Low Cost Solar Thermal Energy Generation for

Developing Economies
Darick W. LaSelle, Robert Liechty,
Hassan Alzamzam; Robert Foster; Jasmin Dzabic; Nathan Clark
Department of Electrical Engineering, University of Colorado
Denver, CO, USA
Abstract Developing economies currently show more potential
benefit from renewable and off-grid technologies when
compared to first world countries like the United States. Basic
humanitarian needs like water pumping, water purification,
refrigeration for inoculations, cooking and heating without
depleting local resources can have a major impact on quality of
living. Once basic hygienic needs are accounted for, villages
and even small towns will have a more legitimate foothold to
develop out of an impoverished state. Unfortunately, the
majority of renewable solutions are complicated, and once
installed can easily fall into disrepair, with little or no local
knowledge base available to maintain complicated photovoltaic
(PV) or wind turbine systems. Additionally, the majority of
renewable energy systems require extensive manufacturing that
are not environmentally friendly. Once these systems are
manufactured, they are bulky and costly to ship. The system
described in this paper will have the ability to produce low cost
solar thermal energy, with a minimal of capital investment. The
system allows for the option to have the solar thermal collector
be built on site, requiring only a minimum of equipment to be
manufactured remotely and shipped. The proposed system will
allow for a much wider spread of renewable energy in
developing economies.

Figure 1: African Power Grid

The difficulty in bringing in renewable solutions is from
two main sources; cost and logistics. A commercially
available PV solution has a cost of $4.22/Wp, and is
extremely bulky to try to ship. [4] Once the solution is
installed, should there be any defects or problems with the
installation, there is rarely a technician qualified to repair it.
Basic knowledge of mechanical pieces and parts is a level at
which these villages can be expected to learn, but a PV
installation is not something a village can be expected to
maintain, especially when many families are struggling to
supply enough food for themselves to survive.
Wind generation can certainly be considered an option.
Wind generation costs have decreased dramatically over the
past 5 years, and the implementation of all except the most
modern of wind turbines is relatively simple. However, as
can be seen from Figure 2, the wind resource for Africa in
particular is minimal. [5] In the United States, a wind farm
would not even be considered without having at least a class 4
site. While the power requirements are certainly different,
there are virtually no sites above a class 2 in the majority of
the African continent, and even those are in coastal or
mountainous regions that traditionally have access to other
forms of energy.

Index Terms off-grid power; renewable energy; solar

power; solar thermal Stirling engine

I. INTRODUCTION

ENEWABLE energy and off-grid power generation can

serve as a vital development point for developing
countries. In places where basic hygiene and water quality
are poor, questions of power generation have very little to do
with grid integration and for more to do with basic
availability. Energy potential due to solar insolation in
countries such as Guinea-Bissau and Malawi is higher than
the best American sites (assumed to be Albuquerque, New
Mexico). [1] However, the traditional solar development in
the United States is predicated on large power inverters and
grid connections. Developing economies rarely have reliable
grids, and in remote villages are often not connected to a
gridl. As an example, in Bangladesh 47% of the population
does not have access to grid electricity at all. [2]
Remote villages in sparsely populated parts of developing
economies are not likely to see an infusion of capital
investment such that it can develop anything remotely
resembling a reliable grid. Off-grid technologies will be the
primary solution for these areas. Figure 1 shows the scarcity
of the existing power grid continent wide in Africa. [3]
978-1-4799-2402-8/13/$31.00 2013 IEEE

IEEE 2013 Global Humanitarian Technology Conference

The profile view of the dish is a perfect parabola. A

parabolic dish is the only shape that will have all vertical rays
converge to a single focal point. The traditional construction
of a parabolic dish concentrating solar power (CSP) is to have
a nearly flat dish reflect to a heat absorption point some
distance away from the dish. Using a steeper reflector allows
the dish to reflect to an internal focal point. This dish has the
profile equation shown in Formula 1. The dimensions given
are all in meters.
(1)

Figure 2: Wind Resource for Africa

The solution proposed here is to use a pre-manufactured
Stirling engine coupled with a permanent-magnet dc engine
and insert it into a parabolic dish that can be built on site with
a variety of materials depending on the locality. This will
serve to significantly reduce the cost of manufacturing and
transportation, as well as provide a method of developing the
knowledge to maintain the generator. This paper will outline
the design of the generator, as well as develop a cost (both
monetary and energy-based) analysis of the generator.

The focal point of this dish sits at 0.125 m. Noting that the
height of this dish is 0.5 m, the focal point is significantly
inside of the dish. This method of heat collecting greatly
simplified the construction of the overall generator. Whereas
a separate structure would have to have been created to accept
the heat from the collector, this simply requires a hole to be
cut to insert the Stirling engine below the dish with only the
heat absorber sticking through into the dish, as shown in the
conceptual rendering in Figure 4.

II. PROOF OF CONCEPT DISH CONSTRUCTION

The proof of concept included a dish with a 1 m diameter.
This size determination was made from a portability
standpoint, as well a local resource standpoint. Specifically,
it was made as big as it possibly could and still fit out the
door of the building. Figure [3] is a picture proof of concept
dish immediately before testing.

Figure 3: Proof of Concept Dish

Figure 4: Parabolic Dish and Stirling Engine

configuration
The stand is of equally simple construction. A basic box
frame was made out of 2x4s, with a hinged top utilizing
standard door hinges. Four casters attached to the bottom
allow for the entire device to be moved with very little effort.
A single adult can easily maneuver the device, and set its
altitude angle (azimuth).

Figure 5: Early Construction with Plywood Rings

The dish itself is made out of a single piece of 8 x 4 x 5/8

plywood. The piece was divided into two 4x4 squares.
Using the width of the plywood as the step size in the y
direction, the corresponding x value is equal to the radius of
each ring. By nailing a T-Square to the center of each half, a
set of concentric circles could be drawn that would be very
accurate for this scale. Each step ring was drawn on
alternating sides, so that by combining the two sides, a
standalone dish could be made that wouldnt collapse in on
itself. For this proof of concept, wood glue proved more than
adequate. Figure 5 shows several of the cut-out rings.
The advantage of this style of construction is that is does
not require any expensive manufacturing setups, and thus can
be completed almost anywhere in the world. Figure 6 shows
one of the authors cutting out the rings with a jigsaw.

will lead to a central hot spot. This makes the heating of the
fluid inhomogeneous, which leads to more loss. This
configuration will allow a more evenly distributed heat
conduction through the absorber. Additionally, once the
cylinder reaches operating temperature, it will have a higher
surface area to volume ratio, which increases the rate of
convection.
More to the point of providing a sustainable solution to
remote villages, the Stirling engine selected for this project is
a very simple device. It is made of mostly black iron, and
notably requires no silicon based manufacturing, which
eliminates a significant environmentally harmful waste
product when compared to the manufacturing of PV Panels.
While the Stirling engine CSP method is not the most
popular method of generating energy from solar insolation, it
is gaining traction due to its more environmentally friendly
manufacturing process. [7] Using the System Advisor Model
(SAM) developed by National Renewable Energy Labs,
thermal solar farms have a mild improvement over PV farms
when comparing output efficiency and capacity factor.
However, the significant cost savings proposed by this paper
are confirmed by the Levelized Cost of Energy. The SAM
simulation results are shown in Table 1. This evaluation
doesnt even take into account the savings in transportation to
a remote village.
Table 1: NREL-SAM Solar Comparison Results

Figure 6: "Manufacturing" the Proof of Concept Dish

III. STIRLING ENGINE SELECTION

The Stirling cycle harnesses a differential temperature to
cyclically change the pressure of a working fluid. Using a
piston and typically a flywheel, this change in pressure can be
converted to rotational mechanical energy and then into
alternating current. In the proposed application, a dual piston
single-cylinder Stirling engine would be used, which is
known as the beta style. Use of the beta style allows for
insertion of the engine through the bottom of the parabolic
dish. This allows the heat absorber part of the Stirling engine
to be contained within the parabola while the heat dissipation
part is outside. The first efficiency gain will be the ability to
use the entire reflective area of a dish. Currently, 10-15% of
the usable thermal reflective area is covered by the apparatus
holding the Stirling engine and the generator.
An additional advantage of this configuration will be the
geometry of the heat absorber. In an extended focal point
configuration, the heat is concentrated on one flat plate which

100
MW
Solar
Farm Boulder, Co

PV

Stirling

Year 1 Annual
Energy (kWh)

145,858,781

153,495,000

LCOE (cents/kWh)

14.5749

10.1961

Capacity factor

17%

18%

DNI to Electric
Grid Output
Correlation

88%

99%

While the Stirling engine technology is a relatively old

concept, harnessing solar energy as the primary fuel source
for the Stirling engine is not. The solar powered Stirling
engine was patented by Roelf J. Meijer in 1987 and provided
the necessary framework for researchers and scientists to
study the techno-economic feasibility of the solar Stirling
technology against other traditional forms of renewable
technologies. There are several different types of solar
renewable technologies readily available today but the solar
powered Stirling engine finds its market adoption most
readily challenged by photovoltaics (PV), both in terms of
total installed cost and deployability.
While PV has enjoyed the continued success of being
recognized as the deployable technology of choice, there are
many lines to be drawn still between where PV may not be
the most economical wise decision when compared with solar

Stirling from an economic value chain perspective. PV, in

recent years, has experienced rapid declining module prices
due in large part to cheaper manufacturing costs and
subsidization from China. China is largely considered to be
the epicenter of efficient manufacturing processes and
materials for high grade electronics and silicon which also
influences the declining price of PV.
While PV may make economic sense in a highly
industrialized country like the U.S. with easy access to ports,
transportation, and infrastructure third world countries may
not enjoy these same luxuries. It has been extensively shown
that there is a direct correlation between a countrys world
standing and the amount of energy per capita that the country
has available to it. Infrastructure, manufacturing facilities and
transmission lines are very energy intensive with high upfront capital costs that third world countries cannot afford.
The energy solution for these countries is easily deployable,
cheap, and technologically simple energy generation that is
located near the fuel source. PV severely lacks in a couple of
these energy defining generation categories, namely, being
technologically complex which requires high levels of
manufacturing processes and precision as well as the
technology itself being expensive for undeveloped countries
due to a lack of infrastructure to transport PV in bulk to
remote areas.
Aside from the economic and logistical considerations
between PV and solar Stirling there are still a few solar
resource considerations to acknowledge as well when
examining third world countries. Most third world and
developing countries are located in regions with high solar
irradiance with minimal cloud cover between the tropics of
Cancer and Capricorn. Using the System Advisor Model
(SAM) developed by the National Renewable Energy
Laboratory (NREL) it can be shown that solar Stirling
performs better in conditions that consist of high ambient
temperatures, and high direct normal solar irradiance.
Using SAM and select locations within the U.S. the below
figures make a strong argument why solar Stirling is more
correlated than PV when comparing energy production and
direct normal irradiance. All systems modeled are for a
nameplate capacity of 100 MW.

IV. EFFICIENCY ESTIMATES

The world record for solar energy to grid power efficiency
is 31.4% [10]. For the sake of comparison, 25% efficiency is
assumed to be good (PV panels typically run 14-20%). The
assumptions below use a baseline of 25% efficiency for
existing Stirling based generators. The percent changes are
applied to that base.
As a basis of comparison, with solar insolation assumed to
be 1000 W/m2, a 25% efficient generator would output 250
W.
Mirror based dishes use a dielectric coating which can
achieve a reflectivity of 99%. With a reflectivity of 95%, the
proposed design is approximately 4% less efficient.
The heat transfer coefficient is inversely linear to the area,
which is, in turn, inverse to the heat convection. Therefore,
the heat losses are magnified by heating a single point of
single surface; as opposed to multiple surfaces, as this design
will. As thermal losses account for the majority of the overall
losses, this change will have the greatest impact on efficiency
approaching a 15% gain.
As discussed earlier, using the footprint more effectively
will also increase the efficiency. For the sake of this analysis,
a conservative 7% efficiency gain will be used.
The overall improvement in efficiency will be:
(3)
Which would make the good system of 25% energy
efficiency improve to 29.5%. Table 1 summarizes the
efficiency calculations.
Parameter

Efficiency Change

Baseline 25%
Reflectivity
Heat Losses
Footprint
Proposed

NA
-4%
+15%
+7%

Multiplication
Factor
0.25
0.96
1.15
1.07
0.295

Table 2: Efficiency Estimates

V. EROEI AND EROI
In objectively considering renewable energy there are two
calculations that need to be taken in account: energy return
on energy investment (EROEI), and energy return on
investment (EROI). EROEI is used to indicate whether an
energy technology is a viable solution for replacement of
fossil fuel technologies. [13] EROI is an economic indicator.
Both calculations allow a comparison of various renewable
technologies to each other.
A. EROEI

Chart 1: Photovoltaic versus Solar Stirling Comparison

The calculation of EROEI is often in dispute, so this paper

will first explicitly define what equation we will use for
comparison [13]. The specific dispute is that the traditional
definition of EROEI is based on fossil fuels, and it is a
calculation of how much energy it takes to extract and
produce that fuel [14]. Because renewable energy sources do

not have an energy cost associated with its source, it uses the
production cost instead. Additionally, an assumed life span
of the device is factored into the equation. A higher EROEI
is better than a lower one, and an EROEI of over 40 would be
considered a technology suitable to replace the burning of
fossil fuel. The calculation used is therefore:
(4)
The energy produced in one year in a simplified format is:
(5)
The approximate solar insolation energy in Denver,
Colorado for one year is 72.33 kW/m^2, which is
approximately 260 MJ when the size of the dish is accounted
for [15]. The assumed energy produced is then 21.3 MJ per
year for our 29.5% efficient estimation.
In order to estimate the energy cost of producing the device
we used the approximate energy cost of the material. The
EPA estimates that woods such as plywood require 6
mmBtu/ton of primary energy to manufacture [16]. That
converts to 6.97 MJ/kg, giving a production energy cost of
this design of 71.6 MJ.
Using this method, the EROEI for this design is 8.94. This
is compared to fossil fuel EROEIs of 15-35, and wind EROEI
of 18 [14].
An additional method to look at the EROEI is to look at the
energy breakeven point. For this design, that point is at
approximately 3.33 years. While neither of these calculations
show this design to be competitive with fossil fuel generation,
it is significantly better than photovoltaic.
B. EROI
EROI measures the economic impact of an energy system by
indicating how much it costs in production to achieve a Watt
of peak power. In contrast to EROEI, a low number is better.
A standard calculation of EROI in renewable energy sources
is:

VI. RESULTS/FINAL DESIGN PROPOSAL

The proof of concept dish did indeed generate usable
electrical power. The measurement in late March in Denver,
CO gave a focal point temperature of 237 C. Using the ideal
Carnot cycle, this gives a theoretical transfer efficiency of
45%, which was right in line with our estimates. However,
due to the starting torque of the DC generator, the initial
trials yielded a 20% efficient system, well below initial
estimates. The lesson learned for this proof of concept was
that small inefficiencies make a large difference in a small
system.
The fix for this is still well within the scope and purpose of
this dish. Specifically, none of our equipment neared a nonlinear state, meaning that a much higher heat could be used.
Therefore, using the exact system we used for the proof of
concept would make for a very effective model with a larger
dish. The thermal energy gained from going to a 2 meter
diameter dish will overcome all of the remaining issues, and
they are all simply related to thermal momentum. A two-fold
increase in thermal energy should result in a 4-fold increase
in power output, effectively doubling the overall system
efficiency.
VII. CONCLUSION
This paper introduced the concept of using a solar thermal
device to generate power in remote areas of developing
economies. It described the effective efficiency, and how it is
theoretically higher than existing, large-scale solutions.
As a group, we designed and built a proof of concept
system using low-cost materials. The explicit instructions on
how to build this dish can be used to reproduce generators
inexpensively in villages and remote outposts, simply by
using materials readily available to the locality. While the
materials used to build the proof of concept consisted of
mainly wood, substitutions can be made to use clay, earth, or
even stone.
A final design was proposed that modified the proof of
concept only in scale. This power generation system should
prove very effective in bringing electricity to developing
economies that need it most.

(6)
VIII. ACKNOWLEDGMENTS
As an attempt to quantify a manufacturing cost, this
analysis takes the cost of the Stirling engine proposed
(~$300) and adds it to a common 1 [m] wide RF Antennae
Dish (-$100). Using earlier assumptions (1000 W/m 2), this
design would output approximately 295 W.
EROI would be $1.36/Wp. $2.5/Wp is considered
economically viable [6]. A comparable PV panel on the
market today costs $400 and generates 250 W, giving an
EROI of $1.60/Wp [17], making the solar Stirling a much
better option.

We would like to gratefully acknowledge Dr. Mark

Golkowski, Dr. Keith Malmedal, and Dr. Jae-Do Park for
their useful conversations and help in reviewing our
published work. We would additionally like thank to the
program advisors for this project, Robert Grabbe and Brian
Atkinson, for their guidance and support.

IX. ABOUT THE AUTHORS

The authors are all currently undergraduate students at the
University of Colorado, in Denver, CO.

X. REFERENCES
[1] Solar Irradiance. Internet:
http://solarelectricityhandbook.com/solar-irradiance.html
[May 2, 2013]
[2] Amin, Narima; Langendoen, Richard, "Grameen Shakti: A
Renewable Energy Social Business Model for Global
Replication," Global Humanitarian Technology Conference
(GHTC), 2012 IEEE , vol., no., pp.324,327, 21-24 Oct. 2012
[3] Global Energy Network Institute, Map of African
Electrical Grid. Internet:
http://www.geni.org/globalenergy/library/national_energy_
grid/africa/africanelectricitygrid.shtml [May 5, 2013]
[4] Solar Panel Kit. Internet:
http://www.harborfreight.com/solar-panel-kit-45-watt68751.html [May 13, 2013]
[5] Global Energy Network Institute, Renewable Energy
Resources.
http://www.geni.org/globalenergy/library/renewableenergy-resources/world/africa/wind-africa/index.shtml [May
5, 2013]
[6] Norton, B.; , "Renewable electricity-what is the true
cost?," Power Engineering Journal , vol.13, no.1, pp.6-12,
Feb. 1999
[7] Faraz, T.; , "Benefits of Concentrating Solar Power over
Solar Photovoltaic for power generation in Bangladesh,"
Developments in Renewable Energy Technology (ICDRET),
2012 2nd International Conference on the , vol., no., pp.1-5,
5-7 Jan. 2012
[8] National Renewable Energy Laboratory, System Advisor
Model.Internet: https://sam.nrel.gov [May 8, 2013]
[9] "Sandia, Stirling Energy Systems Set New World Record
for Solar-to-grid Conversion Efficiency - February 12, 2008."
Sandia, Stirling Energy Systems Set New World Record for
Solar-to-grid Conversion Efficiency. Sandia National Labs, 12
Nov. 2008. Web. 13 Dec. 2012.
[10] "Optical Mirror Selection Guide." Optical Mirror
Selection Guide. Newport Corporation, 2012. Web. 13 Dec.
2012.
[11] Mancini, T; Heller, P; Butler, B; et. Al; , "Dish-Stirling
Systems: An Overview of Development and Status," ASME,
Journal of Solar Energy Engineering, vol. 125, pp. 135-151,

May 2003.
[12] Murphy, T. "The Energy Trap." Do the Math. UC San
Diego, 18 Oct. 2011. Web. 14 Dec. 2012.
[13] Murphy, D.J.; Hall, C.A.S. (2010). "Year in review EROI or
energy return on (energy) invested". Annals of the New York
Academy of Sciences 1185: 102118.
[14] "Surface Meteorology and Solar Energy." Surface
Meteorology and Solar Energy. National Aeronautics and
Space Administration, n.d. Web. 14 Dec. 2012.
[15] Biddle, M. "Recycling the Hard Stuff." Recycling the
Hard Stuff. Environmental Protection Agency, 1 July 2002.
Web.
[16] "250-Watt Monocrystalline Solar Panel." 250-Watt
Monocrystalline Solar Panel-GS-S-250-Fab5 at The Home
Depot. The Home Depot, n.d. Web. 14 Dec. 2012.