The following article was published in ASHRAE Journal, July, 2001.

© Copyright 2001 American Society of Heating, Refrigerating and Air-

Conditioning Engineers, Inc. It is presented for educational purposes only. This article may not be copied and/or distributed electronically or in
paper form without permission of ASHRAE.

ASHRAE Journal Chillers

Deregulation and
Chiller Plant Load
By William Ryan, Ph.D., P.E. of using low capital cost electric equip-

T
he world of the electric customer is undergoing drastic change. The
chiller plant decisions made today will affect the owner’s costs for
the next 20 years. To understand the risks owners face, it is neces- Efficiency
sary to look at the electric industry as it has existed and its future.
The Customer and the Supplier
The relationship between the electric
supplier and the customer will undergo a
radical change in the next few years. In
the past, under complete regulation:
• The electric customer had only one
possible source for all electric services.
• Electric service was comprehensive
(electricity, billing, delivery, etc., were all
provided in one uniform package).
• The regulated electric supplier had
an obligation to serve the customer, no
matter what the characteristics of the load.
• The local regulatory commission
controlled all components of the price of
electricity.
Under complete deregulation of elec-
tricity, the customer should see:
• Many vendors offering electric sup-
ply contracts,
• Electric supply contracts that offer
many options and features,
• Deregulated electric vendors vying
for the “customer relationship,”
• Local regulated electric systems sup-
plying the delivery service as part of any
vendor’s package, and
• Deregulated vendors that do not
have an “obligation to serve.”
Although electric deregulation may of-
fer the customer new options and lower
overall prices, the end of the “obligation
to serve” is the quid pro quo in this
change. The economics of electric deregu-
lation means vendors have little incentive
to give the best prices to undesirable cus-
18 ASHRAE Journal
Improvements
It is unlikely that most common effi-
ciency improvements will enhance the
tomers (customers whose facilities have
low load factors, which lowers the load
factor and the profitability of generation).
In the past, such customers were car-
ried under the overall regulated rate base.
However, this effectively transferred costs
from low-load factor customers, such as
office buildings and other daytime, week-
day-only commercial loads to high-load
factor customers, such as industrial facili-
ties. It is not surprising that industrial elec-
tric customers have been a major force in
promoting electric deregulation.
Out of tradition and habit, some elec-
tric suppliers may temporarily continue
to market to all customer classes regard-
less of load factor. However, in a com-
petitive environment, other suppliers will
move in quickly and “skim” their high-
load factor customers—leaving behind
only the customers that are expensive to
serve. The tendency is clear. Different
classes of customers will pay differing
rates depending on the desirability of their
load to a free market supplier.
The bottom line is that customers must
make their load profile desirable by im-
proving poor load factors.
Designing During Regulation
In the past, customers exploited this
unconditional obligation to serve, and
unless demand charges were high, largely
ignored load factor as an element in their
costs. Customers developed poor load
profile habits spurred by the convenience
www.ashraejournal.org
ment even on the most intermittent of
loads, and the dominating intermittent
load is cooling.
load factor.
• Most commercial buildings are
equipped with fluorescent lighting, al-
though electronic ballasts, high-efficiency
fixtures, and LED exit signs may reduce
overall lighting loads. However, in gen-
eral, load reduction will not be large.
• Occupancy sensors that control
lights are useful in reducing total energy
consumption. However, they tend to only
reduce the load factor further because
they mainly ensure that lighting is shut
off at night.
• Tenants expect to be able to plug a
load into a receptacle and have it avail-
able at any time.
• Tenants in buildings tall enough to
require elevators are sensitive to any el-
evator shutdown.
• Air distribution is likely handled by
a reasonably efficient VAV system. Re-
ducing ventilation minimums or outdoor
air intake is not popular due to indoor air
quality concerns.
Overall, the only remaining target for
improved load factor is the chiller load.
As long as the air system continues to
provide cooling and dehumidification,
tenants are not concerned about the in-
ner workings of the chiller system. Be-
cause chiller alternatives exist, the chiller
About the Author
William Ryan, Ph.D., P.E., is a se-
nior engineer at the Energy Resources
Center at the University of Illinois at
Chicago.
July 2001

Figure 1: Note the dramatic rise in July/Aug. price. This trend
has appeared consistently over the last four years. Prices are
wholesale at the exchange point, and do not include subse-
quent transmission and delivery costs. COB is California –
Oregon Border, PJM is the Pennsylvania, New Jersey, Mary-
land Power Exchange. Data from the New York Mercantile
Exchange as of 1/2001.
plant design and the cooling designer are on the line to provide
a solution to the problem of load factor.
What Does this Mean for Cooling Plant Design?
Let us focus on the chiller plant load for a typical 1,000 ton (3517
kW) office building using two 500-ton (1759 kW) chillers.
The designer can choose from an array of chiller alternatives
to control load factors in the chiller plant including:
• High-efficiency electric chillers,
• Electric chillers with improved load-following capability,
• Gas-fired absorption chillers,
• Gas-fired engine chillers,
• Chilled water storage, and
• Ice storage.
The following describes these options.
High-Efficiency Electric Chillers
High-efficiency electric chillers are available with electric con-
sumption values as low as 0.48 kW/ton (0.14 kW/kW), as com-
Advertisement for the print edition
formerly in this space.
July 2001
Chillers
Figure 2: Electric peak demand for a typical 1,000 ton office
building. Unlike energy use, the chiller has a large effect on
monthly peak demand.
pared to the 0.6 to 0.68 kW/ton (0.17 to 0.19 kW/kW) of stan-
dard efficiency chillers. These chillers can cost as much as 50%
more than standard efficiency chillers.
Given the expense, high-efficiency chillers are best used as
lead chillers in a multi-chiller plant to maximize operating hours
and electric energy savings. Unfortunately, peak demand reduc-
tion will only be in the 20% to 25% range for the lead chiller. For
load leveling or demand charge avoidance, high-efficiency elec-
tric chillers are not effective per unit of cost. If a high-efficiency
chiller (0.54 kW/ton [0.15 kW/kW]) were used to replace stan-
dard efficiency chillers (0.65 kW/ton [0.18 kW/kW]) the $70 to
$80/ton premium would amount to roughly $75,000. Chiller
power demand would drop 110 kW, and all other factors being
equal, demand reduction would cost $680/kW, higher than any
other alternative posed. The results of similar calculations are
listed on Tables 1 through 3.
Electric Chillers with Improved Load Following
Given the variable nature of the cooling load, improved
efficiency at low load will significantly decrease overall elec-
tric energy use. Technical solutions include variable-speed
drives, screw compressors with slide valves, and multi-com-
pressor chillers. Although significant electric energy savings
may be achieved, these chillers are not a solution to reducing
ASHRAE Journal 19
ASHRAE Journal
Plant Type Total Demand Cost of Demand Reduction Cost of Demand
Lead Chiller Lag Chiller Demand
All Electric 800 kW1 Base Case Reduction
High-Efficiency 500-ton SFSE
690 kW $680/kW 500-ton Electric 525 kW $289/kW
Electric Absorber
500-ton DFDE
All DFDE1 500-ton Electric 525 kW $408/kW
250 kW1 $413/kW Absorber
Absorption
500-ton Engine
All Engine 200 kW1 $475/kW 500-ton Electric 475 kW $488/kW
Chiller
Full Ice Storage 0 kW $611/kW 150-ton Ice
500-ton Electric 550 kW $485/kW
Cost of Onsite Storage Chiller1,2
— >500/kW 500-ton DFDE 150-ton Ice
Generation 270 kW $448/kW
Absorber Storage Chiller
Table 1: Comparative cost of reducing demand with various
500-ton Engine 150-ton Ice
chiller types. Each technology is assumed to be used through- Chiller Storage Chiller
230 kW $486/kW
out the chiller plant to meet all cooling loads for a 1,000-ton Cost of Onsite
— — >$500/kW
(3517 kW) office building. Usage includes allocations for tower Generation
pumps, fans and other auxiliaries. The chilled water system Table 2: Total cost of reducing demand with various chiller
is assumed to be identical in all cases and chilled water pumps types. In this table, two different types of chillers are as-
are not included. DFDE = Direct-Fired Double Effect.* sumed in a hybrid plant to meet all cooling loads. Note that
the ice storage option has been optimized 1 but still does not
demand, as they will have similar and in some cases, higher surpass the lower cost of using absorption chillers.*
electric demand at full load than a standard chiller.
Cost of Operating
Absorption Chillers Lead Chiller Lag Chiller Demand Cost Today3
Reduction ($/Yr)
Eliminating the electric drive from the chiller will eliminate 500-ton Electric 500-ton Electric Base Case $92,722
the chiller’s electric demand. Absorption chillers need electricity 500-ton Absorber 500-ton Electric $408/kW $63,022
to operate solution pumps, in the 0.05 kW/ton (0.01 kW/kW) 500-ton Engine
500-ton Electric $488/kW $53,026
range. Traditionally, absorbers also used larger cooling towers Chiller
and larger condenser water flows, in the 4.5 gpm/ton (0.08 mL/J) 150-ton Ice Storage
500-ton Electric $485/kW $66,482
Chiller
range (double-effect absorbers). However, some newer double ef- 150-ton Ice Storage
fect absorber designs are advertised at their capacity at 3.6 gpm/ 500-ton Absorber $448/kW $36,782
Chiller
ton (0.06 mL/J) and can operate as low as 3 gpm/ton (0.05 mL/J) 500-ton Engine 150-ton Ice Storage
$486/kW $27,020
with little loss of capacity. Auxiliaries for heat rejection are gener- Chiller Chiller
ally on the order of 0.2 kW/ton (0.06 kW/kW) rather than the 0.15 Table 3: Total cost of reducing demand and operating cost
kW/ton (0.04 kW/kW) assumed for electric chillers. Absorption estimate with various chiller types. In this table, two differ-
chillers are more expensive than electric chillers. ent types of chillers are assumed in a hybrid plant to meet all
cooling loads. Note that the use of a mix of technologies
Engine Chillers reduces demand at a lower cost than using one technology.*
Engine chillers are gaining popularity because of their low
operating costs and demand avoidance. Heat rejection from an is at such a high temperature that condenser water flows may be
engine chiller is higher than for electric chillers, due to heat left at electric chiller levels, once again with higher temperature
removed from the engines cooling jacket. However, engine heat water sent to the tower. Of course, this high-grade heat from the
engine, typically above 200°F (93°C) also may be rejected to
* (ton x 3.517 = kW) water heating or some other practical application.

Advertisement for the print edition

formerly in this space.

20 ASHRAE Journal w w w. a s h r a e j o u r n a l . o r g July 2001


Chilled Water Storage
Storage technologies have been gaining popularity as they ad-
dress the load leveling issue head on, focusing entirely on demand
reduction. Storage of chiller water allows the chiller to operate at
the same efficiency level as a standard chiller. Unfortunately, chiller
water storage requires sizable storage tanks, and significant con-
struction costs, making these systems expensive.
Ice Storage
Ice storage reduces the size of the storage tanks needed for
thermal storage and, like chilled water storage, it is tailored to
reducing demand costs. However, the plant must be operated
properly to guard against running out of ice on days when capac-
ity is needed. Also, the chillers themselves will consume 0.9 to
1.0 kW/ton (0.25 to 0.28 kW/kW),2,4 substantially more power
than the 0.65 kW/ton (0.18 kW/kW) of standard chillers. Finally,
the low evaporator temperatures will reduce the output capacity
of chiller equipment to 60% to 70% of normal rating for 45°F
(7°C) chilled water production.1
Distributed Generation
The obvious solution to the demand problem may seem to
include generation into the facility to shave load as needed.
Unfortunately, this tends to be a high-cost solution—in the $500
to $600 per kW range when the required paralleling gear is in-
Advertisement for the print edition
formerly in this space.
July 2001
Chillers
cluded. Overall, using a generator to shave load while maintain-
ing the full complement of electric chillers is not practical.
The cost-effective use of on-site generation involves integrat-
ing the generator into the facility. The generator serves the elec-
tric base load (lights, plug loads, etc.), which has high operating
hours per year. Heating and cooling can be accomplished with
waste heat from the prime mover directly.
Assessing Risks
In a traditional energy analysis, costs are calculated, and a
payback or rate-of-return is found for differing options. How-
ever, the operating cost assessment is dependent on local en-
ergy prices making the analysis limited to a specific geo-
graphic area and to the energy prices available from today’s
regulated, or in some cases partly deregulated energy market.
To look to the deregulated future, the following analysis takes
a standard all-electric chiller plant as a base and then analyzes
differing technologies for their cost of peak demand avoid-
ance. What is the cost of using differing approaches to reduce
peak demand in dollars per kilowatt? This gives plant design-
ers an indication of the cost structure of various approaches
and is not limited to a specific geographic or utility area. The
results of these calculations are shown in Tables 1 and 2 for a
1,000-ton office-building load. Table 3 includes a sample op-
erating cost for a specific utility rate.
ASHRAE Journal 21
ASHRAE Journal
Conclusions plant away from the strict electric chiller fo- this chiller on a 24-hour basis to cover the 500-
The high-demand profile of today’s cus of the past to a larger universe including ton (1759 kW) lag chillers duty, thus reducing the
commercial buildings leaves the facility widely available but generally less often cost of the system.
2. Dorgan, C.E. and J.S. Elleson. 1993. De-
owner at considerable risk in the deregu- used thermally-driven equipment. sign Guide for Cool Thermal Storage. Atlanta:
lated future. Many options can reduce that ASHRAE.
risk. However, some of the most cost-ef- Notes 3. Assumes $0.06/kWh, $15/kW/month de-
fective of these options are available equip- 1. The ice storage system was optimized in mand charge, and $0.30/therm charge for gas.
ment used in new ways rather than radi- cost and demand reduction by installing a 150- 4. 1995 ASHRAE Handbook—HVAC Appli-
cally new systems. Some specific conclu- ton (528 kW) ice-storage chiller and operating cations p. 40.13.
sions include:
• The cooling plant is the most impor-
tant and controllable factor in the overall
load factor of a commercial building.
• The cost of electric generating sys-
tems including the needed paralleling gear
($500 to $600/kW) suggests that these sys-
tems be used to reduce the demand of those
loads that must be electrically driven such
as lights and plug loads.
• The use of full ice storage systems—
recharged only at night—will reduce cool-
ing demand dramatically. However, the
cost is high—more than $600/kW.
• If ice storage systems are used only as
a lagging or peaking chiller and are opti-
mized in first cost by reducing chiller size
and charging continuously, the cost of de-
mand reduction still remains in the $480/
kW range.
• Whether used for all or part of the load,
direct-fired double-effect absorption chill-
ers (DFDE) can reduce demand in the $400–
$420/kW range, making these systems a Advertisement for the print edition
logical choice for demand reduction. These formerly in this space.
systems also provide comparable or lower
operating cost in most areas, which makes
them more attractive.
• Although engine chillers are more ex-
pensive for demand reduction purposes
than absorption chillers or an optimized ice
storage system, their high efficiency and
the low cost of gas fuel used generally
makes the operating cost of engine chillers
low. This combination of low operating cost
and demand reduction makes engine chill-
ers desirable in many locations.
• The cheapest mechanism for demand
avoidance is the single-effect absorber in
the $280–$300/kW range. However, due
to the higher operating cost of this option,
a single-effect absorber should be used as
a lag chiller, unless the chiller is powered
by waste heat.
Overall, these conclusions indicate that
finding a technical solution to the problem
of risk exposure for the large commercial
customer will require diversifying the chiller
22 ASHRAE Journal w w w. a s h r a e j o u r n a l . o r g July 2001