a different device, waves passing a pipe standing vertically in the water and
equipped with an internal flap valve that allows water to move upward in the pipe but not
downward, cause a water column to rise in the pipe, to a height many times the wave
height. The water can be released from an elevated port in a controlled manner, to drive
a conventional hydro turbine [21] . Another class of mechanism employs two floats with
different dynamic response characteristics. The differential motion is used to operate
a pump or mechanical engine. For example, a long vertical cylinder extending far below
the surface will remain nearly stationary as the waves pass by. A toroidal float surround-
ing it will rise and fall with the waves, and can be made to push and pull on pump
plungers attached to the cylinder.
444 Applications of Solar Energy
Another differential motion concept, called the "nodding duck," employs large
cam-shaped "ducks" mounted on a very long, floating frame. The "ducks" oscillate or
"nod" relative to the frame, driving hydraulic pumps. This "nodding duck" scheme [22-
24] , though only in the model testing stage, has been investigated more than most of the
other concepts, and is currently one of the leading contenders in the British wave energy
program. In a rather serious criticism it was claimed by a wind-power proponent that the
energy payback time for wave power systems in general and the nodding duck scheme in
particular may be excessively long, owing to the large amount of steel and concrete
required. This claim was disputed by the inventor of the nodding duct concept [25-26].
Other proposed wave converter concepts include sea bottom structures [27],
which presumably could be designed for lower dynamic stresses since they are removed
from the wave zone, converting the fluctuating overpressure as the waves pass by into
pressure differences in separate chambers; and floats supporting water turbines on long
vertical shafts. The turbines, located in undisturbed water, turn as the floats move them
up and down [28]. There are many others [29] . Some theoretical analysis has been
done on wave energy extraction [30-32].
It is interesting to recall that S. J. Savonius, some 50 years ago, used a rotor of the
type generally known as a Savonius rotor and generally used as a windmill, to extract
energy from the waves of the Mediterranean [33]. The rotor was fixed with its axis
horizontal and perpendicular to the direction of wave propagation (near the shore, to
which it was anchored rigidly, in this case). Water passing the rotor flowed both forward
and back at different times and depths; the asymmetric Savonius rotor can take power
from both the forward and back flows. Power outputs of a few kilowats per square meter
were obtained, with equipment that was far from perfected.
Engineering experience with wave converters up to now has been too limited to
constrain and focus the inventive efforts of the host of individuals interested in the
subject. Conceptual designs put forth cover a very broad range.
C. Tides
The first thing to appreciate about tidal power is that the total resource is not very large.
This is primarily because there are only a small number of possible sites in the world.
To a certain extent the number of possible sites depends on the value of energy.
To a limited extent also, the number of sites can be increased by advances in technology,
such as improved turbine technology to use lower hydraulic heads, or new construction
methods to reduce costs. Primarily, however, the suitability of a site for tidal power
development depends on the coastal topography and the height of the tides, both factors
outside human control. Lack of remoteness from centers of power consumption is also
important. Possible tidal power sites are enumerated in Table 40.1.
To understand the problems of tidal power development it is informative to con-
sider specific schemes, past, present, and proposed [34-38]. These harness the tidal
energy in different ways.
It is possible simply to hang an undershot water wheel in a tidal race, and in olden
days some mills were run this way. It is said that the water supply of the City of London
was pumped to a tower by a paddle wheel operated by tidal flow; the wheel was mounted
on a raft and moored between two of the piers of London Bridge. This tidal wheel is
Electric Power Generation: Wind, Waves, Tides 445
said to have operated until 1824. However, blocking the tidal water up in basins has great
advantages over simply hanging a wheel in the flowing water. A larger fraction of energy
available can be recovered, and some control is possible over when the energy is delivered.
The simple single-basin, single-effect scheme has been used for centuries (though on
a very small scale). The sea enters the basin through open sluices, then when the level has
reached its peak the sluices are closed and the water is impounded in the basin at high
tide level. When the level of the sea has fallen the water in the basin is allowed to go out
through the powerhouse. Power is generated on the ebb tide only. Alternatively, the
generator can be turned around and power generated on the rising tide, but this gives less
energy. The civil works required are only a single dam, with movable sluice gates and a
powerhouse.
Substantially increased utility with only minor increase in cost is possible today
with a single-basin, double-effect scheme. This is the system that has been used in the
only large tidal power station ever built, at the mouth of the Rance River, in France [39].
The civil works are the same as for the single-effect scheme, the difference being in the
turbines. The turbines used are different in design from the Kaplan turbines usually used
in low-head hydro schemes in that the turbine axis, and the direction of water flow, are
horizontal. This allows utilization of very low head in the presence of a strong tidal
current. The vane orientations are controllable hydraulically, and the unit can be
operated as either a pump or a turbine in either direction. In fact, the vanes can be
moved out of the way so that the unit is simply an orifice. Thus there are five modes of
operation: two directions of pumping, two directions of flow to generate power, and an
open orifice mode. (The orifice can also be blocked.) The multiple-mode operating
flexibility gives the possibility of generating power on both the incoming and the out-
going tide. It is also possible to pump water in or out of the basin in order to increase
generating capability a few hours hence. Using this capability, the Rance power station
has been operated to maximize the value of power delivered, not to extract maximum
energy from the tides [40].
Single-basin schemes can generate power only intermittently, but a double-basin
scheme can provide power continuously, or on demand, which is a tremendous advantage.
The drawback is that the civil works become more extensive. In the simplest double-
basin scheme there must be a dam between each basin and the sea, and also a dam
between the basins, containing the powerhouse. One basin is maintained always at a
lower level than the other. The lower reservoir empties at low tide, and the upper
reservoir is replenished at high tide. If the generating capacity is to be large, the reservoirs
must be large, which usually means that the dams will be long. The Passamaquoddy
project, described below, has been proposed as a double-basin scheme.
There are presently two tidal power installations operating, one in France and one
in the U.S.S.R. The one in France is a full commercial station. The one in the U.S.S.R.
is more in the nature of a pilot plant or engineering experiment, and is rather small in size.
The French project at Rance, near St. Malo in Brittany, uses a dam that goes
straight across the estuary of the Rance River. The dam is not too long-750 m, shore
to shore. The depth is never more than 12 m below mean sea level, there is an above-
surface rock part way across, the climate is moderate, and all in all the physical topogra-
phy of the site is about as favorable as can be found anywhere. The average tidal range is
8.4 m (±4.2 m), and the maximum is 13.5 m. Effective basin surface area is 22 km2;
Electric Power Generation: Wind, Waves, Tides 447
basin volume is 184,000,000 m3 . There are no special problems with this site, and it
was a very sensible choice for the world's first tidal power station.
Construction began in 1961 and the project was completed in 1966/1967. The
dam contains a small lock for commercial vessels, six movable sluice gates, and 24 turbo-
generator units of 10 MW each. The movable gates are used to accelerate filling and
emptying of the storage basin at small differential water levels. Installed capacity is 240
MW, but of course the average power-generating capability is less, because power can be
generated only intermittently. Maximum electric energy production capability is
544,000 MWh/yr, which gives a plant capacity factor, without any allowances for mainte-
nance or downtime, of 0.26. Maximum utilization of the stored hydraulic potential
energy is 18 percent, which may be increased to 24 percent in future by addition of 80
MW more generating capacity.
As already mentioned, the plant is not normally operated to maximize electric
energy production, but rather to maximize financial return, considering the time-of-day
variation of the value of energy. In 1974, the year of highest energy production up to
that time, electricity delivered totaled 507,000 MWh. Gross output from the turbo-
generators was 598,000 MWh; 91,000 MWh was used to operate pumps.
General belief is that the Rance project cost considerably more to build than its
proponents said it would at the time the decision was made to go ahead. Thus it has
generally been regarded as an engineering success and an economic discouragement. A
noticeable slackening of interest in constructing more tidal power facilities did occur in
1966-1967 after the costs incurred at Rance became clear [41]. The threefold increase
in world oil prices that occurred in late 1972 greatly improved the economic attractive-
ness of the Rance project, as did the further two-fold increase that occurred in 1979.
Still by early 1980, no new tidal power projects have been undertaken.
On the other hand, according to an official of the utility company operating the
Rance facility, the final cost, measured in inflation-corrected francs, exceeded the
planned cost by only 14 percent. It is also pointed out that at no time was the under-
taking of the project justified solely in terms of the value of fuel savings. A major
factor in the decision to go ahead was the fear of power shortage, caused partly by the
Suez crisis and partly because the future of nuclear power appeared uncertain. A full
breakdown of construction cost, operating costs, amortization schedules, and so on, is
given in this paper [42].
The tidal power station in the U.S.S.R. is at Kislaya Guba, on the Barents Sea,
70 km north of Murmansk. It is small and was intended as a pilot project. It has one
generator of 400 kW rating and delivers 700 to 800 MWh/yr to the local electric grid.
Operations began in 1968 [42].
The imaginative aspect of Kislaya Guba is the way it was built. The powerhouse
was constructed of precast buoyant concrete, and the single turbine (horizontal-axis
type) was installed in a shipyard far to the south. The structure was then towed about
100 km, positioned in the narrow tidal inlet at slack tide, flooded and dropped on a
previously prepared sand-aggregate foundation, and quickly anchored down before the
tide rose again. There was no cofferdam and no construction below sea level at the site.
Considering that nearly 30 percent of the cost of the Rance station is attributable to
the cofferdam, the Kislaya Guba construction technique is certainly of interest. The
site of Kislaya Guba is a narrow inlet, only 30 to 50 m across. The powerhouse itself
448 Applications of Solar Energy
blocks the central 18 m of the inlet; the rest is blocked by earth-stone dams. The power-
house dimensions are 36 X 18.3 X 15 m, length, width, and height, respectively.
Rance and Kislaya Guba are the only tidal power projects that have been con-
structed. After these, Passamaquoddy has received more study than, probably, all others
combined [43] .
Passamaquoddy Bay is located on the United States-Canada border, where the Bay
of Fundy cuts in from the Atlantic. It is not the best site on the Bay of Fundy, but the
better sites are completely within Canada, which still has undeveloped hydro in substan-
tial amount and thus lacks motivation to undertake tidal power development. Extrapola-
tion of past trends leads to very large electricity need projected for Canada's future,
however, and on this basis a positive recommendation for a large (7000-MW) tidal power
installation on the Bay of Fundy was recently put forth by the Canadian National Energy
Board [44] .
Passamaquoddy is the southernmost site in the Bay of Fundy, which puts it
closest to the region where the energy is required. The closest large market is Boston,
almost 300 mi away.
Studies of the Passamaquoddy project began seriously in the 1920s, were aban-
doned for a time, revived suddenly in the middle 1930s under the leadership of President
Franklin Roosevelt, and were then abandoned in the late 1930s after some preliminary
construction. The prospects were reviewed by an official commission in the late 1940s,
and detailed plans and surveys were made in the late 1950s. The project received
additional impetus from President John Kennedy in the early 1960s. It is presently
being studied again.
Technologically, Passamaquoddy is a bigger project than Rance, and far more
difficult. It would be constructed, according to one favored version, as a double-basin
scheme, with 7 mi (11.2 km) of dams, 160 movable sluice gates, four large locks for
ocean shipping, and installed capacity of 1000 MW. The cost estimated by friends of
the project in 1961 was around $1.2 billion. (At that time a slightly favorable benefit/
cost ratio was calculated based on an interest rate of 2.87 percent). With a two-basin
scheme the power would be available more or less on demand, and its value was calcu-
lated at peaking power rates. The depths at which the civil construction would have to
be carried out are much greater than at Rance, ranging up to nearly 100 m. Cofferdams
have never been emplaced in strong currents at such depths. The volume of excavation
is many times that of the Rance project. Moreover, the weather is certainly less favorable
than at Rance, the site is more remote from the market, and the construction likely to
be much more expensive.
A high level of international cooperation would be essential for the Passamaquoddy
project, not only during construction but also for operation, since most of both reservoirs
and all the water-control gates would be in Canada. Only the powerhouses would be in
the United States.
Of the remaining possible sites (Table 40.1), a number of French sites have been
under study for many years, but the French government, deterred perhaps by the high
cost of Rance, has given a higher priority to atomic energy. In England, the River Severn
scheme has been studied from time to time, and is again being put forward by proponents
[45, 46] . It is less favorable than Rance, but not nearly as difficult as Passamaquoddy.
The seriousness of the Soviet interest in developing available sites is difficult to evaluate.
Electric Power Generation: Wind, Waves, Tides 449
Although Kislaya Guba was intended as a pilot project for much larger tidal power
stations to be located along the northern and eastern coasts of the U.S.S.R., none of these
larger projects has been taken up.
Canada is not likely to push tidal very seriously until its hydro resources are more
completely developed. The Argentine site is subject to the same consideration.
Sites in Australia [47] , Canada [48] , and elsewhere [37, 38] have been described
that appear to have the physical requirements for tidal power sites but not the economic
ones. In particular, most of the places described are remote from electric power demand
centers. These sites have not been included in Table 40.1.
It is of interest to attempt to estimate the possible contribution of wind, wave, and tidal
energy to world energy supply at various future dates, to the year 2020. The accuracy
of such estimates is low. However, some kind of estimate is necessary to give perspective
for assessing the possible importance of these unconventional sources of energy. Note
that what is being estimated is the possible contribution, the amount of energy that
could be taken from the indicated sources in the indicated time frame, not the amount
that in fact will be taken. The two differ because of many considerations. The most
important (in order of importance) are the following:
The possible contribution of each of the three sources being considered here is,
then, the largest amount of energy that could reasonably be produced, considering
450 Applications of Solar Energy
A. Wind
The total energy in the global atmospheric circulation has been estimated by various
authors, but is not relevant. Average energy in the near-surface wind over some large
geographic area is also not relevant. What is relevant is that average wind velocity, at a
height of 20 to 30 m above ground level, is sometimes high enough so that the energy
flux ("power in the wind") through a properly oriented vertical section may reach
500 W/m2 (annual average). A wind machine located at such a place can convert to
Electric Power Generation: Wind, Waves, Tides 451
electricity perhaps 175 W of the 500 W. It is interesting to note that the annual average
solar power that may be intercepted at a good site is about 250 W/m2 and that from this
not more than 25 or 30 W of electricity may be recovered.
The question arises, how densely may the wind machines be arrayed at the good
high-wind locations? How many units per square kilometer? Or, more exactly, since
the answer depends on the size of machine, how many square meters of wind may be
intercepted per square kilometer of ground surface? The other question is, How much
high wind land area can be found?
There are two different physical effects of importance bearing on the first question.
One is wake interference. Downwind of a large operating wind machine is a wake, where
the wind is altered in form (made more turbulent) and reduced in speed (energy has been
removed from it). If a second machine is located in the wake interference region, it will
not deliver as much energy as it would if the first machine were absent.
The other physical limitation is overall depletion of the wind energy in the lowest
100 m or so of the atmosphere, which is the layer from which the wind machines extract
energy. If an entire plain is covered with large wind machines, those in the central
portion of the array and on the downwind edge will not be able to deliver as much energy
as those on the upwind edge, unless the machines are spaced far enough apart to corre-
spond to the rate at which energy is coupled into the lowest layer of the atmosphere
from the winds above.
The first of these problems, wake interference, has been investigated with model
wind turbines in a wind tunnel [49], and also by field measurements behind the NASA/
ERDA 100 kW Sandusky, Ohio, machine. More work remains to be done, but the
problem is reasonably well under control. It appears that a spacing of five rotor diame-
ters is adequate to avoid excessive wake interference, provided only that the wind
turbines are not laid out in straight rows [49].
The second problem is not completely understood. At present the only approach
is through calculation. The calculations become quite involved and depend on models of
the terrain and the atmosphere that may or may not correspond to the actual situation.
Railly [50] has estimated that the allowable density of wind machines on a large, flat
plain in order that the undisturbed wind speed be available to all the machines is in the
neighborhood of 1500 m2 of wind turbine area per square kilometer of land. For
machines with 60-m-diameter rotors, that density is only 0.5 machine/km2 , a separation
of 35 rotor diameters. This is much greater than the wake interference limit, and, if true,
says that available wind energy is likely to be limited by the ability of the winds at greater
heights to couple energy into the layer of moving air next to the ground.
The second question, what area is available for siting wind machines, is difficult to
answer with confidence because of lack of data. To estimate the energy available from
wind power in a certain region one should know the mean annual wind speeds at, say,
25 m elevation at many sites in the region. The total land area underlying the high-wind
zones can be computed, and the amount of extractable wind energy calculated. This type
of wind-speed data is nowhere available. The two types of data available are (a) at a very
few scattered places, at most a few hundred, worldwide, wind speeds have been measured
at heights and at sites appropriate for wind generators, for periods of several years or
more; and (b) at many more places, several thousand worldwide, wind-speed data have
been accumulated for aviation or general weather purposes, usually at a height of 10 m
452 Applications of Solar Energy
or less. The sites are usually at airports or atop office buildings in cities, or otherwise
close to places of high population density. There is a definite correlation between high
population density and low wind; consequently, the data are biased toward low wind
speed and low power density. For most populated places, mean wind speeds are likely to
be in the range 2 to 5 m/s (5 to 10 mph), too low for economic utilization.
In a recent study, Coty and co-workers [51] estimated the wind energy resource in
the western United States. Their procedure was as follows: Data were analyzed from 348
existing wind-observing stations scattered over the region. A small percentage of these
(2 percent) had mean wind speeds exceeding 7 m/s (15.7 mph), corresponding to mean
annual wind power densities greater than 400 W/m2 . [Note: Eq. (40.1), with V = 7 m/s,
gives less than 400 W/m2 . On an annual basis, however, the higher winds contribute
more to the power than to the mean wind speed. Also, the wind speeds from the wind-
observing stations are at 10 m above ground, and wind generator blades are higher.] It
was then argued that the same small percentage of the total land area of the region—that
is, 2 percent—experienced mean annual wind power densities of 400 W/m2 , or more.
The fraction of this area that would be "off limits" to windmills, such as national parks,
national defense areas, urbanized areas, and so on, was estimated. The remaining part of
the high-wind territory was supposed to be available for wind generators, at a density
determined by aerodynamic interference considerations. In general, wind generator use
does not exclude other use , such as agriculture or mining.
Though inexact, this method seems rational and likely to estimate the wind poten-
tial conservatively. Another approach is simply to search for good wind sites. When a
site is suspected to have high wind speeds, an anemometer is installed, and when the
wind power is verified by several years of measurements, the site is regarded as a resource,
much like a mineral resource, to be exploited when economics and convenience warrant.
The only large-scale, persistently pursued program of this sort was carried out in the
1950s in Britain [8]. Several dozen sites with mean wind speeds over 9 m/s (20 mph)
were found.
Two large federally sponsored studies [51, 52] recently concluded that about
1012 kWh/yr of electricity could be produced in the United States (excluding Alaska)
from wind generators in high windareas ("high wind" was defined as mean wind speed
greater than 7 m/s at 10 m height) using land not now used or reserved for conflicting
purposes. The annual electricity consumption of the United States today is about 2 X
1012 kWh/yr, so on the scale of electric energy use the potential wind contribution is
very appreciable. On the scale of total energy use, the potential wind contribution does
not look quite so large, since only about one-quarter of U.S. energy consumption is for
the purpose of making electricity. Also, if electricity use, or energy use, is supposed to
increase several percent per year compounded, the percentage contribution possible
from wind in the twenty-first century is not so large.
If the threshold for a "high-wind" area is lowered to, say, 6 m/s (13.5 mph), the
amount of wind energy potentially available increases by several times. Of course, the
energy extracted from the lower-speed winds will cost more.
The lack of extensive wind data is not actually a major barrier for making gross
worldwide estimates of possible wind contribution. This is because it appears that the
contribution will not be limited by the amount of wind. Although not nearly as large
as solar, there seems little doubt that the total amount of wind energy over land is quite
Electric Power Generation: Wind, Waves, Tides 453
large. Even if inaccessible locations and all locations with mean wind speed below 6 m/s
are excluded, the total energy is still large. The wind contribution will be limited by cost
of the wind energy as compared to cost and availability of other energy sources.
The schedule of possible future contributions by wind is controlled almost entirely
by this cost comparison. Wind turbines are fairly simple mechanical devices, certainly less
complex than airplanes. Their future introduction is not likely to be limited by the rate
at which they can be manufactured. Neither is it likely to be limited by environmental
or political considertions.
B. Waves
The power in the waves is known accurately at a few locations, where wave heights and
periods have been measured systematically for years. The power in all the waves in all
the world's oceans is known only very approximately, estimated from wave data intended
for other purposes, such as that in the Atlas for Mariners published by the U.S. Navy.
We are concerned, moreover, not with the energy in the waves, but with the rate at
which energy may be extracted from the waves [21, 53, 54]. This maximum energy
extraction rate is much greater, in principle, than the product of the energy stored in
all the waves of the world's oceans multiplied by their velocity. The power in all the
waves in all the oceans has been estimated [21] at 2.7 X 1012 W, and the maximum
power that could be extracted (in principle) at 10 to 100 times this.
Much more accurate, and considerably more relevant too, are the measurements of
power density at a site in the northeast Atlantic, 59°N 199W, west of Scotland [55] . At
this site, which is probably typical of a large section of the North Atlantic ocean, annual
power in the waves crossing a line perpendicular to the average propagation direction
averages 91 kW/m. This is a prodigious amount of power. Even if it proves desirable to
make use of sites of lower power density, because, for example, they might be closer to
shore, the power density is still enormous.
As already noted, the efficiency of the wave converters is very uncertain, but will
likely lie between 10 and 80 percent. Moreover, there will be transmission losses, and,
if chemical transmission is used, conversion losses. A plausible estimate of overall
efficiency would be 20 percent; it could be as high as 40 percent or as low as 10 percent.
Consider wave converters coupled together in strings 10 km long. A considerable
number of these strings—say, 100—could be sited in the northeast Atlantic alone. Thus
we are looking at a wave-power potential for Western Europe of about 2 X 10' kW,
average power, or 1.75 X 1011 kWh/yr. For electrical transmission the waves have to
be within 1000 km, and preferably within 500 km of populated land where the power
can be used; this will limit the worldwide potential to perhaps 108 kW, or five times the
previous number. If chemical transmission (H2 , NH3 , NH4 NO3 ) is envisioned, with
tankers or freighters visiting the wave converters to take off the chemicals, then the
winds in the southern oceans, the north Pacific, the far northern Atlantic, and all other
windy sites far from population centers become available and the potential power is
multiplied perhaps 10 times, to 109 kW.
There is the question of time scale. Presumably wave converters would be built
in shipyards; they would be less complex than most ships. Materials or construction
labor constraints seem improbable. Almost certainly the constraint will be money. The
454 Applications of Solar Energy
extent to which resources are channeled into building wave converters will depend on the
operating experience and apparent economic prospects of those already in use. The
present situation is that there are a number of designs to be explored, all of which will
probably work, but some will work better and more cost effectively than others. The
time scale of introduction of wave converters seems, at this point, likely to be limited
by the rate at which prototypes are funded. A variety of full-scale prototype units could
be in the ocean by the mid-1980s. The beginning of commercial operation in 1990 with
appreciable amounts of wave energy being extracted by 2000 seems possible.
C. Tides
To harness the tides for power, two physical conditions are necessary. First, the tidal
amplitude must be large—several meters at the very least. Second, the coastal topography
must be such as to allow the impoundment of a substantial amount of water with a
manageable volume of civil works. This means a bay with a narrow inlet, or a river
estuary, or a set of strategically located islands. When the coastlines of the world are
examined for locations meeting both of the above conditions, only a few dozen locations
are found. Some of these are not practical, because of inaccessibility or for some other
reason
The above considerations were noted previously, and form the basis for the enumer-
ation of possible sites in Table 40.1. There are no doubt other possible tidal energy sites
in the world, but there are not likely to be many.
For each site in Table 40.1, the maximum possible amount of energy that could, in
principle, be taken from the sea in one tidal cycle, ignoring engineering limitations, is
given by a simple formula:
E = pgR2 S
where p = sea water density, g = acceleration of gravity, R= tidal range, and S = basin
surface area.
Computing E for each of the sites in Table 40.1 and summing gives 13 X 10"
kWh/yr for the total hydraulic energy available at suitable tidal sites, worldwide. Only
10 to 25 percent of the hydraulic energy can be recovered as electrical energy, consider-
ing engineering realities, primarily minimum hydrostatic head requirements. Considering
also that many of the sites enumerated are inaccessible or otherwise impractical, it is
clear that only a small fraction, perhaps a few percent, of the total hydraulic energy will
ever be harvested. World energy use presently is about 40 X 1012 kWh/yr. It seems
certain that tidal power can never make more than a minor contribution to the world
energy picture. Particular tidal sites will presumably be developed when economic and
other considerations are favorable.
Table 40.3 summarizes the characteristics of the three energy sources and assesses the
present situation in qualitative terms.
Electric Power Generation: Wind, Waves, Tides 455
A. Wind
The present status of wind energy is rapidly changing. The largest national program is in
the United States. A test unit with a two-blade propeller sweeping a 38-m-diameter
circle has been erected, and operated sporadically since late 1975. Three more units of
the same size, driving 200-kW generators, have been installed in parallel with existing
diesel generating sets of electric utility lines serving isolated communities. A fourth unit
of the same type has been installed (1980) in parallel with oil-fired steam turbine genera-
tors. The locations of these four units; designated MOD-0-A, are, in chronological order
of installation, Clayton, New Mexico; Culebra, Puerto Rico; Block Island, Rhode Island;
Oahu, Hawaii. A larger machine MOD-1, also a two-blade propeller, 60-m-diameter
swept circle, 2000-kW generator, has been installed (1979) at Boone, North Carolina.
A cluster of three very large machines, called MOD-2, will be installed near Goldendale,
Washington in 1980-81. MOD-2 is a two-blade propeller with a 95-m diameter; the
generator is 2500 kW.
Although the main thrust of the U.S. program is directed toward the development
of large machines for use on electric utility networks, some work on design improvement
and applications engineering of small machines, for agricultural and isolated location use,
is being sponsored. In addition, a number of research analyses and experimental studies
on novel technologies, wind resource studies, systems studies, and others are included in
the program.
In addition to the federal program there are a number of private sector machines
ranging in size from 200 kW to 4000 kW on order. None of these are in operation as of
early 1980.
In Canada, the only country not emphasizing horizontal axis propellers, a two-
blade Darrieus rotor 37 m high, driving a 230-kVA generator, has been installed in a high-
456 Applications of Solar Energy
wind location in the Magdalen Islands, Gulf of St. Lawrence, Province of Quebec. This
machine is by far the largest Darrieus rotor ever erected. It will be operated in parallel
with existing diesel capacity by the provincial utility company presently serving the area
(Hydro Quebec); and, hopefully, will show favorable economics as a fuel saver from the
outset.
Other currently active programs, nationally sponsored, are in Denmark, Sweden,
West Germany, Netherlands, England, and New Zealand. Of these, only Denmark,
Germany, and Sweden are involved with actual large machines at the present time. In
Denmark two machines, each 630 kW, were erected at Nibe, in Jutland, in 1979.
In addition to the nationally sponsored programs, there are nongovernment projects
worthy of mention. In the United States a private company* formed for the purpose of
buiding and selling large wind generators, has erected a large 200 kW horizontal axis
machine on Cuttyhunk Island, off the Massachusetts coast in 1977. Recently a copy of
this prototype was sold to the Nova Scotia Power Board for use at a Cape Breton site.
Alcoa, a large corporation, has begun marketing Darrieus rotors. Several 500-kW units
have been sold, though none are yet in operation. Several utility companies in the U.S.
have purchased large wind generators without direct federal subsidy. In Denmark, an
independent school at Tvind, in Jutland, constructed a two-blade propeller of 54 m
diameter. Operation began in 1978. This is the largest machine ever constructed in
Denmark.
B. Waves
The present status of wave energy development is that the technology is at a much more
rudimentary stage than either wind or tidal. Except for small-scale applications, mostly
marine navigational aids, no operating wave power machines exist at present. Engineering
is in the study, design, and experimentation stages. The problem of transmitting the
energy from the point of extraction to the point of use is (at best) a matter for specula-
tion and study.
The efficiency of conversion of mechanical energy in the waves to electrical energy
at the generator is quite uncertain at present; considering all the proposed schemes, it
can be said only that the overall conversion efficiency of a commercial wave-power
machine, when built, will probably lie in the range 10 to 80 percent. (The 80 percent
number is based on wave tank experiments with "nodding duck" models.)
The present center of world activity in wave-power utilization is Great Britain.
Some work is also in progress in Japan, Norway, Sweden, and the United States. The
main development work in Britain at present is embodied in a 2-year program to build
and study scale models (1:100) of four different types of devices judged to have promise.
Of the four, the "nodding duck" and the torus and cylinder appear to be the most
developed at this time.
In Japan a large number of wave-powered buoys and navigational aids have been
operating for a number of years. The electrical wave-powered generators to operate the
lights and horns are in the range from a few tens of watts up to a kilowatt or so. Some
study and design work has been undertaken toward larger power generators.
C. Tides
The present status of tidal power is somewhat quiescent. The Rance station continues to
operate routinely and successfully. Little is heard from the small station at Kislaya Guba,
but presumably it is also continuing to operate. In any case it has long since completed
its primary purpose, of serving as an engineering experiment, demonstration, and pilot
plant. Proposals and studies appear from time to time and usually recommend positive
action. However, none of the recommended projects seems to have any immediate
prospect of actually being constructed. This is for several reasons. First, and most
important, tidal power projects are large—large in terms of money required, large in terms
of site impact and land-use planning, large in terms of engineering challenge. Unlike
wind, there is no way to implement tidal power projects on a modular or incremental
basis. It is to be expected that extensive and numerous studies will precede action.
Second, the economics of the proposed projects are all somewhat marginal, and often
the engineering challenges are large enough to introduce great uncertainties in the cost
estimates. Third, there is no prospect that tidal power will ever be more than an incre-
mental contribution to total national energy needs. The positive imperative to overcome
the reluctance of leaders to take political and financial risks is thus somewhat limited.
Winds, waves, and tides are sometimes said to be "energy sources," rather than "power
sources." This confusing terminology expresses the fact that a wind-wave-tide installation
delivers energy according to a schedule controlled by the source, not by the energy user.
Thus a wind-wave-tide station cannot replace a conventional electric power station. The
usual terminology describing electric power stations ("peak load," "base load," etc.)
does not fit.
There are several ways, conceptually, to make use of an intermittent energy source,
such as wind, waves, or tides. The most obvious is to provide storage, the storage facility
being charged from the source during periods of low demand and supplying the load
during periods when demand cannot be met from the source alone.
Putting the discussion in terms of wind, storage of some sort may be provided for
each wind generator, or each cluster of generators, or for the entire wind component of
the system. In this case all the wind-generated energy is stored until needed. If delivered
at a time of peak electricity demand, the wind-generated electricity can be valued at the
fuel cost of the peaking generators. Storage large enough to absorb the output of the
wind generators for 6 to 12 hours is required. According to Sorensen [56], a wind
generator equipped with 10-hour storage will be able to provide power at or above the
annual average level for 64 percent of the time; for 24-hour storage the figure goes to
72 percent. Without storage the power output of the wind generator exceeds the annual
average 42 percent of the time, and is zero 31 percent of the time.
The usual and established way to store energy on a large scale is with a hydro
installation. In a grid with existing hydro capacity, the storage can sometimes be accom-
plished by displacement, that is, simply by not using as much water to generate electricity
at times when the wind generators are delivering power. The water not used is held
behind the dam for future use. The large Smith-Putnam wind turbine, connected into a
458 Applications of Solar Energy
system with abundant hydro power, was operated in this manner. However, there are
many utility systems with no hydro. Moreover, in some cases installed hydro capacity
in a utility network cannot be used to provide displacement storage. For example, if
the network already gets a large portion of its energy from sources other than hydro,
what hydro there is may be already scheduled as peaking power. In that case, when the
wind generators deliver energy in the middle of the night, there is nothing to displace.
Systems with installed pump storage, rather than ordinary hydro, may be able to make
good use of wind energy if sufficient pump capacity is available, since wind energy can
then be converted into lifted water without disturbing the remaining part of the network.
Systems with ordinary hydro can handle wind energy in the same way if separate pump-
ing loops, to lift water from the lower pool to the upper reservoir when wind energy is
available, are installed. Of course, this costs more than displacement storage, which
doesn't cost anything. It is also less efficient than using the wind electricity at the time
it is generated. Run-of-river hydro cannot be used for displacement storage.
Another type of large-scale storage suited to use with wind generators is pneumatic
—compressed air in underground caverns. This is not yet a well-established and standard
technology. Other possibilities, such as flywheel or electrochemical storage at the sub-
station level, are still in the future.
Without storage, it may be argued that a large number of wind, wave, or tidal
generators, geographically dispersed, will still displace a certain amount of system capac-
ity, because the wind or waves or tides are not likely to disappear in all spots simultane-
ously. Whether or not this is true in any particular case, with sufficient certainty to
satisfy power network reliability requirements, is not such an easy question.
Another way to utilize an intermittent source is to put it in parallel with a power
source that has the capability of delivering energy on demand. The function of the inter-
mittent source is then to save fuel (or water, in the case of hydro) that would otherwise
be consumed in the conventional source. The conventional source must have the capabil-
ity of responding on the time scale required according to the characteristics of the inter-
mittent source and the load variations in the system—within minutes, in the case of wind.
The value of the energy delivered by the intermittent source when used in this way
depends on the portion of the cost of power from the conventional source that is attrib-
utable to the fuel. If the conventional source is a high-fuel-cost, low-capital-cost type of
unit, such as a diesel set or a gas turbine, the value of the fuel-saving energy delivered by
the intermittent source will be larger than if the conventional source has a low fuel cost
and a high capital cost. Thus, for example, electricity generated from oil has a fuel cost,
currently of 4-6 cents/kWh. Electricity generated from coal or uranium has a fuel cost
less than half this. The value of an intermittent source used as a fuel saver depends on the
nature of the other generators in the system.
When the conventional source in parallel with the intermittent natural source is a
hydro unit, the value to be assigned to the "fuel" saved is less obvious. This case was
considered carefully by Putnam [1]. It depends on the hydrographic and techno-
economic details of the water and energy system in which the hydro station is embedded.
For example, a very useful hydro installation for paralleling an intermittent natural
source is one that (a) is limited by water availability, (b) has substantial ponding capacity
—that is, large reservoir capacity for storing water behind the dam, (c) has no overriding
water-release requirements arising from navigation, irrigation, recreation, fisheries,
conservation, or other reasons. A less useful kind of hydro installation is one like Niagara
Electric Power Generation: Wind, Waves, Tides 459
or Bonneville, located on a river with a very large flow, so that the flow cannot be inter-
rupted—if the power is not required the water is allowed to flow through anyhow. In
these cases the power station is usually operated base load at its full capacity in order to
produce electric energy at the lowest possible cost.
A difficulty in using wind (or wave) energy to save fuel in electric generation
stations in large grids is that the energy produced by the intermittent source may not be
usable at all during off-peak periods. This is because base-load power plants are typically
very large, and cannot be turned on and off rapidly. Perhaps half an hour to an hour, or
even more, is required. (Peaking plants have response times of only a minute or two, and
hydro used as peaking power can be on line in seconds. However, at off-peak periods the
peaking plants are not operating.) Not only is the wind energy less valuable during off-
peak periods because the power stations operating then have low fuel cons, it may have
no value at all.
In order to be able to make use of wind-generated electricity at least to save fuel in
the base-load plants, "grid stability storage," in the form of electro-chemical storage
batteries, has been proposed [51]. The idea is to provide a small amount of storage—
enough to absorb wind generation for perhaps one-half hour. If the wind generators are
carrying part of the system base load and the wind dies, the grid stability storage provides
the needed energy while the conventional (coal/nuclear) capacity gets operating. If the
wind returns, the storage is recharged while the conventional capacity is being shut down.
The studies showed that the amount of wind energy that could be utilized was seven
times greater with grid stability storage than without it. Considering the low fuel cost
of base-load power, however, and the rather considerable cost of storage, it remains to
be established that the idea is worthwhile.
Another approach to using the energy from an intermittent source is to give up
the idea of connecting the source into the electricity grid, and to recover the energy in
some other way. The most common proposal is to recover the energy in chemical form,
by electrolyzing water to make hydrogen gas. Ammonia, or ammonium nitrate fertilizer,
can also be made using water and air as raw materials. A proposal has been made to
harvest oceanic wind energy in the form of methane, using pelagic carbonates as the
source of carbon. Sometimes the proposal to recover the energy in chemical form is
motivated by obvious technical problems in connecting to an electric grid—for example,
from a wind or wave energy installation far out at sea. Sometimes it is proposed because
a need exists for the chemical end product, such as for fertilizer in the low-income, oil-
poor countries. The economics of these chemical energy schemes are in all cases uncer-
tain, but not necessarily hopeless.
Heating of structures and of water consumes a great deal of energy, more than half
the total in some countries. At latitudes higher than 50°, heating with wind energy may
be a more practical proposition than solar heating. Whether this should be classified as
fuel saving or as interruptible load depends on the thermal mass of the building in relation
to its heat loss rate.
A considerable amount of the energy consumption of certain regions is for pumping
water. In some places much of this load could be taken over by wind machines, since it
is to some extent an interruptible load. Other possible interruptible loads include sewage
aeration, heating or cooling of large thermal masses, such as refrigeration plants, the pro-
duction of ice, and perhaps some agricultural processing operations.
460 Applications of Solar Energy
REFERENCES
1. P. C. Putnam, 1948. Power from the Wind, Van Nostrand, New York, 1948. Re-
printed by Van Nostrand Reinhold, New York, 1974.
2. U. Rutter, "Optimum Wind Energy Conversion Systems," Annual Review of Fluid
Mechanics 9, 399-419 (1977).
3. E. N. Fales, "Windmills," in Standard Handbook for Mechanical Engineers, T.
Baumeister and L. S. Marks, eds., McGraw-Hill, New York. The same article
appears in many editions, for example, 7th ed., 1967.
4. P. T. Smulders, "Physical Aspects of Windmill Design," Physics in Technology 7,
no. 5, 208-214 (1976).
5. B. L. Burke and R. N. Meroney, Energy from the Wind: Annotated Bibliography,
Colorado State University, 1975, First Supplement, April 1977.
6. Wind Energy Utilization-A Bibliography with Abstracts, Application Center,
University of New Mexico, Albuquerque, April 1975, TAC W 75-700. A complete
bibliography, cumulative 1944-1974. Title, keyword, author, index, 493 pp.
7. R. Van Steyn, Wind Energy: A Bibliography with Abstracts and Keywords, 2 vols.,
Library Administration, University of Technology, Post Box 513, Eindhoven,
Netherlands.
8. E. W. Golding, The Generation of Electricity by Wind Power, Philosophical Library,
1955. Reprinted by Halstead-Wiley, New York, 1976.
9. V. Torrey, Wind-Catchers, Stephen Greene Press, Brattleboro, Vermont, 1976.
10. D. M. Simmons, Wind Power, Noyes Data Corporation, New Jersey, 1975.
11. Proceedings of the United Nations Conference on New Sources of Energy, Vol. 7,
Wind Energy, Rome, 1961, United Nations, New York, 1964; reprinted 1974.
12. J. M. Savino, ed., Wind Energy Conversion Systems-Workshop Proceedings,
Washington, D.C., 1973, National Science Foundation Document NSF/RA/W-73-
006. Available from National Technical Information Service, U.S. Dept. of Com-
merce, Springfield, Virginia.
13. F. Eldridge, ed., Proceedings of the Second Workshop on Wind Energy Conversion
Systems, Washington, D.C., 1975, Mitre Corporation, McLean, Virginia, Document
NSF-RA-N-75-050; MTR-6970. Available from National Technical Information
Service, U.S. Dept. of Commerce, Springfield, Virginia.
14. F. R. Eldridge, Wind Machines, 1975. Available from Supt. of Documents, U.S.
Government Printing Office, stock number 038-000-00272-4.
15. L. Weatherhold, ed,, Vertical-Axis Wind Turbine Technology Workshop, Albuquer-
que, New Mexico, 1976, Document SAND 76-5586. Available from National
Technical Information Service, U.S. Dept. of Commerce, Springfield, Virginia.
16. E. W. Golding, "Electrical Energy from the Wind," Proceedings of the Institution
of Electrical Engineers 102A, 667-695 (1955).
17. G. Gimpel, "The Windmill Today," Engineering 185, no. 4812, 686-690 (1958).
18. J. M. Drees, "Blade Twist, Droop Snoot and Forward Spars Vertiflite," American
Helicopter Society 22, no. 5, 4-9 (1976).
19. K. Budar and J. Falnes. "A Resonant Point Absorber of Ocean-Wave Power,"
Nature 256, 478-479 (1975).
20. M. E. McCormick, "An Analysis of a Stationary Pneumatic Wave Energy Con-
verter," Journal of Engineering for Industry-Transactions, ASME, 1015-1019
(1975).
21. J. D. Isaacs, D. Castel, and G. L. Wick, "Utilization of the Energy in Ocean Waves,"
Ocean Engineering 3, 175-187 (1976).
Electric Power Generation: Wind, Waves, Tides 461
49. 0. Ljungstrom, "Large Scale Wind Energy Conversion System (WECS) Design and
Installation as Affected by Site Wind Energy Characteristics, Grouping Arrange-
ment and Social Acceptance," Wind Engineering 1, no. 1, 36-56 (1977).
50. J. W. Railly, "A Possible Saturation Criterion for Wind Energy Extraction," Wind
Engineering 1, no. 1, 23-25 (1977).
51. U. A. Coty, Wind Energy Mission Analysis—Final Report, Lockheed-California
Company, Report LR27611, 3 vols., 1977. Available from National Technical
Information Service, U.S. Dept. of Commerce, Springfield, Virginia.
52. J. A. Garate, Wind Energy Mission Analysis—Final Report, Executive Summary,
General Electric Company-Space Division, Report C00/2578-1/1, 1977. Available
from National Technical Information Service, U.S. Department of Commerce,
Springfield, Virginia.
53. N. N. Panicker, "Power Resource Estimate of Ocean Surface Waves," Ocean
Engineering 3, no. 6, 429-439 (1976).
54. J. D. Isaacs and R. J. Seymour, "The Ocean as a Power Resource," International
Journal of Environmental Studies 4, 201-205 (1973).
55. D. Mollison, P. 0. Buneman, and S. H. Salter, "Wave Power Availability in the NE
Atlantic," Nature 263, 223-226 (1976).
56. B. SOrensen, "Dependability of Wind Energy Generators with Short Term Energy
Storage," Science 194, 935-937 (1976).
chapter 41
41.1 INTRODUCTION
The objective of this discussion is to acquaint those working in the solar energy field
with some of the techniques available for collecting, storing, and presenting data. This
will generally lead into the fields of instrumentation engineering, and data processing.
These are well-developed disciplines from which the solar investigator can draw so as to
understand better the operation of a particular solar installation.
The selection of the hardware to instrument a particular solar installation depends
on the complexity of the installation and the objectives of the investigator. For example,
a one-time measurement of the stagnation of a collector might be accomplished with a
single thermometer, whereas a year's evaluation of a solar-heated house would need a
small computer and many sensors to handle the data.
Three fundamental factors govern the selection of the instrumentation: (a) the
complexity of the system being studied, (b) the length of time over which the study
will be conducted, and (c) the accuracy of results required.
These factors directly affect the number of points that will be instrumented, the
quantity of data that must be handled, and the cost of the system. It is generally wise
to minimize all as much as possible.
The following paragraphs will discuss briefly the hierarchy of data collection
techniques available to place the hardware complexity in perspective. This will be fol-
463
464 Applications of Solar Energy
The type of solar energy-related system and the end use for which the collected data
are intended will strongly affect the data acquisition approach to be used. Although
there are many special systems with special data acquisition problems, the majority
of systems can be considered within three general classes: (a) climatic data collection,
(b) component evaluation, and (c) total system evaluation. The spectrum of data col-
lection covered by the three classes might range from a single to a multitude of data
points and a time span ranging from instantaneous to several years. The types of data
acquisition systems that cover the above spectrum will be discussed.
Climatic data collection systems usually are required to operate for a long period of
time. Frequently the objective is to collect sufficient data to provide a statistical base
for future solar system sizing design work. Under some conditions, it may be desired to
record only a_limited set of simultaneous readings as a data base for later calculations.
An example of such a climatic requirement might be solar insolation (I) as a function
of time. A rather simple set of instruments and a recording system can meet this ob-
jective. Figure 41.1 shows a simple technique for handling insolation data. The strip-
chart recorders can be calibrated with the pyranometer to give a reading in W/m2 or
Btuat2 h. The instantaneous insolation is available by simply reading the chart. The
Data Acquisition Systems 465
LOW-SENSITIVITY
PHOTOVOLTAIC RECORDER
(HIGH-OUTPUT) (100 mV)
PYRANOMETER
150 nW/100 Btu/ft2/h
SCALE ADJUSTMENT
THERMOPILE
(LOW-OUTPUT)
PYRANOMETER
3 mV/100 Btu/ft2/h
0 0 0
AMPLIFIER (IN RECORDER)
HIGH-SENSITIVITY
RECORDER
(1-10 my)
historical record is kept by the chart recorder. In order to compute the total solar flux
density for a given period, the area under the curve must be integrated. This is done
either by counting the squares on the chart proper or by using a planimeter. Both
techniques require proper setting of the chart recorder scale factors. This technique
can be utilized for a wide range of parameters, and multipoint recorders enable up to
24 data points to be recorded simultaneously.
Some of the advantages of the chart recording technique for a data acquisition
system are that (a) the chart recorder is cost effective and provides a good data point
per dollar ratio for up to approximately 24 data points; and (b) data are provided in
a format that allows for constant visual scanning of data inputs.
Some of the disadvantages of a chart recording system are as follows. (a) The
total amount of data acquired when the system runs for a long period of time be-
comes unwieldly to handle manually even if only two or three channels of data are
being collected. A recorder that has a speed of 1 in./h will generate 60 ft of chart per
month if run on an around-the-clock schedule. (b) Each channel must be calibrated for
the individual sensor. In most cases, an amplifier must be incorporated to raise the
analog voltage level to one that can be used by the recorder. An example would be a
thermopile-type pyranometer that has a typical output range of 0 to 11 mV. The range
of the strip-chart recorder might be 0 to 100 mV, in which case a X10 amplifier would
have to be used. Amplifiers must have excellent long-term stability, since any drift
introduces errors in the recorded signal.
The data collection volume problem discussed previously also leads to the require-
ment for long-term, machine-readable mass storage. Figure 41.2 illustrates a data logger
acquisition system being used for the collection of climatic data.
The basic data logger is designed to accept analog and discrete inputs. The data
loggers can monitor temperatures via thermocouple, insolation (pyranameter), and
466 Applications of Solar Energy
other data types directly. The data logger outputs normally consist of a display of all
inputs converted to engineering units, a hard copy of all inputs, and an optional mag-
netic tape output. The computations have to be done in an off-line mode using the tape
output and a centralized computer or the hard copy and manual calculations. All of the
data are recorded relative to real time and can be correlated with time. The complete
system as shown in Fig. 41.2 consists of the sensors, a front end or signal conditioner,
the data logger, and the external, off-line processing computer if tape storage is used.
The sensors may be of various types, depending on the signal conditioner that is
used. Normally thermistors, RTDs, or thermocouples are used for temperature, a pyra-
nometer for solar insolation, an anenometer for wind speed, and a wind vane for wind
direction.
The signal conditioner will consist of amplifiers when the signal level is lower than
the active input of the data logger, a resistor network if the signal level is larger, and a
voltage- or current-regulated source if resistive elements are used for sensors (i.e., ther-
mistors or RTDs).
Most data loggers are equipped with a display that monitors all input channels and
shows their value in engineering units. This allows for visual monitoring of the instan-
taneous parameters.
For normal climatic data collection, printers would not be used because of the
volume of paper created by the long duration of the collection cycle. Magnetic tape
or a digital tape cassette would be used to store the input data correlated with time.
Filled cassettes would be taken to an off-line computer and read. This process is referred
to as batch processing. The data, besides being read, can be compiled, plotted, and used
for computations depending on the program in the main computer.
ANENOMETER
FREQUENCY-
TO VOLTAGE
CONVERTER
1 MULTIPLEXER
(CHANNEL
SCANNER)
ANALOG-
.-.-• TO-DIGITAL
CONVERTER
11 DRIVER
0
TAPE CASSETTE
BRIDGE
CIRCUIT
I t t
AMBIENT
TEMPERATURE 4-- REAL-TIME CLOCK
AND CONTROL
r
(SIGNAL
I SENSORS CONDITIONER DATA LOGGER
REMOTE PROCESSING
TAPE READER ... TIME DELAY PRINTER
0 0 MAIN-FRAME
COMPUTER
Figure 41.2 Climatic data collection system doing long-term mass-storage tape system
and remote computer processing (batch processing).
Data Acquisition Systems 467
1000
100
DAYS ON LINE
10
Figure 41.3 is calculated for a typical data logger showing the time a cassette tape
will last as a function of the number of channels being used and the frequency (scan
period) with which these channels are sampled.
Some of the advantages of a data logger DAS are the following. (a) The capacity
of handling many data points is available. Most data loggers can be equipped to accept
100 or more sensor inputs (b) Data acquisition is automatic. Once the data logger is
scaled for the appropriate sensor, it can be left unattended; changing cassettes is the only
labor involved at the site. (c) The LED display on the data logger allows for monitoring
all sensor inputs without affecting the normal collection cycle. (d) The data dump rate
or scan rate is normally variable and can be altered for various applications.
Some of the disadvantages of this type of data acquisition are as follows: (a) One
needs an off-line or central computer or some other device to read the tape storage.
(b) A sensor might malfunction and not be discovered until the tape is read, which might
result in a month of erroneous data collected. This could be overcome by monitoring all
input sensors on the display at regular intervals. (c) All calculations and formatting are
not available "on line." There is a time delay between the acquisition and the processing
cycle. (d) The dollars-per-data point is high for a small number of data points.
B. Component Evaluation
Typically, component evaluation calls for relatively short-term evaluation of a collector,
heat exchanger, storage system, controls, or similar such devices. If the work is to be
done on a long-term basis (weeks or months of data gathering), then a similar approach
468 Applications of Solar Energy
TIN CHANNEL #2
11.
ET AC ROSS
T COLLECTOR CHANNEL #3 ►
ENERGY COLLECTED
FREQ-TO- (Btu's) CHANNEL #4 ►
VOLTAGE MULTIPL ER
CONVERTER
ISOLATION (H) CHANNEL N5
AMBIENT
TEMPERATURE CHANNEL N7 ►
F igure 41.4 Analog data collection system for collector evaluation using seven-track
strip-chart recorder.
to that used for the climatic DAS should be considered. However, as is the case with
much design engineering work, where operation is to be observed over a period of hours
rather than weeks or months, results should be displayed in real time for observation by
the person conducting the tests. Three viable options are available at reasonable equip-
ment cost. The first to be considered is a manual system comprised of local indicating
sensors. The temperatures, flows, insolation, and other parameters can be read and
recorded from sensors equipped with readout dials. The analysis and computations
necessary are carried out manually.
The second type of DAS is the strip-recorder technique, as shown in Fig. 41.4.
The collection system shown monitors seven channels. The multiplier shown in channel
4 and the AT amplifier in channel 3 are not standard devices, but would have to be
designed for the specific system. In the system shown, the area under the curve plotted
on chart recorder channel 4 is the solution to the equation
Q = K _rot rit(T2 - dt
where
dt = time increment
Data Acquisition Systems 469
Dividing the area under this curve by the area under the curve plotted on channel
5, both taken over equal periods of time, will give the collector efficiency = Q/H for
the time period taken.* Instantaneous efficiencies can be calculated by simply reading
the amplitudes of the respective curves (channels 4 and 5) at the same points in time
and dividing. Where long-term average efficiencies are desired, the areas under the res-
pective curves can be obtained by use of a planimeter, or counting squares on the chart
paper. Both techniques require proper setting of chart-recorder scale factors. If the tests
are being done to NBS standards, consideration to changes in fluid density and viscosity
must also be given if appropriate.
The third type of system that can be considered is a data logger similar to the
type discussed in the climatic data collection section. The data collection would be a
real-time collection versus the batch processing as was discussed with regard to the
climatic technique. The time delay of processing the data would be eliminated. The
use of a hard-copy digital printer is impractical for a long-term climatic data collection
because of the volume of paper that is generated—even with a relatively infrequent print
cycle. Where only several hours of data are to be collected, printing can be set for a
shorter period, say, each 10 min. This provides reasonable resolution and a manageable
quantity of data printout.
Comparing the three systems, the indicating sensor approach is the least expen-
sive, but provides only instantaneous readout with manual records being taken. The
chart recorder provides instantaneous values and a historic record, but needs to be
extrapolated manually from the charts. The data logger provides a display of instan-
taneous values and a hard copy of the values printed at a preset scan rate.
*Q is the total energy collected by the system and is the output. H is the insolation and represents
the total energy input to the system (small amounts of energy used by pumps are neglected here).
Efficiency, n, is defined as energy out divided by energy in: ri --= Q/H.
470 Applications of Solar Energy
ROM 1K RAM
BO Hz
sec MICROPROCESSOR
Figure 41.5 Solar energy hot water system with microprocessor data acquisition
system.
The last two systems to consider make use of either a microprocessor or a mini-
computer as a dedicated on-site computer, and are referred to as a programmable DAS
or PDAS. Price reductions in both types of equipment over the last several years make
such an approach practical for monitoring total solar energy and backup energy systems
in buildings. The microprocessor and minicomputer, in addition to storing all raw data
parameters (insolation, temperature, flow, etc.), are also capable of carrying out a wide
range of calculations on-site and in real time. Results can be stored on magnetic or paper
tape and/or displayed on demand with access from front panel control. Since both sys-
tems contain a stored program, the computer can linearize sensor responses by use of
curve-fitting techniques. This can add to the accuracy of the overall results. The processor
is also able to identify failures in the system through use of internal fault analysis pro-
grams available with most machines.
Figure 41.5 shows a block diagram of a typical microprocessor in relation to a
solar hot water system application. Systems such as this are able to accept analog and
digital inputs from sensors. Sensors may be sampled as frequently as 10 times per second.
Solutions to the stored equations can be accomplished about once a second, the exact
frequency being determined by the total number of channels being scanned. Results
may be printed at frequent intervals and/or averaged over a given time base, say, an hour,
and then displayed. Generally, it is also desirable to keep a running total over a 24-h
period. This allows calculations of the percentage of load carried by solar during that
period,
Q
% solar = SO (41.1)
Q solar + Qbackup
Data Acquisition Systems 471
where
1112 = F2 X e
where
F2 = flow rate
e = density
C = specific heat
Q = area f I dt (41.3)
where
Collector efficiency,
EFC = Q co (41.4)
Q
Energy out of storage,
. 3 (T3 - Ta ) dt
Qstg = C111 (41.5)
where
m3 = F3 X e
Backup energy,
()backup = K f dt (41.6)
where
mG = FG
Table 41.1 DAS Comparison
System Type Characteristics Advantages Disadvantages
Total energy,
where
= F1 X e
The programmable DAS can also output to a magnetic tape storage. The tape
storage can contain the sensor inputs corrected to engineering units and the solutions
to the programmed equations. The programmable DAS contains a stored program that
can be modified as the application changes.
The basic differences between the microprocessor and minicomputer-based pro-
grammable DAS are (a) cost—the cost of the minicomputer is generally a factor of two
greater than that of the microprocessor— and (b) versatility—the program for the data
collection and equation solution is stored in memory in both systems. The minicom-
puter lends itself to changes more readily in that the program can be altered via the
front panel or teletype whereas the microprocessor normally requires a change to the
memory element.
The programmable DAS also lends itself to all of the previous discussed appli-
cations, but its cost effectiveness for simple data collection is questionable.
D. Applications Summary
A summary of the advantages and disadvantages of the different types of DAS is given
in Table 41.1. Figure 41.6 is a graph showing the cost of implementing a system by one
100,000
MINICOMPUTER
COST -DOLL ARS
10,000
PDAS
ATA LOGGERS
1000
ANALOG PLOTTERS
100 1
10 100 1000
Figure 41.6 Channel capacity versus cost
CHANNELS
for the four classes of systems treated.
475
of the four types of data acquisition systems: analog plotters, data loggers, programmable
data acquisition systems, and minicomputer systems.
The cost of the monitoring system is related directly to the number of channels
and the level of automation required. Data loggers and programmable data acquisition
systems become economically feasible when a system of 25 or more channels is being
considered.
By utilizing Table 41.1 and Fig. 41.6, the user should be able to select the type
of system most suited for a particular application.
All of the automated systems previously discussed—chart recorders, data loggers, and
PDAS—have a common design. Figure 41.7 is a block diagram of the basic system. The
three basic components are (a) sensors, (b) signal conditioning and/or processing, and
(c) storage and/or readout. Each of these will be discussed in general.
A. Sensors
When selecting a DAS, the type and cost of the sensors to measure the parameters must
be considered. Four general types of sensors will be discussed with several options
available. The four general types of sensors are (a) temperature; (b) flow, air or liquid;
(c) backup energies, gas or electric; and (d) climate conditions, including wind, ambient
temperatures, and insolation.
Temperature
Thermometer-type indicating sensors are available for the manual DAS. They
offer a readout dial as an output. Accuracies of approximately 2 percent of dial reading
can be attained.
Thermistors, resistor temperature devices (RTD), and thermocouples can be utilized
for the other types of DAS. Thermistors and RTD-type sensors provide a known resis-
tance for a specified temperature, whereas thermocouple sensors provide a voltage output
for a specified temperature.
The RTD is considered the standard and offers the greatest accuracy of the three
types of devices, but also the highest cost. Table 41.2 lists the three devices with some
of the parameters to be considered.
SIGNAL OUTPUT
SENSORS CONDITIONING (READOUT
(PROCESSING) or
STORAGE)
INTERFACE
The thermistor offers the best accuracy with the greatest sensitivity for the price.
However, all data loggers and PDAS are not equipped with the signal conditioning for
thermistors and RTDs. In such cases, thermocouples are the only devices that can be
used.
Flow
There are two types of flow, liquid and air, that are of concern to the data acqui-
sition system designer.
The liquid flow for the manual DAS can be measured by flow meters with built-
in registers or rotometers with sight gauges. For the automated types of DAS, an output
is required from the sensing device. A wide variety of flowmeters are available that pro-
vide either a pulse or analog voltage output.
The types of flow meters available are too numerous to discuss in detail, but the
following are some of the more common.
Positive Displacement. Positive displacement meters use a cavity of a fixed volume
that meters the fluid in small increments. The nutating disk and rotary piston are two
examples of this type of meter. Generally the positive displacement meters are acceptable
for low to medium flow rates.
Turbine. The turbine meter uses a turbine wheel which turns as fluid passes. The
turbine generates a signal that is picked up by either a magnetic, photo, or mechanical
device. This meter is used for medium to high flow rates.
In general, for all types of flow meters, accuracies of better than 2 percent over the
specified flow range is common. Attention should be paid to the specified flow range
when selecting a meter. Generally, the minimum range is 5 to 10 percent of the maxi-
mum. If 20 gpm is maximum, then 1 to 2 gpm will be minimum. Some meters will
simply not provide an output below this minimum flow, whereas others will be less
accurate. The accuracy, size, and temperature range are the key parameters that affect
the cost of the flow meters. A positive displacement, plastic, 5/8-in. NPT flow meter
with a contact closure output (1 contact/0.5 gal) that will operate up to 60°C (140°F)
can be purchased for as low as $100. A turbine-type meter, with a 3-in, flange and an
operating temperature of 150°C (300°F) can cost as much as $1800.
Data Acquisition Systems 477
Backup Energies
The amount of gas or electricity consumed by backup and auxiliary systems can
be measured by watt-hour and gas meters for the manual DAS with readings being re-
corded at desired intervals. For the other types of DAS, instrumented watt-hour and gas
meters that provide a contact closure for a given amount of consumption can be used
with a DAS that can accept pulse inputs. For systems that can monitor only an analog
voltage, a Hall effect transducer sensor can be used to measure the electrical consumption
and a differential pressure device can be used for the gas consumption.
Climatic Conditions
The ambient temperature can be measured by a similar device as used for other
temperature measurements. A radiation shield should be used to assure an accurate
reading. Wind direction and velocity can be measured by wind vanes and anemometers
that provide outputs which can be read on a meter for the manual DAS or an analog
voltage or pulses for the other DAS types.
Insolation can be measured by a pyranometer whose analog voltage output can also
be read directly by a metering device or by the DAS and converted to the proper engin-
eering units.
Table 41.3 lists some of the climatic condition instruments available.
One important point the DAS designer must keep in mind is that the output signal from
the sensor must be compatible with the input specification of the DAS. If a 0-100 mV
chart recorder is used, the 0-300 mV output from a photovoltaic pyranometer cannot be
applied directly to the recorder. A resistor network must be used and the recorder cali-
brated for the modified input. The same condition holds for the data logger and pro-
grammable systems. Likewise, a passive temperature sensor, thermistor, or RTD cannot
be used unless there is a known, regulated voltage source that is compatible with both
the sensor and the DAS.
The processing of the input data by a data logger is for the most part accomplished
by an analog-to-digital (A/D) converter and the signals then recorded in digital form.
The resolution is determined by the A/D conversion. The analog signal may be digitized
into 8, 10, 12, or more bits. The conversion resolution may be calculated by using the
following formula:
Resolution = 1
2"
where n = A/D bits. As an example, a 10-bit A/D converter will have a resolution of 1
part in 1024 (210 = 1024) or approximately 0.1 percent. When designing the DAS, the
resolution is an important specification to determine the overall system accuracy.
41.4 CONCLUSION
The options available for a DAS are numerous. For proper design, the solar system to
be monitored must be analyzed and the parameters to be recorded should be selected.
The cost of the system might necessitate the elimination of some parameters if cost is
critical. The types of sensors would be determined by the accuracy needed for the
analysis, and a cost trade-off might be necessary.
Data Acquisition Systems 479
Besides the data acquisition systems discussed here, there are other options availa-
ble. There are solar control systems available that have a limited monitoring capability
by way of LED display. This allows the designer to combine the control and monitoring
functions into one unit. This approach, however, dedicates the monitoring equipment on
a long-term basis, since it is integral with the controller. It cannot be removed without
removing the control functions as well. Such approaches are intended primarily for large
commercial installations. Information on such systems is mentioned in the Bibliography.
They are relatively new.
BIBLIOGRAPHY
C. BYRON WINN
Colorado State University
Fort Collins, Colorado
42.1 INTRODUCTION
Computer simulation programs can be extremely useful tools in the design and analysis of
physical systems. A designer or analyst can very rapidly determine, at relatively low cost,
the effects of various parameter changes on the performance of a given system with the
aid of a valid computer simulation model. Several computer simulation models have
been developed for the design and analysis of solar heating and/or cooling systems with
the above-mentioned objective in mind. This chapter discusses five of the programs that
have been developed and used rather extensively in the design and analysis of solar
heating and cooling systems. Of these five programs, two are used primarily as analysis
tools and the remaining three are used primarily as design tools. These five programs are
the TRNSYS, SIMSHAC, FCHART, SOLCOST, and the Relative Areas programs.
481
482 Applications of Solar Energy
TRNSYS (acronym for "A Transient System Simulation Program") was developed
at the Solar Energy Laboratory at the University of Wisconsin—Madison. An early version
of the program (actually a predecessor of the TRNSYS program) was used for the design
of the first solar house built at Colorado State University. This analysis was conducted
during 1973. The TRNSYS program has been updated several times since the initial
version was released in March 1975. Model validation studies have been performed using
the TRNSYS program and data from the Colorado State University experimental houses,
and these studies have demonstrated that the TRNSYS program may be safely used in the
analysis of a solar heating and/or cooling system that is similar in design to those projects.
Some of these validation results will be presented in this chapter.
At about the same time that the TRNSYS program was being developed, a similar
program, SIMSHAC, was under development. SIMSHAC and TRNSYS both consist of a
collection of subroutines that model subsystem components. These subroutines are
under the control of an executive program in each case. The executive program calls the
subroutines for the analysis according to the system description supplied by the user.
Results from the model validation studies conducted with respect to SIMSHAC are also
included in this chapter. In addition, comparisons between TRNSYS and SIMSHAC are
presented.
The two above-mentioned programs are extremely useful as analysis tools, but are
not very useful as design tools because they are relatively expensive to operate, require a
major computational facility, and require some expertise in programming. Not all
designers of solar systems have ready access to a computational facility, nor the required
expertise in programming, nor can afford the expense incurred in operating these pro-
grams for the design of a solar system. If these three obstacles could be removed, the
simulation programs could prove to be useful design tools as well as analysis tools.
With this in mind, work was initiated in 1974 by the staff of the Solar Energy
Laboratory at the University of Wisconsin—Madison to correlate results obtained from
the TRNSYS program in the simulation of space heating systems using hydronic-type
collectors. The result of this study was the development of the FCHART program.
The FCHART program represents a correlation between the monthly solar fraction
provided by a solar system and two dimensionless parameters that represent, respec-
tively, the ratios between the collector losses and the load and the collector gains and
the load. The FCHART procedure for liquid-based solar heating systems was presented
at the International Solar Energy Society Conference that was held in Los Angeles,
California, in 1975 [1] . The FCHART procedure for solar air heating systems was
presented at the International Solar Energy Society Conference held in Winnipeg,
Manitoba, Canada, in August 1976 [2]. The above-mentioned papers dealt with the
application of the FCHART procedure to a basic configuration for a liquid-based solar
heating system and a basic configuration for an air-based solar heating system. The
results were obtained from application of the TRNSYS program to the analysis of these
specific configurations and were not deemed to be valid for other solar system configura-
tions, such as solar cooling systems, solar-assisted heat pump systems, or low-temperature
applications such as swimming pool heating. Therefore, work on the FCHART procedure
has continued with the idea of developing correction factors that may be applied so that
the results from the program may be used for other solar system types. Results from this
study were reported at the International Solar Energy Society Conference held in
Solar Simulation Computer Programs 483
Orlando, Florida, in 1977 [3]. This chapter includes a brief description of the FCHART
program. For more details the reader is referred to Ref. 4.
Early in 1976 it was recognized that the situation with respect to computer simula-
tion programs was that, on the one hand there were available detailed computer simula-
tion programs such as TRNSYS and SIMSHAC which could be used for modeling virtual-
ly any type of solar system, and on the other hand there was available the FCHART
program which, at that time, was applicable only to two specific configurations of solar
systems. This left a considerable gap between the two extremes, as illustrated in
Fig. 42.1. Therefore, the Energy Research and Development Administration (ERDA)
contracted with the Martin-Marietta Company for the development of the SOLCOST
program. This program was written for the purpose of design of various configurations
of solar systems. Specifically, it includes the design of both solar heating and cooling
systems using flat-plate air or liquid collectors, focusing collectors, evacuated tubular
collectors, and tracking collectors. It also includes the design of solar-assisted heat pump
systems in its library of system types. It lies between the FCHART program and the
TRNSYS and SIMSHAC programs in terms of operational costs and ease of utilization.
This chapter gives a brief description of the SOLCOST program and includes comparisons
between the SOLCOST and FCHART programs.
The final program to be discussed in this chapter is the Relative Areas program [5].
Although the FCHART program eliminated the need for a large-scale computing facility
and significantly reduced the analysis costs and the requirements for an experienced pro-
grammer to operate the earlier TRNSYS program, it still suffers from the requirement
that either a medium-size computational facility be available or else detailed and tedious
computations be conducted by the system designer. Therefore, work was initiated in
1976 at Colorado State University to develop a simplification to the FCHART program
that would enable a solar system designer to complete a system design in just a few
minutes with the aid of nothing more than pencil and paper. This would entirely alleviate
the need for any computational facilities. The Relative Areas method was developed as a
FCHART TRNSYS
SIMSHAC
correlation to results obtained from the FCHART program, and therefore represents a
correlation to a correlation. A discussion of the Relative Areas method is included in
this chapter, and comparisons are presented between results obtained from the Relative
Areas method and the FCHART program.
There are several other computer simulation programs that are available but that
will not be discussed in any detail in this chapter. These programs include the Booz-
Allen program, the Cal-ERDA program, and the Stolar program. The Booz-Allen program
utilizes the FCHART program for the solar system performance analysis and has the
added capability of generating many different types of economic data. The Cal-ERDA
program essentially uses the TRNSYS program and has a much improved executor
program so that simulation time is significantly reduced. It also has a much more detailed
loads calculation program. Finally, the Stolar program is based on stochastic procedures
rather than deterministic procedures and has not been used nearly as extensively as the
other programs mentioned in this chapter. For additional information and details on
these programs the reader is referred to Refs. 6, 7, and 8.
42.2 TRNSYS
This section gives a description of the TRNSYS program and results from model valida-
tion studies conducted with respect to the TRNSYS program.
A. Program Description
Valve No. I
Hea t Ex c hanger
Auxiliary
Storage Boiler
Tank
S2
Cr cZ cr
Pump No. I Pump No.2 Pump No .3
Pump No. 4
Service
Cold Pump No. 5
Hot
Water
Water
Supply
Pre-heat
Tank Hot
Water
®S4 Heater
Figure 42.2 Schematic representation of a typical liquid system. Adapted from Solar
Heating and Cooling of Residential Buildings, Design of Systems, U.S. Dept. of Com-
merce, 1977.
via the pump subroutine. The inputs to the collector subroutine consist of outputs from
the solar radiation process or the time-dependent forcing functions subroutines for wind
speed, and so on, and the pump subroutine. An information flow diagram for the
collector model could be as shown in Fig. 42.3 [10].
PARAMETERS
mode, A, FR , U L ,
Cp,
To m
S = f1 (time)
Ti = f2 (time)
Ta = f3 (time)
Following construction of the information flow diagram, the analyst should associ-
ate type numbers with the system components. A type number for a component merely
relates that component to the subroutine that was developed to model that component.
If the user wishes to provide a new subroutine, then that subroutine must be coded in a
manner that conforms with the requirements imposed by the TRNSYS program. That is,
new subroutines may be added to the TRNSYS program provided that they are coded
in the same manner as the existing subroutines.
A type 1 system component, for example, is a flat-plate solar collector. The
Hottel, Whillier, and Bliss model was used to simulate the type 1 flat-plate solar collector.
For a detailed description of this type of system, and the other systems modeled in the
TRNSYS program, the reader is referred to Ref. 10.
The next step in setting up a TRNSYS simulation is to assign unit numbers to the
system components. The unit number references the particular component. It is com-
mon to have two or more system components having the same type number, whereas
each component must have a unique unit number.
The next step in setting up the TRNSYS simulation is to identify and sequence the
component inputs, outputs, and parameters. The information flow diagram for each
component model typically contains three types of information: inputs, outputs, and
parameters. For example, for the information flow diagram for the collector models
shown in Fig. 42.3, the inputs are represented by Ti, S, and Ta , whereas the outputs are
represented by To and m. The parameters are the collector area (A), the heat removal
factor (FR), the collector loss coefficient (UL), and so on. These must be numbered ac-
cording to the conventions established in the subroutines for the subsystem components.
The next, and final, step in setting up the TRNSYS simulation is to identify and
initialize all time-dependent variables. The TRNSYS program makes a distinction
between integration required to solve first-order ordinary differential equations and
integration used to sum a quantity over time. If a subsystem model involves differential
equations, then the TRNSYS program must be provided with the number of differential
equations and the initial values for each of the state variables.
A complete system flow diagram for a system similar to that shown in Fig. 42.2
is quite complex and usually takes on the order of an hour to develop.
B. Validation Results
Validation studies of solar house design programs have been conducted since 1976, and
some of the results have been reported in Refs. 11 and 12. These procedures have been
followed in examining the performance of the TRNSYS program in simulating the
Colorado State University solar houses. Results obtained from the checking periods that
were used for the heating season operation are shown in Table 42.1. The first row of
Table 42.1 TRNSYS Validation Results Checking Periods
Month
Winter
November December January February Average
Component Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev.
Storage tank (°C) 13.5 6.1 0.6 1.4 0.4 1.9 1.6 3.1 4.1 3.2
Enclosure (°C) -1.4 0.7 0.3 0.9 0.6 0.6 -0.2 0.7 -0.2 0.7
Collector inlet (°C) 0.7 4.0 1.1 0.7 1.1 1.2 -0.2 2.6 0.7 2.2
Collector outlet (°C) 0.1 4.5 0.2 1.2 0.6 0.7 0.7 1.7 0.4 2.1
Collector flow rate
(kg/h) 57.3 62.7 16.4 49.1 27.3 35.5 -32.7 90.0 16.4 60.0
Energy collected
(kJ/d) 84,496 78,258 30,398 27,257 30,940 13,375 -27,723 8,198 29,527 31,772
Heating load (kJ/d) -3,985 32,223 81,258 23,836 150,846 12,389 -69,147 46,755 39,743 28,800
Aux. energy
12,500 16,666 24,719 32,958 a a 12,043 17,054
supplied (kJ/d) -1,089 1,539
figures shows the mean and standard deviation for the residuals of the storage tank
temperature. The residuals were formed by taking the difference between the simulation
results and the actual measured results for the Colorado State University Solar House I.
The checking periods were periods that were selected from the data tapes to provide
periods during which the average process characteristics cluster closely together. That is,
in validating a model, one must select training and checking sequences from the data for
model calibration and model validation purposes. The training sequences are selected to
include data points that are far apart in order to take into account all process character-
istics of the model, whereas the checking sequences are selected to represent average
process characteristics. From the results presented in Table 42.1 it is quite apparent that
the TRNSYS program provides a valid tool for the analysis of solar systems that are
similar to the Colorado State University Solar House I. The schematic diagram for such
a solar system was shown earlier in Fig. 42.2.
42.3 SIMSHAC
The SIMSHAC program will be discussed in this section. The program will be described
and results obtained from the validation studies will be given. Finally, SIMSHAC and
TRNSYS will be compared.
A. Program Description
The SIMSHAC program is very similar to the TRNSYS program in that it is modular and
operates under the control of an executor program. The principal differences between
the SIMSHAC program and the TRNSYS program are with respect to the integration
package and the controls subroutines. The SIMSHAC program models all subsystem
components by systems of first-order ordinary differential equations. In those instances
where the natural model is an algebraic relationship this is still represented by a rate
equation with the rate term equal to zero. The second major difference lies in the
handling of the control subroutines. The SIMSHAC program was developed primarily
as a research tool for control systems studies. Therefore, it was considered during the
development of the SIMSHAC program that the user would ordinarily write a control
system subroutine for the analysis of a particular solar system. The program does con-
tain, however, a default control subroutine that is modeled after the controller in the
Colorado State University Solar House I. In the event the user desires to investigate
alternative controllers such as controllers that might result from various optimization
procedures, the user must write a specific control subroutine. This is similar to the
case of the TRNSYS program, where the user may wish to model a new subsystem
component.
The objectives in the development of the SIMSHAC program were as follows [13] :
1. To develop a general model that can be used for any specific system configura-
tion for solar heating and/or cooling of buildings
2. To modularize the program so that subsystems can be introduced or replaced
as more sophisticated versions become available.
Table 42.2 SIMSHAC Validation Results: Checking Periods
Month
Winter
November December January February Average
Component Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev.
Storage tank (°C) 7.0 3.7 1.9 2.4 1.0 2.3 0.5 2.9 2.6 2.8
Enclosure (°C) -0.7 0.9 0.6 0.7 0.4 0.6 0.3 2.2 0.2 1.1
Collector inlet (°C) -0.4 2.6 -0.4 0.9 1.1 1.8 -0.9 1.2 -0.2 1.7
Collector outlet (°C) 1.1 2.3 -0.9 0.7 -0.3 0.9 0.1 0.9 0. 1.2
Collector flow rate
(kg/h) 95.5 226.4 -16.4 76.4 49.1 27.3 40.9 87.3 43.6 103.6
Energy collected
(kJ/d) 22,858 16,730 3,744 12,507 27,421 14,002 -14,625 43,218 9,849 21,615
Heating load (kJ/d) -3,940 22,436 16,910 17,858 17,576 11,388 7,513 26,796 9,517 19,620
Aux. energy
supplied (kJ/d) -1,089 1,539 a a -281 397 23,561 11,018 7,398 4,318
Solar insol. 45°
(kJ/h • m2 ) 394 1,067 601 1,133 273 1,079 368 1,105 409 1,096
a No auxiliary energy used.
Table 42.3 Comparative Data for TRNSYS and SIMSHAC (Compiled Under FTN Compiler)
Compile Run Run Number Time Lack of
System Core, Compile Core, Time of days Printout Step Closure
Program Type 1000 Words Time (s) 1000 Words (s/d) in Run Time (min) (%)
SIMSHAC Water 56 55.3 53.5 4.67 7 15 min 15 0
TRNSYS Water 55 36.3 73.5 9.12 7 1h 15 5
SIMSHAC Air 56 58.2 53.5 5.11 7 15 min 15 0.3
TRNSYS Air 55 37.0 73.5 8.71 7 1h 15 3
SIMSHAC Water 56 55.9 53.5 2.78 62 1h 15 0
SIMSHAC Air 56 58.4 53.5 3.24 62 1h 15 0.4
TRNSYS a Water 54.6 12.7 71.5 5.16 62 4h 15 7
TRNSYS a Air 54.6 13.3 71.5 5.17 62 4h 15 4
TRNSYS a Water 54.6 12.9 71.5 6.99 7 1h 15 6
TRNSYS a Air 54.6 13.1 71.5 7.46 7 1h 15 4
In order to use the SIMSHAC program, the user merely specifies the components
included in the system and the manner in which they are connected. This is very similar
to the maner in which the TRNSYS program is designed.
B. Validation Results
The data that have been collected from Colorado State University's Solar House I were
also used to conduct validation studies with respect to the SIMSHAC program. These
results are shown in Table 42.2. It is apparent from the results shown in this table and
the previous table of the TRNSYS program results that there is very little difference
between the two programs in their application to a solar system as represented by the
Colorado State University Solar House I. The slight improvement exhibited by the
SIMSHAC program over the TRNSYS program is attributed to the difference between
the modeling of the control subsystems in the two programs. One could feel confident
in applying either one of these programs to the analysis of a solar system of the type that
was used for the validation studies.
42.4 FCHART
As was previously mentioned, the FCHART program was developed for the purpose of
making available for solar heating systems a design tool that would not require the use
of a major computational facility, expensive computer charges, or require expertise in
492 Applications of Solar Energy
programming. The FCHART program was based on correlations to the TRNSYS analyses
conducted for air heating and hydronic heating systems in the Madison, Wisconsin, area.
It was later tested for other climatological locations and the results were quite satisfac-
tory. The standard deviations were typically less than 2 percent.
The correlation equations that were developed are given below for water and air
systems:
where
At
X = FRUL A(TREF TA)
L
, SA
Y = F R Toi —
and where
A = collector area
TA = ambient temperature
TREF = 100°C
L = total space heating and hot water load per month
The equation for the solar fraction for a water system may be used for service hot water
systems, where X is given as
-66.2 + 1.18T, + 3.86Tm - 2.32TA
X = AFRUL At
LSHW
where
A. Program Description
The systems that were simulated by the TRNSYS program in order to develop these
correlation equations are shown in Figs. 42.4 and 42.5 for the liquid and air systems,
respectively. The service hot water systems are shown in Figs. 42.6 and 42.7 for the
liquid and air systems, respectively.
The Hottel, Whillier, Bliss collector model was used in this analysis. The equation
for the useful energy collected is
A = collector area
ra = transmissivity absorptivity
TA = ambient temperature
Service
Hot Water..
Relief Tank
Valve
Yo. Auxiliary
Solar
House
—0—.---I
Water
Supply Auxiliary
Figure 42.4 Schematic diagram of a solar liquid heating system. Adapted from S. E.
Huck, Design Charts for Solar Heating Systems, M.S. thesis, Department of Mechanical
Engineering, Colorado State University, 1976.
494 Applications of Solar Energy
To Taps
0
O
WPT
Heated Space ' DWH
Warm
Return Air ;
Air Mains
Supply
Pebble Bed '
Figure 42.5 Schematic diagram of a solar air heating system. DWH = domestic water
heater, WPT = water preheat tank. Adapted from S.E. Huck, Design Charts for Solar
Heating Systems, M.S. thesis, Department of Mechanical Engineering, Colorado State
University, 1976.
Relief
Valve
To Hot
Water
Load
Figure 42.6 Water system schematic for
service hot water. Adapted from S. E. Huck,
Design Charts for Solar Heating Systems,
M.S. thesis, Department of Mechanical Engi-
neering, Colorado State University, 1976.
To ;a ps
Domes- Water
Vo tic
Pre-
Water heat
Heater
Tank 0
Figure 42.7 Air system schematic for service
hot water. Adapted from S. E. Huck, Design
Supply
Main
Charts for Solar Heating Systems, M.S. thesis,
Department of Mechanical Engineering,
Colorado State University, 1976.
Solar Simulation Computer Programs 495
The automated version of the program uses the FR rot and the FR
' UL products. The
correction for FR may be obtained from the equation
FR
= [1 + F R U L A ( Cc
FR Cc e Cmin
where e is the heat exchanger effectiveness and C min is the minimum fluid capacitance
rate. This term is considered to be a heat exchanger factor, and its value lies between 0
and 1 and represents a penalty paid for the use of a heat exchanger between the collector
and the storage tank. For air systems this would typically be equal to 1.
The transmissivity-absorptivity product is entered in the program for normal
incidence. That is, the program assumes that the ra determination was made at normal
incidence on the collector and then the program corrects for other than normal incidence
throughout the day.
The storage model is represented by a first-order differential equation,
Cs dT,
dt QD (QL - EL) (Qw - Ew)
where
Ts = storage temperature
This is a model for a well-mixed storage tank and does not consider stratification. If
stratification measures are to be applied the results obtained from this model would be
somewhat in error.
The load model assumed in the program is of a UA AT load; that is, the heating
load is calculated by the product of the UA factor for the building and the number of
°C hours per month. The UA factor for the building is an input to the program and
the number of degree hours per month is calculated from the data in the climatological
data included for each city contained in the program. The UA factor for a building may
be readily determined by taking the ratio between the building heat loss at winter design
conditions and the difference between the desired room temperature and the winter
design temperature.
496 Applications of Solar Energy
The service hot water load was modeled according to the product of the amount
used per day times the temperature difference between the desired water temperature
and the temperature from the mains and the number of days per month. The program
does not allow the user to vary the temperature from the mains as a function of the time
of year.
The FCHART correlation curves were developed using nominal values for storage
size, collector fluid flow rates, and load heat exchanger. The nominal values that were
used for these parameters are listed below.
Storage size:
Air system: 0.75 ft3 of pebbles per square foot of collector
Water system: 2 gal of water per square foot of collector
Collector flow rate:
Air system: 2 SCFM per square foot of collector
Water system: 0.02 gal/min per square foot of collector
Load heat exchanger:
cLk"'plmin = 2
UA
In this expression el, represents the effectiveness of the load heat exchanger,
(mcp)min represents the minimum of the fluid capacitance rate of the fluids in
the heat exchanger, and UA represents the UA factor of the building.
The user may apply the FCHART correlation equations to systems that have
parameters that differ from the above-mentioned values by applying the correction
factors shown in Figs. 42.8, 42.9, and 42.10.
16
15
1.4
13
1.2
x lx°
1.0
0.9
0.8
0.7
0.6
2 3 4 5 6 7 8 9 10
SCFM/SQ. FT.
COLLECTOR
Figure 42.8 Collector air flow rate correction factor. Adapted from Solar Space and
Water Heating for Residential Buildings, a training course for the U.S. Department of
Housing and Urban Development, prepared by Colorado State University, 1977.
Solar Simulation Computer Programs 497
16
15
1.4
13
0.8
0.7 - (AIR SYSTEMS)
CUBIC FT. PEBBLES/SQ. FT COLLECTOR
0.6
0.375 0.75 1.125 1.5 1.875 2.25 2.625 3.0 3.375
1 1 1 1 1 I
0.5
2 3 4 5 6 7 8 9 10
GALLONS OF WATER /SQ. FT COLLECTOR (LIQUID SYSTEMS)
Figure 42.9 Storage size correction factor. Adapted from Solar Space and Water
Heating for Residential Buildings, a training course for the U.S. Department of Housing
and Urban Development, prepared by Colorado State University, 1977.
1.10
1.0 5
1.00
0.95
0 90
0.85
0 80
0.75 8 10
2 3 4 5 6 7 9
E (ri1C )
L p MIN
UA
Figure 42.10 Load heat exchanger correction factor. Adapted from Solar Space and
Water Heating for Residential Buildings, a training course for the U.S. Department of
Housing and Urban Development, prepared by Colorado State University, 1977.
The reader should bear in mind that the FCHART program represents a correlation
to results that were obtained from the use of the TRNSYS program to simulate systems
using Madison, Wisconsin, climatological data. If one were to use the TRNSYS program
to simulate a particular system in any one of the given 172 locations included in the
FCHART program and then were to repeat the study using the FCHART program, the
results would not agree exactly. The standard deviations in the errors will be small,
498 Applications of Solar Energy
however. The maximum error that the author has experienced in conducting that
numerical experiment has been on the order of 10 percent.
42.5 SOLCOST
A. Program Description
A simplified flowchart of the SOLCOST program is shown in Fig. 42.11. Many options
are available to the SOLCOST user. The solar system itself can be modeled in many
ways, including using either liquid or air as the fluid in the collectors. Economic data
may also be supplied by the user, including data such as the solar system costs—both
fixed costs of installation and component cost per square foot of collector. Conventional
fuel types and costs can also be input into the simulation.
Several types of solar systems can be evaluated with the SOLCOST program, includ-
ing (a) space and service hot water heating systems, (b) absorption cycle air conditioning
systems, and (c) solar-assisted heat pump systems. All of these systems can be modeled
with any combination of the above options, leading to a large array of possible system
types to be modeled. A brief outline of some of the possible options to the user is given
in Fig. 42.12.
For each model that is simulated, then, the SOLCOST program determines the
optimum size for the collector, the portion of the load supplied by the optimum collector
area, and the payback period for the solar system investment compared to a conventional
system. Typical curves showing these relationships are shown in Fig. 42.13.
Solar Simulation Computer Programs 499
Fraction f of
Input
Load Supplied
Routine
by Solar
END
Figure 42.11 SOLCOST flowchart. Adapted from SOLCOST User's Guide for
Cybernet, U.S. Dept. of Energy, Solar Technology Transfer Program, 1977.
Major outputs from the program consist of (a) prediction of annual fraction of
heating or cooling load supplied by the solar system, (b) the optimum collector area as
well as tilt angle for the installation and location, (c) a detailed cash flow summary, and
(d) heating and/or cooling loads computed by the program.
The computer output consists of three major sections. These sections include first
a concise listing of the input and default parameters, which is in effect a description of
the system. An example of this section is shown in Fig. 42.14.
Figure 42.15 is the second section of the output. This part consists of the thermal
information, or performance of the system. The optimized collector area is specified
together with a monthly summary of its performance. Other system components are also
sized here, such as the storage tank capacity. The current version of the SOLCOST
program also sizes the heat exchangers, pumps or blowers, and pipes or ducts.
The last section, shown in Fig. 42.16, is the detailed economical analysis of the
system. The net cash flow is basically the fuel savings realized by the solar system minus
all the costs of the system. Of course, fuel escalation and inflation rates, as specified by
the user, affect these values. Notice that the program output includes the net cash flow
(undiscounted) as well as payback time for fuel savings to equal total investment,
payback time for net cash flow to offset the down payment, net present worth of solar
savings, and rate of return on net cash flow. These different outputs were included
SOLCOST Analysis options 0
0
Given the appropriate input data, SOLCOST can make the following anayses:
Standard solar analysis A) Physical description of building • Cost optimum collector area
B) Cost and performance details for the • Tilt angle
proposed solar system • Storage size
• Heating and cooling loads
• Payback
• Rate of return
• Detailed cash flow for cost optimum area
Collector Trade Study Analysis User specifies up to three collector types • Program outputs cost optimum area for
each type (payback or internal rate of
return determines best choice)
Financial analysis Complete solar system definition • SOLCOST computes system performance
• Homeowners including loads and collector area and generates a detailed life cycle cost
• Business/Investor owners analysis for the input collector size
• Non-profit institutions • Calculations include tax and depreciation
deduction in the cases of business and
non-profit organizations
Passive structure analysis Passive structure description • Program estimates annual fuel require-
ments based on a transient analysis which
uses "average" conditions for the 15th
day of each month
/Mau] Rios Jo suo!veo!iddv
AA-
_I
c\
<Lif1-
Z
3+
LL.
Z ti)
.SZ° .
z 0
2 5 a 1%saimaso/
t., _
D< Payback
WA F.
ui
z
COLLECTOR AREA COLLECTOR AREA YEARS
Figure 42.13 SOLCOST outputs. Adapted from SOLCOST Users' Guide for Cybernet,
U.S. Dept. of Energy, Solar Technology Transfer Program, 1977.
because of the fact that some analysts prefer to look at net cash flow whereas other
analysts prefer to look at rate of return, or payback periods, or present worth of solar
savings.
In these three pages of output are included all of the results and information
necessary to the user in his decision on whether or not to implement such a system.
The output, as well as the input, is easy to understand, and is quickly and inexpensively
obtained.
Figure 42.15
COLLECTOR TYPE = USER-DEFINED FLAT PLATE LIQUID
BEST SOLAR COLLECTOR SIZE FOR TILT ANGLE OF 55 DEGREES IS 385 SQ. FT.
SOLAR SYSTEMCOSTS (DOLLARS) -- 1722 FIXED, 7700 COLLECTOR, 577 STORAGE
FINANCIAL SCENARIO -- RESIDENCE
-1000
1 700 100 200 810 303 986 -282
2 770 106 212 794 302 986 -231
3 847 112 225 777 300 986 -175
4 931 119 238 758 299 986 -112
5 1025 126 252 738 297 986 -42
6 1127 134 268 715 295 986 34
7 1240 142 284 691 292 986 121
8 1364 150 301 664 290 986 216
9 1500 159 319 635 286 986 322
A2J au3 Rios jo su oRP3H d de
PAYBACK TIME FOR FUEL SAVINGS TO EQUAL TOTAL INVESTMENT 9.3 YEARS
Solar SimulationComputer Programs
PAYBACK TIME FOR NET CASH FLOW TO OFFSET DOWN PAYMENT 11.2 YEARS
NET PRESENT WORTH OF SOLAR SAVINGS 1826.77 DOLLARS
RATE OF RETURN ON NET CASH FLOW 16.0 PERCENT
ANNUAL PORTION OF LOAD PROVIDED BY SOLAR 69.9 PERCENT
ANNUAL ENERGY SAVINGS WITH SOLAR SYSTEM 70.0 MILLION BTUS
So!cost
100
fchart
0 fchart with
So!cost data
80
_1 60
0 QHwc0.534 MMBtu /day
40
20
COLLECTOR AREA , ft 2
shown to be smaller when both programs base their results on the same weather data.
Also, the differences are greater in the high end of the collector area array, that is, for
collector areas that provide greater than 70 percent solar. Within the range of interest,
say, between 40 and 70 percent of the load suppled by solar, the discrepancies between
the two programs are even less significant.
One disadvantage of both the FCHART and SOLCOST programs is that they still require
that the analyst have access to a computational facility capable of having approximately
43000a words of memory. The curve-fit correlation equations used in the FCHART
program can, of course, be calculated by paper and pencil or hand-held calculator, and
several programs have been developed for programmable hand-held calculators. However,
for the user who does not wish to purchase a programmable hand-held calculator and
who desires to have a quick and reasonably accurate means of designing solar systems,
the Relative Areas method may be applied. The Relative Areas method is based on a cor-
relation to results obtained from the FCHART program and is reported in detail in Ref. 5.
By use of the method one may calculate directly, without iteration, the optimal collector
area based on a life cycle cost analysis, the corresponding solar fraction, and the corre-
sponding life cycle cost and life cycle savings of the solar system with respect to a conven-
tional system. This section contains a brief summary of the Relative Areas method, an
example problem, and a comparison of results obtained from the Relative Areas and
FCHART methods.
Solar Simulation Computer Programs 507
A. Description
The Relative Areas approach was suggested by an examination of curves such as those
shown on Fig. 42.18. This figure shows the performance of a given solar space heating
system as a function of collector area for six different collectors as determined from the
FCHART program. It is apparent that these curves all have approximately the same
shape and that the quantitative differences exhibited between different systems may be
attributed to differences in scaling of the abscissa. The same results have been plotted
in Fig. 42.19, where, in this case, the abscissa has been defined as the relative area, where
relative area is defined as the ratio between the collector area and the collector area
required to give a 0.5 load fraction. That is, all of the curves in Fig. 42.19 have been
forced to go through the point Ar = 1, F = 0.5. In this example, it is apparent that the
spread between the curves has been significantly reduced as a result of plotting the solar
fraction as a function of relative collector area. The spread remains quite small even out
to a value of 3 for the relative collector area. This suggests the possibility of being able
to use one curve, or equivalently, an analytical representation for the curve, as a general-
ized performance curve. Unfortunately, this does not turn out to be a realistic goal,
because when one changes to different locations and repeats the analysis there is a rather
significant spread in the results, as shown in Fig. 42.20. However, one can still develop
the analytical representation for the performance curve where the coefficients in the
analytical representation would be dependent on the location. This has been done, and
the results collected in a table showing the location-dependent design data for each one
of the locations included in the FCHART program. In order to apply the method, one
20 40 60 80
A, m2
Figure 42.18 Performance as a function of collector area for six liquid systems in
Boulder, Colorado. Adapted from C. D. Barley and C. B. Winn, The Relative Areas
Method of Optimal Collector Sizing of Solar Systems, Colorado State University, 1977.
508 Applications of Solar Energy
Figure 42.19 Performance as a function of relative collector area for six liquid systems
in Boulder, Colorado. Adapted from C. D. Barley and C. B. Winn, The Relative Areas
Method of Optimal Collector Sizing of Solar Systems, Colorado State University, 1977.
I0
05
. • • I • . . 1 ..
I 2 3 4
A r = A/A 0
Figure 42.20 Performance as a function of relative collector area for liquid systems in
16 different cities. Adapted from C. D. Barley and C. B. Winn, The Relative Areas
Method of Optimal Collector Sizing of Solar Systems, Colorado State University, 1977.
Solar Simulation Computer Programs 509
has merely to read the location dependent parameters from the table and use them in the
equations for optimal collector area and for the solar fraction. These equations are
described in the following paragraphs.
The total heating expense over the lifetime of the system may be expressed by the
equation
where
E, = an economic factor for converting future cash flows for all but operating,
maintenance, and fuel costs to present worth
E4 = an economic factor for converting future fuel cash flows to present worth
The optimal collector area may be determined by setting the first derivative of the
CT equation taken with respect to collector area to zero and solving for the resulting
collector area. This requires that an analytical representation for the solar fraction as a
function of area be developed. The analytical representation that was selected is given as
F= C1 + C2 ln ( A )
A0.5
On using this equation in the partial derivative of life cycle costs with respect to collector
area, one may derive the following expression for the optimal collector area:
C2 LCfE4
AopLf — CaE1 + Co E2
This equation may be applied to determine quickly the optimal collector area for
given economic considerations.
510 Applications of Solar Energy
The economic factors may be calculated from the following equations. If one
considers an exponentially increasing cash flow with an annual rate of increase of r and
a market discount rate d for a period of n years, then the present worth of the sequence
is given as [5]
n X(1 +
P=
i=1 (1 + d)i
where X represents the first year cash flow. This summation has the analytical represen-
tation
- d r
P/A(d, r, n) — (1 +
(1 + (d — r)
and
n
P/A (d, r, n) — d=r
(1 + r)
The P/A notation was adopted to conform with the notation commonly used for the
present worth of an annuity. The E2 factor is then given by
E2 = P/A (d, ro , n)
E3 = P/A (d, rm , n)
E4 = P/A (d, rf , n)
where ro represents the annual escalation rate of operating expenses, rm represents the
annual escalation rate of maintenance expenses, and rf represents the annual escalation
rate of fuel costs. If the expenses are income tax deductible, then each of these factors
should be multiplied by 1 — T where T represents the effective federal-state income tax
rate.
The cash flow related to the first costs consists of a down payment, periodic
mortgage payments, periodic property tax payments, periodic insurance payments, and
possibly an investment tax credit, income tax credits for interest payments, depreciation,
and property tax payments, and possibly a salvage value. These factors may all be
included in the analysis and the E1 factor may be calculated from the following equation
[5] (see p. 513 for definition of terms):
Solar Simulation Computer Programs 511
1 -F g )n
E1 = a - p- 0 (1 + d + [(1 - r) p + FI]P/A (d, g, n)
The depreciation factor may be calculated from the following three equations:
r(1 - a)
BSL - P/A (d, o, k)
k
T5
BDB = — P/A (d,- 5/k, k)
k
As (UA)
AO.5 - ,
FR ra- FR ULZ
As (UA)
A0 .5 -
riE
512 Applications of Solar Energy
where i7E is defined as follows. The Hottel, Whillier, Bliss equation for instantaneous
collector efficiency may be written as
'CI - TA
n = FR Ta - FRUL
HT
T1 - TA
HT
' UL Z
riE = FIR Ta - Fit
Z - FR Ta - As (UA/Ao.$)
FR UL
From the above equation, A, may be calculated for any particular system. These calcula-
tions have been performed and the results indicate that Z is practically collector
independent. The standard deviation in A0.5 , as determined from the FCHART program
calculated with an average value of Z for any particular location, was on the order of
1 percent. Z is a location-dependent parameter only. By determining As and Z from a
table, and by specifying the UA factor for the building and the collector parameters, the
analyst may quickly determine the collector area required for a 0.5 load fraction.
Tables of location-dependent curve-fit parameters include data for both air and
liquid heaters. Domestic hot water loads may be included in the total load when using
these curve-fit parameters. For stand-alone domestic hot water systems a separate set
of location-dependent curve fit parameters is required. For the case of hot water systems
the collector area is proportional to the hot water demand instead of the UA factor for
the building. In this case, AD was defined as A0.5 for a domestic hot water system having
Ta = 1, FR = 1, UL = 0, and a load of 10 13 • °C/d. The resulting equation for A0.5 was
where
and
B. Example
Consider the case of a liquid-based solar space and DHW heating system for a home in
Boulder, Colorado, under the following conditions [14] :
2 .5 27.8
-+ (1 - 0.1) F l - 0.2) + (0.2) 512 - 0 = 1.55
13
*These unrealistic assumptions are made to allow a direct comparison with results from the FCHART
program.
514 Applications of Solar Energy
Then
Also
(9.62) (1.5)
A05 = - 30.8 m2 (331 ft2 )
(0.716) (0.97) - (5.47) (0.97) (0.04251)
Therefore
65.4
F = 0531 + 0.301 1n ( ) = 0.76.
30.8
For comparison, the deviation of Aut from FCHART value is about -14 percent. The
corresponding deviation of total cost from the minimum is only about 0.5 percent.
A comparison was made between the results obtained from the FCHART program and
the results obtained from the Relative Areas method. The results of this comparison were
that the Relative Areas method gave an optimal collector area that was within 10 percent
of the optimal area as determined by the FCHART method in 73 percent of the cases
that were examined. For a detailed discussion of these results the reader should refer to
Ref. 4.
REFERENCES
EDITORS' NOTE
Several simplified methods have been developed that do not require the use of a computer
to predict the performance of a residential or commercial solar system and determine the
appropriate size of the collector array. The "solar load ratio method," developed by the
517
518 Applications of Solar Energy
solar group at the Los Alamos Scientific Laboratory, is described in Chap. 44 of this
handbook. Perhaps the most widely accepted and used method is the "F chart" proce-
dure developed at the University of Wisconsin. An excellent description of this method,
complete with worksheets to fill out, is contained in U.S. Department of Housing and
Urban Development, Intermediate Minimum Property Standards Supplement, 1977 ed.,
Solar Heating and Domestic Hot Water Systems, 4930.2. This has been adapted and
reproduced here. Table A-4 of the original, giving values of H and Kt for 80 U.S.
locations from the Liu and Jordan presentation [10] , has been eliminated. One should
instead refer to the more recent and improved set of insolation data in Table 5.1 of
Chap. 5 of this handbook.
43.1 INTRODUCTION
The performance of any solar energy system is directly related to the amount of solar
radiation available, the outdoor conditions, the heat load thermal characteristics, and the
solar energy system characteristics. The energy flow diagram for one type of a simplified
active solar energy system (using a flat-plate collector) is shown in Fig. 43.1. Specific
systems for providing heat and/or hot water using air and liquid collectors are illustrated
in Fig. 433.
Starting on the left side of Fig. 43.1, solar energy is incident on the solar collector.
Some energy is lost from the collector due to reflection from the collector and heat
transfer to the ambient surroundings. The amount of energy actually collected by a solar
collector is termed useful energy and is directly proportional to the collector area.
Collector losses
Heating Load
Insolation
Auxiliary
Energy
Figure 43.1 Energy flow diagram for a typical active solar energy system.
Procedures for Determining Thermal Performance 519
Useful energy is transferred directly to the heated space, hot water load, or thermal
storage device (TSD) by a working fluid. The TSD is necessary due to the intermittent
nature of solar energy and is characterized by the TSD heat capacity.
Energy from the TSD to the heated space or hot water load is transferred as needed
to meet the load. If the TSD cannot meet the load, auxiliary energy from other sources
is necessary. Heat exchangers and pumping systems must be properly sized to meet
energy flow requirements.
Precise prediction of performance of solar heating systems is difficult because of
the random variation in climatic conditions and the mathematically complex relationships
between the system components.
The procedure for estimating performance of a solar energy system used in this
document is called the "F chart" method and was developed by Klein, Beckman, and
Duffie of the University of Wisconsin [1, 17, 18]. Their approach was to use a detailed
hour-by-hour computer simulation for several typical solar energy systems covering a
wide range of system parameters at several geographic locations in order to develop a
generalized correlation ("F chart") which is useful in evaluating long-term system per-
formance.
Of the methods presently in use, this method is widely accepted and offers flexibil-
ity in the range of design parameters that can be evaluated.
Although differences were found between the simulated and estimated performance
using the generalized chart for specific monthly periods, the correlation, determined by a
least-squares fit, was found to be quite satisfactory in most cases for predicting year-long
performance. It should be emphasized, however, that the F chart procedure is not
intended to provide an accurate estimate of system performance for any particular
month, but rather for the long term. The differences between the simulated and esti-
mated yearly performance of systems in different locations were also found to be small.
The standard deviation in the fraction of load to be met by solar predicted for four
different cities was 0.018, an error judged to be substantially lower than the errors
inherent in the simulation model and in the recorded data.
The F chart procedure has been checked using the data obtained from existing solar
heating systems and the difference between estimated and actual performance is small.
For example, for M.I.T. Solar House IV [2], the difference between actual performance
and estimated performance using the F chart procedure over two complete heating
seasons was about 8 percent.
The procedure can accommodate various nontracking flat-plate collector designs
when the thermal performance is available in terms of the effective collection and conver-
sion of solar radiation into useful thermal energy for a range of operating conditions.
Simulations used to derive the F chart correlation include typical ranges for heat transfer
fluid flow rates, heat exchangers, thermal storage capacity and configuration, auxiliary
energy integration, and control modes. Methods of estimating the performance of systems
having characteristics other than those used to generate the F chart are included in some
instances. The procedure does not apply to passive systems and is generally limited to
geographic regions below 60° latitude. Preliminary design considerations for passive
systems have been studied [3] , but no procedure is currently available. The method does
not apply to systems using heat pumps.
520 Applications of Solar Energy
The procedure is intended for use in the evaluation of the capability of a solar
energy system to provide a portion of the heating and hot water requirements for a
residential dwelling. The method is not intended to perform the thermal design of a
unique system or to optimize the solar energy system performance for a unique thermal
load profile. The calculation can be performed using a slide rule, hand calculator, or
simplified computerized program. Another method is available to size collectors for a
specific type system using hand calculations [4] , and a somewhat more detailed method
has been described using a computer [5] . Each of these methods can be used to estimate
the cost and fuel savings for local environments and conditions.
A. Description
The procedure for calculating the performance of solar heating and/or hot water systems
allows the estimation of long-term solar heating system performance applicable for space
heating and domestic hot water systems, either individually or combined. The method is
applicable to those air and liquid heat transfer systems in which solar energy and auxiliary
energy are added directly to the heating and/or hot water loads. Solar energy systems
using heat pumps are not covered. Examples of each application are presented in
Sec. 43.4. The system evaluation procedure is not intended to provide an accurate
prediction of system performance for any particular month, but rather for the long-term
average.
The procedure is based on a computer-simulated systems analysis. Solar energy
system performance is characterized by the term F, which denotes the monthly fraction
of the total heating load supplied by the solar energy system and is determined from a
single parametric chart, the F chart, Figs. 43.4 and 43.5. The systems analysis has shown
the performance of the system to be well correlated to two systems performance parame-
ters D1 and D2 , the coordinates used on the F chart. D1 is a function of solar insolation
and system heat load and D2 is a function of solar collector heat losses and system heat
load. Several other design parameters were found to have only a small effect on system
performance, and thus the method can be used for a wide range of these design parame-
ters by means of correction factors, K1 , K2 K3 , and K4 •
The calculation method is applicable to systems schematically identical to Fig. 43.3
or similar simplified models. The basis of the calculation procedure is the F chart and the
two equations that define D I and D2 . The equations, given below, have been labeled
with the section numbers that describe the indicated elements. These sections are further
listed in the order in which they appear in the text.
43.3.0
Procedures for Determining Thermal Performance 521
43.3.B
43.3.0
43.3.A Determination of the portion of the total heating load supplied by solar
energy (f1 and Fann u al)
43.3.0 Determination of the total building heating and/or domestic hot water
load (L)
Project Data Worksheet A is provided at the end of this section to record the basic
information necessary to carry out the calculation procedure.
B. Assumptions
1. Thermal storage is contained within the heated structure and all storage heat
losses are considered to supplement the space heating load.
2. Auxiliary heat sources are provided to supply energy for both the space and
water heating when the energy in storage is depleted, or the rate at which solar
energy can be supplied is less than the total heating load.
3. For heating systems utilizing a liquid heat transfer fluid, a heat exchanger can
be used between the collector and the storage tank. When an antifreeze
solution is circulated through the collector to avoid the problems of freezing
and corrosion, the use of a heat exchanger in conjunction with water storage
may be more economical than using the antifreeze solution as the energy
storage medium.
522 Applications of Solar Energy
w
V
<W-
on
UJ
ti Cr
....1 UJ
•IC I-
r At
CL 31t
0
Z /a ...n
M 0 ._._/ III I I I 1 1 1 I I I I I I I I 1 I 1 I 1
0 I 2 3 4 5 6 7 8 9 10 II 12 1 2 3 4 5 6 7 8 9 10 11 12
MIDNIGHT A.M. NOON P.M. MIDNIGHT
4. The heat transfer fluid is circulated through the collector whenever a positive
energy gain can be achieved. During periods of low radiation (when the energy
gain becomes zero or negative), the collector pump or blower is turned off.
5. For the space heating load determination, the energy per degree day method, is
adequate. Load calculations are provided using the "Manual J" method [14],
but any procedure that reasonably predicts the building thermal load is
acceptable.
6. The average domestic hot water demand as a function of the time of day and
family size was established in the development of the procedure. In general,
hot water consumption is highly dependent on the habits of the occupants.
However, it has been determined that the actual time distribution of the water
heating load will have only a small effect on the long-term performance for
solar heating systems combining domestic hot water and space heating.
7. For those solar energy systems used only for hot water heating, the procedure
has been developed using an assumed hourly hot water consumption pattern
that is repeated every day, Fig. 43.2. It is not known how deviations from this
consumption pattern effect the performance of the hot water system.
8. Since in most instances flat-plate collectors are utilized for heating buildings,
the collector component parameters are valid only for modeling flat-plate
collectors. Concentrating collectors or tracking collectors cannot be incorpo-
rated into the solar heating system evaluation as it is presently written in this
chapter.
Figure 43.3 (a) Liquid system: space heating and domestic hot water. (b) Liquid
system: domestic hot water only. (c) Air system: space heating and domestic hot water.
Supply Loop
Pump Supply
E
l-4 Air
AuxiliarY
Heat
Collector
I
T
Hot Ii
Collector / Main Water .— Load
Storage Hot Water
Storag Preheat Auxiliary Heat
Heat Storage Exchanger
Heat
Exchanger t
Hot Water
Loop Pump
(incp).= Ca
(thcp)c=C. (Mcp).=-C8 (Mcp) =C
h h
Return
CO 4— Air
Collector Storage
Loop Pump Loop Pump Building
Supply
Fan
(a) Liquid System: Space Heating and Domestic Hot Water
Hot Water
Collector Auxiliary Heat
1
T/
Collector/ Hot
Storage Water
Heat Preheat
Exchanger Storage
m
(ificp)c=Cc
ancos=c,
Collector Hot Water
Loop Pump Loop Pump
r—IWater
Hot
Preheat
Storage
Main Storage
Hot Water
Loop Pump
.n
(Mcp)c=Cc Ii
Return
Air
3 2
523
524 Applications of Solar Energy
PROJECT
Location Latitude
Building Heating and/or Hot Water Load
Design Heat Loss Rate, qd = Btu/h
oF
Winter Design Temperature (97 1/2%), tom,
Average Hot Water Consumption = gal/day
(may vary on a monthly basis)
Average Cold Water Supply (main) Temp., tm = °F
(may vary on a monthly basis)
Hot Water Supply Temp., is = °F
Collector Subsystem Data
Collector Type
Selective or nonselective, no cover plates
Collector Area, A, ft2
0
Tilt Angle
Azimuth Angle
Collector Shading (ay. % month of Dec.)
Collector Efficiency Data
(from manufacturer): F R(ra),..
FRUL Btu/h• ft 2 • ° F
t in 4. tout
Reference Temp. Basis:mt, , tout
2
Fluid:
Composition•
Specific Heat, cp = Btu/lb•°F
Specific Gravity (if applicable) = lb/lb
Volumetric Flow Rate = gal/min or ft3 /min
Storage Subsystem Data
Volume = gal or ft3
Storage Medium
Specific Heat, cp = Btu/lb•°F
Specific Gravity/or Density = lb/lb or lb/ft3
Circulation Loop Volumetric Flow Rate = gal/min or ft3 /min
Collector/Storage Heat Exchange Effectiveness, ces
Hot Water Preheat Storage Volume = gal
Load Subsystem Data
Load Heat Exchanger Effectiveness, EL
Supply Loop Volumetric Flow Rate = gal/min
Building Air Supply Volumetric Flow Rate = ft3 /min
Procedures for Determining Thermal Performance 525
43.3 PROCEDURE
The fraction of the heating load supplied by solar energy is the measure of system per-
formance that is calculated. The solar contribution is calculated first on a monthly basis
and then summed to provide the yearly total.
The monthly fraction of the heating load supplied by solar energy is determined
from the F charts, Fig. 43.4 for air systems and Fig. 43.5 for liquid systems. Worksheet B
has been provided for tabulating monthly values off and calculating Fannual • On Fig.
43.4 or Fig. 43.5 locate the two system parameters D1 and D2 for each month and read
off the corresponding value of f. D1 and D2 are determined by the procedure described
in Sec. 43 .3 .B .
D
1
Q 4.0 6.0 8.0 10.0 12.0 14.0 16.0 8.
3.00.0 2.0
,..r I / 1 111 TIT l i I i i T r I. 1 i If i IIII, 0.00
=
=
=
=
2.60
- 2.50
=
f=0.9
2.00
=
2.00
f=0.8 =
f=0.7 =
1.60
f=0.6 -- E' 1.50
_
f=0.5 -=_
1.00 f -0.4 =_
_, 1.00
f=0.3 =
.=
0.50 f=0.2
f=0.I == 0 .50
I =
0.0$,0
'tit tit 11 titit ill tit tit tit tit =
2.0 4•0 6.0 8.0 10.0 12.0 14.0 16.0 18.00.00
D2
30 - 1 1 1 1 1 I 1 1 T111111 TITI111111111111111
-
f=0.9 -
-
25 fro e _
f .07
20
I _
_........................„--- f = 0i6 -,
D I _ 0-5
1
_ f = oigi
10 -
f=o 3
.......,.....„-- f = 0 2 -
05
7,----. f = oi
1 _
11I,1 111111111111111111111111 111111111-
5 10 15 20
D
2
estimate for long-term performance such as a year. f is calculated on a yearly basis in the
following steps:
The actual solar energy supplied for each month is calculated as follows:
Elan fJanLJan
EFeb fFebLFeb
(43.1)
EDec = fDecLDec
The total solar energy supplied for the year is calculated by summing the contribu-
tions from each month,
The total heating load for the year is calculated by summing the contributions for
each month.
_ Etotal
Fannual (43.4)
Lt otal
WORKSHEET B Fraction of Total Heating Load Supplied by Solar Energy, Fonnual
PROJECT
1 2 3 4 5
Actual
Tot. Mo. Solar Solar
Htg. Load System System Fraction/ en/mo
L Parameter Parameter mo. E
Month (Btu/mo.) D1 D2 f (Btu/mo.)
Jan.
Feb.
March
April
May
June
July
Procedures for DeterminingThermal Performance
Aug.
Sept.
Oct.
Nov.
Dec.
E t ot
Lt„, E t ot = Fannual —
Lot
1. From Worksheet D.
2. From Worksheet C.
3. From Worksheet C.
4. From F chart.
5. E = f x L.
528 Applications of Solar Energy
The system performance parameters D, and D2 characterize the entire solar energy
system thermal performance. Worksheet C is included for tabulating the system per-
formance parameters. The two parameters are calculated for each month using the
following equations:
energy absorbed by collector plate
D1 -
total heating load
(r) S
_ AYR
x K4
(43.5)
L
ref. collector plate energy losses
D2 - total heating load
AerR UL (tref - to)Atime
X Ki X K2 X K3 (43.6)
L
where:
Ao = collector aperture area (ft2 ). This must be consistent with the collector
manufacturers performance data and in most instances equals Aa used in
the collector thermal performance test, ASHRAE 93-77.
S = total incidenct solar radiation normal to the surface of the collector for
an average month (Btu/month • ft2 ). (A detailed calculation procedure
and data are provided in Sec. 43.3.D, sample calculations are in Sec. 43.4,
and worksheets are in Sec. 53.5.)
L = total heating and hot water load for the particular month (Btu/month).
In hot water only systems this term equals Qµ,. (A detailed calculation
procedure is provided in Sec. 43.3.C, sample calculations are in Sec. 43.4,
and worksheets are in Sec. 43.5.)
c = collector combined performance characteristics obtained from the experi-
Fit(F:7)
and mentally determined efficiency plot for the collector combined with heat
F'RUL exchanger performance data. The Fit (FE) product is dimensionless and
the FR ' UL product has units (Btu/h • °F • ft2 ). (A detailed calculation
procedure and data are provided in Sec. 43.3.E, sample calculations are
in Sec. 43.4, and worksheets are in Sec. 43.5.)
trot- 212°F, reference temperature (arbitrarily chosen).
to = monthly average day-time temperature (°F). (Table 43.4 in Sec. 43.5
provides tables of these temperatures for many locations.)
Atime = total number of hours for the particular month.
K, = air collector flow capacitance rate factor.
K2 = storage mass capacitance factor, for liquid and air heating and hot water
systems (dimensionless).
K3 = hot water factor, for liquid hot water only systems (dimensionless).
K4 = load heat exchanger factor, for liquid heating systems (dimensionless).
(A detailed calculation procedure for K, , K2 , K3 , and K4 is provided in
Sec. 43.3.F, sample calculations are in Sec. 43.4, and worksheets are in
Sec. 43 5 .)
WORKSHEET C System Performance Parameters D I , D2 PROJECT
1 2 3 4 5 6
Tot. Monthly Total Mo. Av.
Radiation Heating Day Time Tot. Hrs
on Tilt Surf. Load Temp. in Mo.
S L to 212- to A time
Month (Btu/mo• ft2 ) (Btu/mo.) S/L D1 (°F) (° F) (h) K3 D2
Jan. 744 Ac = Given data
Feb. 672 FR (Ta) = Worksheet F
March 744 F'R U L = Worksheet F
April 720 K1 = Worksheet G
May 744 K2 = Worksheet G
;tine 720 K4 = Worksheet G
July 744
Aug. 744
P roced u resfor Determi ni n gTh ermal Performan ce
Sept. 720
Oct. 744
Nov. 720
Dec. 744
1. From Worksheet E
2. From Worksheet D
[A Fiji:07)S
3. D, L X K4 = (D1 prod.) =
Where D, prod. = [A,Fk(ra)] x K, =
4. From Sec. 43.5, Table 43.4
5. From Table 43.14 and Worksheet G
A F' U(t t )Atime to) time
6. 132 - [ c R L ref o X K1 X K, X K3 (D2 prod.) x K3
L
Where D, prod. = (AcFkUL) X K, X K. =
530 Applications of Solar Energy
The total heating load is determined on a monthly basis for space and domestic hot water
heating. The space heating and domestic hot water heating loads are calculated separate-
ly, and for combined systems the monthly individual loads are added to get a monthly
total load. Worksheet D is included for tabulating the heating load calculation.
It is recommended that the space heating load for each month be calculated using
the degree-day method. It is based on the assumption that over the long term, solar and
internal heat gains will offset the residential heat loss when the mean daily outdoor air
temperature is 65°F and that the long-term heating load will be proportional to the
difference between the mean daily temperature and 65°F. Tables of degree-days with a
base of 65°F have been constructed and are published for a large number of cities in
Chapter 43 of the 1976 ASHRAE Systems Handbook [9]. However, in many current
buildings larger internal heat gains coupled with the increased insulation levels are suffi-
cient to offset a home's heat loss to a mean daily temperature as low as 56°F (reported
in Ref. 9). Therefore, a modified degree-day procedure is suggested. It consists of the
use of proportionality factor (PF). This factor can range in value from 0.60 to 0.80
depending on the insulation level, weather patterns, internal gains, and so on. However,
it is recommended that a value of 0.75 be used unless practical experience in the particu-
lar locality would dictate the use of a different value.
The ASHRAE Handbook of Fundamentals [8] describes the basic method for
calculating heat losses in a chapter entitled "Heating Load." In addition, NESCA Manual
J [14] IBR Guide H-21 [15] , and ARI Standard 230 present slightly different methods,
including examples, tables, and preprinted calculation forms which simplify the calculat-
ing process. For multifamily installations NESCA Manual M may be used instead of
Manual J. Although the various methods differ somewhat in format and details, the
principles and the overall approach are essentially the same.
The following outline of the main steps in calculating the Building Design Heat Loss
Rate, qd, are adopted from ASHRAE's Fundamentals [8] and a typical calculation form.
Figure 43.6 is shown from NESCA Manual J. Note that the Manual J form also contains
spaces for use only in cooling load calculation, notably the columns labeled Clg and lines
7 and 16-21.
Select the design outdoor weather conditions. The data on winter climatic condi-
tions are given in the ASHRAE Handbook of Fundamentals (select from the
97 1/2 column) or Table 45.3 in Sec. 43.5.
Select 70 F for family and 75 F for care-type and elderly housing as the indoor air
temperature to be maintained in living spaces during winter.
Estimate temperatures in adjacent unheated spaces. Consult ASHRAE's Funda-
mentals for formulas, examples, and tables.
Select or compute the heat transmission coefficients for outside walls and glass,
and for inside walls, nonbasement floors, and ceilings, if these are next to
unheated spaces; include roof if next to heated space. On the Manual J form,
WORKSHEET D Heating and/or Domestic Hot Water Load, L PROJECT
1 2 3 4 5 6
Monthly Monthly Temp. DHW Monthly Total Btu/h
qd =
Degree Space No. of Vol. of Water Temp. DHW Heating (Given data or calculate
Days, Htg Load, Days/ DHW Main Sup., Rise, Load, Load, from Manual J or equiva-
DD Qs Mo., Used/Mo., tm ts - tm Qw lent.)
Month (°F • d) (Btu/mo.) N (gal/mo.) (°F) (°F) (Btu/mo.) (Btu/mo.)
Atd = 70 - tw
Jan. 31 = 70 -
Feb. 28 where tw = 97 1/2%
March 31 winter design temperature
(From ASHRAE Funda-
April 30
mentals, Table 43.3,
May 31
Sec. 43.5 or known
June 30 weather data.)
July 31 70° = indoor design
Aug. 31 temperature
Proced ures for Determi ni n gTh ermal Perform ance
Sept. 30 qd =
UA =
Oct. 31 Atd
Nov. 30 ts =
Dec. 31
1. From ASHRAE Systems, Climatic Atlas, or Table 43.3, Sec. 43.5.
2. Qs = (PF) (24) (UA) (degree day) = X (degree day) =
where PF = 0.75 or more appropriate value.
3. (Vol/day) (no. days/mo.) - (gal/d) (no. days/mo.)
4. May be constant or may vary.
5. Qw = (vol. of water) X 8.33 X 1 X (ts - tm).
6. L = Qs + Qw.
1 Nana al Ram Latin Hera 1 3 3 4 3
I Raul./ Ft Rapand Ha
Dean( Htir a
/4- 4/
7 Wardaara North -1
II
and Glam [AM or NC • HMI
'4 I I
0....1C11) smiths SR • SW
-4
I Maw Doan
11 Not
Eaprasil ►
Wall. and
Panama's
311 C.d)
11 71sen •
1•Vanliaala.
a 4 aa
IS Sala T.441 Saab Lam
71
16 Dan Ink Lam
1••• ,1 aramaa
IS TataJ Ina 14ms
a.- / 1 .:7
-'1
IS People 0 30/ sad Applaasa MA
a I anima.. 14 4,
17 Somalia Stab Gan 1116nartunr)
11111•••• 4 Ranamr•
ISICaan 110.3 Gala
1S• 0•4‘ &mango Fr.-
II Sam al Laan 17•40 II lam
-• • 0••••••• r
111 Total Stub Gan Maw HI a 111 1.3
II Ina la At Qtaatalea .., -a
Figure 43.6 Building design heat loss calculations form. Source: From Ref. 14.
anui Jelos Jo suopro!i ddv
Procedures for Determining Thermal Performance 533
a HTM (heat transfer multiplier) is used which equals the heat transmission
coefficient multiplied by the temperature difference.
Determine net area of outside wall, glass, and roof next to heated spaces, as well as
any cold walls, floors, or ceilings next to unheated space. Such measurements
are made from building plans, using inside dimensions.
Compute the heat transmission losses for each kind of wall, glass, floor, ceiling, and
roof in the building by multiplying the heat transmission coefficient in each case
by the area of the surface in square feet and the temperature difference between
the indoor and the unheated space or outdoor air. Note: In the Manual J form
heat transmission coefficient X temperature difference is replaced by the HTM.
Compute heat losses from floors over crawl space or basement floors or grade-level
slab floors.
Estimate infiltration rate (number of air changes per hour) and compute the heat
equivalent of the infiltration of cold air taking place around outside doors and
windows. The infiltration rate depends on the type of or width of cracks, wind
speed, and the temperature difference between the indoor and outdoor air; the
result expresses the heat required to warm the cold air infiltrating into the
building. Refer to ASHRAE's Fundamentals.
The sum of the transmission losses or heat transmitted through the confining walls,
floor, ceiling, glass, and other surfaces, plus the heat equivalent of the cold air
entering by infiltration represents the total heating load. Since these are for the
design heat loss conditions and are expressed as an hourly rate, this value repre-
sents qd , the building design heat loss rate.
Obtain the monthly total Degree Days from the ASHRAE Systems Handbook [9],
Climate Atlas of the United States [13] , or Table 43.3 in Sec. 43.5 for the
particular location for each month.
Calculate the monthly space heating load using the equation
City Source'Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
2. Miami W 70 70 70 70 70 70 70 70 70 70 70 70
4. Albuquerque w 72 72 72 72 72 72 72 72 72 72 72 72
5. Las Vegas W 73 73 73 73 73 73 73 73 73 73 73 73
6. Denver Ri 39 40 43 49 55 60 63 64 63 56 45 37
7. Ft. Worth L 56 49 57 70 75 81 79 83 81 72 56 46
8. Nashville Ri 46 46 53 66 63 69 71 75 75 71 58 53
9. Washington, DC Ri 42 42 52 56 63 67 67 78 79 68 55 46
11. Seattle Ri 39 37 43 45 48 57 60 68 66 57 48 43
12. Boston Re 32 36 39 52 58 71 74 67 60 56 48 45
13. Chicago 1. 32 32 34 42 51 57 65 67 62 57 45 35
14. New York City Re 36 35 36 39 47 54 58 60 61 57 48 45
Figure 43.7 Monthly temperature (t1) in °F at source for city water in 14 selected cities.
Determine the domestic hot water supply temperature (ts) or assume that is =
140°F.
Calculate the monthly domestic hot water heating load using the following
equation:
(gal)
(DHW consumed) (8.33 lb
w - x gal ) X (specific heat of water)
month
The monthly total heating and hot water load is the sum of the space heating and
domestic hot water load for each month. Where only domestic hot water or space heat-
ing is provided, the appropriate single load is used.
The solar radiation incident on the collector is determined for a particular collector tilt
and orientation, on a monthly basis, by modifying known insolation values for a given
Procedures for Determining Thermal Performance 535
geographic location with factors that relate this to the specific collector geometry. The
relationship between the various solar radiation variables is presented in Figure 43.8.
Terminology and methods developed by Liu and Jordan [10] are used. Worksheet E is
included for tabulating the incident solar radiation calculation.
For cases where local detailed horizontal insolation is available, these data should
be used to calculate IH . Where detailed horizontal radiation is not available, an estimate
of TH can be made using the values listed in Chapter 5, Table 5.1 of this handbook, or the
mean daily insolation maps shown in Sec. 435, Fig. 43.28 [11].
(43.11)
The factor fk accounts for the three components of radiation incident on the col-
lector surface, beam, sky diffuse, and ground reflected. In accounting for ground-
reflected radiation, a diffuse reflectance of 02 has been used for ground surface condi-
tions. If it is desired to use other values, a computer procedure such as Ref. 19 should
be used.
In
To
/7. _ Tilted
Horizontal
IT
Extraterrestrial Terrestrial
TH
Jan. 31
Feb. 28
March 31
April 30
May 31
June 30
July 31
Aug. 31
Sept. 30
Oct. 31
Nov. 30
Dec. 31
1. From Fig. 43.28, Sec. 43.5, or known data.
2. From Table 43.4, Sec. 43.5, used only for Eq. (43.11).
3. From Eq. (43.11).
4. From Table 43.5, Sec. 43.5, latitude (0) = °, with collector tilt (0) °, and latitude - tilt -
5. From Eq. (43.12), TT = (11).
6. From Eq. (43.13), S = (IT) (N).
A2Jau3 Rios jo suoproHdde
Procedures for Determining Thermal Performance 537
k is determined from Table 43.6 in Sec. 43.5 by entering with the known latitude
of the installation (0), tilt of the collector (0) (expressed in relation to the latitude), and
ratio of extraterrestrial to terrestrial radiation (IC- t) (calculated in the preceeding step).
For tilt angles other than those listed, values may be interpolated. These tables assume
that the collector is oriented due south (y = 1800 ).
For exact evaluation of the effects of collector azimuth angle (y) on incident
radiation, a computer procedure such as Ref. 19 should be used. However, for collector
tilt angles approximately equal to the latitude and for latitudes of 45° or less, Ref. 16
states that the total annual beam radiation will not vary by more than 2 percent for
collectors oriented up to 22 1/2 degrees east or west of due south.
Shading
(varying angle) radiation. This gives FR (Fo7). Both of the characteristics are then cor-
rected for the effects of the collector/storage heat exchanger by the heat exchanger
modifier factor, FR
' /FR , which includes corrections for both the heat exchanger effective-
ness and capacitance rates. This step gives FR' (7-67) and FRUL .
Worksheet F is provided for these calculations.
The performance characteristics for the collector alone are determined from manu-
facturers data (which may be presented in any of several forms) or if this is not available,
from typical characteristics curves such as those included.t
Collector Efficiency Data Available from Test. Determine the FR(ra)n and FR UL
factors from the thermal performance efficiency curves provided by the manufacturer
covering the appropriate range of operational temperature, insolation, tilt angle, and fluid
flow rates. The collector performance efficiency curve must be plotted such that the
y axis is the thermal efficiency OD and the x axis is the temperature difference between
a reference fluid temperature (t*) and the ambient temperature divided by the incident
solar radiation [(t* - ta)/It ] as shown in Fig. 43.9.
tASHRAE 93-77 defines the collector efficiency using the gross collector area whereas previous data
was generally reported using aperture area. Figure 43.27 presents data for several liquid collector
types with the efficiency based on gross area and aperture area. Calculations for this procedure can be
performed with either definition, but the efficiency and area must be consistent in all portions of the
calculation.
1.0
76 0.4 Slope = -m
E
a) 0.3
0.2
0.1
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1
t*-t h •ft2
a oF•
IT Btu
PROJECT:
Collector efficiency data from test:
Intercept, b = FR(Ta)n
Slope, m = FRUL
tin 1- tout
Reference temperature basis: 1. tin, 2. , 3. tout
2
Collector area, Ac
Collector volumetric flow rate
Correction to tin basis:
c
Case 2: FR(Ta)n = b 1
1 + mA /2Cc \
FR UL =m
1 + mAc /2C,
Cc = nicp= (volumetric) time \ /specific)
(density) (
flow rate conversion/ \ heat
F R UL = m
1 + mAc
1 /Cc
FR
— = 1
FR
Capacitance rate:
Cc = (from above)
Cs = (calc. as for Cc above)
Cm in = (lesser of Cc or Cs)
540 Applications of Solar Energy
WORKSHEET F (Continued)
Ac(FR UL )
Cc
1
= from Figure 43.10 or =
R 1+ y( x- 1)
ri F4
F R (T 6 ) = FR( 70 )n (Toon ( FR )
FR)
FRUL = FRUL
R
y = -mx+b (43.14)
where
- m = the slope of the straight line
b = the y axis intercept (value of y when x = 0)
Usually, the reference temperature t* in the plot of experimental data is the fluid inlet
temperature, in which cases the collector characteristics FR(ra)n and FR UL are obtained
directly from the efficiency graph. However, in some cases the data is based on other
reference temperatures. The following procedure describes the determination of FR(ra)n
and FR UL for three different values of reference temperature.
Case 1:
then
FR(Ta), = b (the y axis intercept of the data line)
Case 2:
t* _ t in + tout
(average of inlet and exit temperature)
2
Procedures for Determining Thermal Performance 541
then
1
FR (Ta)n = b ( (43.15)
1 + mAc/2C,)
and
1
FRUL = m 1 (43.16)
1 + mAc/2Cc
Case 3:
then
( 1 )
FR(Ta)n (43.17)
b 1 + mAc /Cc
FR UL = m( 1 + m
1Acicc (43.18)
It should be noted that the fluid mass flow rate of air collectors can significantly
influence the performance of the collector. Therefore, the efficiency data used for air
collectors in this procedure must have been obtained for the specific range of flow rates
anticipated for collector operation.
No Measured Efficiency Data for Collector. When collector efficiency data are not
available for the specific collector, performance characteristics may be estimated from the
data provided in Figs. 43.26 and 43.27 in Sec. 43.5 for liquid and air systems, respective-
ly. Note that in Fig. 43.26 for liquid collectors, the primary parameters affecting per-
formance are the number of cover glass sheets and the radiative properties of the absorber
coating (whether selective or nonselective). Figure 43.27 shows that air collectors with
flat black absorbers have performance characteristics that are sensitive to air flow rate,
number of cover glass sheets, and the specific configuration of the absorber.
Alternatively, for flat-plate collectors constructed with relatively simple flow path
geometry, the analytical equations described in Ref. 16 can be used to estimate perform-
ance. For this purpose, the optical properties of several glazing materials and absorber
coatings are provided in Tables 43.1 and 43.2 in Sec. 43.5.
Capacitance Rates. Calculate the collector loop and storage loop capacitance rates
and determine the minimum capacitance rate. The loop capacitance rate is the product
of the mass flow rate, m (lb/h) and fluid specific heat, cp (Btu/lb • °F).
Heat Exchange Modifier Factor, YR /FR . The heat exchanger modifier is deter-
mined by first calculating the following dimensionless parameters:
Cc
x - (43.22)
cc sCmin
Ac(FRUL)
(43.23)
C,
The heat exchanger modifier is then determined either from the following equation or
from Fig. 43.10.
Fit 1
(43.24)
FR 1 y(x- 1)
For air systems and liquid systems without a collector/storage heat exchanger, FR
' /FR = 1.
The values for the combined collector performance characteristics are calculated as
follows:
1.0
0.8
0.6
0.4
0.2
0 10
0.08
0.06
0.04
0.02
001
ccs Cmin
F it
Fict UL = FR UL (43.26)
FR
This section develops the correction factors to be applied to D1 and D2 in Eqs. (43.5)
and (43.6) to account for the effect of variation of several design parameters. The
validity of the procedure for values outside of the ranges indicated on accompanying
Figs. 43.11 through 43.14 is not known, therefore evaluation of systems outside this
range is not recommended. Worksheet G is provided for these calculations.
K1 modifies D2 in Eq. (43.6) and accounts for the effect of flow capacitance rate
on the performance of air collectors. For liquid collectors K1 = 1. Figure 43.11 presents
a plot of K1 as a function of the ratio of collector flow capacitance rate Cc to collector
area for the range of values noted. Cc is calculated using Eq. (43.20).
544 Applications of Solar Energy
PROJECT
Cc
K1 = (from Fig. 43.11) =
(Mcp ),
K2 = (from Fig. 43.12) =
A,
Hot water factor, K3
For liquid systems providing heating only or heating and
hot water, and for all air systems, K3 = 1.0
For DHW only systems: is = (from Worksheet A) °F
= (from Worksheet A or if variable see Worksheet D) °F
K3 is taken from Fig. 43.13 and tabulated on a
monthly basis on Worksheet C
K3 = (tabulated or const.) =
1.3 I t r AI
1.2
1.1
11-
1.0
0.9
0.8
0 • I I T
0 10 20 30 40 50
Cc
Btu/h °F ft 2
A
c
Figure 43.11 Collector capacitance rate factor (air), K1 .
storage mass capacitance (Mcp), is the product of the storage mass M(1b) times the
specific heat of the storage medium, cp (Btu/lb • °F).
K3 modifies D2 in Eq. (43.6) and permits the use of the correlations originally
developed for combined heating and hot water systems to be used for systems that pro-
vide domestic hot water only. For liquid systems that provide heating only or combined
heating and hot water and for all air systems K3 = 1. Figure 43.13 presents a plot of K3
as a function of average day-time temperature to . In this plot the cold water supply
temperature, tm , and the hot water delivery temperature, ts, are parameters that must
be supplied.
546 Applications of Solar Energy
1.4
1.2
1.0
Liquid Systems
0.8
Air Systems
0.64
0
0 10 20 30 40 50 60
(Nlys Btuill.°F.ft2
Ac
1C4 modifies D1 in Eq. (43.5) and accounts for the effect of the liquid to air load
heat exchanger sizing on the performance of liquid systems for heating or combined
heating and hot water. In air systems or domestic hot water only systems where there
is no load heat exchanger K4 = 1. Figure 43.14 presents a plot of K4 as a function of
eLCmin/AUbidg , where
Cmin = capacitance rate of the fluid with the minimum mass flow rate-specific
heat product (nicp) and is usually the air-side fluid capacitance rate
I I
ER
--
HW Supply Temp. •. 140°F
........
HW Supply Temp. • 120° F
1.6
-- -- -.. -.
... — 80°F
....... -.....
1.2
".... CW Suppl y
-.....s.
... ..
..
.. N. Temp 70°F
N. N. N.
111...
4
1 0.8 60°F ---
N. N.`
N.
Ns
1!
11
\10k. 50°F __,
0.4
N,\
\ - 40°F
0 \
0 20 40 60 80 100
When information on the load heat exchanger effectiveness is not provided, Sec. 43.3.G
describes several alternative methods for estimating a value.
In the evaluation of solar energy system performance using the F chart method, there
are two locations in the calculation procedure at which an explicit value for heat
exchanger effectiveness, efix , must be known. This section describes two approaches
to calculation of effectiveness, based on different ways design data may be provided
by a heat exchanger manufacturer. In situations when no manufacturer design data are
available, a method is described to estimate heat exchanger effectiveness.
Definitions are as follows:
Overall conductance (AU)HX for a heat exchanger is the product of the overall
heat transfer coefficient U (which depends on the thermal properties of each
fluid, the fluid mass flow rate, and the heat exchanger geometry) and the
associated heat transfer surface area.
548 Applications of Solar Energy
t
...............„,...............,....--.--6-
1.0
0.9
0.8
0.7
0.6
0.5 I t i
0.1 10 10.0
c l.Cmin
(UA)bldg
Effectiveness efix is the ratio of the actual rate of heat transfer in the exchanger
to the theoretical maximum rate of heat transfer that would occur only in a
counterflow exchanger with infinite surface area.
Log-mean temperature difference A tui is the effective temperature difference
between the inlet and outlet fluids such that the product At Lm and (AU)HX
equals the actual heat transfer rate.
tc , in tc, out
Figure 43.15 Schematic heat exchanger.
Cc and inlet temperature tc,in both enter a heat exchanger that has an overall conduc-
tance (AU)Hx . The actual heat transfer rate is given by
The maximum theoretical heat transfer rate would occur if the fluid having the
minimum capacitance rate were heated (or cooled) to the inlet temperature of the maxi-
mum capaictance rate fluid; that is,
If Cmin = CH,
Or if Cmin = Cc ,
The log mean temperature difference is a complex function of the hot and cold
fluid inlet and outlet temperatures and depends on the heat exchanger flow arrangement.
The term AtLm is defined in terms of a reference log mean temperature difference for a
counterflow heat exchanger (AtLm)c f. and a correction factor K as follows:
where
(th,in tc,out) - (th,out tc,in)
(43.34)
('6't1M)cf) loge Kth,in tc,out)/(th,out tc,inA
and K depends on the particular heat exchanger flow geometry and two dimensionless
temperature ratios R1 and R2 such that
_ tc,out tc,in
R1 (43.35)
th,in tc,in
th,in th,out
R2 - (43.36)
tc,out- tc,tn
•
Tabular or graphical values of (A tLm )cf and K are usually provided by heat exchanger
manufacturers.
Example 1
Calculate the effectiveness of a counterflow heat exchanger located in the collector/
storage circulation loop. Assume that the collector circulation loop flow rate is 2 gpm
and the storage loop flow rate is 6 gpm and that fluid leaves the collector at 120°F and
fluid leaves the storage tank at 100° F. The heat exchanger manufacturers performance
data for the specific heat exchanger and fluid properties are given in Fig. 43.16.
40
8 30
O ,
>-
I- 20
< 10
o
8
GPM(COLD )
Entering Fig. 43.16 with 2.0 gpm (cold) and 6.0 gpm (hot), with a 120° F - 100° =
20°F At, the actual heat transfer rate is determined to be 11,000 Btu/h. The maximum
rate of heat transfer is
Qmax = Cmin At
min Bt u
= 2.0 gal X 8.33 .-g---
lb X 60 X 1.0 X 20°F
mM al h lb • ° F
= 20,000 Btu/h
_ Qact 11 000
- ' = 0.55
€HX
Qmax 20,000
Example 2
Calculate the effectiveness of a two-pass shell-and-tube heat exchanger transferring
heat between water and a heat transfer fluid. Water enters the heat exchanger at 110°F
and 10 gpm and the heat transfer fluid enters the heat exchanger at 160°F and 15 gpm.
The properties of the heat transfer fluid are cp = 0.65 Btu/lb • °F and density = 55 lb/ft3.
The manufacturer's data consist of the following:
R1
Since neither of the fluid exit temperatures is known, it is necessary to use a trial-
and-error procedure.
The capacitance rate for the heat transfer fluid (hot) is given by
Btu
= 4295
h • °F
Q 99,960
th,out = th,in - -111 = 160 - 4,295 - 136.7°F
The reference log mean temperature difference for a counterflow heat exchanger is
calculated from Eq. (43.34):
Since the heat exchanger is not of the counterflow arrangement, a correction factor
K must be determined from the manufacturer's supplied data, in this example Fig. 43.17,
based on the computed values of R1 and R2 as follows:
From Eq. (43.35),
_ 130 - 110 _
0.4
'
1 160 - 110
160 - 136.7
R2 - 1.16
130 - 110
K is then determined from Fig. 43.17 (by extrapolation) as K = 0.90. Then from
Eqs. (43.29) and (43.33),
and
114 615
tc,out = 110 + '
4,998 - 132.9°F
Since the original exit temperature of the water was estimated to be 130°F, the
procedure is repeated until the estimated water temperature and the calculated water
temperature are the same. For this problem, the exit water temperature is then calcu-
lated to be
tc,out 131.2°F
The maximum possible heat transfer rate would occur if the minimum capacitance
rate fluid (the hot fluid in this example) were cooled to the cold fluid inlet temperature.
= 105,960 -
6HX 0.49
214,750
The heat exchanger effectiveness can be assumed to be a constant for a given heat
exchanger and fluid mass flow, provided that the thermal properties of the fluids do not
vary substantially and provided that no change of phase occurs. Thus, the calculated
value of effectiveness based on an assumed set of inlet temperatures would still be valid
over a range of hot fluid and cold fluid inlet conditions usually found in most solar
heating applications.
For the case where there are not performance data available for the particular heat
exchanger, an estimate of effectiveness can be made from the data of Ref. 7 for a range
of different heat exchanger designs, provided that the overall conductance (AU)HX is
known. Figure 43.18 shows the effectiveness of counterflow, parallel-flow, and cross-
flow heat exchangers as a function of the ratio (AU)Hx /Cmin with the ratio Cmin/Cm„
as the parameter. For the cross-flow arrangement usually found in the liquid-to-air load
heat exchangers, Cmin is usually the air-side capacitance rate and the parameter Cmixed/
Cunmixed shown in Fig. 43.18 for this arrangement is equivalent to Cmin /Cmax .
If the particular flow arrangement of a heat exchanger is unknown, the effective-
ness can be estimated by taking an average of the calculated effectiveness for the counter-
flow and parallel-flow arrangements for the given (AU)HX/Cmin and Cmin/Cmax design
parameters.
For situations in which a heat exchanger consists of a coil of tubing submerged in a
tank of water, Cmin is the capacitance rate of the fluid circulating through the coil and
554 Applications of Solar Energy
Z
COLD FLUID (Wc)c=Cc COLD FLUID (Wc)c
ZHEAT-TRANSFER HEAT-TRANSFER
SURFACE SURFACE
100 100
1
C /C =0 —C
min max miniCma 0 11111
025 0111/111/—
p
p
• 50 C 80
80 1 pip.- 0.2,5
l00 0 50
r
in60 w 60 ,--
0.7:5—
w I w I
ii iLoo
w
1
A
> 40 I- 40
1- w
(.) -
w
Lut 2 Lu 20
_,
0 3
0 2 4 5 00 2 4 5
NO. OF TRANSFER UNITS, NTU = AU/C NO. OF TRANSFER UNITS, NTUmax = AU/Cmm
max min
;ROSS- FONEXCEMEGULyani
ONE FLUID MIXER
.14
4.1111.„.
, r
FLUID
NV,
1./N121220 r I. ta a
100
i 1 . 0 25
L
crnfixed .
0.00 '::-.....:
. '''' 05
BO unmixed . ...... -
E FFECTI VENESSE%
‘
-------- I33
60 C -.---
_MiXed____
cunmi. ed
40
20
0
1 2 3 4 5
NO. OF TRANSFER UNITS, NTUmax = AU/Cmm
Cmax is essentially infinite. Thus, the ratio Cmin/Cmax = 0 is used to determine effective-
ness from any of the three arrangements shown since for Cmin /Cmax = 0, the expressions
for effectiveness as a function of (AU)Hx /Cmin for all flow arrangements reduce to the
same expression.
The following examples demonstrate the utilization of the calculation method for evalua-
tion of solar heating and hot water systems. Examples using air and water as working
fluids were chosen to represent existing typical solar heating systems.
Auxiliary
Heat
is
I I —I
Hot .---t.—...
Collector Water Load
Storage Main Hot Water Heat
Storage Preheat Auxiliary
Heat Exchanger
Storage Heat
Exchanger
glim.. tm
I I,.............
Hot Water
Loop Pump
(mc p)a= Ca
The values and calculations developed in this section are tabulated on Worksheet B.
Monthly Fraction, (f). By locating the system parameters D1 and D2 on Fig. 43.5,
(f) can be determined on a monthly basis. D1 and D2 are taken from Worksheet C, L is
from Worksheet D, and S is from Worksheet E. For February, the value for (f) was 0.48
and is tabulated in Worksheet B.
Annual Fraction ( Fannual). The solar energy supplied for the example month of
February is
EFeb = (fFeb)(LFeb)
The total solar energy supplied for the entire year is calculated on Worksheet B.
The total heating load for the entire year is calculated on Worksheet B.
Fannual for the entire year is equal to
The values and calculations developed in this section are calculated on Worksheet C.
The parameters D1 and D2 are obtained from Eqs. (43.5) and (43.6). For the month of
February this is done as follows:
[A,F'R er7i) S] S S
D1 (860)(0.69)(0.95)—
K.4 = (D1 prod) — = (860)(0.69)(0.95)
L L L
S 35.8 X 103
= (563.7) — = 563.7 = 0.86
L 23.5 X 106
[(212 - 5) 6721
= (775.7) X 1.0 = 4.2
23.5 X 106
WORKSHEET B Fraction of Total Heating Load Supplied by Solar Energy, Fannuai
1 2 3 4 5
Actual PROJECT: Liquid H & DHW Example
Tot. Mo. Solar Solar Madison, Wis.
Htg. Load, System System Fraction/ en/mo,
L Parameter Parameter mo., E
Month (Btu/mo.) D1 D2 f (Btu/mo.)
1. From Worksheet D.
2. From Worksheet C.
3. From Worksheet C.
4. From F chart.
5. E = f X L.
WORKSHEET C System Performance Parameters D1 , D2
1 2 3 4 5 6 PROJECT: Liquid H&
DHW Example
Tot. Monthly Total Mo. Av.
Madison, Wis.
Radiation Heating Daytime Tot. Hrs.
on Tilt Surf., Load, Temp., in Mo.
S L to 212 - to Atime
Month (Btu/mo • ft2 ) (Btu/mo.) S/L D1 (°F) (°F) (h) K3 D2
Jan. 34.3 X 103 27.1 X 106 1.26 X 10-3 0.71 22 190 744 1.0 4.1 Ac = 860 ft2 Given data
Feb. 35.9 23.5 1.53 0.86 25 187 672 1.0 4.2 F'R (roe)= 0.69 Worksheet F
March 49.3 20.9 2.36 1.33 35 177 744 1.0 4.9 PR UL = 0.82 Worksheet F
1.0 7.4 K1 = 1.0 Worksheet G
April 45.0 12.3 3.66 2.06 49 163 720
K2 = 1.1 Worksheet G
May 48.7 7.1 6.86 3.87 61 151 744 1.0 12.3
K4 = 0.95 Worksheet G
June 51.8 3.5 14.8 8.34 71 141 720 1.0 22.5
July 55.2 2.2 25.1 14.1 77 135 744 1.0 35.4
Aug. 53.0 2.5 21.2 12.0 74 138 744 1.0 31.8
Sept. 50.7 4.7 10.8 6.09 66 146 720 1.0 17.3
Oct. 45.9 10.0 4.6 2.59 54 158 744 1.0 9.1
Nov. 28.7 17.7 1.62 0.91 38 174 720 1.0 5.5
Dec. 30.9 24.6 r.23 0.69 25 187 744 1.0 4.4
1. From Worksheet E.
2. From Worksheet D.
[Acreic7) Sl s
3. D1 = L X K4 = (D i prod.) - = 563.7 s
where D1 prod. = [AcFit (TN)] X K4 = (860)(0.69)(0.95) = 563.7.
4. From Sec. 43.5, Table 43.4.
5. From Table 43.14 and Worksheet G
[A,VRUL (tref - to) A time] [
( 212 - to) A time]
6. D2 = L x K1 X K2 X K3 - (D2 prod.) X K3 = 775.7 [ ] X K3.
Qs = (PF)(24)(UA)(Degree Days)
where
PF = 0.75
Values obtained are tabulated in Worksheet D for the example problem. In February, for
example,
Q, = mcp - tm)
6. L = Qs + Qw.
Procedures for Determining Thermal Performance 563
Total Heating Load (L). Total heating load L is then the sum of space heating
(Qs) and water heating (C),):
L = Q, + Q,
Values for L for each month are tabulated in Worksheet D for the example problem.
Monthly averages of the daily radiation incident on a horizontal surface (TH) were
taken from Table 43.4, Sec. 435, and are tabulated in Worksheet E for the example
problem.
Monthly values for Rt were taken from Ref. 10 and are also tabulated in Work-
sheet E.
Knowing the collector tilt (0 = 43°), the latitutde (0 = 430), and monthly Et values,
monthly il values were taken from Table 43 .5, Sec. 43.5, and tabulated in Worksheet E.
For instance, in February, K = 0.47. Referring to Table 43.5 (Kt = 0.5) and determining
the latitude minus tilt difference (0 - 0 = 43° - 43° = 0°), the k value under the Febru-
ary column opposite latitude 43° is R = 157.
Monthly average daily radiation on a tilted surface (TT ) is calculated using Eq.
(43.12),
IT = (111)(11)
Total average insolation per month (S) is calculated using Eq. (43.13),
S = (I r)(N)
where N is the number of days in the month. S is then tabulated in Worksheet E. For
example, in February ,
1.0
0.9
0.8
0.6
0.5
0.3
0.2
t -t
i a F•11.ft2
IT Btu
Figure 43.20
566 Applications of Solar Energy
tin + tout
Reference temperature basis: 3 . tout
2
1
Case 2: FR (rct)n — b(
1 + mAc /2Cn ) =
F R UL — m
lc /2Cc
(1 + mA
gal (8.33 lb ) Bu \
= (20 (1.05 lb ) (60 min) (0.82
min ) gal lb h lb °F)
t
Btu
= 8610
h • °F
Case 3: FR(Ta)n = b
ln /Cc )
(1 + mA
FR UL =
l c /Cc
(1 + mA
WORKSHEET F (Continued)
Collector loop heat exchanger modifier, FR/F R
For air systems and liquid systems without a collector/storage heat exchanger,
FR
Capacitance rate:
Btu
Cc = (from above) = 8610 h oF
gal B tu
Cs = (calc. as for Cc = (16 ) (8.33)(1)(60)(1) = 8000
min h • °F
above)
Ac(F R U L ) (860)(0.85)
Y - 0.085
Cc 8610
FR 1 1
= from Fig. 43.10 or = - 0.96
FR 1 + y(x - 1) 1 +0.085(1.537- 1)
( F
FR(FE) = FR (rce)n [(roc)„ (FR) = (0.79)(0.91)(0.96) = 0.69
Btu
F4 )
FRUL = F R U L — = (0.85)(0.96) = 0.82 h • ft
FR 2 • °F
WORKSHEET G (Continued)
Storage mass capacitance factor, K2
M = (vol. storage media) (density) = (1000 + 80)(8.33) = 8996 lb
Note: M includes hot water storage
volume where it is solar heated.
Btu
c = (from Worksheet A) = 1.0
lb • °F
(Mcp ), (8996)(1) _
10.46 K2 = (from Fig. 43.12) = 1.1
A0 860
(1 Btu ) Btu
— 5000
lb • °F lb • °F
Btu ) Btu
(0.24 = 1296
lb • °F lb • °F
Btu
Crain = lesser of CH or CA = 1296
lb • °F
Btu
UA bldg = (from Worksheet D) = 956 0
h • F
UA 956
— 0.92 K4 = (from Fig. 43.14) = 0.95
Crain el_ (1296)(0.80)
Procedures for Determining Thermal Performance 569
Our example is the solar air heating system shown in Fig. 43.21. The system is designed
to provide space heating and domestic hot water. This system has three modes of opera-
tion. Mode 1 occurs when solar energy is available for collection and there is a space
heating load. Then, room temperature air is drawn through the solar collectors, heated,
and returned to the building. The dampers will be in position 1. Mode 2 occurs when
solar energy is available for collection at times when there is no space heating load. Air
from the bottom of the pebble bed is drawn through the solar collectors, heated, and
returned to the top of the storage unit. The hot air moving down through the bed heats
the pebbles resulting in sensible heat storage. The dampers will be in position 2. Mode 3
occurs when no solar energy can be collected but there is a space heating load. Hot air is
drawn from the top of the pebble bed into the house and room temperature air is
returned to the bottom of the bed. The dampers will be in position 3. In modes 1 and 3,
auxiliary energy from a conventional furnace may supplement the solar contribution.
The diagram indicates the dampers in the mode 1 position.
The mode of operation of the solar air heating system is determined by the position
of the dampers. Whenever the collector is operating, damper A is in position 1, 2; other-
wise, it is in position 3. The collector operation is controlled by an on-off differential
controller monitoring the temperatures of the air in the collector outlet manifold, to ,
and in the bottom of the pebble bed, tn, as indicated in Eq. (43.37)
Atiand Ott are controller deadbands* ideally chosen so that the energy collected is at
least equivalent to the energy required to operate the blower. Both Atiand Ott have
been chosen to be 9°F in the examples noted here.
Dampers B and C are controlled by the building thermostat. Whenever the building
needs heat, damper B is in position 1 or 3 and damper C is in position 1, 3 (fig. 43.21);
otherwise they are in position 2. The modeling method used, which is suitable for long-
term simulation, does not follow the system mode changes exactly, but rather assumes
that during any time period, the system operates in whatever modes necessary to main-
tain the building temperature at the desired level. Then by comparing energy rates, it is
possible to determine the fraction of the time period in which the system operated in
each mode.
During each time period, the rate of energy collection is compared with the rate of
energy required to meet the heating load. If there is zero energy collection, the system is
assumed to be in mode 3 operation. If the rate of energy collection is nonzero, but
smaller than the rate of energy required by the load, it is assumed that the system is
in mode 1 operation. If the rate of energy collection is greater than the rate at which
energy is required, the fraction of the time period which the system would have to be in
*Deadband = temperature range over which the controller does not initiate or terminate operation of
the fan.
0
3 SuPPly
Air
Hot Water
Loop Pump
Figure 43.21 Air system: space heating and domestic hot water.
A 2Jau3 mos Jo suopn!i ddv
Procedures for Determining Thermal Performance 571
mode 1 operation to just satisfy the load is calculated; the system is assumed to be in
mode 2 operation during the remainder of the period. This method of calculating system
performance allows the simulation to use time steps on the order of an hour without a
sacrifice in the accuracy of the calculated long-term system performance.
Grand Junction, Colorado, has been selected as the location of the example
problem, as it represents a city with major winter heating requirements and a significantly
different set of solar radiation conditions from the previous example. Project data for the
example are presented in Worksheet A.
The values and calculations developed in this section are tabulated on Worksheet B.
Monthly Fraction ( f). By locating the system parameters D1 and D2 on Fig. 43.4,
(f) can be determined on a monthly basis. D1 and D2 are taken from Worksheet C,
L is from Worksheet D, and S from Worksheet E. For February, the value for (f) was
0.76 and is tabulated in Worksheet B.
Annual Fraction (-Fan nuad• The solar energy supplied for the example month of
February is
The total solar energy supplied for the entire year is calculated on Worksheet B.
The values and calculations developed in this section are calculated on Worksheet C.
The parameters D1 and D2 are obtained from Eqs. (43.5) and (43.6). For the month of
February this is done as follows:
D1 = (AcFF
R a0 S) LS
X K4 = (D1 prod) ES = (700)(0.48)(1.0) —
L
s 56.3 X 10 3
= 322 — = 322 — 1.12
L 16.2 X 106
572 Applications of Solar Energy
+ tout
Reference temp. basis: tin, tin
2
Fluid:
Composition: Air
Specific heat, cp = 0.24 Btu/lb • °F
Specific gravity (if applicable) = N/A lb/lb
Volumetric flow rate = 1500 gal/min o
Storage subsystem data
Volume = 400 gal or
Storage medium: Rock
Specific heat, cp = 0.2 Btu/lb • °
Siaesifis-gra-vity/-oP density = 100 lb/lb o
Circulation loop volumetric flow rate = 1500 gal/min orfCt3 /mir
Collector/storage heat exchange effectiveness, ecs = 1.0
Hot water preheat storage volume = 80 gal
Load subsystem data
Load heat exchanger effectiveness, EL = N/A
Supply loop volumetric flow rate = N/A gal/min
Building air supply volumetric flow rate = 1500 ft3 /min
WORKSHEET B Fraction of Total Heating Load Supplied by Solar Energy, Fannuai
1 2 3 4 5
Actual PROJECT: Air H & DHW Example
Tot. Mo. Solar Solar Grand Junction, Colo.
Htg. Load, System System Fraction/ en/mo,
L Parameter Parameter mo., E
Month (Btu/mo.) D1 D2 f (Btu/mo.)
Etot 87.17 X 10 6
Ltot = 112.5 X 106 Etot = 87.17 X 106 Fannual T - 0.77
-tot 112.5 X 106
1. From Worksheet D.
2. From Worksheet C.
3. From Worksheet C.
4. From F chart.
5. E = f x L.
WORKSHEET C System Performance Parameters D1 , D2
1 2 3 4 5 6 PROJECT: Air H & DHW
Example,
Tot. Monthly Total Mo. Av.
Tot. Hrs. Grand Junc-
Radiation Heating Daytime
in Mo. tion, Colo.
on Tilt Surf., Load, Temp.,
S L to 212 - to time
Month (Btufmo • ft2 ) (Btu/mo.) S/L D1 (°F) (° F) (h) K3 D2
Jan. 50.5 X 103 21.3 X 106 2.37 X 10 -3 0.76 27 185 744 1.0 2.52 Ac = 700 ft2 Given data
Feb. 54.2 16.2 3.34 1.07 35 177 672 1.0 2.86 FiR (77:0= 0.46 Worksheet F
March 62.4 13.6 4.59 1.48 45 167 744 1.0 3.56 VR UL = 0.51 Worksheet F
April 57.1 8.0 7.14 2.30 56 156 720 1.0 5.47 K1 = 1.04 Worksheet G
May 56.3 4.2 13.40 4.31 66 146 744 1.0 10.08 K2 = 1.05 Worksheet G
June 52.9 2.1 25.57 8.88 76 136 720 1.0 18.18 K4 = 1.0 Worksheet G
July 58.5 1.9 30.79 9.91 82 130 744 1.0 19.84
Aug. 58.8 1.9 30.95 9.96 80 132 744 1.0 20.15
Sept. 67.0 2.3 29.13 9.38 71 141 720 1.0 17.21
Oct. 64.0 6.9 9.27 2.98 58 154 744 1.0 6.47
Nov. 54.4 14.4 3.77 1.21 42 170 720 1.0 3.31
Dec. 52.1 19.7 2.64 0.85 31 181 744 1.0 2.66
1. From Worksheet E.
2. From Worksheet D.
[AorR(F-67)
3. D I = X K4 = (D1 prod.) s- = 322 s-
L
where D1 prod. = [AorR (-7a)] x K4 = (700)(0.46)(1.0) = 322.
4. From Sec. 43.5, Table 43.4.
5. From Table 43.14 and Worksheet G.
[AYRUL
0 (tref - to) A time o) A time ]
6. D2 = x K1 X K2 X K3 - (D2 prod.) [(212 - t X K3 = 389.8 [ X K3.
L L
/Mau] reps Jo suoppo!iddv
Space Heating Load. Knowing the building thermal characteristics (heat transfer
coefficients and areas of surfaces exposed to the outside) and using Manual J, the design
temperature difference (Ltd) and design heat loss (qd) were found to be 59° F and 53,000
Btu/h, respectively. These are recorded on Worksheet A. An example of the Manual J
calculation is presented in Fig. 43.22 for the example house of 1400 ft2 floor area.
Monthly total degree days (DD) for Grand Junction were taken from Table 43.3
in Sec. 43.5.
Monthly space heating load (Qs) was then calculated using Eqs. (43.7) and (43.8).
In February, for example,
Qs = (PF)(24)(UA)(degree days)
where
PF = 0.75
Domestic Hot Water Heating Load. In this example, the hot water requirements
are 85 gal/d. Monthly requirements are then
gal X N days
85
day month
WORKSHEET D Heating and/or Domestic Hot Water Load, L
1 2 3 4 5 6 PROJECT: Air H& DHW
Example
Monthly Monthly Temp. DHW Monthly Total
Grand Junction, Colo.
Degree Space No. of Vol. of Water Temp. DHW Heating
Days Htg. Load, Days/ DHW Main Sup., Rise, Load, Load,
qd = 53,000 Btu/h
DD Qs Mo., Used/mo., tm ts - tm Qw L
(given data or calculate
Month (°F • d) (Btu/mo.) N (gal/mo.) (°F) (°F) (Btu/mo.) (Btu/mo.)
from Manual J or equiva-
Jan. 1209 19.4 X 106 31 2635 55 85 1.86 X 106 21.3 X 106 lent)
Atd = 70 - tw
Feb. 907 14.5 28 2380 55 85 1.68 16.2 = 70 - 11 =59°
March 729 11.7 31 2635 55 85 1.86 13.6 where tw = 971/2% winter
design temperature
April 387 6.2 30 2550 55 85 1.80 8.0
(from ASHRAE Funda-
May 146 2.3 31 2635 55 85 1.86 4.2 mentals, Table 43.3,
Sec. 43.5, or known
June 21 0.3 30 2550 55 85 1.80 2.1
weather data)
July 0 0 31 2635 55 85 1.86 1.9 70° = indoor design
Aug. 31 1.86 1.9 temperature
0 0 2635 55 85
qd _ 53,000 Btu/h
Sept. 30 0.5 30 2550 55 85 1.80 2.3 UA -
Ltd 59°F
Oct. 313 5.0 31 2635 55 85 1.86 6.9
Nov. 786 12.6 30 2550 55 85 1.80 14.4 = 898 Btu
h • °F
Dec. 1113 17.8 31 2635 55 85 1.86 19.7 ts = 140°F
1. From ASHRAE Systems, Climatic Atlas, or Table 43.3, Sec. 43.5.
2. Qs = (PF)(24)(UA)(degree day) = (0.75)(24)(898) X (degree day) = 16,164X DD
where PF = 0.75 or snore appropriate value.
3. (Vol/day) (no. days/mo.) = 85 (gal/d) (no. days/mo.)
4. May be constant or may vary.
5. Qw = (vol. of water) X 8.33 X 1 X (ts - tm).
A2JOLI1 J EIOS 40SUOI1E0Il ddy
6. L = Qs + Qw.
Procedures for Determining Thermal Performance 577
2.46 227.
Oro 370
22. I, 4n 28 I 4_ i ) 8 404 28 4:
lc 7.470 4
2.0 .4 In
;:3 164 i/` 2, 32 13 D 204 '.) 3Cz 140
712. 62.13 Jo
IS
IB
l 1f
37
2
Values obtained are tabulated in Worksheet D for the example problem. In February,
for example,
Qw = mcp tm )
8.33 lb Btu
Qw = ( 2380 gal (140°F - 55°F) = 1.68 X 106 Btu
mo. gal) 1 lb • °F
Total Heating Load (L). The total heating load L is then the sum of space heating
(Qs) and water heating (Qw):
L = Qs + Qw
Values of L for each month are tabulated in Worksheet D for the example problem.
Monthly averages of the daily radiation incident on a horizontal surface (TO were
taken from Ref. 10, and are tabulated in Worksheet E for the example problem.
Monthly values for Kt were taken from Ref. 10, and are also tabulated in Work-
sheet E.
Knowing the collector tilt (0 = 54°), the latitude (0 = 39°), and monthly Kt values,
monthly R values were taken from Table 43.5, Sec. 43.5, and tabulated in Worksheet E.
For instance, in February, IZ = 0.63. Referring to Table 43.5, (Kt= 0.6) and determining
the latitude minus tilt difference (0 - 0 = 39° - 54° = -15°), the R value under the
February column opposite latitude 40° is R. = 1.68.
Interpolating for Kt = 0.63 and (1) = 39° gives a value of k = 1.66.
Monthly average daily radiation on a tilted surface (TT) is calculated using Eq.
(43.12),
IT = (IH)(R)
WORKSHEET E Total Monthly Solar Radiation Available, S 0
0
PROJECT: Air H & DHW Example, Grand Junction, Colo.
(7D'
1 2 3 4 5 6 -h
0
Extra- Ratio Monthly Avg. Tot. Monthly
Horizontal terrestrial Horizontal Ratio Daily Rad. Radiation (-7)
.--,
CD
Insolation, Insolation, to Extra- Horizontal on Tilt Surf., No. of Days on Tilt Surf., 7.
TH To terrestrial, to Tilt, IT in Month S 5:
Month Btu/(d • ft2 ) Btu/(d • ft2) 1(t R Btu/(d • ft2 ) N Btu/(mo • ft2 ) a
Jan. 848 0.60 1.92 1628 31
-1z-
CD
N/A 50.5 X 103
Feb. 1211 N/A 0.63 1.60 1938 28 54.2 -g'
,,
March 1623 N/A 0.64 1.24 2012 31 62.4 -0
CD
Total average insolation per month (S) is calculated using Eq. (43.13),
S = (Tr)(N)
where N is the number of days in the month. S is then tabulated in Worksheet E. For
example, in February,
S = 2010 Btu 28 d
d • ft2 mo.
Btu
= 56.3 X 103
mo. • ft2
1.0
0.9
0.8
0.6
n 0.5
0.2
0.1
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1
t -t
• a t °F.Ivft2 ,
,
IT ' Btu
Figure 43.23
Procedures for Determining Thermal Performance 581
tin + tout
Reference temperature basis: 1. tin, 2. 3 . toi:t)
2
1
Case 2: FR (ra)n = b(
1 + mAc /2Cc
1
FRUL
m(1 + mAc /2Cc
) (0 75 Lb \ min Btu )
= (1500 ft3 0 h ) (0.24 lb OF
min • ft3
Btu
= 1620
h - °F
1
Case 3: FR(ra)n = b = 0.64 — 0.50
(1 + mA
lc /Cc ) 1 +(0.65)(700)/1620
1
FRUL = m = 0.65
l c /Cc )
( I + mA 1 +(0.65)(700)/1620
Btu
= 0.51
h • ft2 • °F
(ra) for two cover plates
Incident angle modified,
0. • 3, for one cover plate
(ra)n
582 Applications of Solar Energy
WORKSHEET F (Continued)
Collector loop heat exchanger modifier, FR/F R
For air systems and liquid systems without a collector/storage heat exchanger,
Capacitance rate:
Cc = (from above)
Cs = (calc. as for Cc
above)
(FFt)
F4 (TO = FR (170,1 [ ]( = (0.50)(0.91)(1.0) = 0.46
(TcOn F R
F iR
FRUL = F R U L (-) = 0.51
FR
WORKSHEET G (Continued)
Storage mass capacitance factor, K2
lb
M = (vol. storage media) (density) = (400 ft 3 ) 100 = 40,000 lb (rock)
ft3
lb
(80 gal) 8.33 — = 666 lb (water)
Note: M includes hot water storage gal
volume where it is solar heated.
Crain = lesser of CH or CA
UA
K4 = (from Fig. 43.14) = 1.0
Crain EL
584 Applications of Solar Energy
Our typical example of solar energy systems for a domestic hot water heating application
is a relatively simple device consisting of a collector, hot water storage or preheater tank,
and associated pumps, piping, and controls as depicted in Fig. 43.24. The design of a
solar energy hot water heater differs from a building air conditioning system because the
demand is not a function of seasonal ambient temperature and the collector operating
temperature can be lower. Either liquid or air collecting systems are available for use
with domestic hot water heating.
The City of Fort Worth was selected for the hot water calculation example because
it represents a climatic region with varying seasonal water main temperatures and it has a
relatively warm climate in contrast to the other examples. Utilizing the average monthly
water main temperature and assuming a constant hot water demand of 70 gal/d for the
year at a design storage temperature of 140°F resulted in average daily and monthly loads
significantly higher in winter than in summer. The large winter load resulted in selecting
a collector tilt angle of latitude plus 15° (0 = 48°) to position the collector more normal
to the sun in the winter.
Hot Water
Auxiliary Heat
I1
1
Collector/ Hot
Storage Water
Heat Preheat
Exchanger Storage
t
I m
A single-glazed flat black collector was chosen because of the relatively high
ambient temperature conditions and the resultant good thermal performance in the range
of 100 to 140°F. The use of corrosion-inhibited water in the collector loop implies
the need for a double-wall heat exchanger to transfer the heat to the stored domestic
hot water. Project data are shown on Worksheet A.
The values and calculations developed in this section are tabulated on Worksheet B.
Monthly Fraction (f). By locating the system parameters D1 and D2 on Fig. 43.5,
(f) can be determined on a monthly basis. D1 and D2 are taken from Worksheet C,
L is from Worksheet D, and S is from Worksheet E. For February, the value for (f)
was 0.76 and is tabulated in Worksheet E.
Annual Fraction (Fannual). The solar energy supplied for the example month of
February is
EFeb = (fFeb)(LFeb)
The total solar energy supplied for the entire year is calculated on Worksheet B.
The total heating load for the entire year is calculated on Worksheet B.
F annual for the entire year is equal to
The values and calculations developed in this section are calculated on Worksheet C.
The parameters D1 and D2 are obtained from Eqs. (43.5) and (43.6). For the month of
February this is done as follows:
[A VR(ra) S LS
D1 = e 1 X K4 =031 — 00)(0.737)(1)—
P r° L
s 3 45.95 X 103
= (40.88)— = 40.88 — 1.26
1.49 X 106
586 Applications of Solar Energy
1. From Worksheet D.
2. From Worksheet C.
3. From Worksheet C.
4. From F chart.
00
5.E=-f X L.
Lri
WORKSHEET C System Performance Parameters D I , D2 00
00
1 2 3 4 5 6 PROJECT: DHW Example
Tot. Monthly Total Fort Worth, Tx.
Mo. Av.
Radiation Heating Daytime Tot. Hrs.
on Tilt Surf., Load, Temp., in Mo.
S L to 212 - to A time
Month (Btu/mo • ft2 ) (Btu/mo.) S/L D1 (°F) (°F) (h) K3 D2
Jan. 47.0 X 103 1.518 X 106 3.10 X 10 -2 1.27 48 164 744 1.30 7.66 A, = 56 ft2 Given data
Feb. 46.0 1.486 3.09 1.26 52 160 672 1.10 5.84 Flt(Tce)= 0.730 Worksheet F
March 56.5 1.500 3.77 1.54 60 152 744 1.22 6.75 FiR UL = 1.31 Worksheet F
April 50.5 1.225 4.12 1.69 69 143 720 1.45 8.94 K1 = 1.0 Worksheet G
May 51.6 1.175 4.39 1.79 76 136 744 1.60 10.11 K2 = 1.0 Worksheet G
June 53.3 1.032 5.17 2.11 84 128 720 1.70 11.14 K4 = 1.0 Worksheet G
July 54.0 1.103 4.90 2.00 88 124 744 1.65 10.12
Aug. 54.1 1.030 5.74 2.35 89 123 744 1.75 11.41
Sept. 59.3 1.032 5.74 2.35 81 131 720 1.72 11.53
Oct. 59.9 1.229 4.88 1.99 72 140 744 1.45 9.02
Nov. 54.1 1.470 3.67 1.50 59 153 720 1.22 6.71
Dec. 48.7 1.694 2.87 1.17 51 161 744 1.02 5.29
1. From Worksheet E.
2. From Worksheet D.
[A,FRO-V) S]
3. Di - X K4 = (D i prod.) s- = 40.88 s-
L L
where D1 prod. = [.A.cF"R (7-5)] X K4 = (56)(0.730)(1.0) = 40.88.
4. From Sec. 43.5, Table 43.4.
5. From Table 43.14 and Worksheet G
reFkUL (tref - to) .6, time] 1(212 - to) Atime
6. D2 = - x K1 X K2 X K3 = (D2 prod.) X K3 = 73.36 [ ] x K3.
L L
anuA mos Jo suopro!i ddv
[(212 - 57)6721
(73.36) = X 1.10 = 5.84
1.49 X 106
The total heating load is equal to the sum of the space heating load and the domes-
tic hot water heating load. Worksheet D is used for load calculations.
Space heating load is not applicable.
Domestic hot water heating load: In this example, the hot water requirements are
70 gal/d. Monthly requirements are then
Values obtained are tabulated in Worksheet D for the example problem. In February, for
example,
The total heating load (L) equals the hot water load (L).
WORKSHEET D Heating and/or Domestic Hot Water Load, L
1 2 3 4 5 6 PROJECT: DHW Example
Ft. Worth, Tx.
Monthly Monthly Temp. DHW Monthly Total
Degree Space No. of Vol. of Water Temp. DHW Heating
qd = N/A Btu/h
Days Htg. Load, Days/ DHW Main Sup., Rise, Load, Load, (given data or calculate
DD Qs Mo., Used/mo., tm is — tm Q, L from Manual J or equiva-
Month (°F • d) (Btu/mo.) N (gal/mo.) (°F) (°F) (Btu/mo.) (Btu/mo.) lent)
Jan. N/A N/A 31 2170 56 84 1.518 X 106 Same Atd = 70 — tom,
= 70 — N/A =
Feb. N/A N/A 28 1960 49 91 1.486 Same where tom, = 97'/z% winter
March N/A N/A 31 2170 57 83 1.500 Same design temperature
(from ASHRAE Funda-
April N/A N/A 30 2100 70 70 1.225 Same mentals, Table 43.3,
May N/A N/A 31 2170 75 65 1.175 Same Sec. 43.5, or known
we
June N/A N/A 30 2100 81 59 1.032 Same 70° = indoor design
July N/A N/A 31 2170 79 61 1.103 Same temperature
Aug. N/A N/A 31 2170 83 57 1.130 Same qd
UA = = N/A
Sept. 30 2100 81 59 1.032 Same Atd
N/A N/A
Oct. N/A N/A 31 2170 72 68 1.229 Same
Nov. N/A N/A 30 2100 56 84 1.470 Same ts = 140°F
Dec. N/A N/A 31 2170 46 94 1.649 Same
1. From ASHRAE Systems, Climatic Atlas, or Table 43.3, Sec. 43.5.
2. Qs = (fin (24)(UA)(degree day) = N/A X (degree day) =
where PF = 0.75 or more appropriate value.
3. (Vol/day) (no. days/mo.) = 70 (gal/d) (no. days/mo.)
4. May be constant or may vary.
5. Qw = (vol. of water) X 8.33 X 1 X (ts — tm).
ASJou3 Ri os Jo suopuou ddv
6. L = Qs + Qw.
Procedures for Determining Thermal Performance 591
Monthly averages of the daily radiation incident on a horizontal surface (TO, were
taken from Ref. 10, and are tabulated in Worksheet E for the example problem.
Monthly values for Kt were taken from Ref. 10, and are also tabulated in Work-
sheet E.
Knowing the collector tilt (0 = 48°), the latitude (0 = 33°), and monthly Kt values,
monthly R values were taken from Table 43.5, Sec. 43.5, and tabulated in Worksheet E.
For instance, in February, K = 0.54. Referring to Table 43.5 (Kt = 0.5) and determining
the latitude minus tilt difference (0 - 0 = 33° - 48° = -15°), the R value under the
February column opposite latitude 33° is R = 1.37. Interpolating for Kt = 0.63 and
0 = 39° gives a value of iz- = 1.66.
Montly average daily radiation on a tilted surface (IT ) is calculated using Eq.
(43.12),
IT = (TH)(k)
Total average insolation per month (S) is calculated using Eq. (43.13),
S = (TT)(N)
where N is the number of days in the month. S is then tabulated in Worksheet E. For
example, in February,
Btu
= 45.9 X 103
mo. • ft2
Collector Loop Heat Exchanger Modifier, FR7FR • The collector loop heat ex-
changer modifier is calculated on Worksheet F.
1.0
0.9
0.8
0.3
0.2
0.1,
JJ 1 tt Ilill
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1
ti..t II .F.h.ft2
f )
' Btu '
IT
Figure 43.25
594 Applications of Solar Energy
F R UL = m
lc /2Cc )
(1 + mA
Case 3: FR (Ta)n = b
+ mA
l c /Cc )
1
F R UL = m
1 + mAc /Cc )
WORKSHEET F (Continued)
FR
=1
FR
Capacitance rate:
Btu
Cc = (from above) = 839.7
h
8.33 1 Btu
Cs = (talc. as for Cc 1.8-1!1--) ( lb ) (60 min
above)
=( min gal h lb
Btu
= 899.6
h
Btu
Cmin = (lesser of C, of Cs) = 839.7 h
Ac(F R U L)(56)(1.35)
y = - 0.090
Cc 839.7
FR 1 1
= from Fig. 43.10 or = - 0.970
FR 1 + y(x - 1) 1 + 0.09(1.339 - 1)
FRU L = F R U L F
R) = (1.35)(0.97) 1.31
R
WORKSHEET G (Continued)
Cc
Ac K1 = (from Fig. 43.11) = 1.0
(Mcds (916.3)(1)
16.36 K2 = (from Fig. 43.12) = 1.0
Pic 56
EL = (from Worksheet A)
Cmin = lesser of CH or CA
UA
K4 = (from Fig. 43.14) = 1.0
Cmin E L
Procedures for Determining Thermal Performance 597
A. Nomenclature
▪ 60
rj 40 4
20
0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 *cwm2
t.- ta
i
I
T
(a)
60
R"
40
4
0.
20
2 1.0 • F'h'ft 2
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0. 9r_____LT_21
0 d
0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 ."
w
ti- ta
I
T
(b)
Figure 43.26 Typical thermal efficiency curves for liquid collectors based on (a) collec-
tor aperture area and (b) collector gross frontal area.
600 Applications of Solar Energy
6=6.67 SCFM/FT2
0.7 Cya=6.67 O
COVER PLATE'
AIR PASSAGE
0.6
6=6.67 C1=6.67
\\ •
•
0.5
• O a=2.0
••
•
-
z
w
0.4 0 •
LL
LL
•
w
cc
I—0
tyi 0.3
•
0 6=3.09
0 •
•• FLAT BLACK ABSORBER
PLATES
0.2
•
•
•
•
0.1
0
0
0
0 0.2 0.04 0.06 0.08 1.0 1.2
t -t (°F•h•ft2 )
a
I Btu
T
Figure 43.27 Typical thermal efficiency curves for flat-plate collectors using air as the
working liquid [12]. Q indicates flow rate in standard cubic feet per minute (scfm) per
unit collector area.
Procedures for Determining Thermal Performance 601
"rartalWIt A
JANUARY
so
k
550
/41541
350
C*41‘k
550
750
750
90n
907
9
r 1100 1100
. 900
1100 900
,()
550 750
750 i444NI6A t
900
900
0
4101 :
1100 %Igbaq% 1100
0 1*
1300
1500 to
13 - 1300
MARCH 1100
tv/V .....m.
900 ArAlraiallti
13 0
1500
APRIL
vo 6,17
130 14111 1500
01"1191r 500
1500
170
1800
700
2000A,
I11,11111;,
2000 20
18 0 IN 1800
)0 qui
00' for
200
,o 2000
T ,220 2$00 1800
0 2000
MAY
1800
rM 1800
800
1700
1800
1800
2000
2000
2200
170
1800
2000
JUNE 200
2 00
1800 1800
41111
1110111111111111
1
1911111tit
1 04
0 :7
4*t
71,*
1700
1800 000
2000 114.1111011
11140
2200
2200 11001.1111,41114W
800
1(
2000
2400
00
o
U17)1!"1111700-/\
2000
2200
2000
180
2200
NLY 22 00
gftEsisii...
4sitier:076000 m41
2000 to
80
AI
rikuerstiv-
0 lr
0 2000
24.0
2400
0
RP7Ai
24 00
2000
24 00
22 00
1500 1 00
1800
AUGUST
SEPTEMBER 1300
15
1300
1300
Alillap".04111
aiwtkatibm
1,
41yr 1300
,30 i it unkaliCII
150 : 1
500
1700
1800
.
1800
"
., el I 111
1800 11111611
16
,0 lJk
. O0
2000
,
700
11800 1700
750
ArsookTvra.
900
::°r VWerftwiirtt
1100
1300
Mairn
1111L041 Cr
a
liM
i
datr#
15100‘011 "tref
1500
1700
00
1800
157N
NOVEMBER
550
550 401%
likk AP
WA IS4 700
,or
m
s4aft lbak,
a I I *11114111100r
410.
700
900 le
1100 900
wr
o
1100
\ 1300 ,0
v11
A
1100
o
D 1300 1300
1100
DECEMBER
350
350
550 AN
800
,
ftiatiroi„g• 800
900
w
irM
wv
1100
,
1100 900
fl 1100
900
Solar
Thickness Transmittance,
Material (in.) (%)
Anodize 0.82-0.99
971/2%
Winter
Design Yearly
State Station Temp.d July Aug. Sept. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May June Total
Ala. Birmingham A 22 0 0 6 93 363 555 592 462 363 108 9 0 2551
Huntsville A 17 0 0 12 127 426 663 694 557 434 138 19 0 3070
Mobile A 29 0 o o 22 213 357 415 300 211 42 0 0 1560
Montgomery A 26 0 o o 68 330 527 543 417 316 90 0 0 2291
Alaska Anchorage A -22 245 291 516 930 1284 1572 1631 1316 1293 879 592 315 10864
Fairbanks A -50 171 332 642 1203 1833 2254 2359 1901 1739 1068 555 222 14279
Juneau A -4 301 338 483 725 921 1135 1237 1070 1073 810 601 381 9075
Nome A -28 481 496 693 1094 1455 1820 1879 1666 1770 1314 930 573 14171
Ariz. Flagstaff A 5 46 68 201 558 867 1073 1169 991 911 651 437 180 7152
Phoenix A 34 0 0 0 22 234 415 474 328 217 75 0 0 1765
Tucson A 32 0 0 0 25 231 406 471 344 242 75 6 0 1800
Winslow A 13 0 0 6 245 711 1008 1054 770 601 291 96 0 4782
Yuma A 40 0 0 0 0 108 264 307 190 90 15 0 0 974
Ark. Fort Smith A 19 0 0 12 127 450 704 781 596 456 144 22 0 3292
Little Rock A 23 0 0 9 127 465 716 756 577 434 126 9 0 3219
Texarkana A 26 0 0 0 78 345 561 626 468 350 105 0 0 2533
-13 65 99 279 639 1065 1420 1476 1162 1020 696 440 168 8529
Colo. Alamosa A
Colorado
4 9 25 132 456 825 1032 1128 938 893 582 319 84 6423
Springs A
6 9 117 428 819 1035 1132 938 887 558 288 66 6283
Denver A 3
0 0 90 366 714 905 1004 851 800 492 254 48 5524
Denver C
Grand
11 0 0 30 313 786 1113 1209 907 729 387 146 21 5641
Junction A
-1 0 0 54 326 750 986 1085 871 772 429 174 15 5462
Pueblo A
0 0 0 28 40 31 9 0 0 0 108
Key West A 58 0 0
0 0 57 164 195 146 99 0 0 0 661
Lakeland C 39 0 0
0 0 0 65 74 56 19 0 0 0 214
Miami A 47 0 0
0 0 0 40 56 36 9 0 0 0 141
Miami Beach C 48 0 0
0 0 0 72 198 220 165 105 6 0 0 766
Orlando A 37 0
0 132 415 792 1017 1113 854 722 438 245 81 5809
Idaho Boise A 10 0
0 123 403 756 933 1063 815 694 426 239 90 5542
Lewiston A 12 0
0 172 493 900 1166 1324 1058 905 555 319 141 7033
Pocatello A -2 0
Ky. Covington A 8 0 0 75 291 669 983 1035 893 756 390 149 24 5265
Lexington A 10 0 0 54 239 609 902 946 818 685 325 105 0 4683
Louisville A 12 0 0 54 248 609 890 930 818 682 315 105 9 4660
Me. Caribou A -14 78 115 336 682 1044 1535 1690 1470 1308 858 468 183 9767
Portland A 0 12 53 195 508 807 1215 1339 1182 1042 675 372 111 7511
Md. Baltimore A 15 0 0 48 264 585 905 936 820 679 327 90 0 4654
Baltimore C 20 0 0 27 189 486 806 859 762 629 288 65 0 4111
Frederich A 11 0 0 66 307 624 955 995 876 741 384 127 12 5087
Mass. Boston A 10 0 9 60 316 603 983 1088 972 846 513 208 36 5634
Nantucket A 12 22 93 332 573 896 992 941 896 621 384 129 5891
Pittsfield A -1 25 59 219 524 831 1231 1339 1196 1063 660 326 105 7578
Worcester A 0 6 34 147 450 774 1172 1271 1123 998 612 304 78 6869
Mich. Alpena A -1 68 105 273 580 912 1268 1404 1299 1218 777 446 156 8506
Detroit (City) A 8 0 0 87 360 738 1088 1181 1058 936 522 220 42 6232
Detroit
/Mau] Ri os Jo suo!voudde
(Wayne) A 0 0 96 353 738 1088 1194 1061 933 534 239 57 6293
Mich. Detroit (Willow
(cont.) Run) A 0 0 90 357 750 1104 1190 1053 921 519 229 45 6258
Escanaba C -3 59 87 243 539 924 1293 1445 1296 1203 777 456 159 8481
Flint A 3 16 40 159 465 843 1212 1330 1198 1066 639 319 90 7377
Grand Rapids A 6 9 28 135 434 804 1147 1259 1134 1011 579 279 75 6894
Lansing A 6 6 22 138 431 813 1163 1262 1142 1011 579 273 69 6909
Marquette C -4 59 81 240 527 936 1268 1411 1268 1187 771 468 177 8393
Muskegon A 8 12 28 120 400 762 1088 1209 1100 995 594 310 78 6696
Sault Ste.
Marie A -8 96 105 279 580 951 1367 1525 1380 1277 810 477 201 9048
Minn. Duluth A -15 71 109 330 632 1131 1581 1745 1518 1355 840 490 198 10000
Minneapolis A -10 22 31 189 505 1014 1454 1631 1380 1166 621 288 81 8382
Rochester A -13 25 34 186 474 1005 1438 1593 1366 1150 630 301 93 8295
Mo. Columbia A 6 0 0 54 251 651 967 1076 874 716 324 121 12 5046
Kansas City A 8 0 0 39 220 612 905 1032 818 682 294 109 0 4711
St. Joseph A 3 0 6 60 285 708 1039 1172 949 769 348 133 15 5484
St. Louis A 8 0 0 60 251 627 936 1026 848 704 312 121 15 4900
St. Louis C 11 0 0 36 202 576 884 977 801 651 270 87 0 4484
Springfield A 10 0 0 45 223 600 877 973 781 660 291 105 6 4900
Mont. Billings A -6 6 15 186 487 897 1135 1296 1100 970 570 285 102 7049
Glasgow A -20 31 47 270 608 1104 1466 1711 1439 1187 648 335 150 8996
Great Falls A -16 28 53 258 543 921 1169 1349 1154 1063 642 384 186 7750
Havre A 28 53 306 595 1065 1367 1584 1364 1181 657 338 162 8700
Havre C -15 19 37 252 539 1014 1321 1528 1305 1116 612 304 135 8182
Table 43.3 (Continued)
97'/2%
Winter
Design Yearly
Station Temp.d July Aug. Sept. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May June Total
State
Helena A -13 31 59 294 601 1002 1265 1438 1170 1042 651 381 195 8129
Kalispell A -3 50 99 321 654 1020 1240 1401 1134 1029 639 397 207 8191
Miles City A -15 6 6 174 502 972 1296 1504 1252 1057 579 276 99 7723
Missoula A -3 34 74 303 651 1035 1287 1420 1120 970 621 391 219 8125
Neb. Grand Island A -2 0 6 108 381 834 1172 1314 1089 908 462 211 45 6530
Lincoln C 0 0 6 75 301 726 1066 1237 1016 834 402 171 30 5864
Norfolk A -7 9 0 111 397 873 1234 1414 1179 983 498 233 48 6979
North Platte A -2 0 6 123 440 885 1166 1271 1039 930 519 248 57 6684
Omaha A -1 0 12 105 357 828 1175 1355 1126 939 465 208 42 6612
Scottsbluff A -4 0 0 138 459 876 1128 1231 1008 921 552 285 75 6673
Valentine A 9 12 165 493 942 1237 1395 1176 1045 579 288 84 7425
Nev. Elko A -7 9 34 225 561 924 1197 1314 1036 911 621 409 192 7433
Ely A -2 28 43 234 592 939 1184 1308 1075 977 672 456 225 7733
Las Vegas A 26 0 0 0 78 387 617 688 487 335 111 6 0 2709
Reno A 7 43 87 204 490 801 1026 1073 823 729 510 357 189 6332
Winnemucca A 5 0 34 210 536 876 1091 1172 916 837 573 363 153 6761
N.H. Concord A -7 6 50 177 505 822 1240 1358 1184 1032 636 298 75 7383
Mt. Washington
Obsv. 493 536 720 1057 1341 1742 1820 1663 1652 1260 930 603 13817
N.J. Atlantic City A 18 0 0 39 251 549 880 936 848 741 420 133 15 4812
Newark A 15 0 0 30 248 573 921 983 876 729 381 118 0 4589
ASJau]_reps Jo suonPo!iddv
Trenton 16 0 0 57 264 576 924 989 885 753 399 121 12 4980
N.M. Albuquerque A 17 0 0 12 229 642 868 930 703 595 288 81 0 4348
Clayton A 0 6 66 310 699 899 986 812 747 429 183 21 5158
Raton A 2 9 28 126 431 825 1048 1116 904 834 543 301 63 6228
Roswell A 19 0 0 18 202 573 806 840 641 481 201 31 0 3793
Silver City A 18 0 0 6 183 525 729 791 605 518 261 87 0 3702
N.Y. Albany A 0 0 19 138 440 777 1194 1311 1156 992 564 239 45 6875
Albany C 5 0 9 102 375 699 1104 1218 1072 908 498 186 30 6201
Binghamton A 22 65 201 471 810 1184 1277 1154 1045 645 313 99 7286
Binghamton C 2 0 28 141 406 732 1107 1190 1081 949 543 229 45 6451
Buffalo A 6 19 37 141 440 777 1156 1256 1145 1039 645 329 78 7062
New York
(Cent. Pk.) C 15 0 0 30 233 540 902 986 885 760 408 118 9 4871
(LaGuardia) A 16 0 0 27 223 528 887 973 879 750 414 124 6 4811
(Kennedy) A 21 0 0 36 248 564 933 1029 935 815 480 167 12 5219
Rochester A 5 9 31 126 415 747 1125 1234 1123 1014 597 279 48 6748
Proced ures forDetermi ni ng Ther m alPerform ance
Schenectady C -1 0 22 123 422 756 1159 1283 1131 970 543 211 30 6650
Syracuse A 2 6 28 132 415 744 1153 1271 1140 1004 570 248 45 6756
N.C. Asheville C 17 0 0 48 245 555 775 784 683 592 273 87 0 4042
Cape Hatteras 0 0 0 78 273 521 580 518 440 177 25 0 2612
Charlotte A 22 0 0 6 124 438 691 691 582 481 156 22 0 3191
Greensboro A 17 0 0 33 192 513 778 784 672 552 234 47 0 3805
Raleigh A 20 0 0 21 164 450 716 725 616 487 180 34 0 3393
Wilmington A 27 0 0 0 74 291 521 546 462 357 96 0 0 2347
Winston-Salem A 17 0 0 21 171 483 747 753 652 524 207 37 0 3595
N.D. Bismarck A -19 34 28 222 577 1083 1463 1708 1442 1203 645 329 117 8851
Devils Lake C -19 40 53 273 642 1191 1634 1872 1579 1345 753 381 138 9901
Fargo A -17 28 37 219 574 1107 1569 1789 1520 1262 690 332 99 9226 os
Williston A -17 31 43 261 601 1122 1513 1758 1473 1262 681 357 141 9243
Table 43.3 (Continued)
971/2%
Winter
Design Yearly
State Station Temp.d July Aug. Sept. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May June Total
Ohio Akron-Canton A 6 0 9 96 381 726 1070 1138 1016 871 489 202 39 6037
Cincinnati C 12 0 0 39 208 558 862 915 790 642 294 96 6 4410
Cleveland A 7 9 25 105 384 738 1088 1159 1047 918 552 260 66 6351
Columbus A 7 0 6 84 347 714 1039 1088 949 809 426 171 27 5660
Columbus C 0 0 57 285 651 977 1032 902 760 396 136 15 5211
Dayton A 6 0 6 78 310 696 1045 1097 955 809 429 167 30 5622
Mansfield A 3 9 22 114 397 768 1110 1169 1042 924 543 245 60 6403
Sandusky C 8 0 6 66 313 684 1032 1107 991 868 495 198 36 5796
Toledo A 5 0 16 117 406 792 1138 1200 1056 924 543 242 60 6494
Youngstown A 6 6 19 120 412 771 1104 1169 1047 921 540 248 60 6417
Okla. Okla City A 15 0 0 15 164 498 766 869 664 527 189 34 0 3725
Tulsa A 16 0 0 18 158 522 787 893 683 539 213 47 0 3860
Ore. Astoria A 30 146 130 210 375 561 679 753 622 636 480 363 231 5186
Burns C 12 37 210 515 867 1113 1246 988 856 570 366 177 6957
Eugene A 26 34 34 129 366 585 719 803 627 589 426 279 135 4726
Meacham A 84 124 288 580 918 1091 1209 1005 983 726 527 339 7874
Medford A 23 0 0 78 372 678 871 918 697 642 432 242 78 5008
Pendleton A 10 0 0 111 350 711 884 1017 773 617 396 205 63 5127
Portland A 24 25 28 114 335 597 735 825 644 586 396 245 105 4635
Portland C 29 12 16 75 267 534 679 769 594 536 351 198 78 4109
Roseburg A 29 22 16 105 329 567 713 766 608 570 405 267 123 4491
A2 u3 Rios d o suo!reoudde
Salem A 25 37 31 111 338 594 729 822 647 611 417 273 144 4754
Allentown A 5 0 0 90 353 693 1045 1116 1002 849 471 167 24 5810
Pa.
Erie A 11 0 25 102 391 714 1063 1169 1081 973 585 288 60 6451
Harrisburg A 13 0 0 63 298 648 992 1045 907 766 396 124 12 5251
Philadelphia A 15 0 0 60 297 620 965 1016 889 747 392 118 40 5144
Philadelphia C 0 0 30 205 513 856 924 823 691 351 93 0 4486
Pittsburgh A 9 0 9 105 375 726 1063 1119 1002 874 480 195 39 5987
Pittsburgh C 11 0 0 60 291 615 930 983 885 763 390 124 12 5053
Reading C 9 0 0 54 257 597 939 1001 885 735 372 105 0 4945
Scranton A 6 0 19 132 434 762 1104 1156 1028 893 498 195 33 6254
Williamsport A 5 0 9 111 375 717 1073 1122 1002 856 468 177 24 5934
R.I. Block Island A 0 16 78 307 594 902 1020 955 877 612 344 99 5804
Providence A 10 0 16 96 372 660 1023 1110 988 868 534 236 51 5954
S.D. Huron A -12 9 12 165 508 1014 1432 1628 1355 1125 600 288 87 8223
Rapid City A -6 22 12 165 481 897 1172 1333 1145 1051 615 326 126 7345
Sioux Falls A -10 19 25 168 462 972 1361 1544 1285 1082 573 270 78 7839
Tenn. Bristol A 16 0 0 51 236 573 828 828 700 598 261 68 0 4143
Chattanooga A 19 0 0 18 143 468 698 722 577 453 150 25 0 3254
Knoxville A 17 0 0 30 171 489 725 732 613 493 198 43 0 3494
Memphis A 21 0 0 18 130 447 698 729 585 456 147 22 0 3232
Memphis C 0 0 12 102 396 648 710 568 434 129 16 0 3015
Nashville A 16 0 0 30 158 495 732 778 644 512 189 40 0 3578
Oak Ridge C 0 0 39 192 531 772 778 669 552 228 56 0 3817
Table 43.3 (Continued)
00
971/2%
Winter
Design Yearly
State Station Temp.d July Aug. Sept. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May June Total
Utah Milford A 0 0 99 443 867 1141 1252 988 822 519 279 87 6497
Salt Lake City A 9 0 0 81 419 849 1082 1172 910 763 459 233 84 6052
A2JOU3 RI GS J o SUOUPOil ddV
Wendover A 0 0 48 372 822 1091 1178 902 729 408 177 51 5778
Vt. Burlington A -7 28 65 207 539 891 1349 1513 1333 1187 714 353 90 8269
Va. Cape Henry C 0 0 0 112 360 645 694 633 536 246 53 0 3279
Lynchburg A 19 0 0 51 223 540 822 849 731 605 267 78 0 4166
Norfolk A 23 0 0 0 136 408 698 738 655 533 216 37 0 3421
Richmond A 18 0 0 36 214 495 784 815 703 546 219 53 0 3865
Roanoke A 18 0 0 51 229 549 825 834 722 614 261 65 0 4150
Wash. Olympia A 25 68 71 198 422 636 753 834 675 645 450 307 177 5236
Seattle-
Tacoma A 24 56 62 162 391 633 750 828 678 657 474 295 159 5145
Seattle C 32 50 47 129 329 543 657 738 599 577 396 242 117 4424
Spokane A 4 9 25 168 493 879 1082 1231 980 834 531 288 135 6655
Walla Walla C 16 0 0 87 310 681 843 986 745 589 342 177 45 4805
Yakima A 10 0 12 144 450 828 1039 1163 868 713 435 220 69 5941
W. Va. Charleston A 14 0 0 63 254 591 865 880 770 648 300 96 9 4476
Proced u res for Deter mi ni ng Th ermal P e rf orm an ce
Elkins A 5 9 25 135 400 729 992 1008 896 791 444 198 48 5675
Huntington A 14 0 0 63 257 585 856 880 764 636 294 99 12 4446
Parkersburg C 12 0 0 60 264 606 905 942 826 691 339 115 6 4754
Wisc. Green Bay A -7 28 50 174 484 924 1333 1494 1313 1141 654 335 99 8029
La Crosse A -8 12 19 153 437 924 1339 1504 1277 1070 540 245 69 7589
Madison A -5 25 40 174 474 930 1330 1473 1274 1113 618 310 102 7863
Milwaukee A -2 43 47 174 471 876 1252 1376 1193 1054 642 372 135 7635
Wyo. Casper A -5 6 16 192 524 942 1169 1290 1084 1020 657 381 129 7410
Cheynne A -2 28 37 219 543 909 1085 1212 1042 1026 702 428 150 7381
Lander A -12 6 19 204 555 1020 1299 1417 1145 1017 654 381 153 7870
Sheridan A -7 25 31 219 539 948 1200 1355 1154 1051 642 366 150 7680
Table 43.3 (Continued)
971/2%
Winter
Design Yearly
State Station Temp.d July Aug. Sept. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May June Total
Alta. Banff C 220 295 498 797 1185 1485 1624 1364 1237 855 589 402 10551
Calgary A -25 109 186 402 719 1110 1389 1575 1379 1268 798 477 291 9703
Edmonton A -26 74 180 411 738 1215 1603 1810 1520 1330 765 400 222 10268
Lethbridge A -24 56 112 318 611 1011 1277 1497 1291 1159 696 403 213 8644
B. C. Kamloops A -10 22 40 189 546 894 1138 1314 1057 818 462 217 102 6799
Prince
Georgee A -31 236 251 444 747 1110 1420 1612 1319 1122 747 468 279 9755
Prince Rupert C 15 273 248 339 539 708 868 936 808 812 648 493 357 7029
Vancouvere A 19 81 87 219 456 657 787 862 723 676 501 310 156 5515
Victoriae A 136 140 225 462 663 775 840 718 691 504 341 204 5699
Victoria 23 172 184 243 426 607 723 805 668 660 487 354 250 5579
Man. Brandone A -26 47 90 357 747 1290 1792 2034 1737 1476 837 431 198 11036
Churchill A -38 360 375 681 1082 1620 2248 2558 2277 2130 1569 1153 675 16728
The Pas C -32 59 127 429 831 1440 1981 2232 1853 1624 969 508 228 12281
Winnipeg A -25 38 71 322 683 1251 1757 2008 1719 1465 813 405 147 10679
N. B. Frederictone A -10 78 68 234 592 915 1392 1541 1379 1172 753 406 141 8671
Moncton C -7 62 105 276 611 891 1342 1482 1336 1194 789 468 171 8727
St. John C -7 109 102 246 527 807 1194 1370 1229 1097 756 490 249 8219
Nfld. Argentia A 260 167 294 564 750 1001 1159 1085 1091 879 707 483 8440
Corner Brook C -5 102 133 324 642 873 1194 1358 1283 1212 885 639 333 8978
Gander A -1 121 152 330 670 909 1231 1370 1266 1243 939 657 366 9254
Goosee A -25 130 205 444 843 1227 1745 1947 1689 1494 1074 741 348 11887
ABa au3 Rios Jo suoppoOdy
St. John'se A 6 186 180 342 651 831 1113 1262 1170 1187 927 710 432 8991
N.W.T. Aklavik C 273 459 807 1414 2064 2530 2632 2336 2282 1674 1063 483 18017
Fort Norman C 164 341 666 1234 1959 2474 2592 2209 2058 1386 732 294 16109
Resolution
C 843 831 900 1113 1311 1724 2021 1850 1817 1488 1181 942 16021
Island
4 58 51 180 457 710 1074 1213 1122 1030 742 487 237 7631
N. S. Halifax C
62 71 219 518 765 1113 1262 1206 1150 840 567 276 8049
Sydney A 5
9 102 115 225 471 696 1029 1156 1065 1004 726 493 258 7340
Yarmouth A
Cochrane C 96 180 405 760 1233 1776 1978 1701 1528 963 570 222 11412
Ont.
Fort. William A -23 90 133 366 694 1140 1597 1792 1557 1380 876 543 237 10405
Kapuskasing C -28 74 171 405 756 1245 1807 2037 1735 1562 978 580 222 11572
Kitchener C 1 16 59 177 505 855 1234 1342 1226 1101 663 322 66 7566
London A 3 12 43 159 477 837 1206 1305 1198 1066 648 332 66 7349
North Bay C -17 37 90 267 608 990 1507 1680 1463 1277 780 400 120 9219
C -13 25 81 222 567 936 1469 1624 1441 1231 708 341 90 8735
Ottawa
18 151 439 760 1111 1233 1119 1013 616 298 62 6827
Proced ures for Determini ng Th er mal Performan ce
Toronto C 1 7
P.E.I. Charlottetown C -3 40 53 198 518 804 1215 1380 1274 1169 813 496 204 8164
Summerside C -3 47 84 216 546 840 1246 1438 1291 1206 841 518 216 8488
C 102 136 327 682 1074 1659 1879 1619 1407 891 521 231 10528
Que. Arvida
A -10 9 43 165 521 882 1392 1566 1381 1175 684 316 69 8203
Montreale
C 16 28 165 496 864 1355 1510 1328 1138 657 288 54 7899
Montreal
A -13 56 84 273 636 996 1516 1665 1477 1296 819 428 126 9372
Quebece
C 40 68 243 592 972 1473 1612 1418 1228 780 400 111 8937
Quebec
Sask. Prince Albert A -35 81 136 414 797 1368 1872 2108 1763 1559 867 446 219 11630
Regina A -29 78 93 360 741 1284 1711 1965 1687 1473 804 409 201 10806
C -30 56 87 372 750 1302 1758 2006 1689 1463 798 403 186 10870
Saskatoon
Table 43.3 (Continued)
971/2%
Winter
Design Yearly
State Station Temp.d July Aug. Sept. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May June Total
Y.T. Dawson 164 326 645 1197 1875 2415 2561 2150 1838 1068 570 258 15067
Mayo Landing C 208 366 648 1135 1794 2325 2427 1992 1665 1020 580 294 14454
Latitude Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec.
20 2349 2671 3019 3301 3421 3445 3423 3332 3106 2763 2421 2246
25 2103 2747 2891 3266 3463 3524 3485 3329 3013 2588 2192 1995
30 1851 2260 2740 3206 3482 3581 3526 3303 2877 2395 1950 1735
35 1590 2030 2570 3124 3479 3619 3546 3254 2759 2184 1698 1468
40 1324 1788 2380 3019 3454 3637 3545 3183 2600 1958 1438 1149
Proced u res forDetermi ni n g Th ermal Performance
45 1056 1535 2172 2892 3409 3636 3525 3090 2421 1720 1174 931
50 791 1275 1948 2746 3346 3621 3489 2979 2225 1470 910 669
55 535 1011 1769 2582 3269 3596 3441 2856 2012 1212 651 422
60 299 747 1459 2403 3185 3571 3389 2709 1784 950 405 200
Table 43.5 Ratio of Monthly Average Daily Radiation on a Tilted Surface to That on a Horizontal Surface
Latitude Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec.
(R for Kt = 0.30)
(R for Kt = 0.40)
55 2.68 1.89 1.41 1.12 0.98 0.93 0.95 1.06 1.27 1.70 2.41 3.06
40 1.73 1.41 1.13 0.92 0.80 0.75 0.77 0.86 1.04 1.31 1.64 1.84
45 2.01 1.56 1.19 0.93 0.79 0.74 0.76 0.87 1.08 1.42 1.87 2.18
50 2.45 1.77 1.27 0.95 0.79 0.73 0.76 0.88 1.12 1.58 2.23 2.74
55 3.24 2.08 1.39 0.98 0.79 0.72 0.75 0.89 1.19 1.81 2.85 3.80
Vertical Surface
25 1.05 0.84 0.63 0.44 0.36 0.34 0.35 0.40 0.54 0.77 0.99 1.12
30 1.18 0.94 0.69 0.49 0.39 0.36 0.37 0.44 0.60 0.85 1.11 1.26
35 1.35 1.05 0.76 0.54 0.43 0.39 0.41 0.49 0.66 0.95 1.26 1.45
40 1.57 1.18 0.84 0.59 0.47 0.42 0.44 0.53 0.73 1.06 1.46 1.71
45 1.88 1.36 0.94 0.65 0.51 0.46 0.48 0.58 0.81 1.21 1.73 2.08
50 2.36 1.60 1.06 0.71 0.55 0.50 0.52 0.63 0.90 1.39 2.12 2.68
55 3.18 1.95 1.21 0.78 0.60 0.54 0.56 0.69 1.00 1.66 2.76 3.78
for Kt = 0.50
40 1.77 1.48 1.22 1.02 0.90 0.86 0.88 0.97 1.13 1.38 1.68 1.87
45 2.08 1.65 1.30 1.04 0.90 0.85 0.87 0.98 1.18 1.52 1.95 2.25
50 2.57 1.89 1.40 1.08 0.91 0.85 0.87 1.00 1.25 1.70 2.36 2.86
55 3.44 2.26 1.54 1.12 0.92 0.85 0.88 1.02 1.34 1.97 3.04 4.02
Vertical Surface
25 1.13 0.89 0.63 0.42 0.32 0.29 0.30 0.37 0.53 0.80 1.06 1.21
30 1.29 1.00 0.71 0.47 0.35 0.32 0.33 0.41 0.60 0.89 1.20 1.38
0.79 0.53 0.40 0.35 0.37 0.47 0.67 1.01 1.38 1.60
Proced ures for Determi ni ng Thermal Performance
35 1.48 1.13
40 1.74 1.29 0.89 0.59 0.44 0.39 0.41 0.52 0.75 1.14 1.61 1.91
45 2.11 1.50 1.00 0.66 0.49 0.44 0.46 0.56 0.84 1.31 1.92 2.34
50 2.67 1.78 1.14 0.73 0.54 0.48 0.51 0.64 0.95 1.54 2.39 3.04
55 3.64 2.19 1.32 0.81 0.60 0.53 0.56 0.71 1.08 1.84 3.15 4.34
R for Kt = 0.60
Latitude Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec.
Vertical Surface
25 1.20 0.92 0.63 0.39 0.28 0.25 0.26 0.34 0.53 0.82 1.12 1.28
30 1.37 1.04 0.72 0.46 0.32 0.28 0.30 0.39 0.60 0.93 1.28 1.48
35 1.59 1.19 0.81 0.52 0.37 0.32 0.34 0.45 0.68 1.06 1.48 1.73
40 1.88 1.37 0.92 0.59 0.42 0.37 0.39 0.51 0.77 1.21 1.73 2.07
45 2.30 1.61 1.05 0.66 0.48 0.42 0.44 0.58 0.87 1.40 2.09 2.56
50 2.93 1.93 1.21 0.75 0.54 0.47 0.50 0.65 0.99 1.65 2.61 3.34
55 4.01 2.39 1.41 0.84 0.60 0.52 0.55 0.72 1.13 2.00 3.46 4.80
R for Kt = 0.70
25 1.17 1.12 1.06 1.01 0.98 0.96 0.97 1.00 1.04 1.10 1.16 1.19
30 1.30 1.20 1.11 1.03 0.97 0.95 0.96 1.00 1.07 1.17 1.27 1.33
35 1.46 1.31 1.17 1.05 0.97 0.94 0.95 1.01 1.12 1.26 1.42 1.51
40 1.69 1.45 1.24 1.08 0.98 0.94 0.95 1.03 1.17 1.38 1.62 1.78
45 2.03 1.65 1.34 1.11 0.99 0.94 0.96 1.06 1.24 1.53 1.92 2.18
50 2.56 1.93 1.47 1.16 1.00 0.94 0.97 1.09 1.33 1.75 2.36 2.83
55 3.50 2.34 1.64 1.22 1.02 0.94 0.98 1.13 1.45 2.06 3.11 4.06
30 1.65 1.39 1.14 0.92 0.78 0.72 0.75 0.86 1.04 1.30 1.58 1.73
35 1.86 1.52 1.20 0.94 0.78 0.72 0.75 0.87 1.09 1.41 1.76 1.98
40 2.15 1.69 1.28 0.96 0.78 0.72 0.74 0.88 1.14 1.54 2.01 2.32
45 2.58 1.91 1.38 0.99 0.79 0.71 0.75 0.90 1.20 1.71 2.37 2.83
50 3.24 2.23 1.51 1.04 0.80 0.72 0.75 0.92 1.29 1.96 2.92 3.66
55 4.41 2.71 1.69 1.09 0.81 0.72 0.76 0.96 1.40 2.31 3.83 5.23
Vertical Surface
25 1.26 0.96 0.64 0.37 0.25 0.21 0.23 0.31 0.52 0.85 1.18 1.36
30 1.46 1.09 0.73 0.44 0.29 0.25 0.27 0.37 0.60 0.97 1.35 1.57
35 1.70 1.26 0.84 0.51 0.35 0.29 0.32 0.43 0.69 1.11 1.57 1.85
40 2.03 1.46 0.96 0.59 0.41 0.34 0.37 0.50 0.79 1.28 1.86 2.23
45 2.48 1.72 1.10 0.67 0.47 0.40 0.43 0.57 0.90 1.49 2.25 2.77
50 3.18 2.07 1.27 0.76 0.53 0.45 0.48 0.65 1.03 1.77 2.83 3.65
55 4.39 2.59 1.50 0.87 0.60 0.51 0.55 0.73 1.19 2.15 3.78 5.26
630 Applications of Solar Energy
PROJECT
Location Latitude
1 2 3 4 5
Actual PROJECT -
Tot. Mo. Solar Solar
Htg. Load System System Fraction/ en mo.,
L, Parameter Parameter mo. E
Month (Btu/mo.) DI D2 f (Btu/mo.)
Jan.
Feb.
March
April
May
June
Proced ures for Determi ni n gThermal Performan ce
July
Aug.
Sept.
Oct.
Nov.
Dec.
Etot
Lot = E tot Fannual =
FtOt
1. From Worksheet D.
2. From Worksheet C.
3. From Worksheet C.
4. From F chart.
5. E = f X L.
WORKSHEET C System Performance Parameters D1 , D2
1 2 13 4 5 6 PROJECT •
Tot. Monthly Total Mo. Av.
Radiation Heating Daytime Tot. Hrs.
on Tilt Surf., Load, Temp., in Mo.
S L to 212 — to A tune
Month (Btu/mo • ft2 ) (Btu/mo.) S/L D1 (°F) (°F) (h) K3 D2
Dec. 31
1. From ASHRAE Systems, Climatic Atlas, or Table 43.3, Sec. 43.5.
2. Qs = (PR)(24)(UA)(degree day) — X (degree day) =
where PF = 0.75 or more appropriate value.
3. (Vol/day) (no. days/mo.) (gay/day) (no. days/mo.)
4. May be constant or may vary.
5. Qw = (vol. of water) X 8.33 X 1 x (ts — tm).
6. L = Qs + Qw.
WORKSHEET E Total Monthly Solar Radiation Available, S
PROJECT
1 2 3 4 5 6
Extra- Ratio Monthly Avg. Tot. Monthly
Horizontal terrestrial Horizontal Ratio Daily Rad. Radiation
Insolation, Insolation, to Extra- Horizontal on Tilt Surf., No. of Days on Tilt Surf.,
Ill To terrestrial, to Tilt, IT in Month S
Month Btu/(d • ft2 ) Btu/(d • ft2 ) Kt R Btu/(d • ft 2 ) N Btu/(mo • ft 2 )
Jan. 31
Feb. 28
March 31
April 30
May 31
June 30
July 31
Aug. 31
Sept. 30
Oct. 31
Nov. 30
Dec. 31
1. From Fig. 43.28, Sec. 43.5 or known data.
2. From Table 43.4, Sec. 43.5, used only for Eq. (43.11).
3. From Eq. (43.11).
4. From Table 43.5, Sec. 43.5, latitude (0) - °,with collector tilt (0) ° , and latitude tilt -
S. From Eq. (43.12), IT =_(EH) (R).
6. From Eq. (43.13), S = (IT) (N).
A2aau3 Rios Jo suo!veo!i d dv
Procedures for Determining Thermal Performance 635
PROJECT:
Collector area, Ac
Collector volumetric flow rate
1
Case 2: FR(ra)n = b
1 + mAc i2Cc )
1
FR UL
1 + mAd2Cc
1
Case 3: FR ercOn = b
(1 + mAc /Cc
FRUL = m 1
1 + mA
lc /Cc
WORKSHEET F (Continued)
Collector loop heat exchanger modifier, FR/F R
For air systems and liquid systems without a collector/storage heat exchanger,
FR
FR
Capacitance rate:
Cc = (from above)
Cc
x—
ecsCmin
Ac(F R uL )
Cc
FIR
= from Fig. 43.10 or =
FR 1 + y(x — 1)
FR
F ( TO() = F R ( ra)n [ -71 (
(1120n FR
Fn U L = F RU (
A R/
FR
Procedures for Determining Thermal Performance 637
PROJECT•
Cc
K1 = (from Fig. 43.11) =
Ac
cp = (from Worksheet A)
(Mcds
K2 = (from Fig. 43.12) =
Ac
Cair loop = MC
Cmin = lesser of CH or C A
UA
K4 = (from Fig. 43.14) =
Cmin eL
638 Applications of Solar Energy
REFERENCES
BRUCE D. HUNN
Los Alamos Scientific Laboratory
Los Alamos, New Mexico
44.1 INTRODUCTION
The optimum collector array size for a solar space heating system in a given geographic
location and for given load characteristics is dependent on economic factors. To deter-
mine the future cost of fuel saved by the solar system, the system performance, in terms
of the fraction of the load met by solar (the solar contribution), must be ascertained.
As discussed in Chap. 29, the results of hourly computer simulations can be gen-
eralized by means of heat delivery performance correlations. These correlations can then
be used to develop simplified design and sizing methods that do not require the use of a
computer [1-9] . The most comprehensive of these simplified sizing methods is the F-
chart method described in the previous chapter and in Ref. 1.
Most of the simplified methods in use, however, have been developed for residential
buildings and do not explicitly take into account variation in design water temperature,
which is a significant parameter in commercial building space heating systems. An hourly
simulation program, developed by the Los Alamos Scientific Laboratory (LASL) [10]
639
640 Applications of Solar Energy
but not documented in detail, has been used to develop the "Monthly Solar Load Ratio"
(SLR) method, a simplified design and sizing procedure [3] . Detailed hour-by-hour
computer simulations have been run for 25 U.S. and Canadian cities for up to 12 years of
weather data for each city. On the basis of these simulations, a method has been devised
to allow the designer to estimate system performance based on monthly data of horizon-
tal solar radiation and heating degree-days. The results of this method agree reasonably
well with the hour-by-hour computer simulations. This SLR method has also been ex-
tended to nonresidential buildings for a range of design water temperatures [9] . Because
of its simplicity compared with the more detailed F-chart method, the SLR method
will be presented in detail here.
The basic difference in approach between the F-chart method and the SLR method
is that while the F-chart method includes the basic solar collector characteristics as
variables in the input parameters, the SLR method presents performance for reference
liquid and air systems. The liquid reference system characteristics are given in Table 44.1.
Thus, the system designer uses the results of parametric studies (see Chap 29), in which
each system and collector characteristic is varied singly, to determine performance of
systems differing from the reference systems.
Tables of sizing results have been developed only for the reference liquid system.
Although other researchers [11] have found that air and liquid system performance can
differ substantially, depending on the climate, systems characteristics, and the fraction
of the load met by solar, the LASL simulations [3] indicated very little difference
between the two reference system types. Thus, the same simplified method presented
here can be used for either system type. However, Ref. 12 recommends that air system
performance be approximated by adjusting the liquid system results using a small correc-
tion factor. The F-chart method may also be used for air system sizing (see previous
chapter).
It should be noted that the SLR method gives comparable results on an annual
basis when compared to the results of the F-chart method, the design method in most
widespread use [13] . Furthermore, in a comparison of several solar system computer-
simulation programs, the parent programs of the F-chart and SLR method were also
shown to be in substantial agreement, although the formulation of the programs is
somewhat different [14] .
Collector array sizing information is presented below, for the reference liquid system,
in two formats. The first is in the form of tables of the ratio of building heating load-
to-collector area required for selected solar fractions in the U.S. and Canadian cities
studied. Since 100 percent solar heating, though feasible in some locations, will rarely
be cost effective, results are presented for 25, 50, and 75 percent solar heating fractions
to cover the range most likely to be useful to the designer. A set of supplementary solar
load ratio maps, for the United States and southern Canada, is presented in Ref. 3. The
results enable one to determine collector sizes for the cities analyzed. The second format
presents the results as a universal performance curve for use in the SLR method, which
Sizing Active Solar Space Heating Systems 641
Solar Collector
1. Orientation Due south
2. Tilt (from horizontal) Latitude + 10°
3. Number of glazings 1
4. Glass transmissivity (at normal 0.86 (6% absorption, 8% reflection)a
incidence)
5. Surface absorptance (solar) 0.98
6. Surface emittance (IR) 0.89
7. Coolant flow capacity rate (rn Cp/mc2 ) 115 W/°C•rn (20 Btu/h.°F•ft2c = 0.04
gpm/ft2c )
8. Effective heat transfer coefficient 170 W/°C.r4 (30 Btu/11.°F•ftD
(UcAim2)
9. Back insulation U value (UbA/m2c) 0.5 W/ °C•m2 2)
, (0.083 Btu/h.°F-ft,
10. Heat capacity 20 k.1/°C•in (1 BturF.ftD
Collector Piping
11. Heat loss coefficient, to ambient 2 (0.04 Btu/h.°F•ft2c =
0.23 W/°C•mc
(UpA/m2c) U "-'• 0.3 Btu/h.°F•ft2 pipe or 1 in.
insulation of k= 0.075 Btu/IP°F•ft)
Thermal Storage
13. Heat capacity 305 kJJ°C•rn2 (15 BturF•fQ =
1.8 gal water/ft2c )
14. Heat loss coefficient (UsA/mD 0 W/°C.14 (0 Btu/h.°F-fQ)
(i.e., assuming all heat loss is to heated
space)
Controls
Residential building system and load profile assumed
Building maintained at 20°C (68°F)c
Collectors on when advantageous
aThese values apply for normal incidence on ordinary double-strength glass, 0.32 cm (1/8 in.). For
ether angles of incidence the Fresnel equation is used.
The coil and air circulation are sized to meet the building load with an outside temperature of -19°C
(-2°F) with 56°C (133°F) water and an air flow rate adequate to make up the space heat losses at an
air discharge temperature of 49°C (120°F). This corresponds to a finned-tube coil effectiveness of 80%.
cIt is assumed that internal sources such as people, lights, and equipment be sufficient to raise the
internal temperature by 2°C (4°F), i.e., up to the comfort standard of 22 C (72°F).
642 Applications of Solar Energy
enables the prediction of system performance for any city for which average monthly
horizontal solar radiation and heating degree-days are available.
Both performance prediction formats require the use of average monthly hori-
zontal insolation and monthly building load. In the tabulated data, the insolation is
built into the entry for each city; in the universal curve format, the monthly insolation
is entered explicitly as a parameter. Methods of determining monthly insolation are
discussed in Sec. 44.2.C. In both cases, the building load is generally entered in terms
of monthly degree-days; however, a number of alternative load determination tech-
niques are available. Note that for buildings with heating, ventilating, and air conditioning
(HVAC) systems, it is the load imposed on the heating coil that must be determined.
Several methods of computing building thermal loads are presented in Sec. 44.2.B.
The SLR method, in the form of universal performance curves, has been extended
to nonresidential buildings and for a range of design water temperatures in Sec. 44.4.
Because accurate monthly load determination procedures are not widely available, load
models at two basic levels of sophistication and accuracy are recommended, depending
on the design phase involved and the analytical tools available.
Steady-state models are based on daily or monthly average weather and insolation
data. Transient models, which account for hourly data variations, are more accurate, but
they require more detailed input for their use. Hourly calculations account for the varia-
Sizing Active Solar Space Heating Systems 643
tion in solar loading, varying occupant and equipment schedules, and the thermal capa-
city of the building and its contents, the effects of which are only averaged in daily or
monthly calculations. Because these effects can introduce time lags in the loads on the
heating system, they must be accounted for in the energy consumption calculation.
This is especially true if thermal storage tanks are available to redistribute the loads.
The two basic model levels are described below. Always it is the load imposed on
the heating coil that must be determined by these methods.
QL = (UA)(TB - TA ) (44.1)
where TB is the building temperature (reduced to account for internal heat generation
and called the base temperature) and TA is the ambient temperature.
Traditionally, for residental-type buildings, TB = 18°C (65°F) has been used.
Recent research, however, indicates that monthly average internal and solar heat gains
offset residential heat loss at a mean daily temperature slightly below 18°C (65°F) [15] .
Furthermore, for commercial buildings characterized by relatively large internal loads,
TB may be more like 10 to 16°C (50 to 60°F). Therefore, use of the modified degree-
day procedure described in Ref. 15 is recommended.
The equivalent heat loss coefficient, UA in Eq. (44.1), is the space heating loss
rate at design conditions, estimated by ASHRAE procedures [16] , divided by the design
temperature difference. Where ventilation is significant, UA should include the ventilation
load based on the design temperature difference, calculated by traditional methods.
The building heat loss parameter may also be expressed in kJ/°C•d•mg
(Btu/°F•d•ftB), where 43 (ftj3) represents the building floor area. Values in these units
for several types of residential construction are given in Ref. 17. Heating degree-days,
referenced to a base temperature of 18°C (65°F), are given by ASHRAE [15] or the
Climatic Atlas of the United States [18] .
For a small, single-story, well-insulated building, the heat loss parameter is gen-
erally in the range of 160 to 200 kJ/°C•d•m2 (8 to 10 Btu/°F•d•ft2 ) of floor area. For
an existing building, the load can be determined from past monthly and annual heating
644 Applications of Solar Energy
bills and degree-day values. Fuel consumption corrections should be made for furnace
efficiency (typically 0.6) and for nonspace-heating energy uses. The latter can be
estimated from summer fuel bills.
Bin Method. Heating loads can also be estimated using the bin method. The bin
(or bin-hour) method consists of making instantaneous energy calculations at many
different outdoor dry-bulb temperatures and multiplying the result by the number of
hours of occurrence of each temperature [15] . The bins are usually 2.8°C (5°F) in size,
and they are collected into three daily 8-h shifts. Because this method is based on hourly
weather data rather than daily averages, it is considerably more accurate than the degree-
day method. In addition, the bin method takes into account both occupied and un-
occupied building conditions, and gives credit for internal loads by adjustment of the
balance point.
Weather data for bin method use are given in Air Force Manual 88-29 [19] .
phase of this work produced a version, called Cal-ERDA/1.3 (now called DOE-2), which
operates on a CDC 6600/7600 computer [23] .
Other programs, some load and some both load and energy analysis programs, are
available privately. ASHRAE has published a bibliography of such programs [24] , for
example, AXCESS, ECUBE, ESP1, and TRACE. Because some are designed to compute
annual heating and cooling energy requirements, they may have to be modified to give
requirements on a monthly basis.
Solar radiation data measured at selected U.S. sites are available in hourly, monthly, and
annual forms. Hourly data are measured at some 30 locations, whereas the National
Oceanic and Atmospheric Administration (NOAA) measures data daily at about 130
locations. Using these data, the Climatic Atlas of the United States [18] presents maps
and tables of mean daily total solar radiation on a horizontal surface for each month of
the year, and provides annual totals.
For locations not represented by the available measured data, there are other
meteorological measurements and solar data that can be used to estimate radiation.
Data on solar angles, percent of possible sunshine, and cloud cover are examples.
These produce calculated values of radiation incident on surfaces oriented at various
tilt angles.
Of the several methods available for determining average monthly energy inci-
dent on tilted collectors, two are recommended here for use in the performance and
sizing procedures that follow.
Ho = 1.025Y - 8200
where
Ho (Btu/mo•ft2 )
Y=
cos(L - declination at mid-month)
Month 34 36 38 40 42 44 46
Where data for the cities discussed above do not adequately represent the location
in question, monthly radiation incident on a horizontal surface can be estimated from the
Climatic A tlas maps [18] of mean daily solar radiation on a monthly basis. These data
can be corrected for tilt angle as described above.
Using the techniques of simulation analysis described in Chap. 29, LASL has calculated
the performance of the reference systems for 25 cities where sufficient hourly weather
and solar data have been available. Using monthly climatic data, the results have been
extended for additional cities.
The results enable one to determine collector (and thus storage) sizes for the cities
analyzed. They are presented in terms of the ratio of building load to collector area for
liquid systems in Tables 44.3a (English units) and 44.3b (SI units). These tables present
the ratio in terms of the solar heating percentage of the total space heating requirement.
The load-collector ratio (LC) is defined as
An alternative form of this parameter, which is based on the design load rather than
degree-days, is also given:
To use the building load/collector area (LC) ratio for determining the size of the
collector for one of the cities listed (or an equivalent location), one must (a) determine
the building thermal load and (b) decide what percentage of solar heating the system is
to provide.
Massachusetts
Blue Hill 42 670 70 6,368 82 239 31 90 15 44
Boston 42 157 70 5,624 86 251 33 96 16 47
East Wareham 42 50 70 5,891 97 283 37 108 18 53
Michigan
East Lansing 43 878 70 6,909 76 222 28 82 13 38
Sault Ste. Marie 46 724 78 9,048 74 241 27 88 12 39
Minnesota
Saint Cloud 46 1,062 95 8,879 71 281 27 107 13 51
Missouri
Columbia 39 814 70 5,046 102 298 38 111 18 53
Montana
Glasgow 48 2,109 90 8,996 105 394 41 154 20 75
Great Falls 47 3,692 86 7,750 93 333 35 125 16 57
Nebraska
Lincoln 41 1,316 70 5,864 104 303 39 114 19 55
North Omaha 41 1,323 71 6,612 89 263 34 101 16 47
Table 44.3a Values of LC for Reference Liquid System for Selected Cities (English Units) (Continued) Cr't
km
0
LC
where solar provides 25%, 50%, 75%
of heating load
25% 50% 75%
Location Latitude Elevation Design Degree BTU/ BTU/ BTU/ BTU/ BTU/ BTU/
(United States) (°N) (ft) AT (°F) Days DD (ft2c ) h (ft2da DD (ftD h (ftDa DD (ftD h (ftDa
Nevada
Ely 39 6,279 72 7,733 119 357 47 141 23 69
Las Vegas 36 2,188 50 2,709 218 454 84 175 42 88
Reno 39 4,400 63 6,632 125 328 47 123 22 58
New Jersey
Seabrook 39 110 55 4,812 97 222 37 85 18 41
New Mexico
Albuquerque 35 5,327 60 4,348 161 403 64 160 31 78
Los Alamos 36 7,200 62 6,600 107 276 41 106 21 54
New York
Ithaca 42 951 70 6,914 68 198 24 70 11 32
New York 41 187 55 4,871 88 202 34 78 16 37
Sayville 41 56 50 4,811 98 204 38 79 18 38
Schenectady 43 490 71 6,650 63 186 24 71 11 33
North Carolina
Greensboro 36 914 60 3,805 128 320 50 125 24 60
Hatteras 35 27 48 2,612 204 408 79 158 39 78
Raleigh 36 440 55 3,393 133 305 52 119 25 57
North Dakota
Bismarck 47 1,677 89 8,851 78 289 29 108 14 52
Ohio
Cleveland 41 871 63 6,351 71 186 26 68 12 32
Columbus 40 760 63 5,211 77 202 29 76 13 34
A2Jau3 reps jo suoppo!i ddv
Tennessee
Nashville 36 614 60 3,578 117 293 44 110 21 53
Oak Ridge 36 940 58 3,817 111 268 42 102 20 48
Texas
Brownsville 26 48 30 600 517 646 218 273 110 138
El Paso 32 3,954 45 2,700 228 428 88 165 44 83
Fort Worth 33 574 50 2,405 186 388 73 152 37 77
Midland 32 2,885 55 2,591 202 463 79 181 39 89
San Antonio 30 818 40 1,546 262 437 103 172 52 87
Utah
Salt Lake City 41 4,238 70 6,052 107 312 40 117 19 55
Flaming Gorge 41 6,273 70 6,929 111 324 43 125 21 61
Vermont
Burlington 44 385 77 8,269 63 202 24 77 11 35
Virginia
Sterling 39 276 57 4,224 111 264 43 102 21 50
Washington
Prosser 46 840 55 4,805 117 268 41 94 18 41
Table 44.3a Values of LC for Reference Liquid System for Selected Cities (English Units) (Continued)
LC
where solar provides 25%, 50%, 75%
of heating load
25% 50% 75%
Location Latitude Elevation Design Degree BTU/ BTU/ BTU/ BTU/ BTU/ BTU/
(United States) (°N) (ft) AT (°F) Days DD (ftD h (ftDa DD (ftD h (ftDa DD (ft) h (ft2c)a
Arizona
Page 37 1,304 28 3,684 2,615 857 981 321 470 154
Phoenix 33 347 25 981 6,129 1,773 2,411 696 1,205 350
Tucson 32 683 21 1,000 6,149 1,503 2,411 589 1,205 293
Arkansas
Little Rock 35 84 31 1,788 2,574 910 981 347 490 173
Si zi ngActive Solar SpaceHeati n g Syste ms
California
Davis 39 15 22 1,390 4,045 1,040 1,471 378 674 173
El Centro 33 4 19 810 11,175 2,514 4,209 945 1,982 444
Fresno 37 102 25 1,384 3,984 1,153 1,430 413 654 189
Inyokern 36 666 23 1,960 4,740 1,279 1,798 485 858 233
Los Angeles 34 165 17 1,145 8,499 1,638 3,208 617 1,532 296
Riverside 34 320 22 1,002 7,988 2,054 3,105 797 1,512 387
Santa Maria 35 88 20 1,648 7,212 1,670 2,901 671 1,369 318
Colorado
Grand Junction 39 1,473 39 3,134 2,431 1,093 940 422 449 202
Granby 40 2,542 42 3,069 2,431 1,172 960 463 470 227
Florida
Apalachicola 30 14 21 727 6,619 1,616 2,635 643 1,328 324
Tallahassee 30 20 25 828 5,782 1,673 2,309 668 1,165 337
Georgia
Atlanta 34 310 31 1,645 3,146 1,112 1,205 425 592 208
Griffin 33 305 27 1,187 4,433 1,367 1,716 529 858 265 of
Table 44.3b Values of LC for Reference Liquid System for Selected Cities (SI Units) (Continued)
LC
where solar provides 25%, 50%, 75%,
of heating load
25% 50% 75%
Location Latitude Elevation Design Degree kJ/ kJ/ kJ/
(United States) (°N) AT (°C) Days wi(mc2 ) a wi(mc2)a wi(nac2 ) a
(m) °C•cl.m °C•d*M2c °C.c1.-4
Idaho
Boise 44 882 33 3,227 2,206 851 797 309 347 135
Illinois
ANL, Lemont 42 229 39 3,419 1,614 725 613 277 286 129
Indiana
Indianapolis 40 250 36 3,166 1,757 721 654 268 306 126
Kansas
Dodge City 38 800 35 2,770 2,574 1,043 1,001 406 490 198
Louisiana
Lake Charles 30 18 21 811 4,985 1,184 1,961 466 981 233
Shreveport 32 67 24 1,213 3,657 1,033 1,430 403 715 202
Maine
Caribou 47 195 47 5,426 1,389 750 531 287 245 132
Maryland
Silver Hill 39 89 30 2,347 2,268 788 878 306 429 148
Massachusetts
Blue Hill 42 204 39 3,538 1,675 753 633 284 306 139
Boston 42 48 39 3,124 1,757 791 674 302 327 148
East Wareham 42 15 39 3,273 1,982 891 756 340 368 167
Michigan
East Lansing 43 268 39 3,838 1,553 699 572 258 266 120
Sault Ste. Marie 46 221 43 5,027 1,512 759 552 277 245 123
Minnesota
ASJau3 Ri os Jo suopEn ddv
Saint Cloud 46 324 53 4,933 1,451 885 552 337 266 161
Missouri
Columbia 39 248 39 2,803 2,084 939 776 350 368 167
Montana
Glasgow 48 643 50 4,998 2,145 1,241 838 485 409 236
Great Falls 47 1,125 48 4,306 1,900 1,049 715 394 327 180
Nebraska
Lincoln 41 401 39 3,258 2,125 954 797 359 388 173
North Omaha 41 403 39 3,673 1,818 828 695 318 327 148
Nevada
Ely 39 1,914 40 4,296 2,431 1,125 960 444 470 217
Las Vegas 36 667 28 1,505 4,454 1,430 1,716 551 858 277
Reno 39 1,341 35 3,684 2,554 1,033 960 387 449 183
New Jersey
Seabrook 39 34 31 2,673 1,982 699 756 268 368 129
New Mexico
Sizi ng ActiveSolar SpaceHeati n g Systems
LC
where solar provides 25%, 50%, 75%,
of heating load
25% 50% 75%
Location Latitude Elevation Design Degree kJ/ kJ/ kJ/
(United States) (°N) (m) AT (°C) Days °Cd•m2, W/(m2)a °C•d•m2, wi(nc2 ) a
° C•c1.4 W1(111) a
Oklahoma
Stillwater 36 277 32 2,069 2,697 989 1,062 391 511 186
Oklahoma City 35 401 33 2,069 2,738 1,055 1,083 419 531 205
Oregon
Astoria 46 7 22 2,881 2,595 668 919 236 388 101
Corvallis 45 72 23 2,626 2,452 662 858 233 368 101
Medford 42 403 31 2,782 2,186 772 776 274 327 117
Pennsylvania
State College 41 375 36 3,297 1,594 655 592 243 286 117
Rhode Island
Newport 41 15 33 3,224 1,982 750 756 287 368 139
South Carolina
Charleston 33 21 22 1,129 4,290 1,103 1,675 432 838 214
South Dakota
Rapid City 44 969 42 4,081 1,982 967 756 369 368 180
Tennessee
Nashville 36 187 33 1,988 2,390 923 899 347 429 167
Oak Ridge 36 286 32 2,121 2,268 844 858 321 409 151
Texas
Brownsville 26 15 17 333 10,562 2,035 4,454 687 2,247 435
El Paso 32 1,205 25 1,500 4,658 1,348 1,798 520 899 261
Fort Worth 33 175 28 1,336 3,800 1,222 1,491 479 756 243
Midland 32 879 31 1,439 4,127 1,458 1,614 570 797 280
ASJou] m os Jo suopEn ddy
San Antonio 30 249 22 859 5,353 1,377 2,104 542 1,062 274
Utah
Salt Lake City 41 1,292 39 3,362 2,186 983 817 369 388 173
Flaming Gorge 41 1,912 39 3,849 2,268 1,021 878 394 429 192
Vermont
Burlington 44 117 43 4,594 1,287 636 490 243 225 110
Virginia
Sterling 39 84 32 2,347 2,268 832 878 321 429 158
Washington
Prosser 46 256 31 2,669 2,390 844 838 296 368 129
Pullman 47 787 36 3,079 2,043 854 735 309 327 135
Richland 47 223 31 3,301 2,043 721 715 252 306 107
Seattle 48 34 31 2,658 1,920 677 674 239 266 95
Spokane 48 718 39 3,697 1,430 828 633 284 286 129
Wisconsin
Madison 43 271 44 4,368 1,553 797 572 293 266 135
Sizi ngActive Solar Space Heati n gSystems
Wyoming
Lander 43 1,699 46 4,372 2,206 1,162 858 454 1,287 227
Laramie 41 2,207 40 4,101 2,166 1,002 858 397 1,287 198
Canada
Toronto 44 135 39 5,905 1,471 662 552 249 266 120
Winnipeg 50 250 56 3,793 1,287 828 470 302 225 145
a At design AT.
658 Applications of Solar Energy
For locations not covered by the cities given in Table 44.3, the SLR method can be
used. It is based on correlations developed by many hour-by-hour computer simulations
for a variety of locations. As mentioned earlier, this method had been shown to give com-
parable results on an annual basis to the more widely used F-chart method.
This method is a five-step process that can be performed on hand calculators with
exponential functions. The steps are as follows:
Step 1. Obtain the following data for each month of the year for the site where
the solar heating system is to be located:
a. Heating degree days, per month;
b. Total solar radiation on a horizontal surface, per month.
Sizing Active Solar Space Heating Systems 659
Step 2. Correct the solar radiation to a collector for tilt angle of latitude plus
10°, using the technique described in Section 44.2. The resulting values
are the monthly solar radiation on a plane surface tilted at an angle equal
to the latitude plus 10° and oriented due south.
Step 3. Determine the building thermal load in units of IcI/°C.d or Btu/°F.d.
This is the total heat required by the building per day for a 1° difference
between the inside and outside temperatures.
Step 4. Determine the solar load ratio (SLR) for each month from the following
formula:
SLR = solar collector area X total radiation per unit area on tilted surface
building thermal load X heating degree-days per month
The solar load ratio is dimensionless. It is the ratio of the total solar energy
incident on the collectors to the total energy required to heat the building.
Step 5. The annual solar heating fraction can then be estimated from the following
formula:
12
(degree-days)(X)
Annual solar heating fraction = mo.= 1
12
E (degree-days)
mo.= 1
1.0
Heat from Solar (X )
08
0.6
Fraction ofTota l
0.2
X=1.06 - 1366e -0.55SLR
+ 0.306e-I.05SLR
00 5
2 3 4
Fig. 44.1 Universal performance curve for solar heating systems: SLR method.
660 Applications of Solar Energy
where
The function X is shown plotted in Fig. 44.1. Note that X is not a very good esti-
mate of the monthly solar heating fraction because it is designed to compensate (on an
annual basis) for the fact that the heating load based on degree-days per month is lower
than the heating load based on hour-by-hour calculations.
The key to the accuracy of the method is the determination of the function X used
in step 5. This function has been carefully determined so that the resulting error in pre-
dicting the solar heating fraction (compared to the simulation result) will be minimized
without using a different function for each locality.
The determination of X is based on hour-by-hour computer simulations of a "test
year" for 25 locations for five different collector sizes in each location. The standard
deviation of the prediction error is 4.4 percent for solar heating. Predicted versus simu-
lated results for the 125 cases studied are given in Ref. 12.
Using hour-by-hour simulations, LASL has extended the SLR sizing procedure described
above to nonresidential (commercial) buildings [9] . Liquid solar heating systems were
considered for combined space and building service hot water. Commercial building space
and BSHW loads were simulated for the reference liquid system (Table 44.1), incor-
porating a heating-coil supply temperature governed by an ambient air schedule.
The building thermal loads were modeled for a "typical" three-story office building
with offices on the north and south sides, a corridor in between, and an unheated stair-
well at each end. The building envelope heat loss coefficient, UA, was estimated by taking
the actual annual envelope heat losses generated by an hour-by-hour analysis of the build-
ing and normalizing them to a per-degree-day-per-building-area basis. Thus, the dynamics
of the building envelope, with respect to the envelope losses, were partially accounted
for. A single-zone model was used in which the load was characterized by three param-
eters: one representing the envelope and ventilation loads that are degree-day dependent;
one, the BSHW load and its appropriate profile of occurrence; and one, the internal
generation load with its profile. The internal generation load was taken as a credit against
heating load at each hour. These three load components were varied independently over
a range expected in most office buildings in a simulation using 1974 hourly weather data
for Los Alamos, New Mexico. The result was a performance characterized by virtually
identical curves of fraction of load met by solar energy versus SLR (see Sec. 44.3.C),
regardless of load characteristics.
Although performance was found to be essentially independent of load charac-
teristics, two system parameters were significant: the average temperature of water
supplied to the heating coil and the collector characteristics. A family of curves was
Sizing Active Solar Space Heating Systems 661
B. The Simulations
The typical office building was modeled for eight U.S. cities: Bismarck, North Dakota;
Fresno, California; Los Alamos, New Mexico; Madison, Wisconsin; Medford, Oregon;
Miami, Florida; Nashville, Tennessee; and New York, New York. The weather data year
used for each city was chosen from 10-year records between 1950 and 1962, except
for Los Alamos, where 1974 data were used. The specific year was that in which the
solar performance was nearest the average for the 10-year period.
The results of these simulations are a set of universal sizing curves. All are ex-
pressed as the fraction of demand met by storage versus the SLR on a monthly basis.
For given collector and system characteristics, the simulation data from the eight cities
all fall very close to the appropriate universal curve.
The effects of the parameter variations shown in Chap. 29 are expected to give a
reasonable approximation of commercial building performance in all climatic zones.
As described above, the typical office building was simulated using a single-zone
model. Hourly heating loads were based on the following nominal values of the load
parameters.
TWO
Q Aux
Storage Tank
TWO
® Pump
TWI
(°C) (°F)
70- 160
lW = 54°C (130°F) ( I ) Legend
150
TS = Storage Temperature
60- 1440 TW1= 52°C(125°F)(2) TWI = Coil Inlet Temperature
TWO=Cod Outlet Temperature
TA = Ambient Temperature
130
TWI= Mean Coil Inlet Temperature
over Operating Range
50-
120
110 - _
TWI =46°C (115°F)(4
40-
100
TWI= 49°c (120°F)(3)
90
30-
80
20- 700 10 20 30 40 50 60
70 (°F) TA
-20 -10 0 10 20 (°C)
The heating coil model used is shown in Fig. 44.2. Solar energy supplies the load
when the storage tank temperature is high enough; otherwise, auxiliary energy brings the
temperature to the required heating-coil supply temperature. Four heating-coil supply
temperature schedules varying with ambient air temperature, Fig. 44.2, were run for a
base case, nonselective, double-glazed collector. Because the heating-coil supply tem-
peratures range from 66 to 27°C (150 to 80°F), the nominal collector operating tem-
perature was 52 to 60°C (125 to 140°F). A solar storage tank capacity of 73 kg
water/mc2 (15 lb/ftc2 = 1.8 gal/ftc2) was assumed. Because of the large collector area avail-
able, all BSHW loads were assumed to be met by solar heating.
Sizing Active Solar Space Heating Systems 663
where a l ,b1 , and c1 are constants for a given curve. Here f is the fraction of demand met
by the solar system, and the ratio of the available solar radiation to the building load
(SLR) is defined by
HAc
SLR = (44.2)
QL
where Ho = monthly insolation per unit area on a collector tilted at angle from
horizontal, Ac = collector area, and QL = monthly thermal load.
Using values of QL and Ho obtained as specified in Sec. 44.2,B, one can compute
the SLR, and Eq. (44.2) reduces to f as a function of collector area, Ac, alone. This
1.0
..-
0.9 ..-
0.8
8 0.7 / Htg Coil Sched
/
/ a TWI = 46°C (I15°F)
cR /
0.6 b--- TWI = 49°C (I20°F)
0 / TWI = 52°C (125°F)
c
0 0.5 . d---- TMV = 54°C (130°F)
0
LI: 0.4 /
Weather - eight cities
0.3 / Collector - two glazinqs, nonselective
Storage size =73.3 I waterhq(15 lb/fq)
0.2
0.1
0
2 3 4 5 6 7 8 9 10
Monthly Solar Load Ratio (SLR)
Curve-fit equations
0.0 <_ SLR <1.2 1.2 <--SLR < 10.0
a y= .362 x y= I-1.173e -0.609x
b y = .349x yr.1.1.159e - 0.575x
c y= .334x y=1-1.146e -0541x
d y= .318x y=1-1.132e-o.504x
Fig. 44.3 Universal performance curves for commercial building space and building service
hot water heating at various design temperatures.
664 Applications of Solar Energy
1.0
...
0.8
•/..•
Collector
cn 0.6 7 // a--- One glazing , selective
I
b-Two g lazings, nonselective
0
0 • /
I c -- One glazing, nonselective
0 I
li 0.4 / Weather - eight cities
/ Htg Coil Sched = 4 (TWI= 46°C (I15°F))
• / 2 2
_c / S orage size = 73.31 water/mc (15 lb/ftc )
C /
0 0.2 / +
i
I
0 7 8 10
3 4 5 6 9
Monthly Solar Load Ratio (SLR)
Curve-fit equations
0.0<SLR< 1.2 1.2 <_ SLR 5_ 10.0
a y=.394x y =1-1.198e -0.685x
b y= .362x y=1-1.173e-0.609x
c y=.304x y=1-1.182e- 0.518"
Fig. 44.4 Universal performance curves for commercial building space and building service
hot water heating by three collector types.
equation can then be used for collector sizing and economic feasibility calculations.
Although the curves were generated assuming a tilt of lat +5°, the insolation on a collec-
tor tilted at lat + 10° is entirely adequate.
Figure 44.3 shows the resulting universal curve for combined space and water
heating, given the four heating-coil schedules and a nonselective, double-glazed collector.
Note that because collectors are more efficient at lower temperatures, operating at lower
heating-coil temperatures improves performance significantly. Thus, heating coils should
be designed to operate at 49 to 60°C (120 to 140°F), rather than the higher temperature
used in conventional systems.
Figure 44.4 shows performance of three generic flat-plate collector types: single-
glazed, nonselective surface; double-glazed, nonselective surface; and single-glazed,
selective surface. These curves apply for a heating coil supply schedule characterized by
46°C (115° F) average supply temperature. The heating-coil supply temperature schedule
and collector types that represent the application under consideration dictate which curve
to use.
Sizing Active Solar Space Heating Systems 665
REFERENCES
13. Unpublished minutes of the Solar Heating and Cooling System Simulation and
Economic Analysis Working Group Meeting, Sponsored by the U.S. Department of
Energy, Office of Conservation and Solar Applications, Solar Heating and Cooling
R&D Branch, Golden, Colorado, November 14-15, 1977.
14. T. L. Freeman, M. W. Maybaum, and S. Chandra, "A Comparison of Four Solar
Simulation Programs in Solving a Solar Heating Problem," Proceedings of the Con-
ference on Systems Simulation and Economic Analysis for Solar Heating and Cooling,
San Diego, California, June 27-29, 1978.
15. ASHRAE Handbook and Product Directory, Systems, ASHRAE, 1976.
16. ASHRAE Handbook of Fundamentals, ASHRAE, 1977.
17. E. J. Beck and R. L. Field, Solar Heating of Buildings and Domestic Hot Water,
U. S. Navy Civil Engineering Laboratory, Port Hueneme, California, Technical
Report R835, January 1975.
18. U.S. Department of Commerce, Environmental Science Services Administration,
Environmental Data Service, Climatic Atlas of the United States, 1968.
19. Air Force Manual (AFM) 88-29, Engineering Weather Data, Pepartments of the
Air Force, Army, and Navy, July 1978.
20. W. W. Treichler, "A Non-Computerized Rational Energy Analysis Procedure,"
Proceedings of Conference on Improving Efficiency and Performance of HVAC
Equipment and Systems for Commercial and Industrial Buildings,
School of Engineering, Purdue University (Purdue Research Foundation, 1976),
Vol. 1, pp. 214-224, April 12-14, 1976.
21. T. Kusuda, NBSLD, Computer Program for Heating and Cooling Loads in Buildings,
National Bureau of Standards Report NBSIR 74-574, November 1974.
22. NECAP—NASA's Energy-Cost Analysis Program, Part I, User's Manual, and Part
II, Engineering Manual, R. H. Henninger, ed., NASA CR-2590, Parts I and II,
September 1975.
23. M. Lokmanhekim, Z. Cumali, F. C. Winkelmann, A. H. Rosenfeld, G. S. Leighton,
and H. D. Ross, "Cal-ERDA, A New State-of-the-Art Computer Program for the
Energy Utilization Analysis of Buildings," paper presented at the Third Inter-
national Symposium on the Use of Computers for Environmental Engineering
Related to Buildings, Banff, Alberta, Canada, May 10-12, 1978.
24. ASHRAE, Bibliography on Available Computer Programs in the General Area of
Heating, Refrigeration, Air Conditioning, and Ventilation, Research Project Report
GRP 153, October 1975.
25. ASHRAE Handbook and Product Directory, Applications, ASHRAE, 1978.
26. S. A. Klein, W. A. Beckman, and J. A. Duffie, "A Design Procedure for Solar
Heating Systems," paper presented at the 1975 International Solar Energy Society
Meeting, Los Angeles, California, 1975.
unit 8
NONTECHNICAL ISSUES
Glossary of Some Economic Terms
Used in Chapters 45 and 46
667
668 Nontechnical Issues
Internal rate of return The rate of return that equates the benefits (fuel savings)
derived from a solar investment with the overall costs of
the investment.
Zero year The year during which the solar system is constructed.
The day the system is "turned on" is defined as the
first day of the first year of operation.
Sensitivity analysis A method of evaluating a proposed solar investment as
a function of a change in the various economic and per-
formance variables, which all contain some uncertainty.
Capital recovery factor The constant fraction of the initial solar investment that
must be accumulated annually to amortize the invest-
ment over the expected system lifetime.
Before-tax cost of energy The total amount of annual revenue required, including
taxes and return on investment, to purchase energy.
After-tax cost of energy The net reduction in after-tax cash flow resulting from
the purchase of energy.
Levelized cost (or Obtained by converting a series of unequal annual costs
annualized cost) to an equivalent present value and then multiplying by
the capital recovery factor (for a given discount rate and
system lifetime) to give a series of uniform annual costs.
Life-cycle costing A method of evaluating a capital investment that takes
into account the sum total of all costs associated with
the investment over the lifetime of the project.
OMRI Annual operation, maintenance, component replacement,
and insurance premium costs for a solar system.
Salvage value (residual Value of the solar system hardware at the end of its
value) economic life. Equal to its scrap value less the cost
incurred for its removal and disposal.
chapter 45
WILLIAM C. DICKINSON
Lawrence Livermore Laboratory
Livermore, California
451 INTRODUCTION
This chapter is aimed primarily at the homeowner or the contractor/builder who wishes
to determine the economic effectiveness of installing a solar water heating or space
heating/water heating system on the roof of a house. It will be assumed that he or she
does not have access to a computer, has little or no knowledge of the intricacies of solar
collectors, and is not acquainted with standard life-cycle costing techniques involving
discount factors, present value, rates of return, and so on. What is wanted is a reasonably
accurate answer to the question, "Does the purchase and installation of a solar heating
system for this house in this location constitute a sound investment?"
The standard method of evaluating a capital investment that results in both costs and
benefits (savings) over a number of years is the life-cycle costing method [1] . The deter-
mination of the life-cycle costs of a solar system must include consideration of (a) system
acquisition costs, including purchase prices, delivery costs, and installation costs; (b)
669
670 Nontechnical Issues
system repair and replacement costs; (c) operating costs; (d) maintenance costs; (e) in-
surance; (f) taxes, and (g) salvage values, net of removal and disposal costs. The principal
subsystems for which costs are required are (a) solar collectors, (b) thermal storage,
(c) domestic hot water system, (d) auxiliary energy subsystems, (e) heating and cooling
distribution subsystems, and (f) the control subsystem. Motors, pumps, fans, blowers,
wiring, and piping are included in these subsystems.
A life-cycle cost comparison needs to be made between a conventional, fossil fuel
or electric, heating system and the alternative solar heating/hot water system which will
include a conventional backup system [1] . Since it is the cost differences between these
two systems that are important, those costs that are identical for the two systems can be
omitted in the comparison. For example, the backup heating system to be used with the
solar system (see Chap. 29 for a discussion of the need of a conventional backup system)
would normally be identical to the heating system that would be used in the absence of
a solar system. Hence the first cost of this system can be omitted in the life-cycle cost
comparison.
The homeowner is generally most interested in the monthly costs of the heating system
as reflected by the monthly utility bills. Therefore, a comparison of the monthly costs of
a solar/conventional heating system with those of a conventional system alone is an
appropriate form of analysis. Actually, economic analyses are usually made on an annual
rather than on a monthly basis. We shall, therefore, be concerned with a comparison of
annual costs between the two options.
We shall present a simplified but accurate tabular method, based on life-cycle
costing, to allow the homeowner to calculate the net costs or net savings each year that
will result from the installation of the solar system [2] . In this way the homeowner can
determine the payback period, that is, the number of years required for the total net
savings to equal the total net costs.
Table 45.1 Annual Sum of Monthly Loan Payments for a $1000 Loan
Term
Interest Rate (%) 5 Years 10 Years 15 Years
8 243 146 115
10 254 159 129
12 267 172 144
14 279 186 160
16 292 201 176
Solar Hot Water and Space Heating Systems 671
Line 1. Design costs should include cost of architectural firm that prepares work-
ing drawings, selects components and installer, and supervises installation.
Line 2. Permits and approval costs include cost of building permits and any
needed structural calculations required to satisfy requirements for permits.
Line 3. Installed system costs should include cost of collectors, and piping, any
modifications required to house, and contractor costs for installation.
Line 4. This is the initial cost of system if no money is borrowed.
Line 5. This is the amount of bank loan taken.
Line 6. This is the actual initial amount of cash required by the homeowner.
Line 7. Loan payments. It is assumed that the loan will be repayed in equal
monthly payments at a particular annual interest rate. In Table 45.1 the
annual sum of these monthly payments is given for three different loan
periods and five different possible interest rates for a $1000 loan.
Line 8.* State income tax credit. Many states are now allowing a direct credit
against state income tax for the installation of a solar system. If such a
credit is allowed, it should be calculated and entered here. If the credit
exceeds the tax, some states allow it to be carried over to succeeding
years; others allow only a one-time credit no larger than the tax.
Note: It will be assumed, for convenience, that deductions are not itemized in
the homeowner's federal and state income tax returns. If, in fact, deductions are item-
ized, then there will be an additional annual savings because interest payments on the
solar loan are deductible. Since the interest payments taper off over the loan period,
*At the time of preparing this chapter there was no federal income tax credit provided. In filling
out Worksheet 1, the appropriate federal tax credit should be added to the state tax credit.
672 Nontechnical Issues
this savings will also taper off. For a particular income tax rate, an estimate of the
average annual tax savings can be made and subtracted from line 7 to determine the net
loan payment.
Operating and maintenance costs are difficult to predict. Operating costs will consist
principally of electric power required to operate system pumps, fans, blowers, and so
on. Most likely, maintenance costs (including repairs and component replacement) will
increase as the system ages. However, a reasonable estimate for average operation and
maintenance costs is 1 to 2 percent of installed system cost. If, as a result of installing
the solar system, an increased assessment of property value results, then the additional
annual property tax should be estimated.* Similarly, if an increase in home insurance
premiums is expected, add the estimated increased premium.
The effect of inflation over the system lifetime needs to be taken into account. If
all costs and savings could be expected to escalate from year to year at the same rate,
this factor could be ignored and final annual costs or savings could be calculated in
dollars of today's purchasing power ("constant dollars"). However, the loan repayments
are of fixed numerical amount each year, independent of any inflation factor. Further,
it can be expected that fossil fuel costs will inflate more rapidly than the average general
inflation rate. Therefore, in order to end up with annual costs or savings in constant
dollars, it is necessary to include estimated inflation rates. In Worksheet 2 we assume that
an average general inflation rate of 6 percent will apply to these costs.
*Legislation now exists in several states exempting solar systems from property tax.
Solar Hot Water and Space Heating Systems 673
fuel to useful heat. Typically, electric resistance heaters have a high conversion efficiency,
say, 95 percent, whereas gas and oil furnaces have average conversion efficiencies of only
about 60 percent. Using these values, the following conversion factors result:
On the basis of current unit fuel prices (these can be determined from a recent
utility bill), multiply the number of fuel units saved by the unit price and enter under
each year in the second column. If more than one type of fuel is used—for example,
electricity for water heating and gas for space heating—the appropriate conversion factor
and unit price must be applied for each use before determining total dollar fuel savings.
The rate at which fuel prices will escalate in future years can only be guessed. This
is one of the major unknowns in attempting to calculate the economic attractiveness of
installing a solar system. The more rapidly fuel prices increase, the greater the accumu-
lated fuel savings resulting from the solar system and the more attractive the economics
become. In Worksheet 3 we provide two sets of fuel escalation factors to use, 6 percent
and 10 percent, so that the effect of this factor on system economics can be determined.
This is the final worksheet, which will allow the calculation of net annual costs or savings
resulting from the installation of the solar system. These costs or savings are to be con-
verted from future dollars back to constant dollars by multiplication by the 6 percent
general inflation factor (actually the reciprocal of the factor used in Worksheet 2).
Note that if no loan was required to purchase the system and also the fuel escalation rate
was assumed to be 6 percent, the same as the general inflation rate, then the effect of
inflation could be ignored and the entire calculation performed in constant dollars.
It is suggested that the final calculation of costs and savings, using Worksheet 4, be
done twice, first using a fuel inflation factor of 6 percent and then using the higher
10 percent rate. It is generally thought that the cost of fossil fuels will escalate over the
next 20 years at a higher average rate than that of general inflation. Therefore, the 10
percent fuel escalation rate will probably be closer to reality. In fact, natural gas will
probably escalate at an even higher rate over the next few years as price regulations are
removed.
When Worksheet 4 has been completed with a net cost or savings listed for each
year in the right-hand column, it is a simple matter to determine the payback period (or
breakeven period) for the system. Add the net cost (-) to the net savings (+) for each
674 Nontechnical Issues
year until the sum is approximately zero. The payback period is the year at which this
occurs. Of course, the savings in subsequent years will continue. They can be thought of
as the net savings (expressed in dollars of today's purchasing power) resulting from in-
stallation of the solar system.
Worksheets showing sample calculations are given as Figs. 45.1 through 45.5. Blank
worksheets appear as Figs. 45.6 through 45.9.
Solar Hot Water and Space Heating Systems 675
WORKSHEET 1
1. Design cost
WORKSHEET 2
1 400 x 1.00 = 60
2 (O x 1.06 = VI-
4 60 x 1.19 = 1/
6 60 x 1.34
8 1.50 = 90
9 1.59 = 95
10 60 x 1.69 = /
12 1.90 = 11 14
13 (9 0 x 2.01
14 60 x 2.13
16 Ca0 x 2.39 = 19 3
17 (ea x 2.54 = 15P,
18 60 x 2.69
20 3.03
WORKSHEET 3
Fuel Savings Fuel escalation Fuel savings Fuel escalation Fuel savings
Year Constant $ factor = 6% Future $ factor = 10% Future $
WORKSHEET 4
SAMPLE CALCULATION
2_,.
.a5Q + 477 + f.,0 = 3787 .A5 0 -2537 x LOD - 3537
+ - 102 3 ` + tog = -959 arcs *lady x o99 4- lIS /
3 + 11 77 + = 544 ego cloy x D = -,A 35
477 + 11.__ = 548 aS' -aso x 0 g 5' = -a 10
5 u 977_ + 7(
_, 553 315 —032 s 07 `i -ifiLL
6 477 + fC) 557 335 --AD. x o 15 -/4 6
7 + 1/77 + 5 . 5(, 355 --ca,57 x 0 70 -/4/5
e + 477 + 90 = 5(7 375 —1 /a x 6) 67 - ia9
9 + _4_2/ + 95 57; 398 - - 1 71 s 64 3 = -- //0
10 + 477 + 1151 675' 4A-a. -)56 x 0 59 -- 12.
11 + + t I = /07 4'4,Y. 1-341 x OS(0 = 1190
12 + + I 1,,i = Ill 4175 I- 34/ x o 53 = 7 1 90
WORKSHEET 4
SAMPLE CALCULATION
19 ye,„, Payback
1
WORKSHEET 1
1. Design cost
5. Money borrowed $
WORKSHEET 2
1 x 1.00
2 x 1.06
3 x 1.12
4 x 1.19
5 x 1.26 =
6 x 1.34 =
7 x 1.42
8 x 1.50
9 x 1.59
10 x 1.69
11 x 1.79
12 x 1.90
13 x 2.01
14 x 2.13 =
15 x 2.26
16 x 2.39 =
17 x 2.54
18 x 2.69 =
19 x 2.85
20 x 3.03
WORKSHEET 3
Fuel Savings Fuel escalation Fuel savings Fuel escalation Fuel saving
Year Constant $ factor = 6% Future $ factor = 10% Future $
1 x 1.00 x 1.00
2 x 1.06 x 1.10
3 x 1.12 1.21
4 x 1.19 1.33
5 x 1.26 x 1.46
6 x 1.34 x 1.61
7 x 1.42 = 1.77
8 x 1.50 x 1.95
9 x 1.59 = x 2.14
10 x 1.69 = x 2.36
11 x 1.79 x 2.59
12 x 1.90 x 2.85
13 x 2.01 x 3.14
14 x 2.13 = x 3.45
15 x 2.26 3.80
16 x 2.39 x 4.18
17 x 2.54 x 4.59
18 x 2.69 x L.05
19 x 2.85 = x 5.56
20 x 3.03 = x 6.12
WORKSHEET 4
1 + +
2* + +
3 + +
4 + +
5 ¤ +
6 + +
7 + +
8 0 +
9 + +
10 + + = x
11 + + = x
12 + + = x
13 +
14 + = x
15 + = x
16 + = x
17 x
18 x
19 x
20 x
REFERENCES
WILLIAM C. DICKINSON
Lawrence Livermore Laboratory
Livermore, California
46.1 INTRODUCTION
In this chapter, standard life-cycle costing techniques are used to establish an economic
methodology for solar thermal systems that provide process heat to industry. Possible
applications of this solar-generated heat include producing process steam, preheating
combusion air and boiler feedwater, providing process hot water, and supplying hot air
for dehydration.
Although the reliability and effective lifetime of solar process heat systems can
only be ascertained by long-term use, the technical feasibility of such systems is not in
doubt. Flat-plate collector systems can provide hot water or hot air. Concentrating
collector systems can supply higher-temperature air or process steam (see Chaps. 9, 10,
and 36).
There is, however, a real concern about the cost effectiveness of solar process heat
systems and whether industry will find it economically attractive to purchase, install,
and operate these large-area, capital-intensive systems. A study of the potential of new
energy sources for process heat [1] concluded that before an industrial user will consider
installing an alternate energy source such as a solar thermal system, one or more of the
following conditions must exist:
The economics of the new energy source will provide an investment return that is
sufficiently attractive to compete for the capital required.
685
686 Nontechnical Issues
The company is expanding facilities to meet product demand and will install a new
energy source on the basis of long-term expectations.
Fuel availability or regulatory restrictions force a change in fuel at an existing plant.
Although many companies will undoubtedly be faced with one or both of the last
two conditions, it is difficult to predict how many of them might decide to install a solar
system unless the first condition is satisfied. A company using direct-fired natural gas to
provide clean, dry , hot air for dehydration may be forced by a reduced gas supply to
install a supplementary energy source. The solar alternative would then have to compete
with the more conventional alternative of installing an oil- or coal-fired boiler, heat
exchanger, and fan system.
Because of the availability of alternative energy options, it seems unlikely that
there will be a large-scale adoption of solar process heat systems by industry unless the
first condition above is satisfied. That is, it must be shown that the solar investment can
be expected to yield a reasonable return to the company in the form of after-tax cash
savings resulting from reduced fuel purchases.
This chapter closely follows a previous paper written by the author with H. J.
Freeman [2]. Because the after-tax cost of process heat to industry is critically depend-
ent on government tax laws, the effect of several possible types of government incentives
to encourage the use of solar energy is also considered.
A typical industrial plant provides its process heat from a fossil fuel system (FFS). The
annual fuel bill is F and the annual cost for the conversion equipment (boiler, gas burners,
etc.) is B.* The plant manager is considering the installation of a solar thermal system
(STS) to provide part of the process heat.t This is estimated to reduce the annual fuel
bill from F to aF but should not affect the annual cost of conversion equipment.
In an economic analysis of these two alternative methods of supplying process heat,
the FFS is referred to as plan A (the "defender") and the proposed FFS/STS as plan B
(the "challenger")4 Irreducible factors—reducing air pollution, potential future short-
ages of fossil fuel, the risk involved in installing an STS where little or no operating
experience exists—will usually influence the choice between plan A and plan B. These
factors differ for each potential application and are not usually subject to a quantitative
analysis, although a replacement analysis that compares the annual costs of process heat
from the two systems is always necessary. If this analysis does not show the economic
feasibility of an FFS/STS, the irreducible factors may never be considered.
*We assume that the annual cost, B, is entirely an operation and maintenance charge and can be
written off for tax purposes in the year incurred.
t Before considering the installation of an STS, all reasonable energy conservation measures, such as
insulating hot water and steam pipes and installing heat-recovery systems, should have been taken.
A single challenger is assumed here. If there are two or more possible alternative plans, a method
of economic analysis that considers multiple alternatives must be used. For example, see Ref. 3.
Solar Industrial Process Heat Systems 687
CA = F + B (46.1)
For plan B, the annualized before-tax cost of energy, assuming equity financing,
can be determined from the following relationship:
where OMRI represents the annual operation, maintenance, component replacement, and
insurance premium costs for the STS; T is the effective income tax rate;- I is the total
STS investment; TC is the allowed investment tax credit; D is the annual allowed depreci-
ation expense, and CRF is the capital recovery factor in the STS investment.t Solving
Eq. (46.2) for CB gives
11
C B = aF + B + OMRI + r [(1 - TC)I • CRF - 7D] (46.3)
-
1 2 3
This illustrates that the annualized before-tax cost of energy for an FFS/STS consists of
three parts:
1. Payment for fossil fuel, boiler costs, and solar OMRI costs
2. Additional income tax caused by the installation of the STS.
3. Repayment to equity for the solar investment
*The before-tax cost is the total amount of revenue required (including taxes and return on capital
investment) to purchase energy. The after-tax cost is equal to (1 — r) X (before-tax cost) where T is
the income tax rate. See Ref. 2 for a more detailed discussion.
-tThe effect of other possible taxes, such as property and sales taxes, will generally be much smaller
than the effect of income tax and hence will be ignored. Most states are expected to follow the lead
of several states that have partially or totally exempted solar systems from property taxes.
fl'he capital recovery factor is the constant fraction of the investment I that must be accumulated
annually to amortize the investment in N years at interest rate i. It is usually written as CRF(i, N)
or A/P(i, N) and is tabulated in texts on engineering economics. See Ref. 3. Its value is shown there
to be
N
CRF(i, N) — i(1 + i)
+ oN - 1
688 Nontechnical Issues
cost of energy requirements.* If the internal rate of return is equal to or greater than
the minimum required rate of return on investment set by the company, then plan B is
favored; if not, plan A is preferable.
To avoid confusion by introducing an anticipated general inflation rate over the
STS lifetime, we shall make all analyses in terms of "constant" dollars (dollars of invest-
ment). (See Sec. 46.5 for a discussion of effects of inflation on a decision to invest in an
STS.) Therefore, the term real internal rate of return (RIROR) will be used hereafter,
corresponding to interest rate i. The corresponding capital recovery factor will be desig-
nated CRF(i).
If we equate CA to CB , we obtain the RIROR from the following expression:
(1 [
WO
r- 4. T
ORM) N
CRF(i) - (46.6)
1 - TC
To convert Eq. (465) to Eq. (46.6) it was necessary to assume that the STS invest-
ment is linearly proportional to the total STS collector area. Although there will always
be fixed costs independent of the collector area, these costs will normally represent only
a small fraction of the total investment for an industrial-sized STS consisting of several
*More precisely, the internal rate of return is that discount rate that equates the present value of the
total expected incremental cash inflow under plan B with the present value of the expected incre-
mental cash outflow for plan B.
1- 1 MMBtu = 1.05 GJ.
t1 MMBtu/ft2 • yr = 11.35 GJ/m2 • yr. For a concentrating collector, the collector area refers to the
projected aperture area.
**This should include the cost of land purchase, engineering design, site preparation, installation of
collectors, storage system, piping and ducting, pumps, valves, electrical wiring, control instrumenta-
tion, hookup, modification of the existing system, and necessary fees and permits.
Solar Industrial Process Heat Systems 689
Equation (46.6) can be used to calculate the RIROR for any set of parameters defining
an FFS and STS. A set of baseline parameters is defined here so that a graphical relation.
ship may be established between RIROR and p, the overall STS cost per unit collector
area. The effect of a fractional change in one of the parameters can then be easily
obtained from the graph. These baseline parameters are as follows:
T= 051. The present incremental U.S. federal income tax is 48 percent for all
corporate income above $25,000. A typical state corporate income tax of
6 percent is deductible for purposes of calculating the federal tax. This gives
an effective tax rate of 51.1 percent. Economic studies normally use a
corporate tax rate of 50 percent; however, we chose to use 51 percent so that
the analysis is less ambiguous-for example, to distinguish between r and 1 - T
in various cost equations.
f = $2.40/GJ. This corresponds to fuel oil at $15/bbl or natural gas at $0.25/
therm.
q = 2,3.5, and 5 GJ/m2 • yr. These values of annual collected energy correspond
to 0.18,0.31 and 0.44 MMBtu/ft2 • yr. Table 46.1 lists values for the average
Table 46.1 Average Annual Solar Radiation Availability for Three Locations
(GJ/m2 yr)a
annual amount of solar energy incident on various fixed and tracking surfaces
at three U.S. locations: Albuquerque, New Mexico; Omaha, Nebraska; and
Blue Hill, Massachusetts [4]. Approximately 50 percent of the annual total
incident solar radiation on a typical flat-plate collector can typically be
collected, that is, converted into hot water or hot air at the collector outlet,
and about 50 percent of the annual direct component of radiation on a
typical concentrating collector can be collected in the form of a hot liquid,
hot air, or perhaps steam. It can be seen from Table 46.1 that both flat-plate
and concentrating collectors can be expected to collect about 3.5 to 4 GJ/
m2 • yr in a southwest location and perhaps 2 to 2.5 GJ/m2 • yr in a mid-
western or northeastern location.
-= 0.02. Because there is no operating experience with the industrial STS, it is
difficult to predict OMRI costs, and large variations can be expected from
one STS to another. We are using an average of values estimated by several
U.S. Department of Energy contractors presently designing and constructing
solar industrial process heat systems.
N = 15 years. Because there is no lifetime experience with the industrial STS, a
conservative estimate of 15 years is made. The RIROR is relatively insensi-
tive to the estimated STS lifetime.
TC = 0.10. The present allowed federal investment tax credit for an industrial
facility such as an STS is 10 percent.*
= 0.70. This includes the conversion efficiency of the boiler or burners and the
heat losses in pipes on the way to the point of use. A typical boiler conver-
sion efficiency is 80 percent; about 10 percent of the generated heat is
typically lost in the piping and ducting leading to the point of use.
= 090. This includes the loss of collected heat in delivery to the point of use
and the loss of solar heat resulting from downtime for repairs and component
replacement. If the STS is sized to deliver more than 100 percent of required
process heat on clear summer days, with no storage capability, this additional
loss should be included.
These parameters were used to construct the set of curves in Fig. 46.1. Each curve
relates RIROR to p (the STS cost per unit collector area) for a particular value of q (the
annual energy collected per unit collector area).t One may also consider each curve in
Fig. 46.1 to represent a certain value of the combined parameter flpiln appearing in
Eq. (46.6), respectively, 6.17, 10.80, and 15.43 (units = $/m2 • yr). To determine the
RIROR for a particular set of STS parameters, calculate the numerical value of fcm1/77,
draw in the appropriate interpolated curve in Fig. 46.1, and use this curve to determine
RIROR versus p.
Determining the effect on RIROR of a projected differential escalation in the price
of fossil fuel over the STS lifetime is of particular interest. An expression, given in
* With the passage of H. R. 3919, the Crude Oil Windfall Profits Tax, an additional investment tax
credit for a solar industrial system of 15% is allowed, bringing the total allowed TC = 0.25.
tThe timing of the STS investment cash flows affects the RIROR, but is of little importance in most
cases (see Ref. 2).
Solar Industrial Process Heat Systems 691
20
Payback period — y r
10
Real
10
5
14
0
50 100 150 200 250
Figure 46.1 The real internal rate of return (RIROR) on a baseline STS investment
versus the installed solar system cost for three representative values of q, the annual
collected energy. The right-hand ordinate shows the payback period—the time required
to accumulate an after-tax savings equal to the net initial investment. Curves represent
straight-line depreciation over system life N and a 10 percent investment tax credit.
Sec. 46.5, allows an assumed differential fuel price escalation rate to be converted into
an equivalent "levelized" annual price of fuel. This levelized price can then be treated as
an equivalent constant fuel price over the STS lifetime.
The payback period for a proposed investment is a useful measuring stick of economic
feasibility. It is defined as the number of years required to accumulate an after-tax cash
savings equal to the net value of the proposed investment.
Several precautions should be considered when using this method as a supplemen-
tary decision criterion.
The payback period does not take into account the time value of money. Two
projects might be incorrectly considered to be of equal investment value—for
692 Nontechnical I ssues
example, project A with annual cash flows of - $1000, +$500, and +$500,
and project B with cash flows of -$1000, 0, and +$1000.
The payback method does not consider any cash flows past the payoff year, giving
those dollars zero weight in the decision.
Projects with longer lives are discriminated against.
The requirements for the payback period may be shortened until few projects
could qualify.
The payback method cannot properly select from among several alternative systems
on a rank-order basis, particularly when the system lives are of different
length.
(1 - TC)I
(1 - r) [(1 - a)F - OMRI] + rI/N
which is just the reciprocal of the expression for CRF(i) given in Eq. (46.5).f
*It should be noted that this term for added income tax is not the same as the term for added income
tax in Eq. (46.4). In obtaining an expression for the payback period, it seems more realistic to deduct
the investment tax credit from the initial investment, giving a net investment of (1 — TC)I. In this
case, the actual added annual income tax, as given above, must be used. However, in the calculation
of a series of equivalent uniform annual costs (referred to as the annualized cost), it is more logical to
annualize the lump sum investment tax credit into a series of equivalent uniform annual credits, each
equal to I • TC • CRF, and to make it part of the income tax term rather than including it as part of
the third term of Eq. (46.4).
t The reciprocal of the capital recovery factor is called the uniform series present worth factor and is
tabulated in compound interest tables alongside the tabulation of CRF. Hence it is a simple matter,
given the RIROR and the expected STS lifetime, to look up the payback period.
Solar Industrial Process Heat Systems 693
Payback periods are shown along the right-hand ordinate of Fig. 46.1. If the
RIROR is equal to zero, the payback period is equal to the STS lifetime, N years. The
payback period for a positive value of RIROR is always less than the STS lifetime.
There are two cases of particular interest for investments in solar process heat systems.
First is the situation where all costs and revenues are expected to escalate at some
constant inflation rate over the life of the project. Second is the situation where fossil
fuel costs are expected to escalate at a higher rate than the general inflation rate. These
will be considered separately [5].
If we assume a 6 percent inflation rate, then a 10 percent real rate of return corresponds
to a market rate of 16.6 percent.
This can be illustrated by considering a nondepreciable investment of $50,000 from
equity funds, with an OMRI equal to $5000 and a 6 percent inflation rate, which is to be
amortized over 3 years at a real rate of return of 10 percent. The expression for annual
required revenue is found from Eq. (46.3) to be C = OMRI + I • CRF/(I - r) = $46,032.
Constant Dollars
Cash Flow
Gross Taxable After Present Value
Year Revenue OMRI Income Tax (51%) Taxes at 10% ROR
0 -50,000 -50,000
1 46,032 5,000 41,032 -20,926 20,106 18,278
2 46,032 5,000 41,032 -20,926 20,106 16,616
3 46,032 5,000 41,032 -20,926 20,106 15,106
There is no advantage in taking a predicted inflation into account in this case. The
real rate of return is unaffected by inflation and should always be used. Unfortunately, it
is generally not made clear in a discussion of required rates of return on investment
whether a real or a market rate is being used. If a company states that it requires a
15 percent rate of return on an investment but sets this value because of an anticipated
5 percent inflation rate, then the 15 percent represents a market rate and their required
real rate of return is only 9.5 percent. It has been our experience that the market rate
is often not understood and the 15 percent value is interpreted as a real rate in making
an economic analysis of a proposed solar system using constant dollars.
Because an investment in a solar system is depreciable, the real rate of return will
be decreased by a general inflation rate. The annual depreciation deduction is based on
the initial dollar investment. If future dollars are worth less than constant dollars, the
constant annual numerical value of the depreciation deduction will correspond to a
decrease from year to year in constant dollars.
If straight-line depreciation over project life N is assumed, then the annual deprecia-
tion expense, I/N, can be replaced by an effective or "levelized" depreciation expense
(I/N)eff. This is given by
+e
\ eff = (—11 ) •CRF• (i
\ NI e\ -+ e ) \I-FA/
(46.7)
where the multiplying factor is derived by converting the series of decreasing depreciation
amounts (in constant dollars) to an equivalent present value and then multiplying by CRF
to obtain the levelized annual depreciation expense.
Note: If we assume an inflation rate e, then -e must be used in Eq. (46.7) since
the depreciation expense in constant dollars is decreasing at an annual rate e. In order to
solve Eq. (46.7), we must know the rate of return i and the corresponding CRF. If we
have calculated a real internal rate of return using Eq. (46.5) or Eq. (46.6) in the text,
we can substitute this RIROR into Eq. (46.7) above to obtain a value for (I/N)eff . This
in turn can be put back into Eq. (46.5) or Eq. (46.6) to obtain a revised value of i. Two
or three iterations should be sufficient to obtain a final effective value for the real
internal rate of return.
Solar Industrial Process Heat Systems 695
The size of this effect can be illustrated by reference to Fig. 46.1, where the real
internal rate of return for a solar system cost of $56/m2 and a q = 3.5 GJ/m2 • yr is seen
to be 10 percent. An assumed inflation rate of 6 percent over the 15-year project lifetime
reduces this real rate of return to 8%%.
If the solar investment comes partly from equity and partly from borrowed funds,
then there are two counteracting effects of inflation. The annual repayments of interest
and principal on the loan are fixed numerically and, therefore, decrease in constant
dollars. This results in an increase in the real rate of return on the equity. However, the
effect of inflation on depreciation, discussed above, decreases the rate of return on
equity. The net effect in most cases is small and depends on several factors: the assumed
inflation rate, the tax rate, the ratio of debt to equity funding, the loan interest rate, the
project life, and the method of depreciation.
( il -1-:;) l_ (1 + e \ N i
feff = f • CRF • (46.8)
\1+i/
C
As discussed above, we can solve Eq. (46.8) together with Eq. (46.6) to determine a new
effective value for the real internal rate of return.
If we return to the cae of p = $56/m2 and q = 3.5 GJ/m2 • yr, we find that an
assumed differential fuel escalation rate, e' = 5 percent (compounded each year over the
15-year project life) increases the 10 percent real internal rate of return to 14.6 percent.
For e' = 10 percent, the RIROR increases to 19.5 percent.
*For the case that i = e', Eq. (46.8) becomes fen.= f • CRF • N.
696 Nontechnical Issues
The additional income tax brought about by installation of an STS could be elimi-
nated if the government treated an STS as an R&D project that could be
written off as an expense item in the year of construction. Since it will be
a number of years before the reliability, performance, and lifetime of solar
thermal systems can be ascertained, it does not seem unreasonable to con-
sider them in the R&D stage. Allowing a write-off of an STS as an expense
would eliminate the added income tax term from the annualized cost (see
Table 46.2). The resulting effect on RIROR verus p is shown in Fig. 46.3.
The allowed STS cost for an RIROR of 10 percent is increased from $56/m2
to $72/m2.
Solar Industrial Process Heat Systems 697
20
15
Five-year depreciation — 25% TC
_ \
\
_ \\ \
\ \ \
Real internal rate of return — %
\ .1 \
\
\
\ \• \
\ \
10 \
\ \\
\\ \‘. \
\
\ \ \
\ \
\ \ \
\\ \
\ \ \
\ . \
5 - \ \• N
\ \ N
- \ '\ N
\ • N
- \ \
\ •
‘\. \ \
N•
.\\
0
50 100 150 200
Overall solar system cost p — $/m2
Figure 46.2 These curves show the effect of accelerated depreciation and increased
investment tax credit in increasing the allowed STS cost for a given RIROR. The solid
curve at left is the baseline curve for q = 3.5 GJ/m2 • yr.
A more powerful incentive would be to actually lower the amount of income tax
paid if an STS is installed. If an operative STS reduces a firm's consumption
of fossil fuel, the firm would have the benefit of a tax deduction equal to the
amount that would have been deducted as an operating expense had they
used the normal amount of fuel and not installed an STS. The additional
income tax term (see Table 46.2) now becomes negative, corresponding to a
Table 46.2 Formulas for the Annualized Before-Tax Cost of Energy and the Real Internal Rate of Return for FFS/STSa
Capital Recovery Factor CRF(i) for Real
Tax Treatment Annualized Before-Tax Cost of Energyb'" Internal Rate of Returnb
Basic 1
oiF + B + OMRI + 1--- [I • CRF(r- TC)- r13] + I • CRF (1 - r) [(1 - a)F- OMRI] + 7D
equations --- ... -r ........„..,-
(100% equity -----.-...."--,--•••'....--- (1 - TC)I
capital) (1) (2) (3)
(3)
V)
1 TC 1A 0
a F + B + OMRI + [aI • CRF (7 —) 7D + r(1 - a - (1 - 7-) [(1 - a) F - OMRI - (1 - anii(1 -
Equity capital 1-T a N G)]
(aI) and debt aI (I - TC/a) 0_
(1)
capital (2)
r.
(1 - a)I at
PT;
interest rate + al • CRF + (1 - a)I + + TD - ( 1 - a) I/N
e N N G/]
al(1 - TC/a) 0
(3) CD
CD
(4)
a CD
To convert CRF(i) to the form of Eq. (46.6), replace (1 - a)F by (f/n)/(p/q7I') and OMRI by and eliminate I.
bStraight-line depreciation can be expressed by D = I/N over system life N, and by D = (I/n) (CRFN/CRF,) over shorter accounting life n. The sum-of-the-
years' digits depreciation over system life N can be expressed as D = 112 (N CRFN - 1)/N (N + 1)i]. vt
vt
cThe terms in the above equations represent, respectively: (1) payment for fuel, boiler costs and solar operation, maintenance, replacement, and insurance; co
(2) additional income tax caused by installation of the STS; (3) repayment to equity for the solar investment; and (4) repayment of principal and interest 9
on a debt.
dMultiply the before-tax cost of energy by (1 - 7) to obtain the after-tax cost.
e The "gradient series factor," A/G A/Gi, N), converts a uniform gradient series to an equivalent uniform annual series (see Ref. 4). The loan principal is
assumed repayable in equal installments over the system lifetime with an annual interest i on the unpaid balance.
700 Nontechnical Issues
20
15
internal rate of return —
10
Real
•
5 •
•
•
0
50 100 150 200
Figure 46.3 These curves show (a) the baseline curve for q = 3.5 GJ/m2 • yr, (b) the
solar investment charged as an expense in year zero, (c) the case of an income tax rebate
on fossil fuel saved, and (d) the case of a tax rebate and a 5 percent differential escalation
in fuel price compounded annually over the STS lifetime.
net tax rebate. Figure 46.3 shows that the effect of such a rebate would be
substantial, increasing the allowed STS cost from $56/m2 to $115/m2 , while
still assuming an RIROR of 10 percent. If, in addition, we assume a 5 percent
differential escalation rate in fuel price per year over the STS lifetime, the
allowed cost is seen in Fig. 46.3 to rise to $160/m2 .
To this point we have assumed for simplicity that the financing of an STS would
come completely from equity funds. However, in many cases a portion of the funds
would be expected to come from debt financing at an interest rate i'. The expressions
for annualized before-tax cost of energy and capital recovery factor for the case of partial
Solar Industrial Process Heat Systems 701
debt financing are given in Table 46.2. In this case, we have based the rate of return, i,
only on the equity portion of the solar investment.
An important possible government incentive would provide industry with low-
interest loans to cover a fraction of its STS investments. The effect of a 6 percent
government-backed loan is shown in Fig. 46.4. Note that the effect of such partial debt
financing for an RIROR greater than 3 percent increases the allowed STS cost. Debt
financing is advantageous to the investor only as long as the after-tax cost of borrowed
money is less than the RIROR on the project for 100 percent equity investment. For
a 6 percent loan, the after-tax cost of borrowed money is (1 - 0.51) X 6 percent = 2.94
percent.
It should be noted that partial debt financing increases the risk in an investment in
the sense that the RIROR on the equity portion of the investment becomes more sensi-
20 1 i i 1 i
I 1 1 I I I
15
75% equity, 25% debt at 6% interest --,
Real internal rate of r eturn — %
1 1 I 1 1 1 1 1 1 i 1 I I 1 1 1 I 1 1 1 1
0
50 100 150 200 250
Figure 46.4 These curves show the effect of partial debt financing of an STS at 6
percent interest. The RIROR is based on the equity portion of the investment. The solid
curve at left is the baseline curve (100 percent equity financing) for q = 3.5 GJ/m2 • yr.
702 Nontechnical Issues
tive to an uncertainty in any system parameter. For an estimated solar system cost of
$80/m2 (see Fig. 2.4), the RIROR for 100 percent equity financing is 5.6 percent; for
75percent debt financing at 6 percent, the RIROR on the 25 percent equity portion is
14.5 percent. However, if the actual STS cost turns out to be $100/m2 , this 14.5 percent
drops sharply to 4.5 percent; with no debt financing, the RIROR drops from 5.6 percent
to 3.4 percent. Therefore, any uncertainty in STS cost or performance is amplified in
terms of its effect on RIROR when partial debt financing is used. The positive aspect of
partial debt financing is the risk of a smaller amount of equity capital.
The combined effect of a tax rebate and partial debt financing at 6 percent interest
is shown in Fig. 46.5. With a tax rebate and 75 percent debt financing, the allowed STS
cost increases from the baseline value of $56/m2 to $180/m2 .
20
15
Real internal rate of return —
10
0 1
Figure 46.5 These curves show the effect of a tax rebate on fossil fuel saved plus partial
debt financing of an STS at 6 percent interest. The solid curve at left is the baseline curve
(100 percent equity financing) for q = 3.5 GJ/m2 • yr.
Solar Industrial Process Heat Systems 703
REFERENCES
1. Evaluation of New Energy Sources for Process Heat, The Dow Chemical Company
and Environmental Research Institute of Michigan, 1975.
2. W. C. Dickinson and H. J. Freeman, An Economic Methodology for Solar-Assisted
Industrial Process Heat Systems: The Effect of Government Incentives, UCRL-
52254, June 1977. (Available from National Technical Information Service.)
3. E. L. Grant and W. G. Ireson, Principles of Engineering Economy, 5th ed., Ronald
Press, New York, 1970.
4. E. C. Boes, et al., Availability of Direct, Total, and Diffuse Solar Radiation to Fixed
and Tracking Collectors in the USA, SAND-77-0885, Sandia Laboratories,
Albuquerque, N.M., August 1977. See also the 1978 Addendum to this report,
which is available from the authors.
5. A thorough discussion on the effect of inflation on investments is found in App. G.
of Gerald W. Smith, Engineering Economy: Analysis of Capital Expenditures, Iowa
State University Press, 1973. See also R. T. Ruegg, Solar Heating and Cooling in
Buildings: Methods of Economic Evaluation, Rept. NBSIR-75-712, 1975.
6. R. E. Crane and H. G. Lorsch, "Solar Collection at Different Temperatures by
Different Collector Types Under Various Orientation Methods," in Proceedings of
the Annual Meeting of the American Section, International Solar Energy Society,
Vol. 1, pp. 3-11, International Solar Energy Society, Orlando, 1977.
chapter 47
705
706 Nontechnical Issues
47.1 INTRODUCTION
Institutions in both the public and private sectors can play a significant role in mitigating
the barriers to the rapid and successful adoption of solar heating and cooling systems. If
solar is to be offered to the public as a viable energy resource, these institutions must
grant the solar industry the same benefits accorded the established energy industry, so
that solar can compete in the energy market that has been nurtured by institutional
considerations and incentives. The barriers to solar energy's acceptance into society
can be traced directly to two obstacles, inherent in the energy market.
Both conventional and solar resources can be used for space heating and cooling
and hot water service , yet the basic differences between the two must be recognized so
that the public and private sectors will know how to approach subsidization and financing
of solar heating and cooling.
Conventional resources represent a stored energy medium, found in a limited
supply at specific and changing locations. Their scarcity and demand create a highly
capitalizable commodity and a relatively eager market. The costs of producing conven-
tional fuel supplies accrue from exploration, mining, transport, refining, conversion to
electricity, and the research and development pursued in each of these areas.
In contrast, solar energy represents an abundance of energy in the form of non-
monopolizable light and heat. Solar radiation itself is not a diminishing scarcity, thus
is not subject to market forces and foreign policy. The market for solar does not develop
from the task needed to deliver an energy as a commodity; rather the market develops
from the production and marketing of hardware that can convert solar radiation into
usable energy.
The institution that may present barriers within the public sector is primarily the
federal government; it can as well provide the greatest incentives to solar in the public
sector. It does not actually create barriers for solar except that its regulation of the other
fuel sources has created an incentive for conventional fuels that makes these fuels more
Barriers and Incentives in Commercialization 707
attractive. Institutions that present barriers in the private sector include public utilities,
the construction industry, construction codes, architects, labor unions, savings and loan
associations, and insurance companies. Another barrier may come from consumers'
perceptions of what energy must look like . This chapter presents some potential solu-
tions for the removal of barriers to solar heating and cooling (SHAC) or how they can
be countered by incentives from either within or without specific instititions. Most
often, incentives from without institutions are provided by action of the federal govern-
ment. This chapter also includes some scenarios combining a number of incentives that
serve to organize and propel a solar energy market.
A major incentive for the energy industry has come from the public sector. It is estimated
that $144 billion have been allocated to the energy industry through direct governmental
subsidy, for technology grants for research and development, incentives to control energy
production (i.e., pollution control, protection from foreign markets, safety controls), and
tax incentives in the form of deductions. The percent of total expenditures in each of
these areas is as follows: tax credits, 40 percent; controls, 39 percent; research and
development, 11 percent; and subsidization, 10 percent. By far, the greatest emphasis
on revenue to industry has been in the area of controls and tax incentives. Another major
incentive has been the establishment of 40 departments and offices for the express
purpose of carrying out and initiating energy policy.
These incentives provide the precedent by which a number of programs could be
designed to nurture and promote the solar energy industry. The programs include the
following.
B. Guaranteed Loans
By guaranteeing loans, the federal government would reassure private lending institutions
of the viability of the new solar technology and would relieve concerns about borrower
default. It places government credit and confidence behind both consumers and industry.
Loans could be guaranteed through the Federal Housing Administration or the Veterans
Administration. Mortgages could be backed by the government-sponsored Government
National Mortgage Association, a public corporation, or the Federal National Mortgage
Association, a private corporation.
Guaranteed loans may incur a barrier arising from the low maximum value of the
loan; that is, loans may not exceed $6000 for a single family, nor may they mature over
a period of more than 15 years.
As with low-interest loans, guaranteed loans benefit low- and middle-income groups
who have difficulty establishing loans. Guaranteed loans may, in fact, achieve maximum
effect in conjunction with low-interest loans.
A precedent for loan guarantees has been set for housing, aerospace, and maritime
loans.
employ solar equipment. Even more equitable would be the subsidization of energy
cooperatives such as those described later in this chapter. Subsidies to a cooperative
would directly benefit coop members by lowering monthly leasing charges. This
arrangement would further facilitate the solar user by not requiring each individual to
file applications for subsidies—a time-consuming and involved process.
H. Federal Compensation for State and Local Property and Sales Tax
The federal government could allow tax deductions for state and local property and sales
taxes. This incentive could also be taken on by individual states.
J. Tax-Exempt Bonds
Bonds could be issued by federal institutions such as GNMA and TVA, which would be
exempt from state and local taxes. Concomitantly, state and local bonds could be issued
with interest paid by local businesses such as public utilities, which would benefit from
bonds. Bonds could also be sold to finance separate solar energy utilities.
K. Demonstration Programs
These incentives would hasten solar's breakeven point with other fuels and thus would be
a powerful tool for encouraging adoption of solar. Although these measures would
stimulate the solar market, they would do so at the expense of low- and middle-income
groups. These incentives would be further encouraged by deregulation of other areas of
the fuel economy.
This is not so much an incentive to solar as it is a disincentive to fossil fuel con-
sumption.
A. Public Utilities
Most solar systems will require auxiliary power, which is most often connected to a
public utility service, that is, electricity or natural gas. A positive relationship between
public utilities and solar users is vital to the success of solar energy's assumption of a
significant role in the energy industry. The barriers presented by public utilities include
the following.
The Hopkinson rate schedule, or demand charge tariff, charges an energy charge,
reflecting the number of kilowatt-hours used and a demand charge for the maximum
kilowatt demand during a 15-minute interval. The latter reflects the generating capacity
at the power plant, not merely the cost of fueling electricity, as is the case with declining
block rates. The Hopkinson schedule encourages the use of a constant, average load
throughout the day.
Both rate schedules discriminate against the solar user. Declining block rates
penalize solar users by charging the highest rates for limited use of auxiliary power.
The Hopkinson schedule forces solar users to pay high rates because auxiliary power
use is concentrated in specific times. For some utilities it has been shown that solar
buildings' maximum monthly demand occurred during system off-peak hours. The
users energy/demand relationship (kWh/kW) would be more in accord with generating
costs if the demand charge were based on the amount of load during system peaks.
A potential incentive to solar energy systems is peak-load pricing. Two alternative
rate structures are possible under this type of rate schedule. The first, marginal cost
pricing, accounts for changes in generation based on demand. It is a more efficient
means of energy utilization, as it encourages base-load generation, which is less expensive
than the use of peaking plants. The other alternative is an interruptible rate schedule.
In this schedule, a reduced rate is offered to the user who foregoes the use of peak-
demand electricity. Feldman and Anderson [1] point out that the interruptible rate
schedule could save a user $600/yr in fuel bills.
Both of these alternatives require utilities to reduce rate charges when the need
for capital equipment and fuel is reduced. This will be an important incentive to solar
users, who will only need auxiliary power at off-peak hours. Thermal storage will
provide heating and cooling during periods of low insolation. Auxiliary power will
be required for periods when thermal storage has been exhausted, generally in early
morning when utilities are using base load. Insolation often coincides with peak-load
periods during the summer, when air conditioners cause heavy drain on utilities. At
these times solar systems use insolation most effectively, thus reducing potential peak
demands.
Shared solar systems prove to be highly economical for community application. The
Public Utilities Commission (PUC) may require that shared solar systems obtain certifica-
tion before beginning operation. The PUC has the power to deny certification. Because
hearings are long and expensive, they act to maintain public utility monopolies. Often,
what is not certified is innovative. This barrier could be overcome if the PUC were
restrained from making decisions on SHAC systems. Legislation showing that public
interest demands shared SHAC systems be permitted to compete with utilities could be
a powerful incentive.
An alternative remedy would have utilities finance and insure solar equipment
required by a PUC-imposed regulation. A precedent for this kind of action exists where
utilities are forced to comply with new air pollution technology.
712 Nontechnical Issues
*This section is a summary of work presented in Chap. IV of S. Feldman, ed., The Economics of Solar
Energy: The Public Utility Interface, Lexington Books, 1980. The model was developed by S.
Feldman, R. Wirtshafter, E. Wessler, and B. Anderson.
Barriers and Incentives in Commercialization 713
Table 47.1 Examination of Load Factors for Varying Buildings; Public Service
Company of New Mexico
Conventional building
Average annual 21,524 kWh/8,760 h = 2.46 kW
Capacity demanda = 4.31 kW
Load factorb = 57%
BUILDING PERFORMANCE
MODEL BUILDING
SECTOR
SUBROUTINE
SOLAR
PERFORMANCE
- -1
DETERMINE INCREMENTAL
CONSTRUCTION COST OVER
CONVENTIONAL BUILDING INCENTIVE/ SOCIAL INCENTIVE/
DISINCENTIVE WELFARE DISINCENTIVE
TO UTILITY
BUILDING OWNER TO CONSUMER
In the flow diagram of the model, the first section is the building performance
section. Building performance can be broken up into two components, average annual
and peak electric. The peak electric consumption is the performance of the building
during the peak demand of the electric utility. The simulation of building performance
is modeled using TRNSYS [4], a solar building simulation program.
Barriers and Incentives in Commercialization 715
The results of the run by TRNSYS of the yearly weather sequences yield the average
yearly performance of both the solar and conventional buildings. Performance of either
building is measured by the amount of electricity that is demanded. Because this
electricity is supplied by the electric utility, it is important to gauge the performance
of these buildings by their effects on the utility's load curves. The above performance
calculations are the average annual consumption broken down into hour-by-hour values.
This consumption represents the demand for electricity by the two buildings. It should
not be thought of, however, as a complete account of the buildings' performance or the
complete effect on the utility, since the values are representative of space conditioning
only.
716 Nontechnical Issues
The performance of the two buildings has been determined as above. The impact of these
buildings on the electric utility can be analyzed by examining expenses incurred by the
utility and broken down into three components. The components reflect the three
separate areas into which costs can be compartmentalized: energy costs, capacity costs,
and customer costs. The model determines the utility's long- and short-run incremental
costs and subsequent rates by examining separately the contribution of each component.
The model examines the effect of an additional building, either the solar or the conven-
tional, on the load of the utility. This direct analysis permits one to ignore any common
expenses other than for heating and cooling. In this manner, the direct contributions of
space conditioning to the expenses of the electric utility are calculated. The first compo-
nent of the electric utility's costs is its energy costs or short-run costs. Concern for
imported fossil fuels and increasing prices has centered much attention to this category.
Short-run costs consist of all costs that vary with the amount of energy produced. This
includes the cost of fuel itself and any operation and maintenance costs that vary with
output. The short-run costs per kilowatt-hour increase as output increases because higher
output requires use of less efficient generating plants. This means a greater use of fuel
and higher cost per kilowatt-hour. Should output drop to zero, however, there would be
no short-run costs.
The second aspect of the utility's costs and the effect of solar buildings on those costs is
the demand or capacity cost. This cost is a result of the demand for additional generating
capacity and subsequent transmission and distribution needed to service the new growth.
The costs are the result of an instantaneous increase in peak demand. The long-run cost
of the utility is the amortized cost of the capacity required by each of the buildings. The
costs are based on utility-supplied data on marginal generating costs.
Electric utilities often have a number of rate schedules that apply to different classes of
customers. As a result, the selection of the proper rate for the heating and cooling
portion of the electric bill is a complicated process. Under average-cost pricing, most
utilities utilize a declining block rate. As consumption increases, the unit price per
kilowatt-hour decreases. Because heating and cooling consumption may be viewed as an
increment to other uses of electricity, the appropriate block to charge off the heating
and cooling demand is at the margin. We assumed that a typical residential house uses
about 700 kWh per month, 24 kWh per day, exclusive of electricity for space condition-
ing. The charge for heating and cooling uses should be the appropriate rate from the
seven hundredth kilowatt-hour on. Marginal cost rates were calculated using the marginal
cost model developed by Cicchetti, Gillen and Smolensky [5]. Under marginal cost
pricing, the unit price per kilowatt-hour is assumed to be the same for all customer classes
for peak and off-peak. In addition, utilities may employ other types of rate schedules.
Barriers and Incentives in Commercialization 717
In each case, the proper rate must be based on the total consumption of the building.
For example, many utilities have in effect a demand-charge rate schedule, or Hopkinson
rate. The highest kilowatt demand in each month or period determines the demand
charge. In this type of rate, heating and cooling consumption must be added to the
other electricity usage to determine the total demand of the building at each hour.
Otherwise, the demand for heating and cooling may not actually be the maximum
demand for the building for any period.
The revenue to the utility is specified as the amount equivalent to the electric bill of the
building owner. This represents only the revenue from heating/cooling use of electricity.
What is important is the comparison of the solar building to its conventional counterpart.
consideration is whether the amortized cost of the solar building is exceeded by the
average annual savings in the heating/cooling bill within a specified payback period.
Theoretically, the total costs incurred to supply energy to any customer should be
recouped by the revenues from sales to that customer. If this is not the case, then either
other utility customers must subsidize that customer, or the utility must devise a rate that
puts costs and revenues in balance. In practice, the administration costs become so
oppressive that rates are derived to balance costs and revenues over broad classes of users.
National economic efficiency will be served if the total costs to society of the solar
construction are less than those of substituting electrical energy for solar energy. These
total costs must include the incremental costs for the solar energy system and the costs
of forestalled utility capacity. Some measure of the social costs and benefits of solar
energy must also be assessed.*
There is no denying the difficulty that ensues when one attempts to collect these
figures. Though some costs and benefits are immeasurable, some indication of the values
is obtainable. In the final analysis, it becomes a political decision as to whether the
perceived benefits are in excess of the known costs. The problem still arises that the solar
investor, with whom the final decision to invest in solar lies, is not influenced by any
external benefits that he produces.
The discrepancy between the rates and costs is in part a regulatory problem. Rates
that approach total societal marginal costs would partially eliminate this problem. Rate
schedule adjustment lacks the practical applications to accompany its theoretical under-
pinnings. One application that is being experimented with at present and that has con-
siderable impact on solar buildings is peak-load pricing. This is an attempt to redesign
electric rates so that the price of electricity approaches the marginal cost of supply. Of
particular concern is the discrepancy in costs by time of day. In most utilities, it costs
at least two to three times as much to generate electricity during peak periods as it does
during periods of less use. As a result, time-of-day rates are being implemented to
account for these different margins of costs.
The argument for marginal cost (time-of-day peak-load pricing) is based on the
stipulation that regulated industries can best approach free-market conditions by regulat-
ing the price at the marginal cost of supply. The problem is one of application, with
difficulties arising from joint and unassignable costs, increasing potential and justifiable
administration costs, inframarginal residuals and hesitant consumer acceptance. Several
methods of calculating marginal costs have been constructed by economists.
The Cicchetti model [5] is a practical approach that is straightforward and requires
easily obtainable data from the utilities. Capacity costs are distributed over those hours
of the year with a high probability of loss of load. The peak period varies from utility
to utility, with some utilities being winter peaking, others summer peaking, and a few
peaking year-round. The number of peak hours within each day also has a wide range.
In fact, the determination of marginal costs is so variant that generalizations among
utilities with similar climates, size, load factors, and fuel costs are difficult.
*One assessment of these costs by Dennis Costello of Midwest Research Institute varies from $55 to
$330/kW. Dennis Costello, "Midwest Research Institute Programs Dealing with Incentives and
Barriers to Commercialization of Solar Energy," Working Paper, Midwest Research Institute, Kansas
City, Missouri, 1976.
Barriers and Incentives in Commercialization 719
Table 47.2 contains a listing of the marginal cost rates as calculated by Feldman et
al. for 10 utilities. As one can see, the rates for on-peak range from 195 mils in Arizona
Public Service to only 25 mils in Sacramento Municipal Utility District. There are quite
marked differences of on/off-peak ratios within utilities: at a rate eight times higher
on-peak than off-peak for Arizona Public Service Company and Georgia Power Company
to only two times as high for the Puerto Rico Water Resources Authority.
As Feldman and Anderson [1] have pointed out, generalities about the effects of
solar buildings on electric utilities are difficult to make. It is necessary to consider not
only the climatic conditions, but the utility load conditions and the building design
conditions as well. The rates become the critical link in determining the scope and
intensity of the interaction of the building and the utility.
A concern of the rate schedules is the degree to which they are discriminatory.*
The methodology developed by Feldman et al. compares the total costs of the building
owner along with the total costs of the servicing electric utility for either a solar building
or a comparable conventional building. In a perfect situation, the electric bill paid by
either building should meet exactly the costs to the utility of supplying that service.
When this is not the case, as it almost certainly never is, then there is discrimination for
or against that particular use of electricity. As a result, no rate can be expected to be
precisely equitable and efficient for each user. The important concern is the degree to
which this discrimination occurs.
In Table 47.3 the actual bills for the heating and cooling portions of the two
buildings are compared. As one can see, the solar building's bills do not cover the costs
under either average or marginal cost pricing. Despite the fact that the cost of service
for the solar building with the 30-m2 collector is $960, the solar consumers' bill is only
$158 or 0.16 as much as the costs under an average-cost pricing schedule, and $353 or
0.37 as much as the costs under a marginal-cost pricing schedule. The case of the solar
building with a collector of 67 m2 is even more disparate. In this case the cost of service
is $188 and the bill is only $3 or 0.02 as much as the costs under both average and
marginal cost rates. It would appear that the proper rate that should be charged the
solar building owner with the 67-m2 collector should be 60 times that now charged.
It is important to note that this type of discrimination occurs for the conventional
building as well. Although the rate is not 60 times too low, but at most two times,
the actual amount of revenue is usually similar or even larger, so that elastic demand
is served with lower rates under present pricing policy. This is at the expense of such
uses as electric lighting, which appears to subsidize the heating component of the bill.
The ability of marginal cost pricing to eliminate discrimination is indeed limited by the
administration costs of employing individual rates.
The above analysis shows the discrepancy between the building owners' decision
and an optimal configuration for society. Average cost pricing encourages the least
efficient use of resources. Time-of-day pricing, a practical attempt to apply marginal
cost pricing, approaches a more efficient utilization of resources for society.
The case study of the New England Electric System produced different results.
In this example, the conventional building was still the most cost effective to the building
owner and society. The larger collector sizes did not become more efficient in the New
*Discriminatory refers to the extent to which costs of services are not reflected in prices.
O
Wisconsin Power & Light 66 28 29 3514 19.1 10.0 37.8 11.0 22.0
Los Angeles Water & Power 84 53 28 1620 31.2 25.8 89.0 28.0 16.9
sanssi reo!upaluoN
Barriers and Incentives in Commercialization 721
Table 47.3 Ratio of Rates to Costs, Conventional and Solar Buildings: Public Service
Company of New Mexico
Cost of Revenue Differences in Ratio of Rates
Building Service A.C. Costs to Revenues to Costs
Average Cost Pricing
Conventional $ 822. $409. $ 413. 0.50
Solar, 15 m2 1627. 390. 1237. 0.24
Solar, 30 m2 960. 158. 802. 0.16
Solar, 50 m2 433. 30. 403. 0.07
Solar, 67 m2 188. 3. 185. 0.02
England case. Under either rate schedule, average or marginal cost, the proper signal to
society is that the smaller collector on the solar building (of this design) is the most
resource efficient. Yet, it is not altogether clear what the effect of marginal cost pricing
will be on alternative building design. Presently, there does appear to be a trade-off
between the sizing of the solar collector and off-peak electric energy. If marginal cost
pricing were to be initiated, off-peak electricity could recharge the storage component
at far cheaper rates than active solar systems could.
It does not necessarily follow that solar energy and electric utilities cannot be
interfaced, as Asbury and Mueller have implied [6] . Solar building design, and particu-
larly passive designs including fenestration, insulation, orientation, and thermal mass, can
serve a crucial role in the interface. The potential for these applications in reducing not
only energy consumption but also peak generation demand is great. Preliminary studies
by Clark University and Total Environmental Action, Inc., have suggested that increased
thermal mass may be one of the most effective methods of sliding through peak demand.
As a result, the optimal investment policy becomes a combination of investments in
energy conservation, decentralized energy collection, centralized utility energy genera-
tion, as well as other applications and technologies.
Cooperation between the solar building and the electric utility requires the under-
standing that the revenue problem must eventually be corrected by the utility, either
through prohibition or higher rates. The ignorance of the problem by the building owner
will no longer be possible. The interaction between the utility rates and the solar building
owner is a dynamic process. The present electric rates that do not account for peak loads
have resulted in buildings, both conventional and solar, that require high demands during
722 Nontechnical Issues
peaks. As utilities begin to account for peak problems in their rates, the building designer
will probably begin to consider the additional problem of controlling peak electric usage.
It should also be remembered that charges that are too high leave the building
owner the option of using some other method of auxiliary. The above point is oftentimes
overlooked or downplayed. If the costs of electricity backup become too high for what-
ever reason, other forms of backup are available. Wood stoves offer quite an inexpensive
alternative in many parts of the country. The added advantages of having complete
control of one's own space conditioning system is meaningful to many users of solar,
especially those solar customers whose reasons for adopting solar include self-efficiency
in their objective function.
Another alternative would be the use of gas as an auxiliary energy source. Because
of the constricted supply of gas, the gas utilities are looking to solar as a supplementary
source. The peaking problem characteristic of electric utilities is not as pronounced in
the gas utility case. The major cost to the gas utility is at the source. Although there are
costs of storage, analogous to electricity-generation costs, the costs may not be significant
due to present and expected future overcapacity.
The alternatives listed above do circumvent the capacity effect of the solar building
at the initiative of the building owner. The costs of capacity are borne by the solar
building owner for the solar and auxiliary system. Unfortunately, these options, wood
and gas, have limited applicability due to limited supply. If solar heating and cooling are
to play an important role in the future, then cooperation is required between solar
building owners and the utilities. The key mechanism to effect this cooperation should
be the rates.
As has been pointed out throughout this section, presently available alternatives for
the use of space conditioning do not represent the most efficient options. The rates that
are now applied do not provide any incentives to correct this situation. Rate schedules
can be applied so that they encourage the more efficient utilization of solar energy and
utility load management.
One viable type of electric rate is an interruptible rate. This would allow the utility
to lower its capacity demand by denying on-peak power to the auxiliary heater. The
customer would benefit by having a lower rate. Because this system is voluntary, no
customer is forced to participate. The extent of the interruption can be set at the entire
peak period or just 15 min during times when the system demand is nearing peak. The
efficacy of this rate to solar buildings is predicated on the fact that solar buildings would
probably forego only a few days of auxiliary energy per year. The savings that could be
accrued were as high as $600 per year for New England Electric System. Under this type
of rate schedule, the solar building owner had the ability to weigh the capacity and costs
of his other system against the duration of the discomfort to be suffered from lack of
auxiliary. The value of interruptible rates is that it represents a rate schedule that
approaches the marginal cost of service. At the same time, it provides the necessary
incentives to the building owner to utilize resources more efficiently.
The analysis has shown that the solar buildings examined do receive a large subsidy
from the electric utilities. The salient point is that large subsidies are also enjoyed by
electric resistance heating. The solar market penetration within the next few years is
such that the total effect on the utilities will be small. The vagaries of the cross-customer
Barriers and Incentives in Commercialization 723
subsidy are lessened when it is considered that more economically efficient solar
buildings, from the point of view of society, would be built if less discriminatory rates
were employed. Rates do not adequately reflect the costs even under the best conditions,
so it is the total long-run costs for each alternative that should be considered.
If the final desire is to heat and cool buildings (or hot water), then many options
exist to this end. The economic analysis should include the external costs and benefits
of each option. In the case of solar, the benefits include a nonpolluting and nonforeign
energy source. In addition, subsidies to present energy sources must be accounted for
in order to assay true solar benefits. If rate schedules are to be used to promote solar
energy usage, they must be directed to the total societal aspects of usage. As time-of-day
pricing schedules begin to be implemented, new designs of solar energy buildings will
become cost effective.
E. Lending Institutions
Lending institutions include savings and loan associations, commercial banks, mortgage
companies, and life insurance companies.
The primary barrier to solar, common to all lending institutions, is the method used
to evaluate a borrower's ability to pay loans. The loan is calculated by projecting housing
costs and comparing these costs to the applicant's income, with the loan not exceeding
25 percent of income. The cost calculation is limited to principal and interest payment
on the loan, property taxes, and hazard insurance premiums. This "PITI" calculation
does not take into account the cost of supplying the home with energy over the period
of the mortgage. The solar home will require a greater mortgage because it includes the
energy cost in construction. The potential borrower's ability to pay the loan is thus
unfairly judged, as the lender's evaluation does not take into account that, although the
mortgage is higher, the monthly fuel bill is significantly reduced. By comparing housing
costs with personal income to determine the size of the loan, fewer potential borrowers
will meet the terms of the credit evaluation. The benefits of SHAC systems can be
assessed by including energy costs in the credit evaluation. A "life-cycle" cost compari-
son of a home with and without solar system would be used to consider an applicant's
ability to pay the loan. With life-cycle costing, both lender and potential solar home
buyer are made aware of capitalization of heating costs, rather than increasing monthly
fuel bills. When compared to conventional heating and cooling, the high mortgage on
the solar home is given a fair comparison. As mentioned earlier, tax incentives, govern-
ment loans, guarantees, and so on, will clarify solar systems as a home improvement.
The dominant and private institutions in the financing of residential mortgages
are savings and loan associations, which furnish more than 50 percent of single-family
dwelling loans and 21 percent of multifamily dwelling loans.
Barriers from savings and loan associations include a limitation on mortgage loans
to $55,000. Loans are regulated by the Homeowners Act of 1933. Not only do the
regulations put a ceiling on loans, but 20 percent of the corporation's assets must be set
aside in a fund. These regulations pose a barrier to solar homes because of added costs
of what may be comparable to 15 years of heating and cooling with conventional fuels.
724 Nontechnical Issues
An incentive from savings and loan associations would be to increase the loan
ceiling to the increment necessary to cover the cost of solar systems. A precedent for
this incentive exists under HUD-sponsored solar home demonstration projects.
Most savings and loan associations are authorized to make first mortgage loans
only. This barrier could be removed by an amendment (for which legislation is pending)
to the National Housing Act, which would permit an increase on first mortgage to include
the cost of retrofitting for solar equipment.
Many lending institutions require 100 percent backup systems. Lenders' fears stem
from general lack of knowledge about the capability of solar systems, and therefore
require a 100 percent conventional system to avoid risk. Incentives to lenders would
include an amendment to the Consolidated Farm and Rural Development Act to include
SHAC systems as a home improvement. Related to the above is the reluctance of lending
institituions to consider solar home loans as a benefit to property improvement.
Second loans have significantly higher costs. This is a major barrier, as cost com-
petitiveness of solar systems is dependent on low interest rates. A remedy to the second
loan barrier is the addition of an "open-end" clause in first mortgage contracts. Such a
clause would allow mortgagees to borrow back funds up to the amount of equity repaid
in loans, at interest rates equal to the original interest rates. Another solution to the
second loan barrier is the "wraparound" mortgage, which extends the original mortgage.
F. Construction Industry
Barriers may be encountered in labor unions, building codes, and architectural designs.
The construction industry has traditionally shunned innovation, as witnessed most
recently by the slow acceptance of plastic piping into building design. The traditional
"craft" nature of the construction industry has concentrated on the use of locally found
building materials and has consistently shunned industrially produced materials. Con-
trary to this trend, solar collectors represent industrialized tools, built away from building
sites. It must be noted that the construction industry has accepted prefabricated energy
devices, particularly air conditioners.
Craft guilds may resist total fabrication of solar components in the factory.
Problems may also arise from conflicts over work assignments; that is, guilds will
represent new skills. The major incentive to the building industry will be the creation of
thousands of jobs directly and indirectly related to SHAC production and installation.
The creation of a Solar Energy Labor Board could lead to recommendations of appropri-
ate job assignments.
Building codes may be biased toward conventional equipment because of the new
technology's unfamiliarity. Building inspectors will not be able to assess new technology
until solar systems are in wide use. These barriers stem from the lack of specific require-
ments and nationally accepted standards for SHAC systems in building, mechanical, and
plumbing codes; hence, solar is subject to building officials' discretionary powers. Left
to local discretion, the solar market could be fragmented and the growth of the solar
industry severely impeded. This barrier will be overcome when guidelines for solar
construction are incorporated into building codes. Standards could arise from federal
government demonstrations.
Barriers and Incentives in Commercialization 725
Equipment for which building code provisions are made or installation regulated
must be granted a special permit. Permits could be denied on the basis of lack of data on
solar systems. Prefabricated components will require local testing to determine durability
when data are lacking. The uncertainty of solar's viability, the extra expense from time
delays, as well as special application processes, will all contribute to fragmentation of the
potential market. Again, nationally recognized standards as well as standardized testing
are mandatory. A national accrediting organization, such as already exists for the electric
and gas industries, could be created for the purpose of certifying solar system compo-
nents.
Architectural barriers arise from the lack of concern in residential architecture for
passive energy design, that is, conservation of energy within the structure by nonmechani-
cal means. Passive solar systems require careful evaluation of the energy conservation
performance of the structure. Passive measures include:
1. Fenestration of windows and shading allow for maximum solar radiation intake
for heating with minimum loss through northern windows in winter and mini-
mization of heating through windows in summer.
2. Design of buildings to accommodate natural convection of energy throughout
the structure to conserve and self-regulate.
3. An understanding of the aerodynamics of the structure so that the effects of
high and low air pressures are minimized.
4. The use of wall construction and insulation materials that permit maximum
conservation of thermal mass.
5. The lowering of load requirements for heating systems in structures designed
with proper passive systems will help offset the initial cost of the active system.
G. Insurance Companies
The only foreseeable barrier to solar systems from insurance companies would be a reluc-
tance to provide coverage for new technology for which little data exist. This barrier will
be remedied by government incentives described above, including loans, procurement,
and sizable demonstration programs, all of which will provide technical and psychological
affirmation of solar energy.
47.4 SCENARIOS
In 1977, solar energy accounted for one-thousandth of 1 percent of U.S. total heating
and cooling needs.* Natural market forces and piecemeal incentives alone will not pro-
vide sufficient impetus for the rapid expansion of the solar industry and its potential
market. If solar is to make a serious impact on residential energy needs in the immediate
future, the fledgling solar industry and its potential customers must find a common
organizational network through which solar technology can be most expediently trans-
acted and dispersed. We discuss below two possible scenarios under which this might be
accomplished.
A. MISEC
We first consider the establishment of a Metropolitan Independent Solar Energy Coopera-
tive (MISEC), which would be organized as an alternative energy cooperative, the purpose
of which would be the creation of a marketplace for coop members to:
MISEC would achieve the latter by creating a marketplace that brings together and
nurtures the diverse members of the solar industry, the banking community, government,
labor unions, utilities, and other consumers and institutions.
Membership in MISEC would commence when a customer leases solar equipment
from the coop. It might also begin when a customer pays a fixed fee for consultation
regarding energy conservation for his home. The coop would furnish a list of government-
certified solar hardware manufacturers. Certification would include specific warranties
covering the hardware's lifetime and servicing. For a nominal fee, members could use the
coop's Resources Section, which would prepare residential solar system plans and assess
the needs of the member's residence. The Resources Section would be comprised of solar
engineering experts, who would make use of computer facilities provided by the coop,
to determine placement and number of collectors, collector slope, placement and size
of storage tank, insulation, window size, and so on. The focus of the Resource Section
would be the ongoing search for the systems that will contribute the most energy at the
least expense to members. The Resource Section would also cooperate with the Solar
Energy Research Institute in providing and delivering services.
MISEC would be a service industry, similar to the telephone service, where the
user/member pays a monthly fee for the leasing of equipment. By leasing, the solar user
would be freed from the burden of the high initial cost of solar systems. The installation
and maintenance would be provided by independent certified professionals in plumbing,
ventilating, and heating industries.
The leasing of solar systems has two advantages over the purchase of such systems
that are crucial to the success of the solary industry. First, leasing would provide a safe-
guard against obsolescence of the solar investment. The rapidly evolving new technology
will undergo radical changes as it matures; mass production will reduce equipment costs,
and new innovations will modify older prototypes. These changes will lead to reduced
initial solar system costs. As these costs decline, monthly bills could be adjusted to
reflect the declining costs so that members who leased the first equipment could share in
the overall savings. Conversely, the buyer of solar equipment may see the high price once
paid for a solar system quickly declining. Upon resale of a home, the consumer may be
Barriers and Incentives in Commercialization 727
stuck with an obsolete system. The coop member would be offered new technology as it
occurs, without the major first-cost, refitting expense incurred by the buyer. The leasee,
upon resale of a home, would find the property value increased as a result of maintaining
the highest state-of-the-art solar technology. Second, the coop would encourage innova-
tion. MISEC would not be a manufacturer, tooled up for a specific solar system model.
Rather, it would seek out the best equipment, thereby encouraging new ideas in industry,
and discouraging the business practice of shelving innovations that rock the corporate
boat.
Through a MISEC credit system, MISEC would finance home insulation and other
energy conservation or improvement repairs. Finally, coop members would benefit from
profits to be disbursed annually. Coop members could vote to reinvest earnings in the
coop, in part or whole, as would be determined at meetings held periodically.
Even with the rising cost of fossil fuels, the solar energy industry must hurdle the
problem of creating a market. The industry can attempt to educate the public to the
individual and societal benefits of using solar, and may show the cost effectiveness of
solar, but the fact remains that capital requirements for a new solar system will put solar
out of reach of most Americans.
MISEC would destroy the "first-cost" capital barrier, and provide market incentives
that the consumer could realize today. With MISEC as a facilitator, the manufacturer
avoids the maladies of a fragmented housing industry. The manufacturer of a solar
collector need not search for water tank manufacturers, nor hire plumbers and installers
for their product, nor wait for any number of private contractors to buy equipment.
Equipment would be sold directly to MISEC, where all parts would be brought together
as parts in a whole system encompassing hardware, installation, and maintenance. It
should be noted that MISEC would not preclude the development of a competitive
marketplace but would be an alternative institution. MISEC would have the potential
for creating thousands of jobs, directly and indirectly. Employees of the solar industry,
installers, and maintenance crews, and consultants would be encouraged to become coop
members, helping to perpetuate the market.
Financing for MISEC would originate with seed money provided by the coop bank.
With these funds, solar equipment would be purchased and then used as chattel
mortgage for coop expansion.* MISEC should be permitted to take advantage of tax
exemptions for borrowing and lending, because of its nonprofit structure. Further,
bonding would be the concern of private investment, and regulated by indenturement
by-laws established for the coop.
It is worth noting that MISEC would not seek federal funding reallocation of Public
Utility Funds to fund new energy development, nor require direct state or legislative
appropriation. Funds would be relatively free from controls and regulations. Regulations
would be determined by coop members. The investment spur for MISEC, however,
should be leasing systems to the government for its own buildings.
The first Board of Directors would act as a steering committee. It might include
representatives from Public Utilities, Solar Energy Industry Associations, labor unions,
ASHRAE, banking and mortgage firms, neighborhood consumer groups, and Chambers
of Commerce. They would set the guidelines for the initial stages of MISEC, and would
hire paid management, subject to board approval.
We conclude that a solar energy cooperative would facilitate a rapid expansion of
the solar energy industry and its pursuant marketplace in the next decade.
B. S.A.G.E.
Another comprehensive plan that has been studied by the Southern California Gas
Company is called S.A.G.E., Solar Assisted Gas Energy. In this scenario, the utility
company would be responsible for the purchase and subsequent leasing of solar equip-
ment. Again, as in the MISEC solution, the leasee would be charged a monthly fee for
leasing solar equipment and for using auxiliary power from the utility. Centralization
under utilities would have the benefit of utilizing an industry well accustomed to a sales
distribution and service network. One drawback of the plan may be some degree of
minimizing the solar system to allow a greater use of the utility power source.
47.5 CONCLUSION
There are real and perceived barriers to the rapid adoption of solar energy into the energy
market presented by the institutions that comprise the energy and construction industries.
From the public sector, the major barrier to solar technology adoption is the unfair
market advantage given to conventional fuel supplies by the federal government's regula-
tion of fossil fuel, hydroelectric, and nuclear fuel sources. This barrier will be removed
when the solar industry and solar buyer are afforded the same support, through loan
credits, tax deferments, and subsidies. These public sector incentives can help remove
the barriers of lack of confidence in the new solar technology, which exist in the private
sector.
Public utilities can play an important role in removing barriers to solar energy's use
by providing more equitable rate structures for solar systems' auxiliary power.
Finally, efforts must be made to organize a comprehensive scenario of all the
elements required for the rapid expansion of the solar energy industry and its acceptance
by the private sector and adoption by consumers.
REFERENCES
1. S. Feldman and B. Anderson, Utility Pricing and Solar Energy Design, Clark Univer-
sity, NSF/RANN APR 75-18006,58,1976.
2. H. Lorsch, Implications of Residential Solar Space Conditioning on Electric Utilities.
Draft of Final Report Prepared for the National Science Foundation Under Contract
No. NSF-C-1033 (AER-75-18270), The Franklin Institute Research Laboratories,
1976.
3. G. Bennington, M. Bohannon, and P. Pewak, An Economic Analysis of Solar Water
and Space Heating, Prepared for Energy Resource and Development Administration,
MITRE Corporation, November, 1976.
Barriers and Incentives in Commercialization 729
BIBLIOGRAPHY
D. Barrett, P. Epstein, and C. M. Haar, Financing the Solar Home: Understanding and
Improving Mortgage Market Receptivity to Energy Conservation and Housing
Innovation, Regional and Urban Planning Implementation, Inc., Cambridge, Mass.,
1976.
D. Barrett, P. Epstein, and C. M. Haar, Home Mortgage Lending and Solar Energy, U.S.
Department of Housing & Urban Affairs, Division of Energy Building Technology
& Standards, 1977.
R. Bezdek, Analysis of Policy Options for Accelerating Commercialization of Solar
Heating and Cooling Systems, The George Washington University, 1977.
T. G. Bolle, "Summary Statement on the Realities and Problems of Financing Residential
Solar Installation," Solar Conference, University of Denver, Center for Public
Issues, 1977.
Booz, Allan, Hamilton, and Shaw, The Effectiveness of Solar Energy Incentives at the
State and Local Level, Federal Energy Administration, 1976.
S. H. Butt, Solar Market Capture and Market Penetration—Solar Heating and Cooling,
Solar Energy Industries Association, 1976.
Cost/Benefit Analysis Report and Supporting Opinion Survey, Commercial Buildings—
National Solar Demonstration Program, Prepared for ERDA Under Contract No.
E(11-1)-2688, InterTechnology Corporation, January 1976.
S. Feldman, et al., Public Utility and Solar Energy Interface: An Assessment of Policy
Options, Energy Research and Development Administration, DSE/2523-1, 1976.
S. Feldman, et al., "Solar Energy and U.S. Public Utilities: The Impact on Rate Structure
and Utilization," Energy Policy, 195-210 (September 1977).
A. Hirshberg, "Barriers to Widespread Utilization of Residential Solar Energy: The
Prospects for Solar in the U.S. Housing Industry," Policy Sciences 5, 4 (1975).
A. D. Little, Solar Heating and Cooling Constraints and Incentives: A State-of-the-Art
Analysis, 1976.
Miller, Hayes, and G. P. Thompson, "Solar Access and Land Use: State of the Law," in
Legal Barriers to Solar Heating and Cooling of Buildings, Environmental Law
Institute, 1977.
MITRE/ERDA, An Economic Analysis of Solar Water and Space Heating, 1976.
J. G. Moore, Solar Energy Legislation in the 94th Congress, Library of Congress, Social
Policy Research Division, 1976.
H. C. Petersen, The Impact of Tax Incentives and Auxiliary Fuel Prices on the Utilization
Rate of Solar Energy Space Conditioning, NSF/RANN Report, 1976.
J. P. Phillips, Assessment of a Single Family Residence Solar Heating System in a Subur-
ban Setting, Annual Research Report NSF/RANN 75078, 1975.
R. T. Ruegg, Designing Effective Incentive Policies for Solar Energy Systems, National
Bureau of Sandards, Center for Building Technology, 1976.
730 Nontechnical Issues
W. Shultze and S. Ben-David, Near Term Prospects for Solar Energy, National Science
Foundation, 1976.
Solar Heating and Cooling of Buildings—Phase 0, NSF/RANN 74-021E, General Electric
Company, Contract Performed for the National Science Foundation, RANN, May
1974.
Solar Heating and Cooling of Buildings—Phase 0, NSF/RANN 74-023C, Westinghouse
Electric Corporation, Baltimore, Maryland, Contract Performed for the National
Science Foundation, RANN, May 1974.
Solar Heating and Cooling of Buildings—Phase 0, NSF/RANN 74-022C, TRW Systems
Group, Redondo Beach, California, Contract Performed for the National Science
Foundation, RANN, May 1974.
Study for Defining the Number of Residential and Non-Residential Projects—National
Solar Demonstration Program, Prepared for ERDA Under Contract No. E(11-1)-
2683, General Electric Company, Space Division, April 1976.
R. H. Timms, Wrap Around Mortgage-Lending. U.S. League of Savings Associations,
1976.
Westinghouse Electric Corporation, Solar Heating and Cooling of Buildings—Phase 0 Final
Report, NSF Report NSF/RANN 74-023D, May 1974.
M. D. White, The Allocation of Sunlight: Solar Rights and the Prior Appropriation
Doctrine, 47 V. Col. L. R. 423,1976.
chapter 48
731
732 Nontechnical Issues
48.1 INTRODUCTION
Solar heating and cooling of buildings is, and likely will be, the solar energy application
most widely used over the near term. Relative to fossil or nuclear energy, current solar
heating and cooling of buildings (SHACOB) technology appears almost benign environ-
mentally; indeed, under most circumstances it can be. However, unlike fossil or nuclear
energy use, SHACOB is highly decentralized in its application, and is intended for use by
the general public. Because of such widespread potential public involvement, any environ-
mental issues associated with SHACOB require the most conservative examination.
Examination of SHACOB safety and health aspects must consider the implications
of widespread public utilization and must take into account all phases of the system's
lifetime—manufacture, installation, operation and maintenance, and decommissioning.
Evaluation of the physical environmental impacts associated with employing SHACOB
must consider the cumulative nature of these impacts resulting from the large-scale and
widespread use of SHACOB technologies. Through these approaches, the overall environ-
mental, health, and safety concerns associated with SHACOB use can be determined on a
realistic, if conservative, scale.
The following sections present the key environmental, health, and safety issues of
SHACOB identified to date. In most cases, the major concerns involve safety during
accidents, system misuse, or system malfunction. Cumulative impacts arising from
SHACOB use primarily concern disposal of working fluids and general system decom-
Environmental, Health, and Safety Issues 733
missioning. Despite the potential occurrence of these impacts, however, it should be kept
in mind that, under normal conditions, SHACOB operation should be relatively safe.
A. Water Contamination
Domestic water supply contamination can occur at heat exchanger interfaces in solar-
heated domestic hot water (DHW) systems. If the contaminated water is ingested, adverse
health consequences could result. This interface between solar heat transfer fluids and
domestic water supplies usually occurs within the domestic hot water storage tank.
Contamination at these interfaces can occur if the heat exchanger ruptures or fails in
such a way that working fluid is allowed to enter the potable water supply. Potential
toxic hazards are associated with the working fluids themselves or, in a water-based
system, the various additives used as corrosion inhibitors, pH controllers, bactericides,
freeze protectants, and so on.
The toxicities of many potential working fluids and additives have been identified,
based on experience in other industrial sectors (e.g., boiler water treatment, heating and
air conditioning systems). Corrosion inhibitors include chromate, borate, nitrate, sulfate,
sulfite, arsenate, and benzoate salts, and various triazole, silicate, and phosphate com-
pounds. The toxicities of these compounds range from the relatively benign to highly
toxic and some could be potentially carcinogenic. Because of the nature of the medium
(i.e., domestic water suppty), most examination focuses on the oral toxicity of the com-
pounds in question. Most toxicity data are based on the pure compound itself. To date,
little is known of the synergistic effects of these substances, or of the relative toxicities
after long-term use under conditions of high temperature and/or pressure. Table 48.1
summarizes the relative simple toxicity of some probable working fluids and additives
and lists references where more information on their uses and toxic properties can be
found.
A number of steps to guard against potable water contamination have been out-
lined [1] . These include (a) separation of circulation loops between nonpotable work-
ing fluids and the domestic potable water system; (b) identification of nonpotable fluid
and potable water systems by color-coded piping or metal tags; and (c) nonpotable fluid
leak indicators, such as harmless vegetable dyes.
Generally, normal pressure differentials between potable water and solar system
fluids work against potable water contamination; however, this does not eliminate poten-
tial hazards entirely, as municipal water pressure is not totally reliable. Employing
double-wall heat exchanger systems can provide a high degree of protection against
potable water contamination by leaking working fluids. Some localities require their
installation in facilities where solar energy systems provide heat to DHW systems. Pro-
tective heat exchanger designs are illustrated in Fig. 48.1.
kg.
Phosphates Corrosion inhibitors Na, PO4 used Low toxicity; Phosphates are mild alkaline irritants and 2, 3, 4, 40
for iron and aluminum in concentra- HEDP (Monsanto rarely dangerous in concentrations
trations from Dequest 2010) normally encountered in solar system
2 to 3 ppm orally administered use.
to rats has resulted
in LD50 of 3.13 g/
kg.
Glycols Freeze protectants in 10 to 100% Ethylene glycol Under ordinary conditions of industrial 2, 3, 5, 40
aqueous-based systems most toxic, LD50 use, most of the commonly used glycols
= 100 ml for oral are considered to be only moderately
ingestion by adult. toxic.
Silicon Heat transfer fluid 100% Low toxicity; Dow The toxicity of these compounds is 2, 6, 40
fluid or Corning 200 has generally low; many studied have no
siloxanes been fed to rats in toxic properties at all, and most have
E nvi r onmental , Heal th ,and Safety I ssu es
1
outer shell heat transfer
(a) medium
Wafintlp
b)
hot water
heat transfer
medium
outer shell
(c )
Fig. 48.1 Heat Exchangers. A heat exchanger is a device for transferring thermal energy
from one fluid to another. In some solar systems, a heat exchanger may be required be-
tween the transfer medium circulated through the collector and the storage medium or
between storage and the distribution component. Three types of heat exchangers that are
most commonly used for these purposes are illustrated. (a) Shell and tube. This type of
heat exchanger is used to transfer heat from a transfer medium containing antifreeze to
736
Environmental, Health, and Safety Issues 737
its behavior under overheat or fire conditions, situations that can cause release of po-
tentially toxic gaseous compounds from some insulation materials prior to their actual
combustion. This phenomenon is known as "outgassing" or "offgassing."
In collectors, outgassing can occur when the absorbing surface reaches tempera-
tures above that which the surrounding insulation can tolerate. Such high-temperature
conditions can occur during stagnation (when working fluids are not circulated through
the collector). However, fumes from outgassing during stagnation rarely prove hazar-
dous; they are parimarily a collector efficiency concern, since their filmy condensation
on the underside of collector glazings can reduce transmittance.
Most fumes released during collector outgassing are simple starches that are used
as binders for many inorganic insulation materials (e.g., fiberglass). Phenolic substances
that sometimes serve as binders also may be released. Phenolic substances may prove
harmful if inhaled in large quantities or over long periods of time [2] . Such exposure
is possible if outgassing occurs within the dwelling, a potential occurrence with some
integrated roof collector structures. However, phenolic offgases from collector materials
rarely would prove injurious in the open air.
Outgassing from collectors and other parts of the SHACOB system and dwelling
would become a major concern only during fires. During fires, temperatures could be
high enough to degrade insulating plastics and synthetics, releasing irritating and highly
hazardous fumes of cyanide, toluene diisocyanurate (TDI), hydrogen chloride, ammo-
nia, nitrogen oxides, and others. Such fumes are the result of both outgassing (i.e.,
pyrolytic decomposition) and actual combustion of the material. Tables 48.2 and 48.3
list potential toxic or hazardous properties of common insulation materials.
Overall, any dangers of outgassing during normal operation can be prevented by
proper system design. This involves the selection of insulation materials that exhibit
thermal and chemical stability at maximum collector temperatures as measured during
stagnant or no-flow conditions. Table 48.4 indicates the upper temperature limits for a
number of common SHACOB collector insulation materials. Maximum no-flow collector
water used for storage. Shell-and-tube heat exchangers consist of an outer casing or shell
surrounding a bundle of tubes. The water to be heated is normally circulated in the tubes
and the hot liquid is circulated in the shell. A single shell-and-tube heat exchanger cannot
be used for heat transfer from the toxic liquid to potable water because double separa-
tion is not provided and the toxic liquid may enter the potable water supply. (b) Shell
and double tube. This type of heat exchanger is similar to the previous one except that
a secondary tube is located in the shell. The heated liquid circulates between the shell and
the second tube. An intermediary nontoxic heat transfer liquid is located between the
two tube circuits. As the heated liquid circulates through the tube, the intermediary is
heated, which in turn heats the potable water supply. The heat exchanger can be
equipped with a sight glass to detect leaks—toxic liquid often contains a dye—or to in-
crease the liquid level in the intermediary chamber. (c) Double wall. Another method of
providing a double separation between the transfer medium and the potable water supply
consists of tubing or a plate coil wrapped around and bonded to a tank. The potable
water is heated as it circulates through the coil. Source: National Bureau of Standards,
Intermediate Minimum Property Standards for Solar Heating and Domestic Hot Water
Systems, Interim Report, April 1976.
00
Material Outgassing Productsb Toxic Constituent Route Local Systemic Local Systemic
temperatures should not exceed the upper temperature limit of the insulation material
used in that collector.
Guarding against hazards from fire involves careful selection of desired dwelling
insulation, considering material flammability, decomposition products, and applicable
building codes. In most cases, a material can be selected that will provide maximum
thermal insulation while remaining relatively safe under most conditions.
C. System Flushing
Many of the liquid working fluids and/or storage media used in solar heating and cooling
systems will degrade over time. Therefore, these fluids will have to be flushed and re-
placed periodically. Release of these fluids may also occur as a result of system failure or
decommissioning. The release of these working fluids into local water bodies could
impact aquatic life significantly, as well as prove harmful to people if such impacted
water bodies are used as sources of drinking water. In addition, fluids that are allowed to
enter conventional sewage systems potentially could reduce the effectiveness of sewage
treatment in areas of major solar system use.
Contamination of water bodies could result from the accumulation of such com-
pounds as chromates, nitrites, and sulfates/sulfites. Significant impacts resulting from
accumulation of these compounds would require large-scale and concentrated release of
contaminated fluids into local water bodies, probably within a short time span. Because
dangerous heavy metals such as chromium are needed only in minute quantities, if at
all, dangers from aquifer contamination due to SHACOB use should never materialize,
However, short-term effects from freeze protectants (glycols) disposed of in municipal
sewage systems or waterways are possible in areas of concentrated SHACOB application.
742 Nontechnical Issues
Solar system working fluids may contain 20 to 100 percent ethylene glycol, a
commonly used heat transfer medium. Since this fluid is relatively unstable, it must be
changed periodically. The common tendency is to dispose of the used fluid by dumping
it on the ground or down the drain. If dumped on the ground, the glycol solution could
be an immediate but short-lived threat to children, animals, and plant life. Disposal
through the sewage system reduces the direct environmental dangers because of the large
dilution factor, but the effects on waterways or sewage treatment facilities may be
significant.
The extent of the dilution can be estimated by using the City of Albuquerque as
an example [10] . If it is assumed that half the city's 74,000 private residences had solar
collectors, each containing 0.09 m3 (20 gal) of ethylene glycol to be disposed of in a
period of 60 days, the average daily disposal of about 56 m3 (12,000 gal) of fluid would
occur for 60 days. The total daily liquid effluent is 160,000 m3 (35,000,000 gal), which
would give a concentration factor of 3 parts fluid to 10,000 parts water by volume if
totally mixed. The actual concentration could be higher, since the fluids normally would
be changed during the daylight hours. It also could be lower if the bacteria in the sewage
treatment plant reduced it to harmless substances.
'A 40-kg child would have to drink approximately 0.19 m3 (50 gal) of this dilute
mixture before receiving the lowest lethal reported dose (LDLO) [3]. Thus, it appears
unlikely that any direct fatalities would result from the disposal of the fluid down the
drain, but cumulative effects might occur in animals and plants as a result of chronic
exposures.
Regarding the effects on sewage treatment facilities resulting from working fluid
disposal, only ethylene glycol is expected to be in concentrations sufficient to affect such
processes. Because ethylene glycol is biodegradable, problems in treatment would involve
facility capacity rather than chemistry. Adequately sized sewage treatment works should
be able to handle large quantities of glycol. If capacity is strained, however, sewage plant
effluent could raise the biochemical oxygen demand (BOD) of receiving waterways, and
thus contribute to pollution or cause short-term pollution episodes.
Solar agricultural and industrial process heat (AIPH) may employ solar energy systems
already in use or being developed for solar heating and cooling of buildings or solar
thermal electricity production. Hence, the environmental, health, and safety impacts of
AIPH will be somewhat similar to the impacts of these two technologies [11] . The cross-
cut impacts, therefore, will be mentioned only briefly; the emphasis of the discussion will
be on the EH&S impacts particular to AIPH systems.
A. Product Contamination
Perhaps the most significant EH&S concern specific to MPH systems is potential con-
tamination of products, particularly foodstuffs, resulting from working fluid leakage or
insufficient product heating [12] . Solar energy used for process heat can result in two
Environmental, Health, and Safety Issues 743
Improper handling and disposal of system working fluids and wastes can result in signi-
ficant water quality impacts. In many solar process heat systems, periodic flushing will
744 Nontechnical Issues
The major environmental, health, and safety (EH&S) impacts associated with solar ther-
mal power systems can be divided into two general classes: system working fluid/storage
media impacts, and ecological impacts. The former represents primarily health and safety
concerns, whereas the latter concentrates on local ecosystem changes brought on by the
construction and operation of a solar thermal power facility. Finally, a safety concern
unique to central receiver systems is misdirected radiation, which will be treated separately
Working fluids for solar thermal power systems generally include high-temperature fluids
and storage media used in power generation, and low-temperature fluids that may be
employed as a means of waste heat utilization in solar total energy system configurations.
The low-temperature fluids generally are water based, hence their impacts were discussed
in the solar heating and cooling section. This discussion will concentrate on the impacts
of high-temperature working fluids that would be used directly in solar thermal power
generation.
In addition to high-pressure water and steam, candidate working/storage fluids for
solar thermal power systems include hydrocarbon oils, eutectic salts composed of sodium
Environmental, Health, and Safety Issues 745
contacts water and forms, through a vigorous reaction, sodium hydroxide and hydrogen
gas. This latter reaction could generate sufficient heat to ignite combustibles; in addition,
the released hydrogen could burn explosively [16] .
Whether in the form of an aerosol, a solid, or in concentrated solution, sodium
hydroxide can be quite harmful. It is an extremely corrosive material; in concentrated
form, NaOH causes bums and deep ulceration, with ultimate scarring. Mists, vapors, and
dusts of this compound cause small bums, and contact with the eyes, either in solid or
solution form, rapidly causes severe damage to the delicate tissue. Inhalation of the dust
or concentrated mist can cause damage to the upper respiratory tract and to lung tissue,
depending on the severity of the exposure. If aerosols of the sodium oxides were inhaled
or contacted by moist skin, similar problems would be encountered, since the oxides
would form sodium hydroxide upon contact with tissue moisture [2] ,
Certain molten salt mixtures proposed as heat storage and transfer media contain
anhydrous sodium hydroxide. Release of these salts could result in the same adverse
effects previously described for NaOH. However, the hazards would be more acute, since
the hydroxide would be present in very concentrated form.
The potential problems associated with working/storage fluid release into area
water supplies are mitigated by the fact that most prospective solar thermal power sys-
tem sites are located in arid desert regions with few permanent surface bodies of water.
In addition, ground water in these areas usually occurs at depths of 30 m or greater, and
thus, chances of contamination by released working/storage fluids are minimal [16] .
Proper and periodic system maintenance should be adequate to prevent major fluid
leakages. Chemical management procedures, such as containment ponds and dikes, may
be employed to control inadvertent leakage and allow for controlled system flushing. The
potential flammability and explosion characteristics of the molten salts, liquid metals,
and oils are well documented, and proper handling and safety procedures have been im-
plemented in most cases [17] . Therefore, working/storage fluid impacts should not be a
major environmental problem if this area of concern is given proper attention during
plant design, site development, and plant operation.
B. Ecological Impacts
The key ecological impacts of solar thermal power system deployment are potential
alteration of the microclimate, and the impacts associated with the heliostat field. The
ecological concerns presented in this discussion are based on the siting of a central re-
ceiver solar thermal power plant in the arid regions of the U.S. Southwest.
A surface exposed to solar radiation absorbs part of the radiation, and reflects
the remainder back into the atmosphere. The percent reflected is called the "albedo."
Desert soils can be expected to reflect about 30 percent of the total incident short-
wave radiation. Solar thermal power plants, especially the heliostat field, will alter the
albedo of desert soils, possibly resulting in significant microclimate alterations. For a
single facility, the range of these effects probably will be limited. However, micro-
climate impacts are a key concern because of the novelty of the impacts and the lack of
adequate observational data.
Environmental, Health, and Safety Issues 747
1—
=
0
Iii
I
TOPS OF
HELIOSTATS
Fig. 48.2 Comparison of approximate wind speed profiles expected for a heliostat field
and the land surrounding a solar thermal power plant. Source: U. S. Energy Research and
Development Administration, Solar Program Assessment: Environmental Factors, Solar
Thermal Electric, ERDA 77-47/4, Washington, D. C., March 197
An array of heliostats can modify significantly the net absorption of direct and
diffuse insolation within the power plant boundaries. The net albedo for a typical helio-
stat field has been estimated to be about 56 percent [16] . This is almost twice as high
as the albedo for desert lands, and close to typical for a several-day-old snow layer [18] .
The increased reflectivity could result in an appreciable cooling of air flowing over the
mirror field during the day.
Heliostats also can disturb low-level air flow patterns. This is indicated in Fig.
48.2. Wind speeds below the tops of the heliostats will decrease, and these changes
could reduce the air/surface temperatures. Light wind speeds and cooler temperatures
beneath the heliostats probably will reduce evapotranspiration within the field [16] .
With the reduction in air/surface temperatures, reduced evapotranspiration, and
shading of the ground beneath the heliostats, the field may stimulate the microclimatic
conditions of a north-facing slope located near the power plant. These conditions might
be favorable to the establishment and growth of shade-tolerant vegetation; native wild-
life populations utilizing such plant cover may increase. Shifts in species diversity and
relative abundance could lead to ecosystem imbalances, and the densities of certain
rare or endangered species (e.g., the Gila monster, the spotted bat, and the spot-tailed
earless lizard) may be reduced [16] .
748 Nontechnical Issues
C. Misdirected Radiation
Reflection and misdirected radiation from the heliostat field is a concern unique to cen-
tral receiver solar thermal power systems. Heliostat reflections can cause severe burns,
eye injury, and fires, as well as create hazardous conditions for nearby ground and air
traffic. During normal operations, personnel and materials within the heliostat field may
be exposed to high-intensity reflected radiation, especially near the central receiver. This
could result in skin burns and cornea/retina damage. Heliostat reflections also could
scorch, char, melt, or ignite materials such as fabrics, wood, plastics, and structure coat-
ings [19] . Birds flying over the field could be burned or blinded. The light intensities
outside the field might induce momentary flash blindness; an unexpected flash could
cause an aircraft or automobile accident. In addition, heliostat glare can be a nuisance,
especially when close to highways or populated areas.
Although heliostats in their normal operational mode may create safety hazards,
potentially the most serious impacts will stem from misdirected solar radiation caused by
Environmental, Health, and Safety Issues
L IGHT I NTE NSITY 749
INTENSITY LEVEL OF
DIRECT SOLAR RADIATION
DISTANCE
Fig. 48.3 Variation in reflected light intensity over distance. Source: U. S. Energy
Research and Development Administration, Solar Program Assessment: Environmental
Factors, Solar Thermal Electric, ERDA 77-47/4, Washington, D. C., March 1977.
a misaligned heliostat field. Misaligned tracking heliostats will focus a beam of concen-
trated radiation at a point a given distance away from the central receiver. This is con-
ceptually illustrated in Fig. 48.3. At the focal point, the intensity of the beam would be
sufficient to cause blinding and severe burns (second to third degree). Any type of com-
bustible material could be ignited easily, including clothing and vegetation. Glare prob-
lems would be intensified; an even greater potential for vehicle accidents would exist.
In certain system designs, heliostats may be individually focused, such that the above
safety impacts could result from misalignment of only one unit. Since these systems may
be deployed in industrial or otherwise populated areas, workers and residents, as well as
plant personnel, may be exposed to the dangers of misdirected solar radiation.
The fabrication, transport, and installation of heliostats may result in the same
impacts discussed above. Since the heliostats may be individually focusing prior to actual
750 Nontechnical Issues
plant operations, a single heliostat could generate a concentrated beam of light and create
serious burn, fire, and glare hazards.
A significant amount of research on this issue already has been completed [20] ,
[21] . In addition, collector system safety criteria have been developed for the Depart-
ment of Energy's 10-MW, solar thermal power baseline system experiment [22] . As a
precaution, heliostat systems should be designed for quick and safe defocus and emer-
gency shutdown and should be kept in a safe "stowed" position when not in use. In
addition, all potentially combustible materials should be stored in places inaccessible to
misdirected radiation. Further, plant building and access roads should be laid out so that
they are not in pathways of possible misdirected radiation. Perimeter fencing of the entire
site, as well as prohibition of overflights, also may be necessary precautions.
48.5 PHOTOVOLTAICS
The EH&S issues associated with photovoltaics generally relate to the health and safety
aspects of materials production and cell fabrication. A number of potentially toxic chem-
ical compounds and derivatives can be released during the several stages of photovoltaic
cell production. In addition, fire conditions in operational cell arrays may result in off-
gassing of toxic substances. The specific impacts of the production of three candidate
materials, silicon (Si), cadmium sulfide (CdS), gallium arsenide (GaAs), and their by-
products will be discussed below.
A. Silicon
The production of silicon cells involves the mining of quartz (or high-purity sandstone),
refining the quartz to silicon metal, purification of the silicon, recrystallization, doping,
grid placement, and encapsulation. Health and safety concerns are most significant in
the silicon production and doping phases.
Inhalation of silicon dust, which is an established cause of respiratory disease
[23] , may occur during several phases of silicon cell production, primarily mining of
quartz and sandstone, and secondarily, silicon refining. Deposition of silicon particles
within the lungs may result in fibrosis and subsequently the lung disease, silicosis. An-
other possible result of inhalation and/or ingestion of silicon, for which there is only
limited evidence, is accumulation within the kidney, followed by development of
systemic hypertension [24] .
Toxic gases from doping silicon with boron and phosphorus may be released to
the workplace or ambient environment. One of the methods of doping silicon with boron
involves the use of boron trichloride (BC13 ) at high temperatures. The dissociation of
boron trichloride and diffusion of boron into the silicon result in the venting of un-
determined amounts of chlorine and boron trichloride. Chlorine gas is irritating to the
eyes and respiratory tract and may result in pulmonary edema. However, whereas con-
centrations of 50 parts per million (ppm) are dangerous even for short exposures and
levels of 1000 ppm may cause death, a detectable odor at 3.5 ppm and throat irritation
at 15 ppm serve as "alarms" and generally prevent severe exposure [2] . Boron trichloride
Environmental, Health, and Safety Issues 751
is, however, extremely toxic, having a lowest reported lethal concentration (LCLo) for
inhalation in rats of 20 ppm [3] , and therefore presents a serious threat in cases of acci-
dental leakage.
Doping with phosphorus also presents several concerns, since phosphine (PH3 ),
phosphorus oxychloride (POC13 ), and phosphorus pentoxide (P2 0) are used as the
diffusing agents. Phosphine is a highly toxic gas with a lowest published lethal concen-
tration (LCLo) for human inhalation of 9 ppm [3] . Chronic exposure to phosphine may
result in gastrointestinal damage, anemia, and nervous system disorders [25] .
B. Cadmium Sulfide
Cadmium is produced exclusively as a by-product of zinc mining and smelting. The sul-
fide generally is produced from cadmium metal [26] . Hence, toxic cadmium compounds
may be released during various phases of CdS cell production.
Inhalation of toxic cadmium compounds during material production phases pre-
sents some major health concerns, including possibly lethal acute poisoning. The cad-
mium extracted from zinc ore is collected in baghouses. While breathing equipment
usually is used, accidental exposure to fugitive emissions during equipment maintenance
or throughout the material production process is possible. Inhalation of dust or fumes
generally affects the respiratory tract, but the inhaled product becomes concentrated in
and may damage the kidney. The initial effect is dryness of the throat, cough, headache,
tightness in the chest, shortness of breath (dyspnea), and vomiting. Acute exposure may
result in changes in the lung similar to those found with bronchopneumonia, accom-
panied by severe chest pain, coughing, severe dyspnea, and death. Significant darkening
of the urine is observed, accompanied by fatty degeneration of the liver and acute in-
flammation of the kidneys. Since onset of these symptoms is delayed several hours, very
high levels of exposure are possible before the worker is alerted to leave the area [2] . The
Occupational Safety and Health Administration (OSHA) standards for airborne cadmium
dust are 200 µg/m3 for a time-weighted average and 600 pig/m3 maximum concentration
[3] . Chronic exposure to levels in the range of 3.2 X 103 to 1.6 X 104 mg/m3 has been
linked to proteinurea [27] and hypertension (elevation of systolic blood pressure) [28] .
Although cadmium sulfide appears to be about 10 times less toxic than cadmium
dust or soluble salts, it also poses a threat to worker health, being a recognized carcino-
gen [3] . The potential for exposure exists during synthesis of the material itself and its
vacuum deposition during cell fabrication. "Safe" levels of exposure to carcinogens are
not established.
C. Gallium Arsenide
Gallium is extracted as a by-product from zinc and aluminum ores. Aluminum ore is the
most probable future source of gallium for photovoltaic cells [26] . A significant increase
in gallium production will result in large amounts of alumina sludge, as well as possible
contamination of surface water with mercury and caustic or acidic effluents [29] . The
quantity of solid alumina waste that would arise is estimated at between 55,000 and
109,000 metric tons per year per 1 GW equivalent of GaAs photocells [26] .
752 Nontechnical Issues
meric concentrating materials (e.g., methyl methacrylate used in Fresnel lenses) are
highly flammable. Offgasses and combustion products from these materials have not
been characterized.
Fires involving gallium arsenide cells may cause vaporization of the arsenic follow-
ed by oxidation to the highly toxic arsenic trioxide (As2 03 ). Dermal irritation and sys-
temic poisoning resulting from exposure to this substance were treated in the previous
discussion on exposure to arsenic compounds.
Although no flammable materials are believed present in cadmium sulfide cells,
the potential for vaporization (at 800°C) and condensation to particulates exists. Car-
cinogenesis, as well as respiratory and renal damage, are serious possible health effects
and have been discussed above.
It is therefore necessary to ensure that adequate fire protection is integral to photo-
voltaic cell deployment. This is especially critical when photocell arrays are mounted on
buildings, since offgas by-products may seep into the structure, causing asphyxiation.
Advance warning systems are necessary to enable prompt evacuation. In addition, it also
may be desirable to have specialized chemical fire-fighting equipment (e.g., self-contained
oxygen gear) on hand to control potential photocell fires.
Since ocean thermal energy conversion (OTEC) plants will operate in an ocean environ-
ment, their effects on ocean ecosystems, climate, and biosystems will be the major envi-
ronmental concerns. However, the secondary effects of construction of OTEC technolo-
gy may impact local institutions (e.g., coastal construction communities) and, more im-
portant, international relations. Since the ocean is one of humanity's greatest resources,
its increased utilization will become a matter of growing concern for all nations.
Current baseline designs project the need for approximately 9.3 m2 of heat exchanger
surface area for each kilowatt of capacity in an OTEC plant [33, 34] . Therefore, a
100-MW, plant would utilize approximately 930,000 m2 of heat exchanger area, all of
which would be subjected continually to the erosive and corrosive forces of sea water
flow, causing metallic elements to be dispersed into the ambient sea water every day.
These metallic trace elements will be largely in solution as ionic or soluble chemical
compounds. Precipitated particulates and particles of elemental metal also may be
discharged.
Through food-chain transfer these metallic elements may build up in local marine
organisms and, depending on concentration and type, may be toxic to certain marine
species. Concentrations of metallic elements should be diluted quickly by ambient sea
water at the point of discharge, partially mitigating potential environmental problems.
However, some metallic elements are concentrated by marine organisms from very
dilute solutions.
Copper/nickel alloy, one candidate for use in OTEC heat exchangers, would pose
the most serious environmental problems. The adverse effects of both copper and nickel
on marine life have been well documented, and EPA has set marine water quality limits
of 0.05 mg/1 and 0.01 mg/1 for copper and nickel, respectively [15] . OTEC copper and
nickel ion and solid discharges are anticipated to result in concentrations significantly
less than these levels.
C. Use of Biocides
Chlorine has been suggested for use as a biocide to prevent biofouling on the sea water
side of heat exchanger surfaces. The U. S. EPA's "Proposed Criteria for Water Quality"
states that concentrations of free residual chlorine in marine waters in excess of 0.01
mg/1 are unacceptable [15] . Proposed initial concentrations in an OTEC plant are 10 to
50 times greater than EPA's acceptable limit for marine water, indicating that dechlori-
nation may be necessary before discharge.
If ammonia is used as the working fluid and the water flowing through the system
is chlorinated, the accidental leakage of ammonia and its subsequent mixture with chlor-
ine in the sea water could prove more toxic than chlorine alone to certain marine species.
However, if the chlorine can be recaptured from the sea-water flow after cleaning the
heat exchangers, its adverse effects on the marine ecosystem will be minimized.
Environmental, health, and safety issues of the fuels from biomass production tech-
nologies—agriculture, silviculture, and mariculture—and biomass conversion technologies—
thermochemical conversion, bioconversion, direct combustion, and biophotolysis—are
based on a review of existing research data, biomass production and conversion experi-
ence, and a certain amount of conjecture associated with future technology application.
The key issues can be identified with a fair degree of confidence because most biomass
Environmental, Health, and Safety Issues 755
production and conversion technologies are similar to other explored or commonly em-
ployed technologies such as agriculture and silviculture (biomass production), fossil
energy conversion processes (direct combustion and thermochemical conversion), and
sewage treatment practices (bioconversion).
The production of biomass as an energy source requires substantial water and acreage.
Both are critical factors that determine biomass production sites and affect the surround-
ing environment.
Water availability is an important factor for both arid and humid regions of the
country. Water requirements vary depending on plant species and soil type; for the
high-yield crops being considered for terrestrial biomass production, the amounts re-
quired are recognized as substantial. However, some of the tree species being considered
can grow adequately on moderate rainfall. In the arid West, water supplies are scarce and
establishment of biomass farms in such regions is unlikely, unless supplies can be channel-
ed from water-abundant areas. However, the humid Eastern region of the country con-
tains ground water that usually is closer to the surface, allowing wells to be drilled at
moderate cost. Consequently, establishment of biomass farms in these regions is more
probable.
Land requirements for biomass energy farms and silvicultural plantations also are
substantial. When viewed as an energy crop, the land areas involved are immense: Total
acreage needed to supply 1 percent of the nation's present energy requirements would
range from 1.5 to 4.5 million acres (present agricultural activity in the United States
accounts for 700 to 750 million acres). This translates to about 32 to 96 square miles of
land per quad of energy grown as biomass.* Furthermore, individual biomass production
tracts would run between 20,000 and 60,000 acres, acreage that is far greater than the
average farm of 100 to 600 acres (depending on the region) [35] .
The potential environmental consequence of land and water use for biomass pro-
duction is essentially one of competition—competition with food, fiber, and other pro-
duction activities requiring land and water. If water is plentiful, competition can be
avoided; if scarcity occurs, the resulting competition would become a significant issue.
On the other hand, the use of productive land to grow biomass as an energy source has
inevitable consequences of competition over the long term. Growing biomass for energy,
rather than for food or fiber, could represent a significant impact during times of food
and fiber shortages. Furthermore, secondary land use impacts could result from displace-
ment of smaller farms to form larger biomass farm tracts.
The U. S. Department of Energy's "Fuels from Biomass Program" recognizes these
concerns and attempts to resolve such issues in mission analysis, site selection, and re-
source studies. Thus, the program of terrestrial biomass production is directed at areas of
extensive water availability and low land-use desirability (so-called marginal land). More-
over, research is directed at plant species that could be capable of growth on marginal
land with little water. However, less desirable land (for agricultural and fiber production
interests) with sufficient water may not be available often enough in the large quantities
necessary for biomass energy production. Therefore, large-scale terrestrial biomass pro-
duction retains the potential to affect land and water use patterns, at least to the degree
that present agricultural and silvicultural activity affect these resources.
Intensive agriculture and silviculture activities affect both air and water quality. In farm-
ing operations, dust and sediment loads are the major concerns, as these impact both air
and water quality. Silviculture also generates air pollutants, but the major concern of this
activity is sediment loads to waterways. Because of the large size of biomass energy plan-
tations, the effects of air and water pollution from biomass production could be
magnified.
Air quality is affected by a number of biomass production activities including soil
cultivation, logging, harvesting, and heavy equipment traffic over unpaved areas. In agri-
cultural tilling operations, dust particles formed by loosening and pulverizing the soil are
injected into the atmosphere as the soil is dropped to the surface. Dust emissions are
greatest for dry soil and during final seedbed preparation; they also are directly propor-
tional to the average traffic speed and to the silt content of the road surface. Aerial appli-
cation of pesticides, fertilizers, and fire retardants also contributes to ambient particulate
levels.
The major water quality impact of intensive crop or forest cultivation is sedimen-
tation to surface water caused by rain erosion and/or irrigation runoff from exposed
soils. Pollutants resulting from silvicultural or agricultural runoff include nutrients,
pesticides, organic loads, and fire retardants. Sediment resulting from soil erosion is the
primary pollutant affecting water quality, cropland being responsible for about 50 per-
cent of the total sediment yield in inland waterways [36] . In forest management areas,
skid lanes and logging and fire roads are the principal source of soil sediment loads to
waterways from forestlands. Large energy plantations could contribute sediment loads at
least equal, and perhaps higher in quantity than those from conventional crop or forest
lands, since residual materials, which protect the soil from wind and water erosion, may
be removed.
Mitigation of fugitive dust and sediment discharge primarily involves mechanisms
that prevent wind and water soil erosion. Advanced cultivation methods, such as "no-
till" farming and contour-type plowing schemes, aid in reducing runoff. Also, when
coppicing species are used, fugitive dust and sediment discharge are reduced because
tilling is done only every fourth to sixth rotation (every 2 to 6 years). However, biomass
production still would employ farming and silvicultural practices that disturb the soil
(e.g., soil discing and unpaved roadway travel); thus, air and water quality impacts could
occur to the extent that they do under current practices.
Environmental, Health, and Safety Issues 757
Thermochemical biomass conversion can produce gases, tars and oils, unconverted residue
(char), and ash, depending on the particular conversion process employed. Pollutants
associated with these products can affect air and water quality, land use, and may present
some health and safety concerns.
Thermochemical reactions generate sulfur-containing [hydrogen sulfide (H2 S),
carbonyl sulfide (COS), carbon disulfide (CS2 ), sulfur oxides (SOX)] and nitrogen-
containing [hydrogen cyanide (HCN), nitrogen oxides (N0x), ammonia (NH3 )] gases.
All of these pollutants are produced in some quantity during biomass thermochemical
processes. However, because of the toxic nature of the compound, H2 S is the most sig-
nificant air pollutant from biomass conversion. Although biomass has an inherently low
sulfur content, the quantities present are sufficient to cause local air standard violations
and possible odor problems from resulting H2 S concentrations if untreated raw offgases
are released. Such concerns may be eliminated by preventing leaks, flaring raw offgas
streams [flaring converts H2 S into harmless quantities of sulfur dioxide (SO2 ) and
water] , or chemically treating product gases to remove H2 S.
Water quality also can be affected by the residuals produced from thermochemical
conversion. Low-molecular-weight oils, phenols, leachates from char and ash residues,
and scrubber solution runoff may enter water bodies by direct discharge and by percola-
tion to subsurface waters from evaporation ponds. Possible impacts may be aggravated
when using water as a reactant, as proposed in some thermochemical schemes. Adverse
water quality effects can be lessened by the use of oil flotation devices, activated sludge
treatment, lagooning, or evaporation/infiltration ponds.
Land use is affected only minimally by disposal of ash in landfills, since biomass
contains small ash quantities (approximately 1 percent). However, land disposal of char
residue—assuming that it is not to be combusted or used further—could require significant
land areas, since char has much greater volume by weight than ash.
Health and safety concerns involve the tar and oil products produced by thermo-
chemical conversion. These products have superficial resemblance to coal tar, a known
carcinogen, and research in this area is not yet sufficient to dispel such concerns asso-
ciated with biomass.
Overall, potential impacts related to biomass thermochemical conversion would be
significantly less than those found in similar coal conversion schemes, because the pri-
mary carbohydrate nature of biomass affords "cleaner" products. Nevertheless, because
of the severity of the thermochemical processes, the production of some undesirable
environmental residuals is difficult to avoid.
ough out the world, few large wind electric conversion (WEC) machines have been
Opdated. Such limited experience requires that an examination of WEC's environmental
issues be based more on careful projection than on field experience. However, even
758 Nontechnical Issue5
limited experience, coupled with a measure of sound logic, suggests that WECS systems
will present few of the traditional air and water pollution problems, even in the produc-
tion phase. The potential environment issues, summarized below, include safety consider-
ations, electromagnetic radiation interference, bird and insect collisions, and esthetic
issues associated with the appearance of wind structures.
A. Safety Considerations
Electromagnetic radiation interference may occur when signals reflected from moving
rotor blades interact with the original signals, causing fluctuations in signal frequency
and amplitude that degrade reception quality. Types of signals that may be affected are
in the FM radio, television, navigational air (navaid), and microwave frequencies at points
where geometries favorable for interference occur among the wind, turbine, transmitter,
and receiver. Other factors affecting the magnitude and severity of this impact include
blade area and speed, direct signal strength, and reflected signal strength relative to the
direct signal. Preliminary studies indicate that reception difficulties may occur as far
away as 400 m (0.25 m) for low-frequency (VHF) TV signals and 4800 m (3 mi) for
higher-frequency (UHF) signals [37] . ERDA's (now DOE) environmental assessment for
its Block Island, Rhode Island, experimental wind turbine site indicated that interference
with area residents' television reception could occur, and that mitigating measures might
be required.
Bird and insect collisions with towers or moving rotors can be a hazard to local or migra-
tory populations. Large bird kills (4000 to 7000 individuals killed in a single night) have
occurred via collision with large (350 m or 1000 ft) television towers (or their guy wires)
under conditions of darkness or poor visibility [39] . Night-flying birds may be disorien-
ted by, or attracted to, lights on tall structures, thereby increasing the possibility of fatal
collisions. The number of birds that may be lost through such collisions is a function of
flock size, pathway and height of migration, weather conditions, and the probability that
Environmental, Health, and Safety Issues 759
an individual bird entering the area swept by the rotor will collide with the blades (this
last item being a function of blade size and velocity and the angle at which the bird enters
the plane of rotation). Consideration of these elements when siting could reduce the pro-
bability of such collisions.
48.9 CONCLUSIONS
In summary, it would appear that, contrary to popular opinion, significant solar energy
environmental concerns do exist. Most of these concerns are not evidenced in the tra-
ditional "emissions" sense, but rather are less noticeable and tend to affect human health
and safety or delicate ecosystems (e.g., deserts). Perspective is important, however, and
it must be kept in mind that solar energy, in almost all cases, has considerably less impact
on the environment than other energy alternatives (e.g., fossil fuel systems).
Where environmental, health, or safety concerns do occur in relation to solar energy
manufacturing and deployment activities, mitigation and control strategies can be em-
ployed in most cases to ensure environmental acceptability. Where possible, these strate-
gies have been presented in the body of this chapter. The essential element is to recognize
the potential environmental, health, and safety impacts of a particular solar energy sys-
tem, and to design mitigation or control into the system before operation. In this way,
truly clean and safe solar energy can contribute significantly to world energy needs while
preserving the quality of the human environment.
A number of studies have been completed recently, or are presently underway to assess
further the environmental implications of solar energy, and to develop mitigating mea-
sures. The U. S. Department of Energy has undertaken the preparation of several "Envi-
ronmental Assessments" covering research, development, and demonstration programs in
solar thermal power systems, solar agricultural and industrial process heat, wind energy
conversion, photovoltaics, and ocean energy systems. Additionally, other agencies and
organizations, including the Environmental Protection Agency, the National Institute of
Occupational Safety and Health (NIOSH), the Solar Energy Research Institute (SERI),
and the National Bureau of Standards (NBS), are conducting studies aimed at the assess-
ment and mitigation of environmental effects of solar energy technologies. A principal
source of federal government research in this area is the Federal Interagency Energy/
Environment Research and Development Program coordinated by the U. S. Environ-
mental Protection Agency.
REFERENCES
1. National Bureau of Standards for the Department of Housing and Urban Develop-
ment, Intermediate Minimum Property Standards for Solar Heating and Domestic
Hot Water Systems, Interim Report, April 1976.
760 Nontechnical I ssues
21. Black and Veatch Consulting Engineers, A Study of Optical Radiation Hazards
Associated with a Central Receiver Solar Thermal Power Facility, prepared for the
Electric Power Research Institute, Palo Alto, California, 1977.
22. Martin Marietta Corporation, System Safety Design Criteria for the Central Receiver
Solar Thermal Power System, prepared for the U. S. Energy Research and Develop-
ment Administration, Washington, D C., 1977.
23. National Institute for Occupational Safety and Health, Criteria for a Recommended
Standard-Occupational Exposure to Crystalline Silica, PB-246-697, Rockville,
Maryland, 1974.
24. Salhana et al., American Journal of Medicine 59, 95 (1975).
25. L. T. Fairhall, Industrial Toxicology, New York, 1969.
26, U, S, Energy Research and Development Administration, Division of Solar Energy,
Solar Program Assessment: Environmental Factors, Photovoltaics, ERDA 77-47/3,
Washington, D. C., March 1977.
27. U. S. Environmental Protection Agency, Scientific and Technical Report on Cad-
mium, Washington, D. C., March 1975.
28, H. Perry et al., Archives of Preventative Medicine 3, 344 (1974).
29. U. S. Energy Research and Development Administration, Division of Solar Energy,
Environmental Development Plan-Photovoltaics, compiled by Energy and Environ-
mental Analysis, Inc., contract no. EG-77-C-01-4054, Washington, D. C., September
30, 1977.
30. National Institute of Occupational Safety and Health, The Registry of Toxic Effects
of Chemical Substances, 1976 ed., Rockville, Maryland, 1976.
31. J. F. Fraumeni et al., Lancet, p. 142 (July 25, 1975).
32. Chemical Rubber Company, CRC Handbook, Cleveland, Ohio, 1968.
33. Lockheed Missiles and Space Company, Ocean Thermal Energy Conversion (OTEC)
Power Plant Economic and Technical Feasibility, 2 vols., prepared for the U. S.
ERDA, April 1975.
34. TRW Systems Group, Ocean Thermal Energy Conversion, 5 Vol., prepared for U. S.
ERDA, June 1975.
35. USDA, Agricultural Research Service, personal communication, March 1977.
36. C. H. Wadleigh, Wastes in Relation to Agriculture and Forestry, USDA Misc. Publi-
cation No. 1065, March 1968.
37. U. S. Energy Research and Development Administration, Solar Program Assessment:
Environmental Factors-Wind Energy Conversion, ERDA 77-4716, March 1977.
38. U. S. ERDA, Division of Solar Energy, Wind Systems Branch, Environmental Impact
Assessments for Large Experimental Wind Turbine Generator Systems at Seventeen
Candidate Sites, 1976-1977.
39. S. E. Rogers, et al., Battelle Columbus Laboratories for NSF/ERDA, The Potential
Environmental Effects of Wind Energy System Development, August 1976.
40. J. Q. Searcy, ed., Hazardous Properties and Environmental Effects of Materials Used
in Solar Heating and Cooling Technologies: Interim Handbook, SAND 78-0842.
Sandia Laboratories, Albuquerque, New Mexico, August 1978,
Appendix A
CONVERSION OF UNITS
Area
1 acre = 4,047 m2 = 43,560 ft2
1 hectare = 10,000 m2 = 2.47 acres
1 mil = 2.59 km2 = 640 acres
1 ft2 = 0.0929 m2
Pressure
1 pascal (Pa) = 1 N/m2
1 lb/in.2 (psi) = 6.895 kPa
1 lb/ft2 = 47.88 Pa
1 atm = 101.3 kPa
1 in. of water = 248.8 Pa = 0.036 psi
1 in. of mercury = 3.377 kPa = 0.489 psi
1 torr (mm Hg, 0°C) = 133 Pa
Volume
1 ft3 = 0.0283 m3
1 gal (U.S. liquid) = 3.785 X 10-3 m3 = 3.785-1
1 gal (U.S. liquid) = 0.1337 ft3
1 acre-foot = 1233 m3
1 barrel (oil, 42 gal) = 0.1590 m3
Flow Rate
1 gal/min (gpm) = 6.309 X 10-5 m3 /s
1 ft3 /min (cfm) = 4.719 X 10-4 m3 /s = 0.472 1/s
1 cfm/ft2 = 0.3048 m3 /min • m2
1 gpm/ft2 = 0.0407 m3 /min • m2
Velocity
1 mi/h (mph) = 0.447 m/s
1 ft/s = 0.3048 m/s
Density
1 lb/ft3 = 16.02 kg/m3 = 0.016 g/cm3
1 lb/in.3 = 27.68 g/cm3
763
ENERGY
Mult iply
Nu giga- mega- kilo- megawatt kilowatt
co By
O ---E3Y -.---al of-f joules joules joules hours hours horsepower kg- g-
ob ain MBtu Btu (GJ) (MJ) (kJ) (MWh) (kWh) hours calories calories
MBtu 1 1.000(-6) 0.9478 9.478(-4) 9.478(-7) 3.412 3.412(-3) 2.544(-3) 3.969(-6) 3.969(-9)
Btu 1.000(6) 1 9.478(5) 947.8 0.9478 3.412(6) 3412 2544 3.969 3.969(-3)
gigajoules 1.055 1.055(-6) 1 0.001 1.000(-6) 3.600 3.600(-3) 2.684(-3) 4.187(-6) 4.187(-9)
megajoules 1055 1.055(-3) 1000 1 0.001 3600 3.600 2.684 4.187(-3) 4.187(-6)
kilojoules 1.055(6) 1.055 1.000(6) 1000 1 3.600(6) 3600 2684 4.187 4.187(-3)
megawatt hrs 0.2930 2.930(-7) 0.2778 2.778(-4) 2.778(-7) 1 0.001 7.457(-4) 1.163(-6) 1.163(-9)
kilowatt hrs 293.0 2.930(-4) 277.8 0.2778 2.778(-4) 1000 1 0.7457 1.163(-3) 1.163(-6)
horsepower hrs 392.9 3.929(-4) 372.5 0.3725 3.725(-4) 1341 1.341 1 1.559(-3) 1.559(-6)
kg-calories 2.520(5) 0.2520 2.389(5) 238.9 0.2389 8.600(5) 860.0 641.3 1 0.001
g-calories 2.520(8) 252.0 2.389(8) 2.389(5) 238.9 8.600(8) 8.600(5) 6.413(5) 1000 1
( 1 langley a 1 g-cal/cm2)
ENERGY DENSITY
2
Btu/ft 1.000(6) 1 3.587(-2) 8.805(4) 88.05 8.805(-2) 0.8805 317.1 1.224(-4) 3.687
gigajoules/meter2 11.35 1.135(-5) 4.071(-7) 1 0.001 1.000(-6) 1.000(-5) 3.600(-3) 1.389(-9) 4.186(-5)
kilowatt-hours
2 277.8 0.2778 2.778(-4) 2.778(-3) 1 3.861(-7) 1.163(-2)
meter 3154 3.154(-3) 1.131(-4)
kilowatt-hours
2 7.200(5) 720 7200 2.590(6) 1 3.012(4)
mile 8.170(9) 8170 293.0 7.200(8)
Multiply
by umber MBtu MBtu Btu megawatts kilowatts watts kilowatt-hrs horsepower q-cal 2:511
to -,,,_ f+ day hr hr (MW) (kW) (W) year day min
obtainy ----4 (kWh/Y)
MBtu/day 1 24.0 2.400(-5) 81.91 8.191(-2) 8.191(-5) 9.350(-6) 6.108(-2) 3.970(-9) 5.717(-6)
MBtu/hr 4.167(-2) 1 1.000(-6) 3.413 3.413(-3) 3.413(-6) 3.896(-7) 2.545(-3) 1.654(-10) 2.380(-7)
Btu/hr 4.167(4) 1.000(6) 1 3.413(6) 3413 3.413 0.3896 2545 1.654(-4) 0.2380
megawatts 1.221(-2) 0.2930 2.930(-7) 1 0.001 1.000(-6) 1.142(-7) 7.457(-4) 4.845(-11) 6.978(-8)
kilowatts 12.21 293.0 2.930(-4) 1000 1 0.001 1.142(-4) 0.7457 4.845(-8) 6.978(-5)
watts 1.221(4) 2.930(5) 0.2930 1.000(6) 1000 1 0.1142 745.7 4.845(-5) 6.978(-2)
kilowatt-hrs
year 1.069(5) 2.567(6) 2.567 8.760(6) 8760 8.760 1 6532 4.244(-4) 6.111(-5)
horsepower 16.37 392.9 3.929(-4) 1341 1.341 1.341(-3) 1.531(-4) 1 6.498(-8) 9.354(-5)
g-cal/day 2.519(8) 6.046(9) 6046 2.064(10) 2.064(7) 2.064(4) 2356 1.539(7) 1 1.440(3)
g-cal/min 1.751(5) 4.202(6) 4.202 1.433(7) 1.433(4) 14.33 1.636(4) 1.069(4) 6.944(-4) 1
POWER DENSITY
( 1 langley a 1 g-cal/cm2 )
Multiply
number MBtu Btu Btu watts megajoules gigajoules kilowatt-hrs kilowatt-hrs g-cal cl-cal
oi----- 2 2 2 2 2 2 2
ft year ft day ft hr meter m day m year m year m month cm day cm hr
to by
2 2
obtain (W/m 2) (MJ/m2 day) (GJ/m y) (kWh/m2y) (kWh/m mo)
4
sl!un Jo uo!SJ anuoj
megajoules/m2day 31.10 1.135(-2) 0.2724 8.640(-2) 1 2.740 9.861(-3) 0.1184 4.188(-2) 1.005
gigajoules/m2yr 11.35 4.142(-3) 9.942(-2) 3.154(-2) 0.3649 1 3.600(-3) 4.320(-2) 1.528(-2) 0.3667
kilowatt-hrs 3154 1.151 27.62 8.760 101.4 277.8 1 12.00 4.244 1.019(2)
2
m year
kilowatt-hrs 262.8 9.592(-2) 2.302 0.7300 8.446 23.15 8.333(-2) 1 0.3537 8.489
2
m month
2 2.064 23.88 65.45 0.2356 2.827 1 24.00
g-cal/cm day 743.1 0.2712 6.510
Multiply
value in_i Btu ' in Btu ' ft watts joules g-cal
ic;";-------_,Ipy 2 2
valuel,in ----------:":"-----
.„...._ ft hr °F ft hr °F cm °K m sec °K cm sec °K
2 0.5780
Btu ' ft/ft hr °F 8.333(-2) 1 57.80 242.0
HEAT CAPACITY
Multiply
value Btu joules g-cal
for lb °F g °K g °K
value inl
----,....itiply
Btu watts kilojoules g-cal
value
2 2 2 2
n ft hr °F m °K m hr °K cm sec °K
for by
2° 2
value in (W/m K) (kJ/m hr °K)
sh un to uo!sJanUo3
2
Btu/ft hr °F 1 0.1761 4.892(-2) 7374
kilojoules
20.44 3.600 1 1.507(5)
2 0
m hr K
g-cal
1.356(-4) 2.389(-5) 6.636(-6) 1
2
cm sec °K
FLUID TO AMBIENT TEMPERATURE DIFFERENCE
AT/I
INSOLATION
Multiply
°F °K °K °K °K
byvalue
for 2 2 2 2 2
Btu/ft hr kW/m W/m W/cm cal/cm min
value in .4, --------16
°F
1 5.675(-3) 5.675 5.675(-4) 8.139(-3)
2
Btu/ft hr
°K
176.2 1 1000 0.1000 1.433
2
kW/m
°K
2 0.1762 0.001 1 1.00(-4) 1.433(-3)
W/m
°K
2 1762 10.00 1.00(4) 1 14.33
W/cm
°K
2 122.9 0.6977 697.7 6.977(-2) 1
cal/cm min
Appendix B
Definitions
= latitude
a = solar altitude; measured from 0° (horizontal) to 90° (vertical).
ys = solar azimuth; measured from 0° (north) clockwise to 180° (south). (In
tables the [+] values apply to a.m. and the [- ] values to p.m.)
= solar incidence angle; 0° for normal incidence, 90° for grazing incidence.
Note: The values of a, ys, and 0 are calculated using the equations presented in
Chapter 3, Table 3.3. Because months with slightly different values of solar declination
are paired, the tabulated values, except for the equinoctial and solstice months, are
accurate only to about ±0.50 . For a tabulation by individual month, including latitudes
of 56° and 64° , the reader is referred to ASHRAE GRP 170, Applications of Solar
Energy for Heating and Cooling of Buildings, ASHRAE, 1977.
771
Table B.1 Latitude 0 = 24°N: Solar Altitude and Azimuth; Solar Incidence Angle for Horizontal and South-Facing Inclined Surfaces
Solar Time Solar Position Incidence Angle 0
Date/Declination a.m. p.m. a .-Lys Horiz. 0 - 10 + 10 0 + 20 Vert.
88.9
11 1 75.6 102.0 14.5 17.6 24.8 33.4 42.6 86.7
12 86.3 180.0 4.0 10.0 20.0 30.0 40.0 86.0
R = C -1 =
Note: The following tables of R values and U values are reproduced from The Solar
Home Book, by Bruce Anderson with Michael Riordan, Cheshire Books, Harrisville, N.H.
All values are in English units. For more complete tables see the ASHRAE Handbook
and Product Directory, 1977 Fundamentals.
781
R-VALUES OF BUILDING MATERIALS R-VALUES OF BUILDING MATERIALS
Material and Description (Ib/ft3 ) per inch r for listed Material and Description (Ib/ft3 ) per inch for listed
thickness thickness thickness thickness
Siding Materials • Representative values intended for use as design values of dry building materials in
normal use.
Shingles IK-values of acoustical tile depend upon the board and the type, size and depth of
Asbestos-cement 120 0 21 perforations, these are average values
Wood, 16" with 7 1/2" exposure - 0.80 2 Roof deck insulation is made in thicknesses to meet these standards, thickness may
Wood, double 16" with vary somewhat with manufacturer.
12" exposure - 1.19 3 Face brick and common brick do not always have these densities and R-values.
4
Wood, plus insulating Weights of blocks approximately 7-5/8" high by 15--3/8" long.
s Vermiculite, perlite, or mineral wool insulation
backer board 5/16" - 1 40
SOURCE ASHRAE Handbook of Fundamentals. 1967. Reprinted by permission.
Siding
Asbestos-cement lapped 1/4" - 0.21
Asphalt roll siding 0 15 R-VALUES OF AIR FILMS
Asphalt insulating siding 1/2" - 1.46 Type and Direction R-value for Air Film On:
Wood, drop (1" X 8") - 0,79
Orientation of Fairly Highly
Wood, drop (1/2" X 8" lapped) 0.81 Non-
Wood, bevel (M" X 10", of Air Film Heat Flow reflective reflective reflective
lapped) 1 05 surface surface surface
Plywood, lapped 3/8" - 0 59
Plywood 1/4" 0 31 Still air:
3/8" 0.47 Horizontal up 0.61 1.10 1.32
1/2" 0 62 Horizontal down 0.92 2.70 4.55
5/8" 0 78 45° slope up 0.62 1.14 1.37
3/4" 0.94 45° slope down 0.76 1.67 2.22
Stucco 116 0.20 - Vertical across 0.68 1.35 1.70
Sheathing, insulating board 1/2" 1.32 Moving air:
(regular density) 25/32" - 2.04 15 mph Wind any: 0.17 -
7% mph wind any' 0.25 -
Woods
Hardwoods (maple, oak) 45 0.91 - 'Winter conditions
t
Softwoods (fir, pine) Summer conditions.
32 1.25
SOURCE ASHRAE, Handbook of Fundamentals, 1972 Reprinted by permission
R-VALUES OF AIR SPACES U-VALUES OF WINDOWS AND SKYLIGHTS
I in units of Btu/hr/ft2 f F
2
double and triple refer to the number of lights of glass
3 nominal dimensions.
4
U-values for horizontal panels are for heat flow up in
winter and down in summer. 00
a based on area of opening, not surface 00
SOURCE ASHRAE, Handbook of Fundamentals, 1972.
Reprinted by permission.
SELECTED TABULAR AND GRAPHICAL DATA
INCLUDED IN HANDBOOK
Part A
786
787 Appendix D
Table 5.10 Mean daily availabilities of total radiation to south facing surface tilted
upward from horizontal at 15, 30 and 45°. 26 U.S. locations
Table 5.11 Mean daily availabilities of total radiation on south, east, and west walls
by season. 26 U.S. locations
Figure 6.1 U.S. map of new NOAA solar radiation network
Table 7.2 Classification of solar radiometers
Table 8.1 Characteristics of infrared measuring instruments
Table 9.4 Solar flat-plate collector performance factors and prices
Figure 10.36a Specular reflectance at 500 nm as a function of collection angular
aperture for several reflector materials
Table 10.5 Characteristics and price of various selective surfaces
Table 10.6 Nusselt number for natural convection between parallel plates
Table 10.9 Specular reflectance properties of several mirror materials
Table 10.10 Thermal and radiative properties of collector cover materials
Table 10.11 Temperature ranges for several heat transfer fluids
Table 12.4 Characteristics of plastics of interest for collector glazings
Figure 13.2 Solar transmittance of glazings as function of incidence angle
Figure 13.3 Effect of dirt on transmittance of glass as function of tilt and exposure
Table 13.1 Relevant properties of glasses for solar collectors
Table 13.2 Relevant properties of plastics for solar collectors
Table 13.3 Selective solar receiver coating types
Figure 16.14 Optical absorption coefficients as a function of photon energy for
a number of materials of interest for solar cells
Figure 16.15 Bandgaps of several semiconductors of interest
Table 16.1 Condensed table of physical properties of silicon
Table 16.3 Forbidden energy gaps for some materials of interest for photovoltaic cells
Table 20.1 Energy content and yields for various plant biomass
Figure 21.6 U.S. map of mean annual wind power
Figure 21.7 U.S. map showing seasons of maximum wind power
Figure 21.8 U.S. maps showing average wind power for winter, spring, summer, and
fall (1.8-1.11)
Table 22.1 Characteristics of some large wind generators
Table 25.2 Characteristics of some promising phase change materials
Part B
Table 34.4 Physical characteristics of fluid sensible heat storage materials
Figure 36.1 Distribution of total 1975 use of primary energy in the U.S.
Table 40.1 Tidal power sites around the world
Selected Tabular and Graphical Data 788
Table 40.2 Quantitative estimates of possible contribution of wind, waves, and tides
to world energy supplies
Figure 43.29 U.S. maps of mean daily insolation for each month of year
Table 43.1 Solar transmittance of various cover materials
Table 43.2 Solar absorptance of several absorber coating materials
Table 43.3 Average monthly and yearly degree days and winter design temperatures
for U.S. locations
Table 46.2 Formulae for the annualized before-tax cost of energy and the real internal
rate of return for industrial fossil fuel/solar thermal systems
Table 48.1 Potential toxic properties of representative solar working fluid additives
Table 48.2 Potential hazardous properties of common insulation materials
INDEX
A B
789
790 Index
K N
[Wind] [Wind]
present status, 454 Sweden, 456
programs, 454 three-blade propeller, 442
Province of Quebec, 455 two-blade propeller, 442, 455, 456
pumping water, 440, 459 Tvind, 456
qualitative characteristics, 455 value of energy from wind, 457
rated wind speed, 442 vertical axis wind machine, 441
sewage aeration, 459 vortex concentrators, 442
single-blade horizontal-axis rotors, West Germany, 456
442 wind machines, 440
Smith-Putnam wind generator, 440,
457 Z
storage, 457
Zomeworks Corp., 143