University of Southern California Dissertations and Theses

Closed landfills to solar energy power plants: estimating the solar potential of closed landfills in California

(USC Thesis Other)

Closed landfills to solar energy power plants: estimating the solar potential of closed landfills in California

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

CLOSED LANDFILLS TO SOLAR
ENERGY POWER PLANTS:
ESTIMATING THE SOLAR POTENTIAL OF
CLOSED LANDFILLS IN CALIFORNIA

by

Devon R. Munsell

____________________________________________________________

A Thesis Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(GEOGRAPHIC INFORMATION SCIENCE AND TECHNOLOGY)

May 2013

i

Acknowlegements
I would like to take a moment to show my appreciation to the chair of my committee, Dr.
Robert Vos, for his supervision throughout this process. Additionally, I would like to thank Dr.
Darren Ruddell and Dr. Jennifer Swift for their help as committee members for this work. This study
was made complete by the guidance and wisdom of each committee member.

ii

Table of Contents
Abstract ........................................................................................................................................... 1
Chapter 1 : Introduction .................................................................................................................. 2
1.1 The Current State of Renewable Electricity ......................................................................... 2
1.2 Solar Electricity .................................................................................................................... 4
1.3 Solar Technologies Used in Model ....................................................................................... 6
1.4 Economic Rationale for Solar Development at Closed Landfills ....................................... 11
1.5 Landfill Gas to Energy Facilities ........................................................................................ 12
1.6 Siting Grid-Connected Solar Power Plants on Closed Landfills ........................................ 14
1.7 Estimating Solar Radiation ................................................................................................. 15
1.8 California as Location for Solar Inventory ......................................................................... 16
1.9 Document Structure ............................................................................................................ 19
Chapter 2 : Literature Review ....................................................................................................... 21
2.1 GIS Models for Solar Analysis ........................................................................................... 21
2.2 Studies using Fuzzy Logic Modeling for Solar Analysis ................................................... 25
2.4 Inventories of Solar Potential.............................................................................................. 28
2.5 EPA and NREL Photovoltaic Potential in Landfill Studies................................................ 34
Chapter 3 : Methodology .............................................................................................................. 38
3.1 Site Identification ................................................................................................................ 39
3.2 Prescreening ........................................................................................................................ 40
iii

3.3 Detailed analysis ................................................................................................................. 45
3.4 Aggregate Estimates for the Total Population .................................................................... 58
Chapter 4 : Results ........................................................................................................................ 59
4.1 Outcomes of Prescreen Analysis ........................................................................................ 59
4.2 Detailed Analysis Results ................................................................................................... 65
4.3 Estimates for the Solar Potential of Landfills in California ................................................ 79
Chapter 5 : Discussion .................................................................................................................. 80
5.1 Analysis Results .................................................................................................................. 80
5.2 Major Findings .................................................................................................................... 82
5.3 Limitations and Assumptions ............................................................................................. 84
5.4 Future Work ........................................................................................................................ 88
References ..................................................................................................................................... 92

iv

List of Figures

Figure 1 - California Direct Normal Irradiance and Analyzed Landfills...................................... 18
Figure 2 - Prescreening Requirements .......................................................................................... 41
Figure 3 - Detailed Analysis Summary Listing Tools Used ......................................................... 45
Figure 4 - ArcMap Model for PV Geomembrane Raster Creation ............................................... 49
Figure 5 - Area Solar Radiation Tool Parameters ......................................................................... 53
Figure 6 - Prescreened Landfills ................................................................................................... 61
Figure 7 - Potential Locations for LFG Energy and CSP Solar Combination Systems ............... 64
Figure 8 - Percentage of Area Remaining After Slope Analysis for Dish-Stirling Installations .. 67
Figure 9 - Acreage Remaining After Slope Analysis for Dish-Stirling Installations ................... 67
Figure 10 - Area Remaining After Aspect Analysis for PV Geomembrane Installations ............ 68
Figure 11 - Miramar Landfill Solar Potential ............................................................................... 74
Figure 12 - BKK Landfill Solar Potential ..................................................................................... 75
Figure 13 - Milliken Landfill Solar Potential .............................................................................. 76
Figure 14 - Clover Flat Landfill Solar Potential ........................................................................... 77
Figure 15 - City of Ukiah Landfill Solar Potential ....................................................................... 78

v

List of Tables
Table 1 - GIS Models for Solar Analysis...................................................................................... 23
Table 2 - Inventories of Solar Potential ........................................................................................ 29
Table 3 - List of Landfills with Natural Gas Supplementation Potential ..................................... 62
Table 4 - Sampled Landfills .......................................................................................................... 66
Table 5 - Landfill and Technology Combinations Tested in Analysis ......................................... 70
Table 6 - Estimated Annual Electricity Potential Summary ......................................................... 71

vi

List of Equations
Equation 1 - Boolean Slope Function for CSP ............................................................................. 48
Equation 2 - SetNull Raster Calculator Function for CSP ............................................................ 49
Equation 3 - Boolean Aspect Function for PV ............................................................................. 50
Equation 4 - SetNull Raster Calculator Function for PV .............................................................. 51
Equation 5 - Solar Analyst Direct Insolation ................................................................................ 55
Equation 6 - Solar Analyst Diffuse Insolation using the Uniform Sky Diffuse Model ................ 56
vii

Appendices
Appendix A - Landfills Passing Prescreen Requirements ............................................................ 98
Appendix B - Solar Calculations ................................................................................................ 102
Appendix C - Maps of Analyzed Landfills ................................................................................. 103

1

Abstract
Solar radiation is a promising source of renewable energy because it is abundant and the
technologies to harvest it are quickly improving. An ongoing challenge is to find suitable and
effective areas to implement solar energy technologies without causing ecological harm. In this
regard, one type of land use that has been largely overlooked for siting solar technologies is
closed or soon to be closed landfills. By utilizing Geographic Information System (GIS) based
solar modeling, this study takes an inventory of solar generation potential for such sites in the
State of California. The study takes account of various site characteristics in relation to the siting
needs of photovoltaic (PV) geomembrane and dish-Stirling technologies (e.g., size, topography,
closing date, solar insolation, presence of landfill gas recovery projects, and proximity to
transmission grids and roads).
This work reaches three principal conclusions. First, with an estimated annual solar
electricity generation potential of 3.7 million megawatt hours (MWh), closed or soon to be
closed landfill sites could provide an amount of power significantly larger than California’s
current solar electric generation. Secondly, the possibility of combining PV geomembrane, dish-
Stirling, and landfill gas (LFG) to energy technologies at particular sites deserves further
investigation. Lastly, there are many necessary assumptions, challenges, and limitations when
conducting inventory studies of solar potential for specific sites, including the difficulty in
finding accurate data regarding the location and attributes of potential landfills to be analyzed in
the study. Furthermore, solar modeling necessarily simplifies a complex phenomenon, namely
incoming solar radiation. Lastly, site visits, while necessary for validating details of the site, are
largely impractical for a large scale study.

2

Chapter 1 : Introduction
This chapter provides an overview of renewable energy, solar electricity, landfills and
solar radiation modeling in order to establish the environment surrounding implementation of
solar technologies at closed landfill sites. The present environment of renewable energy and
solar electricity is explored. Photovoltaic and solar thermal technologies are also introduced.
Issues surrounding landfills are discussed along with the rational of collecting solar energy at
such locations. Finally, basic theory for solar modeling is described.
1.1 The Current State of Renewable Electricity
To contextualize the potential found for harvesting solar power at landfills, present global
and renewable energy environments are explored. The world’s energy demands and the types of
energy used to fulfill this demand are discussed to highlight the need for local and renewable
energy sources. The challenges and potential of renewable energy are also discussed.
Today, the world’s energy outlook is increasingly dim; although the pool of resources is
shrinking, total demand for energy is rising. From a 1998 baseline, the world’s energy
consumption is predicted to double by 2035 if present trends continue (Demirbas 2009, 213).
High energy costs and global warming concerns have influenced some nations to incentivize
alternative energy sources (U.S. Energy Information Administration 2012a, 74).
Renewable energy sources like solar, wind, biomass, hydroelectric, and geothermal are
often looked to as a solution to the planet’s energy situation, yet the technologies make up a
small fraction of energy consumption. The world uses renewables for only 14 percent of its total
energy consumption (Demirbas 2009, 215). In the United States, renewables make up an even
smaller share, about 8% of total energy consumption (U.S. Energy Information Administration

3

2012a, 76). With the exception of hydroelectric and wind energy, there are significant
technological, political, and economic challenges associated with new energy technologies
(Bravo, Casals, and Pascua 2007, 4879). These obstacles must be overcome in order to make the
technology more appealing to investors.
In spite of these setbacks, the estimated potential of renewables is increasing. Wind
energy, for instance, was once thought to only have the potential to contribute 5 percent of the
energy demand, but today contributions of 25 percent seem possible. Denmark has shown the
viability of an electric generation system where wind power could supply 50 percent of their
total energy consumption (Bravo, Casals, and Pascua 2007, 4880).
Predictions concerning the future energy contribution of renewable energy vary.
Globally, Demirbas (2011, 218) estimates a renewable energy contribution of 50 percent by
2040. Bravo, Casals, and Pascua (2007, 4892) demonstrate the possibility of renewable energy
as the sole energy source for Spain by 2050. In the United States, the Annual Energy Outlook
predicts that 14 percent of national energy consumption will be derived from renewable sources
by 2035 (U.S. Energy Information Administration 2012a, 76). Although these figures were
derived from different scopes and timeframes, they nonetheless illustrate the disparity between
predictions for the future of renewable energy.
The outlook of renewable energy is dependent on somewhat unpredictable factors like
policy, the private market, and technology development; even so, most agree that developing
renewable energy is both inevitable and necessary. Sufficient, inexpensive, and environmentally
benevolent energy sources contribute to a nations’ sustainable development (International
Atomic Energy Agency and the United Nations Department of Economic and Social Affairs
2007, 2).

4

1.2 Solar Electricity
There are numerous advantages and challenges associated with using solar power and
several types of solar power collecting technologies. While solar power is abundant, clean, and
versatile, it provides only a small fraction of U.S. energy needs at the present. Solar power can
be derived from photovoltaic or solar thermal technologies. These technologies can be passive
or active, and concentrated or non-concentrated.
Solar energy has many advantages, the largest of which is the abundance of the resource.
Solar radiation is the most plentiful energy source on Earth and is said to be sufficient to meet
the present global energy needs “thousands of times over” (Byrne and Kurdglashvili 2010);
Zweibel (2008, 64) states that the sunlight that strikes the earth in forty minutes is equal to the
world’s annual anthropogenic energy demand. It should be emphasized, however, that this
figure does not account for radiation needed to power animal and plant processes. Furthermore,
capturing a significant portion of this energy is extremely difficult using current technologies
given existing efficiency factors. Although solar energy is not a panacea to the world’s energy
problems, solar radiation has the potential to be an abundant resource contributing to the world’s
renewable energy.
In addition to the abundance of the energy source, solar power has further advantages.
Technologies used for converting solar radiation into energy have the advantages of
decentralization, modularity, and high potential for integration (Bravo, Casals, and Pascula 2007,
4885). These characteristics allow solar devices to be installed in a variety of environments,
including those with variable terrain such as landfills.

5

In spite of these benefits the current U.S. solar industry is small, but is gaining
momentum. Presently, solar energy produces only 0.4 percent of total renewable energy and
achieves an average annualized power increase of 11.7 percent. Solar power is the fastest
growing source of renewable energy according to the 2012 Annual Energy Outlook report (U.S.
Energy Information Administration 2012a, 75). Byrne and Kurdglashvili (2010, paragraph 5)
project that these technologies will double in efficiency within five to eight years and that the
cost of production will continue to fall.
There are a variety of solar device types, each with different installation requirements,
efficiency factors, and other characteristics vital to determining it’s energy production potential.
Probably the most basic division in solar power technologies is the difference between solar
thermal and photovoltaic (PV) technology. Solar thermal technology uses heat captured from
irradiation to drive thermal engines for electricity production or heat water for household use.
The present study focuses on the former type of thermal energy. Contrastingly, photovoltaic
cells convert electricity directly into electricity (Byrne and Kurdglashvili 2010). More about
these types of solar power can be found in Section 1.3.
Solar collectors can also be passive or active. Passive solar applications harvest solar
radiation without the use of mechanical devices and are therefore static. This is in contrast to
active solar technology, which utilizes mechanics to dynamically reposition solar collection
devices.
Finally, some solar power collecting devices are concentrated while others are not.
Concentrated solar applications use lenses, dishes and other systems to concentrate radiation,
directing waves to one concentrated area. Solar devices that are not concentrating do not use
such systems and absorb incoming radiation without modifying the irradiation.

6

1.3 Solar Technologies Used in Model
To estimate the potential of harvesting solar radiation at landfills, specific technologies
most suitable to these sites must be chosen. This is needed to determine values used for energy
production calculation for the specified technology, and to find areas suitable for the installation
of the chosen devices. The following section describes various technologies used to harvest
solar energy while rationalizing this study’s use of dish-Stirling and PV geomembrane
technologies.
1.31 Photovoltaic Technologies
One technology used in the present study is a type of photovoltaic system; specifically
PV geomembrane installations are explored. The following paragraphs describe various types of
photovoltaic technologies and justify the use of PV geomembrane technology in this study.
The first photovoltaic cell was made of both p- and n- type silicon by Bell Telephone
Laboratories more than fifty years ago (Beek and Janssen 2009, 321). When the sun shines on a
solar cell the radiant energy is converted into direct current (DC), which is then inverted to
alternating current (AC) for use. The efficiency of this operation is dependent on technical
factors like the technology used, economic constraints, and the suitability of the site to the
particular solar device.
Photovoltaic technology can be divided into three generations. The first uses crystalline
materials to form ridged solar panels. Second generation cells use materials such as Copper
Indiium Deselenide, Cadmium Telluride, and Gallium Arsenide to form a thin film solar
collector. Third generation cells are made of very efficient material, but have not made it past
research in laboratories (Patel 1999, 27).

7

First generation PV technologies are generally more efficient than second generation
ones. Compared to first generation cells, thin film technology is relatively new, as the first real
wave of thin film solar cells did not occur until 2007. First generation technologies have had
significantly more time for development.
The lack in efficiency for second generation cells is often offset by the fact that thin cells
are significantly cheaper to produce (Yang 2011, 335 - 336). The relatively little material used
in second generation thin film cells significantly reduces the fabrication costs of these devices.
A series of studies by the U.S. National Renewable Energy Laboratory (NREL) and
Environmental Protection Agency (EPA) determined that second generation cells provided more
energy per dollar invested than first generation cells (Lisell and Mosey 2010; Salasovich and
Mosey 2011a; Salasovich and Mosey 2011b; Salasovich and Mosey 2011c; Stafford, Robichaud,
and Mosey 2011; Salasovich and Mosey 2012). More on these studies can be found in Chapter 2
of the present work. Based on these results, the current study explored the potential of second
generation PV collectors rather than first generation devices.
Second generation PV systems seem to be gaining momentum in the photovoltaic market.
Yang (2011, 335) reports that thin film system efficiencies rose from 4 percent efficiency in
1995 to 11 percent by 2010. This efficiency factor indicates the technology produces electricity
equivalent to approximately 11 percent of incoming solar radiation. Green (2007, 15) shows
efficiencies of 4 to10 percent; however, later Green (2010, 88) explored a variety of materials to
be used for thin film PV cells and showed the possibility of using SulnGaSe
2
modules to achieve
a 27.6 percent efficiency rate. In 2011, Lee, Chen, and Kang (2011, 1271) state that conversion
efficiencies are currently less than 12 percent for silicon thin film technologies. Considering

8

these numbers, this study used 11 percent as an efficiency factor for thin film photovoltaic
systems.
PV geomembrane has several advantages over other types of PV technologies. Instead of
using ridged panels, the membrane lays flat over a surface; therefore, the flexible thin film can be
installed in areas of high slope and adaptable to the constantly shifting terrain found in closed
landfills. Also, since these devices are not raised off the ground, they are not as visible as rigid
panels and therefore the geomembrane has an inherent aesthetic advantage. Lastly, PV
geomembrane weighs less than ballasted systems, which can be an important factor when siting
such technologies on landfills. Because of these advantages, PV geomembrane was used to
determine the solar energy production of closed landfills in this study.
1.32 Solar Thermal Technologies
Solar thermal technologies are very different from PV systems. The following section
provides reasoning for using dish-Stirling technologies to estimate the solar thermal energy
production at closed landfills. Solar thermal technologies utilize a different method of
converting the sun’s rays to electricity. Rather than directly producing electricity from sunlight
like photovoltaic cells, thermal power must heat up a medium to indirectly power an engine.
Solar thermal power has a few key differences compared to photovoltaic technology.
Thermal systems are cheaper compared to energy output and have reduced the problem of
intermittency common in solar applications. Unlike photovoltaic systems, solar thermal
technology has the potential to store power in the form of heat for times when there is little or no
sun; thermal solar plants also address the issue of intermittency because the technology can be
supplemented with natural gas to generate power when solar radiation is absent (Lehman 2011,
5). A disadvantage of the technology is that it only takes advantage of direct solar radiation

9

while photovoltaic systems take advantage of global radiation (Price and Margolis 2010, 53;
Lehman 2011, 15).
Solar thermal technology has experienced success using lenses and reflectors to
concentrate the sun’s rays into high-temperature heat in what is known as concentrated solar
thermal power (CSP) (Lehman 2011, 5). A CSP has three main components. The solar collector
field is a set of mirrors and reflectors that focus the radiation to the receiver. The solar receiver
transforms the radiation to heat. Lastly, the energy conversion system transforms the heat into
usable energy (Schild 2004, 8).
Several types of CSP systems exist. One such technology is parabolic trough CSP; the
parabolic shaped trough uses mirrors and an integral receiver tube. Contrastingly, central tower
systems receive solar radiation through hundreds of sun tracking flat plane mirrors known as
heliostats. Molten salt or synthetic oil is pumped through a solar receiver located within the
tower where the hot liquid produces steam to power an engine.
Parabolic dishes are a third type of solar thermal collection device. The dishes are
covered in mirrors to reflect radiation into a receiver, which uses heat to power a thermal engine,
usually a Stirling type. A Stirling engine is a type of thermal engine powered by the
compression and expansion of a gas. This type of CSP, previously and hereon referred to as
dish-Stirling, was be explored in the current work because of the many advantages of this
technology.
As a result of low costs and high efficiency, dish-Stirling systems are particularly suitable
for decentralized use (Gabriel 2006, 76). Smaller installations of dish-Stirling systems are
possible when compared to central tower and parabolic trough systems (Schild 2004, 12). The

10

ability to install small-scale, high efficiency systems makes dish-Stirling systems best for CSP
installation on landfills.
Additionally, dish-Stirling systems do not require large amounts of water. Water can be a
key factor in determining the feasibility of a solar thermal plant, since it is necessary for cooling
in the Rankine cycle which is used the generation of electricity for these technologies (Lehman
2011, 22). Unlike other solar thermal technologies, dish-Stirling systems use a Stirling engine,
which does not require large quantities of water to produce electricity (Dahle 2008, 31).
Furthermore, dish-Stirling systems have a unique advantage over other types of CSP
when used at closed landfills. Stirling engines have the potential to be powered by another
resource available at landfills, landfill gas (LFG). Although still under development, Stirling
engines are being developed capable of running on LFG (SCS Engineers, 1995, 2.27 – 2.29;
Tsatsarelis et al., 2006, 5). Although an example of a combination dish-Stirling and LFG system
could not be found, this potential makes LFG to energy facilities pair particularly well with dish-
Stirling CSP.
Dish-Stirling technology is generally the most efficient of solar thermal systems.
Currently, parabolic trough systems range from 15 - 21 percent efficiency, power towers range
from 18 to 20 percent, and dish-Stirling systems operate at 25 - 30 percent efficiency (Gastli and
Charabi 2009, 794). Gabriel (2006, 76) states that the technology has achieved 30 percent
efficiency. Schild (2004, 9) shows an efficiency of 29 percent for parabolic dish systems linked
to Stirling engines. For the purposes of this study, an efficiency rating of 25 percent was
assumed for dish-Stirling systems.

11

1.4 Economic Rationale for Solar Development at Closed Landfills
Landfills are a necessity for waste disposal in most modern societies, yet they may result
in numerous environmental and health risks. Once closed, landfills represent environmentally
damaged land that is difficult to develop into lucrative facilities. Nonetheless, this study is
largely motivated by the economic and environmental justification for the viable development of
landfill sites to harvest solar energy.
Beginning in October, 1991, the U.S. Environmental Protection Agency set criteria for
the closure and regulation of municipal solid waste landfills. These laws can be found under the
EPA’s Resource Conservation and Recovery Act’s Subtitle D (EPA 2012b; EPA 1993, 3 - 4).
Subtitle D regulations structured landfill closures and made their redevelopment safer
(O’Connell 2001, 46 - 50). Landfills that closed before EPA Subtitle D regulations controlled
closing procedures are undesirable for reuse, since unknown hazards and structural
complications may be present.
As defined by Subtitle D, cover systems may cost a total of ten to hundreds of thousands
of dollars per acre (O’Connell 2001, 47). Typical uses for closed landfills may include a park,
animal refuge, golf course, parking lot, and commercial or industrial buildings; these land uses
usually do not contribute enough income to absorb these closing costs (O’Leary and Walsh 2003,
44). By converting landfills to sources for renewable energy, owners may be more likely to
cover these costs. Solar facilities sited on landfills represent a profitable use on land that is
difficult to develop. Because of their economic and environmental advantages, the U.S. EPA
supports building renewable energy facilities on closed landfills and other contaminated lands
(Sampson 2009, 1).

12

In order to determine how much solar potential exists on landfills, the number of sites fit
for particular types of solar installations must be known. However, because of the absence of
modern regulations before 1991, the actual number of closed landfills in the United States is
uncertain. O’Connell (2001, 47) found the number of landfills closed in the past decade to be
between 4,000 and 7,400 and Suflita et al. (1992, 1486) indicate there are as many as 100,000
closed landfills in the United States. As can be seen in the large variation in these figures, there
is great difficulty in obtaining accurate data on closed landfills in the United States. A
significant challenge for this study was to determine the number, location, and extent of recently
closed landfills in order to further assess for solar energy potential.
1.5 Landfill Gas to Energy Facilities
Landfills produce LFG, which can then be turned into energy. This process is described
below. Additionally, the extent to which LFG to energy facilities have been developed in the
United States is discussed.
Landfills produce methane through a three stage anaerobic digestion, where the gas is
produced by methanogenic bacteria (Themelis and Ulloa 2006, 1247) which can be used to
harvest energy. LFG is typically composed of predominately methane and carbon dioxide and
therefore can be used for a source of energy. Once a landfill ceases to receive refuse, the gas
production rate reaches its peak in one or two years, and LFG can be produced for five to
eighteen years (Malik, Lerner, and Maclean 1987, 78). Some landfills have been equipped with
systems to capture these gases and convert them to energy, where LFG is collected and sent to
combustion engines where it is transferred to energy (Malik, Lerner, and Maclean 1987, 77 - 79).

13

Willumsen estimates that as of 2001, around 955 landfills recover landfill gas worldwide,
with capacities ranging mostly between 0.3 to 4 MW as quoted in Themelis and Ulloa (2006,
1244). Of these, 325 reside in the United States, the most facilities in the world. This number
rose to 380 by 2004 according to the U.S. EPA. The U.S. EPA also estimates that LFG to energy
facilities in America have a total nameplate capacity of 1.07 GW (Themelis and Ulloa 2006,
1244 - 1255). Nameplate capacity refers to the maximum energy that may be generated by LFG
to energy facilities as rated by manufacturer specifications.
Landfill gas to energy facilities are well established in the United States. For example, in
1993 the Metro Waste Authority based in Des Moines, Iowa partnered with private businesses to
develop a LFG to energy facility. The site produces the energy equivalent of 112,000 barrels of
oil annually (Rasmussen 2005, paragraph 3). The largest landfill gas to energy facility in the
Nation is located just outside of Los Angeles, California at the Puente Hills landfill. This site
uses biogas to fuel a 50 MW turbine generator (Themelis and Ulloa 2006, 1244).
The presence of landfill gas collection for energy facilities at closed landfills may provide
an additional incentive to solar development because a complementary, established energy
source would already be available. LFG to energy systems have the potential to power a Stirling
engine and could therefore be complimentary to dish-Stirling systems (SCS Engineers 1995, 2.27
– 2.29; Tsatsarelis et al. 2006, 5). This potential shared infrastructure between the two
technologies may result in higher return on investment (ROI) for these facilities. Additionally,
LFG power could provide energy to the grid when solar power cannot be harvested at night or
under overcast conditions. However, since the engineering for this shared Stirling engine has not
yet been completed, benefits from increased ROI and decreased intermittency cannot be
quantitatively measured and thus are not included in the present study.

14

1.6 Siting Grid-Connected Solar Power Plants on Closed Landfills
Closed landfill sites have already been used to successfully harvest solar radiation for
energy using two separate types of PV systems: ballasted first or second generation panels and
PV geomembrane film.
In the more traditional type of solar power development on landfills, PV panels are
placed on the site using a ballasted rack tilted to receive optimal radiation. These systems can be
fixed in tilt or installed with a single-axis tracking system at the price of additional monetary cost
and weight. An example of a fixed tilt ballast system can be found in Middleton, Wisconsin; the
landfill supports a 10 kilowatt (kW) crystalline silicon PV system that covers an area of roughly
3,400 ft
2
, which is 0.9 percent of the 378,384 ft
2
site (Salasovich and Mosey 2011a, 12).
The second PV technology type used on landfills involves PV thin film placed directly
over the landfill. Known as solar covered landfills, these sites utilize flexible thin film PV
geomembrane to cover the landfill. The system serves a dual purpose, working as a cover to cap
the landfill and as a PV collector to generate electricity. The Hickory Ridge solar covered
landfill near Atlanta, Georgia is an example of one of these landfills. There are only a handful of
these facilities in the United States and the technology is still being developed (Salasovich and
Mosey 2012, 4). Nonetheless, the current study used PV geomembrane to estimate the inventory
of potential power generation because of the inherent benefits of the technology.

15

1.7 Estimating Solar Radiation
The measurement of solar potential is a complex undertaking involving many
interconnected variables. This process is described below, emphasizing how the phenomenon
varies over space.
Solar radiation reaches the earth in three ways. Direct irradiance originates from the sun
while diffuse sky irradiance is scattered by atmospheric particles before reaching the ground.
Additionally, reflected radiation coming from both diffuse and direct irradiance may be reflected
off of nearby terrain (Dubayah and Rich 1995, 406). A solar model ideally accounts for all three
of these radiation types, which vary greatly with geography (Hetrick et al. 1993, 132 - 133;
Dubayah and Rich 1995, 406 - 408; Kumar, Skidmore, and Knowles 1997, 475).
Byrne and Kurdgelashvili (2010, paragraph 1) point out that this solar influx is affected
both by geographic variation and diurnal, or daily, processes. Solar radiation varies over time
and space in the following ways: the Earth’s geography including declination, latitude and solar
angle; the location’s terrain including elevation, surface inclination, orientation, and shadows;
and atmospheric attenuation which includes gas in the atmosphere, atmospheric particles, and
cloud cover (Hofierka and Súri 2002, 2).
The atmospheric scattering of solar radiation takes two forms. Rayleigh scattering refers
to the dispersion of radiation via atmospheric gas molecules. Turbidity from water vapor and
pollution further separates the sun’s rays (Hetrick et al. 1993, 134). The air mass through which
radiation must pass changes throughout the day since atmospheric attenuation is greater in the
morning than it is in the afternoon (Kumar, Skidmore, and Knowles 1997, 478).

16

1.8 California as Location for Solar Inventory
California was chosen in the current study as a location to perform a first inventory
measuring the potential of dish-Stirling and PV geomembrane solar facilities on closed landfills.
The State has policies favorable to renewable energy development and aims to receive a third of
consumed energy from renewable sources by 2020 (Office of Planning and Research 2013).
While other states may receive greater solar radiation, California’s mix of this resource and
dense population makes the State particularly suitable for this study. Additionally, California is
the leader in U.S. solar energy production, generating a larger proportion of electricity from solar
power than any other EPA Emissions and Generation Resource Integrated Database (eGRID)
region according to EPA’s 2009 eGRID Summary Tables (EPA 2012a). Finally, California was
chosen because EPA national landfill data is limited, and the State’s California Energy
Commission (CEC) provided supplementary data on landfills to be analyzed in this study.
Although landfills analyzed in the present study are spread throughout California, this
analysis provides an inventory of individual sites at a small scale. With this in mind, the overall
solar environment of California must also be kept in mind. Figure 1 illustrates the general
picture of overall solar resources available at California landfills by displaying sites analyzed in
the present study with incoming solar direct normal irradiance (DNI) from NREL data (2012).
These sites were taken from EPA and CEC databases, and closed, or will close, between 1992
and 2022; the location data for each site was also verified to produce the landfills shown in
Figure 1.
Other inventory studies for solar radiation, discussed in Section 2.4 of this work, are
either site specific or measure an entire region; the current study is similar to these site specific
studies, which generally measure the solar potential of rooftops. By limiting the present study to

17

landfills, the proposed solar facilities would exist on already ecologically disturbed land; often
California’s desert regions are looked to as a location for solar facilities, however this landscape
is ecologically sensitive and solar installations disrupt virgin landscape and habitats (Abbasi and
Abbasi 2002, 132).

18

Figure 1 - California Direct Normal Irradiance and Analyzed Landfills

19

1.9 Document Structure
The remainder of this report is organized as follows. Chapter 2 discusses literature
concerning models and studies relevant to estimating solar radiation. Next, methods used in the
current study are identified and explained. Results from this analysis are then presented in
Chapter 4. Finally, Chapter 5 discusses these findings and contextualizes them.
In the literature review of the present work, past studies and models are reviewed to
illustrate various methods for modeling solar radiation. Models using GIS to estimate irradiance
are explored in addition to studies using fuzzy membership to predict the phenomenon.
Furthermore, studies that take an inventory of solar potential of a given area are summarized
since they share a common goal with the current work. Lastly, a series of studies estimating PV
potential in specific landfills conducted by NREL and EPA are described. These studies are
particularly relevant because, like the present work, they estimate solar power production
potential at landfill sites.
The methodology of this study’s analysis is broken up into a three part process. It was
first necessary to compose a list identifying landfills with potential for harvesting solar radiation
collection methods; these sites were then prescreened based on the age of the landfill, presence
of accurate location data size, and proximity to roads and transmission lines. A sample was
taken from sites that met prescreening requirements, which was then analyzed for solar potential
for both PV geomembrane and dish-Stirling using ArcMap. ArcMap was utilized to analyze
terrain and estimate solar irradiance. Results from this sample were then generalized to estimate
the solar production capacity of landfills in California.

20

The results chapter reviews the outcomes from the methodology described above. Both
prescreening and detailed analysis results are examined. Furthermore, sampling theory is used to
generalize results from sampled landfills to the total population of California.
These outcomes, and the study in its entirety, are discussed in this work’s final chapter.
Results are contextualized by comparing them to the current energy environment in California
and past studies. Chief findings from this study are also illustrated here. Assumptions and
limitations of this work’s methods are then discussed. Finally, areas where future work is
needed are explained.

21

Chapter 2 : Literature Review
The following literature review discusses past studies and models in order to take
advantage of modern advances in the field of modeling for solar potential. Various GIS models
designed to estimate solar radiation are described. Because of its extensive use in the field of
spatial solar modeling, studies using fuzzy logic are also explored along with the variables used
in these models that go beyond solar radiation.
The present work takes an inventory of solar radiation potential, specifically at landfill
sites. Therefore studies that also take inventory of such potential are reviewed. Additionally, a
series of studies conducted by EPA and NREL that explore the potential of landfill sites
producing solar power from various PV technologies are summarized.
2.1 GIS Models for Solar Analysis
GIS models have been developed to estimate a location’s incoming solar radiation, and
these models are outlined in Table 1. The following spatial models utilize different software and
parameters, and a brief introduction to such systems is provided. The present analysis utilizes
one such model, Esri’s Solar Analyst, for incoming solar radiation estimates.
One of the first GIS based solar radiation models was SolarFlux, which was developed
for Esri’s ArcInfo. Hetrick et al. (1993) created a model to integrate the many influences on
solar radiation utilizing a GIS framework. SolarFlux determines solar potential from an area’s
surface orientation, solar angle, horizon shading, and atmospheric conditions (Hetrick et al.
1993, 133).
Similarly, solar analysis was performed using automation mark-up language (AML)
script with the commercial GIS software, GIS Genasys. GIS Genasys solar radiation algorithms

22

worked similarly to the Solar Flux model, but uses different software (Hofierka and Súri 2002, 1-
2).
Using Microsoft Windows, a standalone model called Solei was used to estimate solar
radiation and was linked to the GIS software, IDRISI, by using identical data formats. The
model differs from the two previously discussed models by accounting for elevation through a
raster Digital Elevation Model (DEM). However, all three models discussed thus far use
spatially averaged parameters according to Hofierka and Súri (2002, 1 - 2).
The r.sun model proposed by Hofierka and Súri (2002) aimed to overcome the limitations
of the aforementioned models. The application is appropriate for large areas, considers the
effects of terrain and shadowing, and can simulate overcast conditions and its effect on
irradiation. R.sun utilizes Geographic Resources Analysis Support System (GRASS) GIS, taking
advantage of advanced interpolation techniques and accounting for land use and environmental
concerns to model the complex process of solar radiation.
Pons and Ninyerola (2008) developed a simple model whose only input is a DEM, yet
this radiation model had demonstrated significant accuracy. The methodology is composed of a
physically based model to determine potential solar radiation and a process that uses
meteorological data to refine this potential. Using MiraMon GIS software, Pons and Ninyerola’s
model estimates solar radiation and accounts for astronomic, atmospheric and geographical
variables. The success of this model highlights the importance of elevation in general, and
DEMs in particular, in estimating solar potential.

23

Table 1 - GIS Models for Solar Analysis
Model’s Name Source Technology Used Summary
SolarFlux Hetrick et al. (1993)
Hofierka and Súri
(2002)
ArcInfo Based on solar and atmospheric
conditions
GIS Genasys
solar radiation
algorithms
Hofierka and Súri
(2002)
GIS Genasys AML
script
Algorithms similar to SolarFlux
Solei Hofierka and Súri
(2002)
Microsoft Windows
and GIS IDRISI
Estimates solar radiation from DEM
R.sun Hofierka and Súri
(2002)
GRASS GIS Considers terrain, shadowing, and
climate using interpolation
techniques
MiraMon GIS
Model by Pons
and Ninyerola
Pons and Ninyerola
2008
MiraMon GIS Accounts for astronomic,
atmospheric and geographical
variables and uses DEM
Solar Analyst Fu and Rich 1999 ArcInfo/ArcGIS Uses DEM and location data with
advanced algorithms

2.11 Esri’s Solar Analyst
The most commonly used GIS program for solar analysis today seems to be Solar
Analyst, available in Esri’s ArcGIS. Solar Analyst uses elevation data to measure solar radiation
over time. This tool was created by Fu and Rich (1999) as part of the Spatial Analyst Extension.
Solar Analyst was utilized in the present work because of its accuracy and ease of use.
Esri’s Solar Analyst model, summarized in Table 1, combines various algorithms to
estimate incoming solar radiation over time. The model uses an input raster, such as a digital
elevation model, a latitude value, time configuration, and additional parameters to model solar
radiation. The tool estimates global, direct, or diffuse radiation for a given period of time. As of
ArcMap 10.1, the Solar Analyst calculations are available via the Area Solar Radiation tool,
which requires the Spatial Analyst Extension (Esri 2012). Additional details and specific

24

methods used by Solar Analyst can be found in the Chapter 3 of the present work and Esri’s help
page for the tool (Esri 2012).
Fu and Rich (1999, 1) created the Solar Analyst extension out of a need for “expanded
functionality, accuracy, and calculation speed” of GIS solar radiation modeling. Optimized
algorithms were created to account for the complexities of solar irradiance including viewshed,
hillside, surface orientation, atmospheric conditions, and elevation calculations. The model’s
accuracy was validated by comparing the outcomes to empirical results within the vicinity of
Rocky Mountain Biological Laboratory (RMBL) located in Gothic, Colorado.
The RMBL hosts four weather stations. The first of which was established in 1989 by
the EPA while the other three were completed in 1997 to monitor climate change and global
solar radiation. Data from these monitoring stations were averaged and recorded in both hourly
and two-hour intervals. An analysis of this empirical data was then compared to those produced
by the model (Fu and Rich 1999, 18).
Solar Analyst has been used by several studies because of its accuracy and usability.
Gastli and Charabi (2009, 793) used it in their study of solar electricity prospects in Oman and
cited several benefits of the model. It enabled them to analyze and map the effects of solar
radiation over time and space while accounting for atmospheric effects, latitude, elevation, slope,
aspect, shifts of the sun angle over time, and the effects of shading from local topography.
Huang and Fu (2009) used the tool to create solar and temperature distribution maps for
Yellowstone National Forest. The Spatial Analyst enabled the team to “efficiently implement
time consuming processing of this data in a timely fashion” (Huang and Fu 2009, 28).

25

A DEM, which serves as the primary input of the tool, is a well-established data format
for elevation. In their work Huang and Fu (2009, 28) state that topography is the major factor in
determining the spatial variability of insolation. It is therefore fitting that a DEM is the primary
input for the Solar Analyst model. Fu and Rich used a 30 meter DEM constructed from United
States Geological Survey (USGS) 7.5’ quadrangles (Fu and Rich 1999, 18), though the model
accepts DEM’s of various sources and resolutions.
2.2 Studies using Fuzzy Logic Modeling for Solar Analysis
Fuzzy logic modeling has widely been used to analyze an area’s solar potential. This
multi-criteria approach is able to find optimal areas for solar energy harvesting and is especially
useful for screening large areas such as counties, states, and nations. Investigating specific,
predetermined and widespread land uses negates the need for this large scale prescreening
analysis and instead dictates a more detailed examination of each site. Nonetheless, the
following studies display the importance of multi-criteria modeling when analyzing an area for
solar power harvesting potential.
Fuzzy membership models have been created at national scales. Badran and Sarhan
(2008) developed a model that uses fuzzy logic to assess Jordan’s solar potential. Parameters
such as solar resources, site capacity, site accessibility, soil condition, water availability, grid
connection distance, land cost, land roughness, and wind speed were used in the model. Salim
(2012) used a fuzzy logic model to determine the viability of solar desalination in Egypt,
considering solar radiation, aquifer depth and salinity, proximity to water sources, and the
presence of hazards or seawater intrusion. Aydin (2009) utilized fuzzy membership rules to
assess Turkey's wind and solar potential considering protected and agricultural areas,
transmission line distance, slope, and other parameters.

26

Janke (2010) studied the solar and wind power potential of Colorado using both GIS and
a multi-criteria membership analysis. The study considered incoming solar radiation,
environmental and land use considerations, and proximity to resources. An area’s distance to
nearby resources, as seen in many of the above studies, was an important part of this study’s
prescreening analysis.
Furthermore, Lehman (2011) created a model to estimate the feasibility of concentrated
solar thermal facilities in San Bernardino County, CA. This study made use of the Weighted
Overlay tool in ArcMap to establish fuzzy membership, while taking into account many
environmental variables. The present study also took advantage of ArcMap software, while
focusing on analyzing specific sites rather than conducting a site selection analysis.
Rylatt, Gadsen and Lomas (2001) developed a decision support model for energy
planning that focused on a specific land use, the rooftop of buildings. The model’s parameters
included latitude, mesoclimatic factors, microclimatic characteristics, building codes, roof angle
and space, and socio-economic characteristics of homes. Socio-economic factors, which
included information on income, residency, and ownership, where used in a fuzzy membership
application. The model was used to find homes with ideal socio-economic conditions, and this
data was inserted into the rest of the model.
2.21 Non-Solar Radiation Variables used in Solar Models
The best models that estimate solar resources consider variables beyond solar radiation.
Rylatt, Gadsen and Lomas (2001) used socio-economic factors to determine find the optimal
homes to install a roof solar system. Lehman (2010) used inputs such as the location of Bureau
of Land Management lands, Wilderness, parks, forests, conservation areas, Critical Habitats, and
proposed Wilderness in order to account for environmental concerns. Aydin (2009) uses similar

27

environmental parameters. Bravo, Casals, and Pascua (2007) supplemented their model with
land use restrictions and local environmental constraints. Badran and Sarhan (2008) used soil
condition, water availability, grid connection distance, and land cost in addition to parameters
that assess solar resources and capacity.
It should be noted that in the present study environmental and land use constraints are
less constricting than in the above models since landfills are already disturbed land. Even so,
there may well be competing potential land uses for the area. Nevertheless, considering these
contending land uses is out of scope for this study because the relevant factors are very site
specific.
Some non-solar radiation variables can be useful when siting renewable energy on
disturbed land. Distances from a site to a usable transmission lines, roads, or natural gas
pipelines for instance, are still important to consider when siting such facilities. There are
precedents for using these factors in past studies measuring solar potential. Badran and Sarhan
(2008) and Janke (2010) used transmission line distance as a parameter in their models.
Additionally, Lehman (2011) used a site’s distance to a natural gas pipeline to estimate the
viability of installing CSP.
A landfill’s proximity to resources like transmission lines and graded roads were used in
the current study to add another dimension to suitability modeling. A site existing nearby
transmission lines greatly simplifies the project development from the standpoint of public
acceptance and lower overall costs. The site also needs to be located nearby a graded road, for
practical maintenance and construction access.

28

A landfill’s distance to natural gas lines was also observed. Natural gas lines are a
resource since a nearby solar power facility could use natural gas to power thermal engines in
times of little or no sunlight without the need to construct costly new gas lines. This could be
used in addition to the LFG, and natural gas could take over when LFG runs low. However,
because of the existence of a secondary power source at a landfill, LFG and the presence of gas
lines are not strict preconditions for suitability.
2.4 Inventories of Solar Potential
Since the current study aims to take an inventory of potential solar power collection
facilities on closed and soon to be closed landfills in California, studies that also take an
inventory of such potential are explored. Some previous solar inventory site studies estimate the
potential of PV systems, while others analyze solar thermal systems. Such inventory studies are
listed in Table 2. Regardless of specific methods used in the following studies, a majority of
them measure solar potential from total energy production, and this same type of measurement
was calculated in the current study.

29

Table 2 - Inventories of Solar Potential

Type Citation Subject Location Method
Photovoltaic
Carrión et al.
(2008)
Large-scale photovoltaic
solar farms
Andalusia,
Spain
R.sun GIS model and
interpolation techniques
Tadlock
(2009)
Photovoltaic collectors for
rooftops
Huntington,
WV
Measurement of rooftop areas
excluding north-facing slope
Janke (2010) Inventory for photovoltaic
potential
Colorado GIS and multi-criteria models
used to rank areas for solar
potential
Van Hoesen
and Letendre
(2010)
Baseline inventory of
incoming solar radiation
from rooftops
Poultney, VT Portion of rooftop multiplied by
average annual solar radiation
Arnette and
Zobel (2011)
Large-scale photovoltaic
solar farms
Appalachian
area of the
United States
GIS modeling with spacing and
derate factors, economic
analysis also conducted
Solar Thermal
Turchi et al.
(2011)
Augmenting existing fossil-
fired power plants with
either power tower or
parabolic trough CSP
United States
southeast and
southwest
ArcMap ranking system with
minimum requirements
Pan, Kao, and
Wong (2012)
The authors investigate the
potential of solar water
heaters (SWHs) in Taiwan
Taiwan SAM and regression analysis
used to estimate capacity

30

2.41 Photovoltaic Inventories
Several studies conduct an inventory of photovoltaic potential using GIS models. First,
studies that take inventory of solar potential of an entire region are discussed. Studies modeling
rooftops are also summarized; like the present work, these studies take inventory of a particular
land use.
Arnette and Zobel (2011) conducted an analysis of renewable energy within the greater
southern Appalachian mountain area of the United States. A GIS model was created to
determine the most suitable locations for large-scale photovoltaic solar farms within the region,
an end goal analogous to the present study. The model considered both slope and aspect,
rejecting areas with unsuitable terrain. Potential locations were narrowed further by only
considering sites with an area greater than ten acres.
Arnette and Zobel (2011) found 477 individual sites within the greater southern
Appalachian Mountains area; the potential of each location was based on associated cost and
generation capacity. The productivity for each site was determined by its average kW/m
2
, with
25 percent of the site removed to address spacing and shading concerns. An efficiency factor of
14 percent was then multiplied by this figure.
A derate factor, representing the energy lost when converting Direct Current (DC) to
Alternating Current (AC), of 77 percent was also used. This derate factor means that the AC
produced after converting it from DC is 77 percent as powerful as the original DC created by the
power source. This factor represents a loss in electricity that some other models do not account
for and is necessary to consider in inventory studies for solar farms, since solar devices produce
DC, not AC which for the transmission grid.

31

The current work used a similar methodology to that used by Arnette and Zobel (2011).
Although different specific criteria were used, the present study’s prescreen analysis too rejected
landfills that did not meet certain requirements. A slope and aspect analysis of the location’s
terrain was conducted in addition to estimating incoming solar radiation from average kW/m
2
.
Spacing and derate factors of 25 and 77 percent, respectively, were borrowed from Arnette and
Zobel (2011) and an assumed efficiency factor was also be applied.
Janke (2010) conducted an inventory for photovoltaic technology, exploring the potential
of solar farms within Colorado. Using weighted multi-criteria GIS modeling, Janke (2010)
considered Watts of solar irradiation per m
2
/day, nearby distance to transmission lines, close
proximity to cities and population, and nearby distance to roads to characterize a favorable area.
Based on this ranking, areas were given a score between 0 and 100. Scores of existing solar
power facilities were noted to compare the model to real world solar production.
Carrión et al. (2008) calculated an inventory for large-scale photovoltaic solar farms in
Andalusia, Spain. The study estimated solar radiation using the r.sun model and interpolation
techniques described by Hofierka and Súri (2002). Since the study was focused on grid-
connected facilities, only lands within 4 km of the outer limits of the city center were accepted.
Additionally, areas with a slope more than 2 percent were eliminated because of shading
concerns. A similar method is used in the present study’s terrain analysis, which rejects areas of
high slope.
For each suitable area, the total electricity generated was calculated from the product of
the maximum power installed in kW, the performance ratio of the technology, the average daily
global irradiation, and 365 days per year. A similar calculation was conducted in the current
study. From this equation, Carrión et al. (2008) developed a map of annual photovoltaic

32

electricity production in MWh within Andalusia, Spain. This map represents areas of suitable
land and estimated power production from grid-connected photovoltaic power plants.
The above study is dissimilar than the present work in that its scope is an entire country,
rather than one specific land use. Because its goal was to conduct a site suitability analysis as a
basis for an inventory, many parameters used to eliminate potential areas differ from the current
work. This inventory is measured, however, in total annual MWh. Such a measurement
provides production potential of a given area. This figure is straightforward and easy to
understand and is also the measurement chosen in the present inventory to represent the solar
harvesting potential for landfills in California.
Van Hoesen and Letendre (2010) study the potential of multiple renewable resources in
Poultney, Vermont, providing a baseline inventory of local and green energy sources. Solar
potential of rooftop systems was calculated using GIS by digitizing rooftops and eliminating
areas that are not south-facing by only considering 25 percent of the rooftop area, a figure was
taken from the proportion of south-facing roof for the average Poultney building. The
simplification of aspect is unfit for modeling specific land uses, including both rooftops and the
landfills examined in the current work.
Tadlock (2009) studied the potential of applying photovoltaic collectors for rooftops in
Huntington, West Virginia. Three neighborhoods were selected within Huntington, each
representing different land use and socioeconomic patterns. The final study area consisted of
four randomly chosen blocks within each of these neighborhoods. Rooftop areas were digitized
and summed to find the usable space for solar installations. Potential power production was
calculated based on these areas matched with various sized photovoltaic systems and irradiation
values.

33

To account for non-south facing slopes, the total digitized rooftop area was divided by
half, a similar method to that used by Van Hoesen and Letendre (2010). Calculating usable area
in this way to account for slope orientation may be too general for estimating usable area at this
small scale; chances are slight that nearly 50 percent of the rooftops analyzed did not face south.
Additionally, this measurement does not account for areas that may be occupied by structures
such as chimney or satellites and those shaded by nearby structures or vegetation. Lessons
learned from these aspect analyses were applied to the methods used in the present work, which
measured aspect and did not consider areas containing obstacles.
2.42 Solar Thermal Inventories
The following studies take inventory of solar potential from solar thermal technologies.
In a study investigating the potential of solar water heaters (SWHs) in Taiwan; Pan, Kao, and
Wong (2012) base solar potential on the annual ratio of effective days to non-effective days and
effective solar radiation measurements. To evaluate the potential of these sites, the National
Renewable Energy Laboratory’s Solar Advisor Model (SAM) and regression analysis were used
to estimate capacity factor and relative cost index. Satellite DNI data were used as irradiation
values for these calculations and the sums of these calculations were added, finding the total
annual Terra Watt hours produced. The study by Pan, Kao, and Wong (2012) is very different
from the current work in that it estimates personal, residential systems that do not provide energy
to the grid. Comparable to the present study, however, Pan, Kao, and Wong (2012) illustrate a
multifaceted inventory of solar thermal systems that considers economic, technical, and solar
environments.
Turchi et al. (2011) observed the potential of augmenting existing fossil-fired power
plants with concentrated solar thermal power for sixteen states in the United States southeast and

34

southwest. The augmentation of solar power to an existing power source is analogous to adding
solar power facilities to landfills, which may have existing LFG to energy facilities. Turchi et al.
(2011) used minimum requirements for a power plant to be considered, and then ranked them.
The age and capacity of the fossil plant, DNI amount of available land, topography, and solar-use
compatibility were all considered in this ranking system. Newer sites with large capacity,
plentiful area, mild topography, and high compatibility for combination systems were ranked the
highest.
The study by Turchi et al. (2011) provides lessons for the current work. The inventory
considers the characteristics of fossil-fired power plants to rank the facilities potential for CSP
augmentation. Some of the same characteristics used for this ranking system, such as the age,
size, and topography were accounted for in the present work when analyzing the potential of
landfills to host solar energy. Additionally, the above study uses minimum requirements to
eliminate unfavorable areas before they are analyzed further. Likewise, the present work utilized
a prescreening analysis that sorts out sites with low potential before a more detailed examination
is conducted.
2.5 EPA and NREL Photovoltaic Potential in Landfill Studies
This study is not the first to analyze the solar potential of closed landfills. The United
States Environmental Protection Agency and the National Renewable Energy Laboratory
conducted a series of studies between 2010 and 2012 that investigated the feasibility of solar
photovoltaic energy collection at brownfield sites, many of which included landfills ( Salasovich
and Mosey 2011a; Salasovich and Mosey 2011b; Salasovich and Mosey 2011c; Stafford,
Robichaud, and Mosey 2011; Salasovich and Mosey 2012). Similarly, Lisell and Mosey (2010)
explore siting renewable energy on a variety of brownfield sites in Nitro, West Virginia. This

35

research was performed to encourage renewable energy at potentially contaminated locations as
part of the EPA’s Re-Powering America’s Land initiative.
The EPA and NREL recognize that PV installation is a “promising and innovative use of
closed landfills” (Salasovich and Mosey 2011a, 1). Because of this potential, the organizations
analyze landfills and brownfield sites in several areas, including West Virginia, Puerto Rico, a
Massachusetts Military Reservation, Wisconsin and Kansas (Lisell and Mosey 2010; Salasovich
and Mosey 2011a; Salasovich and Mosey 2011b; Salasovich and Mosey 2011c; Stafford,
Robichaud, and Mosey 2011; Salasovich and Mosey 2012).
2.51 General Methods
Some of these studies began with a site selection analysis. Salasovich and Mosey
(2011b) and (2011c) preformed such an analysis to select landfills most suitable for PV
installation in Puerto Rico. Sites passed the analysis that had a minimum size of 14 acres, were
within one mile from 38kV transmission lines, and were near graded roads. The Puerto Rico
study also screened the slope of sites, rejecting sites with more than 20 percent slope (Salasovich
and Mosey 2011b, 8 - 10). A similar prescreening process was performed in the current work,
considering size and proximity to transmission lines and roads before a more detailed analysis
was performed.
After prescreening, many of the studies used on-site visits to find the total usable area for
each site. Sometimes Google Earth was used in order to discover obstacles and other spacing
concerns in place of a physical visit. The present study used the latter method for this task.
In order to predict the possible system size and production for each landfill or brownfield,
the NREL used a combination of PVWatts and either SolOpt or the Solar Advisor Model (SAM).

36

PVWatts is a NREL PV system performance calculator used to estimate the performance data of
a PV system and these data were used to determine annual revenue for each site. SAM is
another NREL application that complements these data and can be used to model impacts of
various costs, system performance, government incentives, return of investment, and other
financial and performance characteristics (Blair, Mehos, Christensen, and Cameron 2008, 1).
Alternatively, the SolOpt Optimization Tool analyzes system production, design and financial
components and provides data regarding the most effective system for a given environment and
tool utilizes unique parameters unavailable in SAM and therefore may be better suited for certain
studies (Lisell, Metzger, and Dean 2011, 23). These tools were used to estimate optimal
performance for each site using various technologies and additional financial calculations were
made to estimate costs and payback periods for each location and technology included in the
study.
2.52 Results
All sites, with the exception of the Johnson County Landfill, were found to be feasible for
PV installation (Salasovich and Mosey 2012). Of the 311 acres analyzed at this landfill, 43
percent of the area was either presently feasible for PV installation or will be viable in the future
when refuse disposal ceases in the area. The remaining 57 percent of the landfill was too sloped
or had an unfavorable orientation for PV installation (Salasovich and Mosey 2012, 1 - 8). This
does not necessarily mean that the Johnson County Landfill is less suited for PV installation;
rather Salasovich and Mosey (2012) represent their results more specifically by indicating exact
areas within the site suitable for solar power harvesting. Because the current study performs a
similar detailed analysis, the Jonson County Landfill study proved to be the most influential of
the EPA and NREL studies to the present work.

37

Each analysis compared fixed and single axis crystalline silicon and fixed axis thin film
ballasted photovoltaic technologies. For each study, the thin film panels were found to provide
the quickest return on investment. Although these systems produced less energy output than
crystalline silicon panels, the cost effective thin film system made up for this fact in terms of
return on investment. The present study relies on these NREL and EPA findings that thin film
PV systems provide the quickest return on investment to reinforce the choice of using the
technology for solar power harvesting on closed landfills. Furthermore, the present study
assumes that PV geomembrane will provide the same type of cost effectiveness as PV thin film
ballast systems.

38

Chapter 3 : Methodology
The methodology for this study is separated into a three part procedure. First, a baseline
population of landfills was created using state and national lists. Next, these landfills were
prescreened for size and age of the landfill in addition to its proximity to transmission lines and
roads. Pre-screened landfills were selected using simple random sampling (Dixon and Leach
1977, 13). For each location sampled, solar irradiation was calculated as appropriate for either or
both PV geomembrane and dish-Stirling systems using Esri’s Solar Analyst. Radiation was
translated to energy using mathematical formulas, and statistics were used to generalize these
results to the state level, providing an estimation of solar potential of California landfills.
This study intends to answer the following question: How much potential solar energy for
electricity generation exists in closed, or soon to be closed, landfill sites in California? This state
was chosen based on data availability, the size of the state, and the variety of climates that exist
there. To inventory solar energy potential, the study considered two types of technologies - PV
geomembrane or dish-Stirling systems, considered best for sites with varying characteristics.

39

3.1 Site Identification
The present study required a list of spatially referenced landfills within California to form
a baseline population for such sites in order to take inventory of solar potential. This database
was composed from two sources, one from the EPA and another from the State of California.
The Landfill Methane Outreach Program (LMOP) database (EPA 2012a) contains a list
of over 2,800 landfills, both open and closed, along with the location’s city, county, and state;
from this list 370 records were found in California. Although this was the most complete
database that EPA has released, this list may not provide a sufficient starting point for this study.
Since national landfill regulations did not go into effect until 1991, an empirical measurement of
such sites is extremely difficult to produce. Estimates for closed landfills in the United States
vary, but have been as large as 100,000 sites (Sampson 2009, 1).
To attempt a more complete list of sites, state data were added to the EPA data. The CEC
CalRecycle database was found to have 314 records listed in California and was the most
complete state list found (CEC 2012). This database was compared and added to EPA’s LMOP
database (EPA 2012a) to create a newly compiled list.
It should be noted that both of these databases were established for the purposes of
identifying potential and existing LFG to energy projects. It is likely that landfills fit for LFG
energy projects share many characteristics as those well suited for solar development. Landfills
potentially useful for either system must be both fairly large and recently closed. The compiled
database of landfills then favors landfills that could potentially be used for solar harvesting,
making these sources particularly relevant to the current work.

40

Each record on the list of landfills was investigated. Although the databases have a
combined total of 684 records, some of these records refer to different LFG projects on the same
landfill. There were also redundancies between the two databases, so duplicates were omitted.
Additionally not all landfills were associated with closing dates; only records with closing
information moved onto the subsequent prescreen tests. A total of 324 landfills constituted the
final list of landfills used for the prescreening portion of this analysis.
3.2 Prescreening
Once a list of California landfills was created it was necessary to conduct a prescreen
analysis of these sites to eliminate those unsuited for solar power development, this method is
summarized in Figure 1. Landfills that did not pass any portion of the test were considered
unsuitable for power generation; those that did underwent a more detailed analysis. This way,
locations that did not meet minimum requirements for practical solar installation were excluded
before more time intensive analyses are performed. The compiled list of landfills from the CEC
and EPA contained information used in this analysis regarding the landfill’s closing date,
location data, and size. Next, the proximity from landfills to resources like roads and
transmission lines were measured using ArcMap. The existence of LFG to energy facilities
nearby natural gas lines were examined as a complementary energy source to solar power,
although these requirements were not prerequisites for a given site to pass this analysis.

41

Figure 2 - Prescreening Requirements
Locations that have been closed too recently, or not recently enough, may be undesirable.
As seen in Figure 2, a landfill must have been closed between 1992 and 2022 to pass the
prescreening portion of this analysis. Landfills that have been recently closed are prone to
settling as waste compacts; this process must be nearly complete before any major structures can
be placed on the area.
In order to give a landfill time to settle and to begin initial construction, the present study
considered landfills that have a planned closing date of 2022 or earlier. This included landfills
that will close within ten years that can start being converted into solar power farms within
fifteen years. Although this number is somewhat arbitrary, it nonetheless establishes a limit to
define landfills available for construction in the near future.
Furthermore, a potential site should be lined or capped properly in order to minimize the
risk of structural and environmental issues. Landfills closed after the landmark Subtitle D
Characteristics
•The landfill must have a closing date between 1992 and 2022
•Landfill must have accurate location data
•Landfill must have a minimum area of 2 acres for PV,10 acres for CSP,
and 12 acres for combination systems
Proximity
•A landfill must be...
• < 1/2 mile from a 38kV transmission line
• < 1 mile from a graded road

42

regulation best fit this criterion, this study considered landfills closed after 9 October 1991 (EPA
2012b, 4). Because the data used in this study only provides the year of closing dates, this date
was rounded up to 1992. This study therefore considered landfills that closed after 1992, but
before 2022.
Many landfills on the list were attached to coordinates or other location data; landfill
locations that could not be found after a comprehensive search were rejected, in addition to sites
being reused or with future reuse plans. Some landfills were, or will be, converted into a park,
golf course, or similar facility. A major reason that closed landfills are attractive sites for power
generation is because they are typically underutilized, not generating revenue or providing a
service to the public. If the landfill has a redevelopment plan or is currently being utilized, this
benefit is lost (EPA and NREL 2012, 5 - 6). Location and land use data were found using a
combination of maps, digital imagery, municipal publications, and newspaper articles.
The size of a landfill is also important to note, because sites that are too small may not
produce enough electricity to make the project economically viable. Most landfill size
information was taken from the compiled landfill list, although a few locations without footprint
data were manually calculated using ArcMap’s measure tool and digital imagery. Landfills that
were smaller than ten acres were considered unfit for dish-Stirling installations while those
smaller than two acres were deemed too small for either dish-Stirling or PV geomembrane
facilities.
A landfill’s distance from resources like roads and transmission lines is important to
consider when predicting the viability of a project. Landfills that are close to the transmission
grid provide relatively cheap access to the transmission grid, reducing the cost of creating the

43

power plant. Likewise, roads provide access to the site for maintenance and initial construction
without the need for creating new roads for these purposes.
According to the EPA and NREL document, “Screening Sites for Solar PV Potential”
(EPA and NREL 2012, 4), the distance from transmission lines should be less than 0.5 miles.
Additionally, these transmission lines must have capacity for a large scale solar project. The
present analysis uses 38kV as the smallest transmission line needed for this task, a figure also
used by Salasovich and Mosey (2011b). Transmission line shapefiles from Federal Emergency
Management Agency (FEMA) were compared to landfill locations using the ArcMap measure
tool (FEMA 2012).
Additionally, a landfill’s distance from a graded road should be less than one mile in
order to maximize accessibility (EPA and NREL 2012, 4). Salasovich and Mosey (2011b, 10)
also used this criterion. To find this attribute, ArcMap’s measure tool was again used, with road
data from Bing Hybrid imagery available at ArcGIS.com.
While the distance between a landfill and natural gas pipeline was not a precondition for
sites to be identified for dish-Stirling potential, all sites passing the prescreen requirements were
analyzed for this useful attribute for dish-Stirling technology. Leman (2011, 23) states that
natural gas lines are about twice as costly as transmission lines. Since this study uses a distance
of 1/2 mile to define a transmission line as being viable, this distance was halved and 1/4 mile
was used to define a landfill with natural gas connection potential. However, because of the
additional power supply found on landfills from LFG, proximity to natural gas lines was not
chosen as a criterion for dish-Stirling installation.

44

The National Pipeline Mapping System is an online GIS application that allows the user to view
natural gas pipelines over a base map a single county at a time (National Pipeline Mapping
System 2012). The application’s measure tool was used to calculate the distance between a
landfill’s border and the closest pipeline.

45

3.3 Detailed analysis
The detailed analysis of the present study determined if a landfill is suited for specific
solar technologies and estimates the annual production of those facilities. Figure 3 outlines the
steps taken in the analysis.

Figure 3 - Detailed Analysis Summary Listing Tools Used

Calculating Energy Production
ArcMap Int and Zonal Statistics as Table Tools and Mathematical Formulas
Estimating Solar Irradiance
ArcMap Area Solar Radiation Tool
Detailed Site Size Comparison
ArcMap Raster Properties
Slope and Aspect Calculations for PV Systems
ArcMap Slope, Aspect, and Raster Calculator Tools
Calculating Remaining Area for PV Systems
ArcMap Int, Raster to Polygon, and Symetrical Difference Tools
Slope Calculation for Dish-Stirling Systems
ArcMap Slope and Raster Calculator Tools
NED Extraction
ArcMap Extract by Mask Tool
Usable Area
ArcMap Imagery
Random Sample of Prescreened Population
Statistics and Microsoft Excel formulas

46

First a random sample was taken from the prescreened population. The usable area from
each site sampled was digitized and elevation data was extracted from this usable area. By
analyzing the terrain in each landfill, areas appropriate for dish-Stirling and PV geomembrane
were determined. The size available for each technology was compared to the size of a viable
system; technologies without ample space available on the landfill were rejected for the site.
Solar Analyst was used to estimate solar radiation, which was converted to energy output using
mathematical formulas.
Sites that passed the prescreen analysis constituted the sampling frame for this study.
The sample size was chosen, noting the sample’s confidence interval and confidence level.
These figures were used in this study’s final calculations to determine how well the sample
analyzed represents the population as a whole. A simple random sample was taken from this
population using an Excel formula to generate random numbers. The locations that were chosen
in this sample proceeded to a more detailed analysis.
Usable areas were found using ArcMap by digitizing a vector layer representing
unobstructed areas of the landfill from Bing imagery, eliminating areas containing trees, shrubs,
structures, or other obstacles. Many of these obstacles could be removed by developers to make
room for solar collection devices, however determining whether or not each structure is essential,
or if removal of such obstacles is worth the costs, is a task beyond the scope of the current study.
Therefore the present work considers areas of the landfills free of obstacles.
National Elevation Dataset (NED) raster files of 1/3 arc second resolution were
downloaded from the USGS National Map Viewer (USGS 2012). NED raster files represent the
primary and most recent digital elevation data from the USGS. Each raster file was clipped

47

using ArcMap’s Extract by Mask tool creating a raster file with elevation data fit to the size,
shape, and location of the landfill.
Sites may be appropriate for: 1) dish-Stirling 2) PV geomembrane 3) a combination of
both technologies or 4) unsuitable for solar installation. Areas appropriate for dish-Stirling
systems have a slope less than 5° (Gastli and Charabi 2009, 794). Furthermore, since parabolic
dishes used in dish-Stirling devices follow the sun as it moves, eliminating non south-facing
slopes to the usable area for these systems was not necessary. Locations fit for thin film PV
geomembrane installation must have a slope less than 60°, and have a southward aspect. For the
purposes of this study southward is defined as any aspect facing southeast, south, and southwest.
As a priority, all areas that could host dish-Stirling installations were assigned as such.
This is because of the superior efficiency of dish-Stirling systems compared to PV
geomembrane. Implementation of dish-Stirling systems also may benefit from natural gas lines
and existing LFG projects. If the area suitable for dish-Stirling systems was large enough for a
viable system (i.e., 10 acres), the location was analyzed for incoming direct solar radiation.
Additionally, landfills that showed potential for dish-Stirling systems were compared to
LFG to energy projects in the State. LFG to energy data was taken from LMOP and CEC
databases and matched with landfills with dish-Stirling potential. Like a location’s proximity to
natural gas pipelines, containing a LFG to energy project was not a prerequisite for a landfill
with solar potential and does not factor into the inventory calculation. Both the presence of
natural gas pipelines and LFG projects are reported separately in the results.
It was assumed that dish-Stirling would yield higher energy output than PV
geomembrane because of its higher efficiency factor and lack of slope orientation requirements.

48

To test this hypothesis, sites assigned entirely as dish-Stirling facilities were also analyzed for
PV. These PV comparisons did not contribute to this study’s inventory and were used only to
back up the assumption that dish-Stirling systems would be more productive.
Areas unfit for dish-Stirling systems were then analyzed for PV systems. In cases where
the area suitable for dish-Stirling systems in a site was too small, the entire landfill was tested for
PV. If the area analyzed for PV geomembrane was larger than 2 acres, the minimum size
required for an economically viable system, the landfill was considered to have potential for PV
installations in these areas.
An analysis of clipped elevation data determined usable slope within each site using
ArcMap’s Slope tool, which converts elevation data to slope values in degrees. Resulting values
were analyzed using the Raster Calculator tool. General Raster Calculator comparisons and the
SetNull function were used to eliminate areas of unfavorable slope. Equations 1 and 2 show two
separate Raster Calculator processes ran to accomplish this.
First, a Boolean raster was created which returned the number one if slopes were equal to
or less than 5°.

Secondly, the SetNull function was used to eliminate areas with unfavorable slope.
Using the Boolean raster from Equation 1 as the conditional raster, the function produced a raster
with values equal to the original clipped NED. However, this dataset only contained areas with
slopes equal to or less than 5°. This formula can be seen in Equation 2.

("Landfill_Slope" <= 5)
Where “Landfill_Slope” = The output of the Slope tool/ usable area raster

Equation 1 - Boolean Slope Function for CSP

49

Small, unconnected areas were removed by digitizing a new polygon feature around large
and connected areas of the raster file. This process was repeated as necessary until all islands of
raster cells were removed.
In order to account for all areas unassigned to dish-Stirling systems, including small
unconnected areas with less than or equal to 5° slope that were excluded, the ArcMap model
pictured in Figure 4 was used to determine areas to be analyzed for PV thin film systems:

Figure 4 - ArcMap Model for PV Geomembrane Raster Creation
First, values for the raster file used for dish-Stirling suitable areas were converted to
integers. This was necessary in order to convert the raster file to a vector format, a conversion
necessary to compare the space with that of the entire landfill. Areas of the entire landfill that
did not overlap with the vector file representing areas analyzed for dish-Stirling installations
were then extracted to create the file produced by the model in Figure 4. The resulting vector
file was used to clip the NED representing terrain data for the landfill. The resulting raster layer
represented the area of the landfill analyzed for PV suitability.
SetNull("Landfill_Slope_RasterCalculator ","Landfill_ UsableRaster”, "VALUE < 1")
Where “Landfill_Slope_RasterCalculator” = The output raster from Equation 1
“Landfill UsableRaster” = The output of the Slope tool/ usable area raster

Equation 2 - SetNull Raster Calculator Function for CSP

50

When considering a site for PV geomembrane technology, the recommended slope is 60°
or below (“Installation Manual for PVL” 2010, 8). The slope of areas being tested for PV
geomembrane potential was analyzed using Raster Calculator comparisons and the SetNull
function in ArcMap. Equations 1 and 2 were used for these calculations, replacing the
requirement of 5° with 60° to account for technology specific requirements of PV geomembrane.
Since this site inventory is for the Northern Hemisphere and PV geomembrane is
stationary, the slope of a landfill must have south, southeast, or southwest orientation.
Inappropriate slope orientations were found and removed using ArcMap’s Aspect and Raster
Calculator tools. First, a raster file representing elevation data for areas suitable for PV
geomembrane installation was used as an input for the Aspect tool, which produces a raster file
containing the slope orientation for each cell, representing a 10 m
2
area, in degrees. Next,
Aspects were kept ranging from southeast (112.5°) to southwest (247.5°), flat areas were also
used (-1°). This was done by using the output from the Aspect tool in the Raster Calculator
formula shown in Equation 3:

(("Landfill_Aspect" >= 112.5) & ("Landfill_Aspect "<= 247.5)) | ("Landfill_Aspect " == -1)
Where “Landfill Aspect” = The output raster from the Aspect tool

Equation 3 - Boolean Aspect Function for PV

51

The output from the above Raster Calculator calculation was modified with another
formula, as seen in Equation 4. This calculation was necessary in order to remove areas without
south-facing or flat slopes.

A solar power harvesting system must be able to provide enough electricity to make it
economically viable; otherwise the system is economically unfeasible and therefore not
considered in the present analysis. Because the size comparison conducted in the prescreen
portion of this analysis was based on the total footprint of each landfill, it was necessary to check
the area of a location again after obstacles and areas of unsuitable slope and aspect were
removed. According to (EPA and NREL 2012, 5) after deducting areas with obstacles present,
the usable land for a PV facility should be greater than two acres.
The minimum size needed for an economically viable dish-Stirling system is 10 acres.
This estimation was taken from the following studies. Schild (2004, 12) states that a dish-
Stirling installation should have a capacity of one MW. Dahle (2008, 25) establishes that dish-
Stirling systems require approximately ten acres of space per one MW of capacity. Although
this figure assumes a certain level of radiation is available to the site, it provides a minimum size
of which a dish-Stirling system could be viable. Locations that do not contain enough usable
land for either PV or CSP systems were not analyzed any further for that technology.
SetNull("Landfill_Aspect_RasterCalculator ","Landfill_ UsableRaster”, "VALUE < 1")
Where “Landfill_Aspect_RasterCalculator” = The output raster from Equation 3
“Landfill_UsableRaster” = The output of the Slope tool/ usable area raster
and "VALUE < 1" is the where clause which sets all cells with non-south facing
aspects to null

Equation 4 - SetNull Raster Calculator Function for PV

52

For each site and technology combination, the usable area raster layer was assigned as an
input raster for the Area Solar Radiation tool. By using the parameters and inputs described
below, Solar Analyst produced a raster layer representing the solar radiation in watt hours (Wh)
that reached each cell of the elevation over a year.
Although the tool only requires elevation data for input, Solar Analyst parameters must
be configured properly to yield results suitable to the study. Such parameters include which
dates to analyze incoming solar radiation since irradiation values fluctuate throughout the year.
The sky size, or resolution of the viewshed, sky map, and sun map, in units of cells per side,
must also be chosen. Since it would be computationally intensive to analyze solar radiation for
each hour of each day of the year, Solar Analyst allows the user to define both daily and hourly
intervals (Fu and Rich 1999).
Because of the variances in solar angle through one year, this study uses Solar Analyst to
survey incoming solar radiation for an entire year. However, the tool also asks for a specific
year to analyze. This study did not use the year of that the analysis was conducted, 2012,
because it is a leap year and is therefore not a good representative of an average year. Instead,
2011 was analyzed in this study since it contains a typical 365 days, a result visible in Figure 5.
It does not matter which specific year is chosen, it only matters how many days are in that year.
The calculation was run from the first to last day of the year to account for seasonal variations.
Daily and hourly intervals need to be optimized to yield the most accurate results from
Solar Analyst. According to Esri (2012), four days is the smallest recommended day interval
suggested, since sun tracks three days apart tend to overlap. A large sky size of 2800 was chosen
to complement the small day interval used as recommended by Esri (2012). This means that the
viewshed, sky map, and sun map used were 2800 x 2800 cells large, each cell representing 10

53

m
2
. These parameters were used to optimize the results of the model. The default hourly
interval, 0.5, was also used and can be seen in Figure 5.
In the tool’s output parameters, the user can chose an output raster indicating direct,
diffuse, or global radiation values (Fu and Rich 1999). Solar thermal technology only utilizes
direct solar radiation while photovoltaic systems take advantage of global radiation (Price and
Margolis 2010, 53; Lehman 2011, 15). Therefore, while evaluating for dish-Stirling systems,
only the direct radiation output from the Solar Analyst was used. Locations suitable for
photovoltaic systems were analyzed using a global radiation raster.

Figure 5 - Area Solar Radiation Tool Parameters

54

Esri’s Solar Analyst is composed of several calculations that together account for direct
and diffuse radiation, and the hemispherical viewshed algorithm is among the most important of
these calculations (Fu and Rich 1999, 4). Solar Analyst calculates a veiwshed for every cell
contained in the input DEM. A viewshed is composed of an area that is visible from a static
vantage point; in this case that viewpoint is a cell in the Digital Elevation Model viewing
upwards to the sky. The calculation finds the maximum angle of sky obstruction, or horizon
angle, in each direction. Each cell is finally assigned a value describing visible and obstructed
sky directions.
Solar Analyst also creates a sun map, where the sun’s location is calculated based on time
and latitude, represented via zenith and azimuth angles. These angles are placed into a two-
dimensional hemispherical projection with the same resolution of the viewshed (Fu and Rich
1999, 4). A map is created for December to June, and another from June to December; both
maps are divided into sectors where time duration, azimuth and zenith angles are calculated for
each sector’s centroid.
A skymap must be calculated for diffuse solar radiation. This map begins with the entire
sky and is then divided into zenith and azimuth angle increments. These angles are calculated
for the centroid of each sector (Fu and Rich 1999, 8).
When complete, the sunmap and skymap are each overlaid with the viewshed and a gap
fraction is calculated. This fraction represents the percentage of obstructed to unobstructed areas
in the overlain map. When the sun is blocked by nearby terrain or other obstacles, the area is
considered obstructed. This figure is found by dividing the number of unobstructed cells in the
viewshed by the total in that sector (Fu and Rich 1999, 9).

55

Once these maps are created and gap fractions are calculated, Solar Analyst can compute
direct radiation. For each unobstructed sunmap sector, direct radiation is found based on the gap
fraction, atmospheric attenuation, sun position, and ground receiving surface orientation. A
transmission model is used that begins with the solar constant and uses transmittivity and air
mass depth to account for atmospheric effects. The total direct insolation for a location equals
the sum of the direct radiation for all sunmap sectors. Fu and Rich (1999, 10) show the direct
insolation value from a sunmap sector, which has a centroid at zenith angle θ and azimuth angle
α, as calculated in Equation 5.

Solar Analyst calculates diffuse radiation for every sky sector, combined over the time
interval, and adjusted by the gap fraction and angle of incidence. Fu and Rich (1999, 11) provide
the diffuse insolation calculation shown in Equation 6:

Dirθ,α = SConst * τm(θ) * SunDurθ,α * SunGapθ,α * cos(AngInθ,α)
Where,
Dirθ,α = Total Direct Insolation
SConst = Solar Flux Constant
τ = Transmittivity of the Atmosphere
m(θ) = Relative Optical Path Length
SunDurθ,α = Duration of Sunlight
SunGapθ,α = Gap Fraction
cos(AngInθ,α) = Cosine of the angle of incidence between the axis normal to the surface and the centroid
of the sky sector

Equation 5 - Solar Analyst Direct Insolation

56

Global solar radiation is the sum of both direct and diffuse radiation. Reflective
radiation, while technically part of global radiation, is not included in Solar Analyst because of
its complexity and the relatively small influence it has on total radiation (Fu and Rich 1999, 29;
Huang and Fu 2009, 30; Gastli and Charabi 2009, 793).
While Solar Analyst is a sophisticated model, it is not without its faults. The most
notable limitation is that the model generalizes overcast conditions. Cloud cover is an important
factor when determining incoming solar radiation of an area and is addressed through radiation
parameters in Solar Analyst by estimating the proportion of radiation that passes through
overcast skies, and the proportion of diffuse radiation. These parameters, however, do not
directly account for the local overcast. The present work uses the default radiation parameter
values, which assume ‘generally clear skies’. This point is addressed further in Chapter 5 of this
work.
In order to estimate energy production at each location, the raw solar radiation produced
by the Area Solar Radiation tool was analyzed using spatial and non-spatial tools. Statistics for
each raster file were found using ArcMap. The cell count and mean values were copied from
Difθ,α = Rglb * Pdif * Dur * SkyGapθ,α * Weightθ,α * cos(AngInθ,α)
Where,
Difθ,α = Total Diffuse Insolation
Rglb = Global Normal Radiation (the sum of direct radiation from each sector without accounting for the
angle of incidence)
Pdif = proportion of Rglb that is diffused
Dur = Time Interval Used
SkyGapθ,α = Gap Fraction
Weightθ,α = Proportion of Diffuse Radiation Coming from a Given Sky Sector

Equation 6 - Solar Analyst Diffuse Insolation using the Uniform Sky Diffuse Model

57

ArcMap tables to an Excel spreadsheet to calculate estimated annual energy output for each
location in MWh.
Raster outputs from the Area Solar Radiation tool provided the total Wh that reached
each cell of the input raster during the year surveyed. The cell count was taken from each raster
and multiplied by the cell size of that raster, 129.60189 m
2
, to calculate the area for the site. 25
percent of this area was removed to account for service roads and other spacing considerations.
This method of accounting for spacing was also used by Arnette and Zobel (2011) and Van
Hoesen and Letendre (2010).
The area of each landfill was multiplied by the mean value of that site, which represents
the average incoming solar radiation in watt-hours (Wh). Following Arnette and Zobel (2011),
radiation values were multiplied by efficiency and derate factors to produce an approximation of
the energy produced at each site. Raw irradiation values were multiplied by an efficiency factor
of 11 percent for PV estimates and 25 percent for CSP installations to find an estimated annual
energy output for each location. The energy calculated represents DC energy, which must still
be converted to alternating current using a derate factor. Arnette and Zobel (2011) and
Salasovich and Mosey (2011a) multiply energy produced by 77 percent to account for this
phenomenon, the derate factor is also used in the current study.
Lastly, combination systems were added together and dish-Stirling and PV geomembrane
outputs were compared. Output values at locations able to support combination PV
geomembrane and dish-Stirling installations were summed to find the total output for the system.
Additionally, locations suited for CSP were compared to a PV installation at the same location,
to ensure that CSP was indeed the most efficient option for the location.

58

3.4 Aggregate Estimates for the Total Population
After each location was analyzed, the output for all sites was summed and divided by
their quantity to find the average output of landfill installations analyzed. To generalize results
from the sampled landfills to the total population, the average annual MWh generated per acre
from analyzed landfills was applied to the total acreage found in landfills that passed the
prescreening portion of this study. Acreage data was taken from CEC and EPA databases. This
calculation represents an estimation of the annual energy output potential for solar power
installations on closed or soon to be closed landfills in California.
Sampling theory, discussed by Dixon and Leach (1977), provided an estimate for the
likelihood of a random sample being a good representative of the total population. Based on
these calculations for simple random sampling, a sample size was drawn of seventeen, which
gives an uncertainty of +/- 20 percent at a 95 percent confidence level. This indicates a 95
percent chance that the actual value of potential solar power from the total lies within 20 percent
of the number determined from a sample.

59

Chapter 4 : Results
The outcomes from this study’s analysis are illustrated below. Results from prescreening
and detailed analyses are discussed and illustrated. Additionally, outcomes from sampled
landfills are generalized to estimate energy contribution from all landfills. This study’s
prescreen analysis began with a list of landfills from CEC and EPA sources. As previously
stated, a total of 324 landfills were included in this study. These sites had an average size of 101
acres. The largest site analyzed was 2,290 acres and the smallest was 0.8 acres.
4.1 Outcomes of Prescreen Analysis
Results from the prescreen portion of the analysis are discussed below. First, outcomes
from analyzing landfill characteristics like closing date information, location data, and landfill
size are presented. Additionally, results from the analysis of a site’s proximity to transmission
lines and roads are shown. Finally, landfills nearby natural gas pipelines and those with LFG to
energy projects are revealed. Parameter ranges, common reasons for exclusion, and the number
of sites that passed each step are summarized herein.
The landfills analyzed displayed a wide range of closing dates. Landfills closed as early
as 1958 and others were estimated to close as far as hundreds of years into the future. A total of
205 landfills had closing dates within the range between 1992 and 2022.
Location data was verified for each of the 205 landfills. Many records had accurate
coordinates or addresses, while others could not be so easily found. Eight landfills were
currently being used as dog parks, public parks, golf courses, and similar facilities. After
searching a variety of sources, forty five landfill locations could not be verified. 152 landfills
were location verified.

60

To determine which landfills are large enough to host either PV or CSP installation, the
size of a landfill was compared to the minimum size of an economically practical facility: 2 or 10
acres depending on the technology. Landfills showed much variation in size, ranging from 0.8 to
600 acres. Four locations were smaller than 2 acres and therefore deemed unfit for PV or CSP
installation. Twenty one locations fell between 2 and 10 acres and were considered viable
options for PV facilities, but not for CSP. 126 locations were feasible for either technology. A
total of 147 sites were found with an area over 2 acres and could therefore support feasible solar
power production.
Locations were also judged based on their vicinity to transmission lines and roads. All
landfills were at least one mile from a graded road and many sites had roads present that lead to
the site directly. After comparing landfills to FEMA transmission line data, some transmission
lines were located on top of landfills while others were as much as 50 miles away. Fifty four
locations passed this final condition of the prescreening process. These sites were location
verified and found to possess favorable closing dates, sufficient size to support a solar power
facility, and a location nearby the transmission grid and road systems.
Landfills that passed this prescreening analysis had an average size of 139 acres. The
largest site that passed prescreening requirements was 600 acres while the smallest was 5 acres.
A map of landfills that passed prescreening requirements can be seen in Figure 6. Sites are
dispersed throughout California with slightly larger concentrations surrounding areas of high
population such as Greater Los Angeles and San Francisco areas.

61

Figure 6 - Prescreened Landfills

62

The distance from natural gas pipelines to landfills that passed the above prescreening
prerequisites was calculated. While not a criterion, a location’s close proximity to natural gas
lines displays the landfill’s potential to integrate the energy source into a solar facility. Table 3
lists landfills that are estimated to have such potential by being at least 1/4 mile from a natural
gas pipeline and therefore have the added benefit of supplementing thermal engines.
Table 3 - List of Landfills with Natural Gas Supplementation Potential
Landfill City County
Distance from Natural
Gas Pipeline (Miles)
Vasco Road Livermore Alameda 0.25
Central Contra Costa Antioch Contra Costa 0.05
Chateau Fresno Fresno Fresno 0.2
Orange Avenue Fresno Fresno 0.25
China Grade Bakersfield Kern 0.15
Boron Boron Kern 0.05
Bradley Sun Valley Los Angeles 0
Miramar San Diego San Diego 0.05
Cold Canyon San Luis Obispo San Luis Obispo 0.16
Newby Island Milpitas Santa Clara 0.2
Fink Rd Crows Landing Stanislaus 0.15
Beale Air Force Base Beale Air Force Base Yuba 0

63

Because of the potential compatibility between LFG to energy and dish-Stirling engines,
landfills that passed the above requirements that also had a footprint larger than 10 acres are
shown in Figure 7 along with LFG to energy project data at those locations. The size of the LFG
project is illustrated to show locations with the greatest potential for dish-Stirling and LFG to
energy combination systems. It should be emphasized that these landfills were only prescreened
and were not necessarily submitted to the more detailed analysis performed later in the study.
Potential dish-Stirling facilities are found throughout California, especially in the western
and central areas. Locations with current LFG energy projects, however, seem to be focused in
the southwest part of the state. Of these seventeen sites, twelve are located in neighboring Los
Angeles, San Bernardino, Riverside, and San Diego Counties of southwest California. The
remaining four landfills are situated to the north in Monterey, Santa Clara, and San Joaquin
Counties in a region nearby San Francisco.

64

Figure 7 - Potential Locations for LFG Energy and CSP Solar Combination Systems

65

4.2 Detailed Analysis Results
The following section discusses the results from the detailed analysis portion of the
present study. Sites that constituted the random sample chosen for the detailed analysis are
introduced. Next, results from slope and aspect analyses are described. Solar Analyst outputs,
representing incoming solar radiation, are also discussed. Finally, results comparing dish-
Stirling systems to PV geomembrane on the same location are presented. Table 4 outlines the
seventeen landfills chosen in this analysis.

66

Table 4 - Sampled Landfills
Landfill Name City County
BKK West Covina Los Angeles
Central Contra Costa Antioch Contra Costa
City of Ukiah Ukiah Mendocino
Clover Flat Calistoga Napa
Cold Canyon San Luis Obispo San Luis Obispo
Echo Gold Fort Irwin (Mil Res) San Bernardino
Exeter Lindsay Tulare
Forward Manteca San Joaquin
Guadalupe San Jose Santa Clara
Hanford Hanford Kings
Lewis Rd Watsonville Monterey
Milliken Ontario San Bernardino
Miramar San Diego San Diego
Oasis Thermal Riverside
Orange Ave. Fresno Fresno
Redding Redding Shasta
Twentynine Palms Twentynine Palms San Bernardino

67

The slope and aspect characteristics of each landfill were unique. When analyzing for
dish-Stirling installations, fourteen of seventeen landfills passed minimum size requirements
after areas of high slope were removed. The City of Ukiah, Clover Flat, and Lewis Road
landfills did not have 10 acres of flat land left to justify an economically viable dish-Stirling
installation. While five landfills lost no usable area during this step of the analysis, the Clover
Flat Landfill lost over 97 percent of potential space for solar installations. The amount of land
lost for dish-Stirling slope requirements are summarized in the histogram shown in Figure 8.
Overall, more land was available with gentle slope for dish-Stirling systems than PV
geomembrane. This was favorable to the estimated potential of solar power projected from the
present study. The size of the remaining areas ranged from just over 10 acres to over 300 acres,
as seen in Figure 9.

Figure 8 - Percentage of Area Remaining After Slope Analysis for Dish-Stirling Installations

Figure 9 - Acreage Remaining After Slope Analysis for Dish-Stirling Installations
0
2
4
6
0 25 50 75 100
Frequency
% Area Lost from Slope Requirements
0
5
10
15
80 160 240 320
Frequency
Acres Remaining after Slope Requirements

68

In the case of combination systems, the residual area unsuitable for dish-Stirling
installations at each landfill was then tested for slope and aspect requirements for PV
geomembrane installations. No slope on any landfill analyzed was greater than 60°; however
prospective PV geomembrane space was lost to dish-Stirling facilities in the case of combination
facilities. At the Central Contra Costa and Milliken landfills, less than 2 acres remained after
dish-Stirling areas were designated. These two landfills were not analyzed further for
combination systems and instead were considered for dish-Stirling only facilities.
Of all requirements used in this detailed analysis, the southern-facing aspect criterion
proved to be the most restricting. An average landfill lost approximately two-thirds of its
previously usable area for PV geomembrane cells from this step. This is compared to the slope
requirement for dish-Stirling systems, which only claimed an average of just over one-third of
previously usable land. The City of Ukiah and Twentynine Palms landfill lost over 99 percent of
potential land for PV installations from this requirement. The histogram presented in Figure 10
display percentages of area lost from PV geomembrane aspect requirements.

Figure 10 - Area Remaining After Aspect Analysis for PV Geomembrane Installations
0
1
2
3
4
5
6
60 90 100
Frequency
% Area Lost from Aspect Requirements

69

Out of three locations being tested for PV only installations, the City of Ukiah landfill
was the only site that did not pass the aspect requirement. The City of Ukiah location was also
the only location found to be unsuitable for any solar development including dish-Stirling, PV
geomembrane, or a combination system.
Potential Oasis and Twentynine Palms landfill combination facilities did not have enough
area remaining for PV geomembrane after the aspect requirement was applied. These landfills,
in addition to the Central Contra Costa and Milliken locations, were found to be unsuitable for
combination systems due to PV geomembrane requirements. Therefore, these four landfills were
tested only for dish-Stirling systems in this inventory.
Table 5 provides a summary of landfill and technology combinations that were found to
be viable through the analysis of a landfill’s slope, aspect, and size. BKK, Cold Canyon,
Guadalupe, Miramar, and Redding landfills illustrated potential as combination systems. Central
Contra Costa, Echo Gold, Exeter, Forward, Hanford, Milliken, Oasis, Orange Avenue, and
Twentynine Palms landfills were analyzed for hosting dish-Stirling facilities, but were also
compared to a PV installation on the same location. Clover Flat and Lewis Road landfills only
showed potential for PV geomembrane systems while the City of Ukiah landfill was found to be
unsuitable for any solar development.

70

Table 5 - Landfill and Technology Combinations Tested in Analysis
Landfill Technology
Acreage
for PV
Acreage
for CSP
BKK CSP/PV Combination

167

20.4
Central Contra Costa CSP Only 0 78.6
City of Ukiah Unsuitable for Solar Development

0

0
Clover Flat PV Only

16.7 0
Cold Canyon CSP/PV Combination 46.5 29.5
Echo Gold CSP Only 0 10.0
Exeter CSP Only 0 43.6
Forward CSP Only 0 113
Guadalupe CSP/PV Combination 23.1 33.0
Hanford CSP Only 0 86.2
Lewis Road PV Only 10.4 0
Milliken CSP Only 0 169.6
Miramar CSP/PV Combination 74.1 310
Oasis CSP Only 0 26.4
Orange Avenue CSP Only 0 32.5
Redding CSP/PV Combination

22.3 40.4
Twentynine Palms CSP Only 0 42.0

Once viable location and technology combinations were found, direct and global
incoming solar irradiance were estimated for each using ArcMap’s Area Solar Radiation tool and
mathematical formulas. Table 6 summarizes these results for each landfill. Appendix B
provides greater detail on these calculations.

71

Table 6 - Estimated Annual Electricity Potential Summary
Landfill Technology
Annual MWh
from PV
Annual MWh
from CSP
Total Annual
MWh
BKK
CSP/PV
Combination 82,900 16,900 99,800
Central Contra Costa CSP Only 0 57,900 57,900
City of Ukiah
Unsuitable for
Either 0 0 0
Clover Flat PV Only 7,810 0 7,810
Cold Canyon
CSP/PV
Combination 22,200 23,400 45,600
Echo Gold CSP Only 0 11,600 11,600
Exeter CSP Only 0 34,000 34,000
Forward CSP Only 0 84,300 84,300
Guadalupe
CSP/PV
Combination 11,400 25,500 36,900
Hanford CSP Only 0 66,900 66,900
Lewis Road PV Only 4,770 0 4,770
Milliken CSP Only 0 141,000 141,000
Miramar
CSP/PV
Combination 36,900 259,000 296,000
Oasis CSP Only 0 21,200 21,200
Orange Avenue CSP Only 0 25,000 25,000
Redding
CSP/PV
Combination 9,790 28,800 38,600
Twentynine Palms CSP Only 0 35,600 35,600

Total 175,770 831,100 1,007,000

72

4.21 Outcomes from Detailed Analysis
Below, landfills characteristic of various outcomes from the solar analysis are shown and
discussed. The Miramar and BKK Landfills have the potential to host large combination solar
energy harvesting facilities. The Milliken Landfill was shown to have the largest potential for
dish-Stirling systems only while the Clover Flat Landfill has the largest potential for PV
geomembrane only facilities. Lastly, the City of Ukiah Landfill was shown to not have any
potential for solar facilities. Maps of solar potential for remaining landfills analyzed in this study
can be found in Appendix C.
The Miramar Landfill is in the City of San Diego and is just south of the Marine Corps
Air Station Miramar. The site is located nearby the San Clemente Canyon Freeway and Convoy
Street in northern San Diego. The location shown in Figure 11 was the largest surveyed, with an
area of nearly 500 acres. Open space, commercial, and residential areas surround the facility and
no significant obstacles were present at the location. The Miramar Landfill was shown to have
the largest potential to harvest solar radiation with an estimated 296,000 MWh of annual output
potential coming from both dish-Stirling and PV geomembrane systems.
Owned by the BKK Corporation, the BKK Landfill is located in West Covina of Los
Angeles County and pictured in Figure 12. The landfill is the third largest analyzed and seems to
host dense vegetation, likely trees or shrubs, on its west side, presenting obstacles to solar
installations. Galster Wilderness Park is located on the north side of the park, commercial and
residential developments surround the western and southern boarders of the location. The BKK
Landfill was shown to have the second largest potential for a combination dish-Stirling and PV
geomembrane systems with an estimated annual 99,800 MWh energy output.

73

The Milliken Landfill, illustrated in Figure 13, is located in Ontario at the junction of
Milliken Avenue and East Mission Boulevard. Ontario is located in San Bernardino County.
The landfill is surrounded by industrial and commercial development. No significant obstacles
were present as seen in the imagery. The Milliken Landfill was estimated to be the largest
potential plant for dish-Stirling systems, with a projected annual output potential of
141,000 MWh.
The landfill illustrated in Figure 14 was built in Calistoga, California of southwest Napa
County. Set at the foot of Clover Flat Road, the Clover Flat Landfill is completely surrounded
by vegetation and open space. The northwest side of the landfill is covered in trees or shrubs.
The surrounding area is mostly composed of open space, farms, vineyards, and wineries. The
Clover Flat Landfill displayed great potential for PV geomembrane installations, with an
estimated 7,810 annual MWh of potential output.
The City of Ukiah Landfill is located in Ukiah, the largest city in Mendocino County.
The landfill pictured in Figure 15 is north of Vichy Springs Road in an area with little
development. The site did not have any significant obstacles present. The facility is surrounded
by open space, although commercial and residential neighborhoods are located within a quarter
mile of the site. The City of Ukiah Landfill was found to be unsuitable for either dish-Stirling or
PV geomembrane systems. The terrain outlined in Figure 15 consisted of terrain with steep
slope (over 5°), most of which was not facing south. Given the terrain requirements for dish-
Stirling and PV geomembrane, this landfill does not contain enough land to host economically
viable facilities for either technology.

74

Figure 11 - Miramar Landfill Solar Potential

75

Figure 12 - BKK Landfill Solar Potential

76

Figure 13 - Milliken Landfill Solar Potential

77

Figure 14 - Clover Flat Landfill Solar Potential

78

Figure 15 - City of Ukiah Landfill Solar Potential

79

4.3 Estimates for the Solar Potential of Landfills in California
This study calculates the annual electrical contribution of solar power from California
landfills. The estimated annual potential of the landfills analyzed in this study totaled
1.01 gigawatt hours (GWh). Using acreage data from the CEC and EPA databases, this figure
was divided by the total size of sampled sites, 2,000 acres, to find the average annual MWh
output per acre. This figure, 0.520 MWh/acre, was multiplied by the acreage of each landfill that
passed the prescreen analysis. From these calculations, it is estimated that landfills in California
have the potential to contribute 3.78 GWh of solar energy to the electric grid annually.
Considering the average California home uses 567 kWh monthly, or 6804 kWh per year, it is
projected that an average of 555 homes could be powered annually from the solar energy
projected in this study (Energy Information Administration 2012b).
It would be illogical to assume that the landfills analyzed were perfect representatives of
the population sampled, as this is rarely the case; the uncertainty of this assumption must be
addressed using sampling theory. Given fifty-four landfills were included in this inventory and
seventeen of them were analyzed in detail, sampling theory suggests that the results from this
analysis are within 20 percent of 3.7 GWh with 95 percent confidence. Therefore, this study
estimates the potential annual solar energy generation from landfills to be between 3.02 GWh
and 4.54 GWh.

80

Chapter 5 : Discussion
This study’s projected findings are explored in the following chapter. First, results are
put into context by comparing them to California energy statistics and past studies. Next, major
findings from this study are defined. Limitations and assumptions of the present study are then
acknowledged and, finally, future areas of work are outlined.
5.1 Analysis Results
The present work estimates that landfills with closing dates between 1992 and 2022 could
generate a potential 3.7 GWh in California annually; this figure is compared to overall energy
generation and consumption. In 2010 California produced 204 GWh and consumed 259 GWh,
by retail sales, according to the U.S. Energy Information Administration State Electricity Profiles
of 2010 (Energy Information Administration 2012c, 25). The energy potential at closed, and
soon to be closed, landfills for California is therefore equivalent to 1.85 percent of the State’s
2010 energy generation and 1.46 percent of consumption based on the these figures. Relative to
California’s overall energy market, the potential contribution of solar power from closed landfills
is fairly small.
The State Electricity Profiles 2010 report shows that renewable energy contributed 28.9
percent of California’s electric power net generation for that year, or 28,793,591 MWh (Energy
Information Administration 2012c, 27). Renewable energies in the report include solar,
hydroelectric power, wind, biomass, LFG, sludge waste, and agricultural byproducts. Using this
figure, the prospective solar power estimated in this study is equivalent to 13 percent to the
State’s 2010 renewable energy production.

81

Overall, the results show significant potential to use landfill sites to expand California’s
solar energy production. According to the EPA’s eGRID, the CAMX subregion, which
encompasses most of California and no other states, solar power contributed 0.3003 percent of
212,768,947 MWh net generation for 2009. This means solar energy added 638,945 MWh to the
electric grid in that year. The potential annual contribution of 3.7 GWh estimated from closed,
and soon to be closed, landfills in this study is about 5.9 times California’s current solar
electricity production. These results show the significant potential found in California landfills
to produce solar energy as estimated in this study when compared to the current solar generation
in the State.
Next, the current study’s result is compared to similar past works. The analysis by
Carrión et al. (2008), estimated that Andalusia, Spain has an annual potential of 38,693 GWh on
406,000 acres of land. With an average annual generation of 95 MWh/acre, results from Carrión
et al. (2008) were exponentially higher than the present study, which estimates 0.520 MWh/acre.
This may be a result of Carrión et al. (2008) using technologies with a very high efficiency
factor, 78%, in their analysis and also because Andalusia receives plentiful incoming solar
radiation. This comparison shows that results found from the present analysis, based on energy
produced per acre, are not incredibly high by comparison.
Arnette and Zobel (2011) found a potential of 6,599,651.7 MWh of solar power annually
in the greater southern Appalachian Mountains area, a region similar in size to California. The
Appalachian study used similar land use requirements to Carrión et al. (2008) to find 405 sites
suitable for solar installation. Arnette and Zobel (2011), however, do not disclose the total area
of these analyzed sites in their report. Therefore, a comparison of the Appalachian study and the
current work based on estimated electricity generation per acre is not possible. This highlights a

82

challenge in conducting studies that take inventory of solar potential: there are few good points
of reference to compare one’s results to past studies, a point that is elaborated in Section 5.4.
5.2 Major Findings
In this section, three principal findings from the present study are discussed. First, while
relatively insignificant to the overall energy production of California, closed landfills have been
shown to be a significant source of solar energy in the State. Secondly, combining dish-Stirling
systems with PV geomembrane and LFG to energy facilities warrants further investigation. The
last major finding of the current analysis is that there are many unique assumptions, challenges,
and limitations when conducting a solar inventory of a specific land use.
As shown in the previous section, solar facilities on closed landfills have the potential to
provide a significant contribution to California’s solar energy generation. In fact, electricity
generation from these sites could potentially represent nearly six times the current solar energy
industry in the State. While it is unlikely that all of these sites could be developed in reality,
only 1/6 of this energy is needed to double California’s solar energy production. Furthermore,
since closed landfills are already sited on disturbed land, it is likely that these sites could provide
this energy with relatively little ecological damage. Other locations often looked to for siting
these projects often displace native species and disrupt virgin land with large scale solar farms;
siting medium sized solar facilities on closed landfills represents a solution to this ecological
damage (Abbasi and Abbasi 2002, 132).
This study has also shown the potential of combination systems for PV geomembrane,
dish-Stirling, and LFG to energy technologies for electricity production on closed landfills. The
first combination system type involves siting both PV geomembrane and dish-Stirling on the

83

same location. Because of the technologies’ varied terrain requirements, one system can often be
located on topography unfit for the other; combing both technologies at a single location results
in more usable land available for that site overall and therefore more electricity production.
Secondly, LFG to energy facilities utilize similar thermal engines to those used by dish-Stirling
systems. LFG to energy systems could potentially power the Stirling engine used in dish-Stirling
systems. Since there exists a large potential in closed landfills for siting solar power, and many
of these sites have LFG to energy facilities in place, these types of combination systems warrant
further technical investigation.
Lastly, this study has called to attention major challenges in conducting an inventory
using a specific land use type; these issues are discussed further in the next section. However,
the following challenge is applicable to a wide range of inventory studies: it is both imperative
and surprisingly difficult to find data to use as a base population for an inventory regarding a
specific land use. An analysis can be extremely detailed, but if potential sites are not included in
this analysis, the study’s results will be inaccurate. This study incorporated the most complete
data sources available to address this issue. However, it is likely that there are numerous
landfills with potential for solar installations that were not contained in this list. The quality of
these data is also important, for instance this study could not consider forty-five landfills because
they were missing spatial data and these locations could not be found.

84

5.3 Limitations and Assumptions
In this section limitations and assumptions of the current study are addressed. First, there
are many necessary simplifications involved in modeling solar radiation. The Solar Analyst tool
used to estimate solar radiation does not account for all factors that influence irradiation, such as
reflective radiation and empirical cloud cover data, and these limitations extend to the present
study. Additionally, there exists a lack of accurate and complete data regarding landfills in
California. Furthermore, this study’s methodology required transformations between raster and
vector data types, which can also yield inaccuracy.
This study encountered additional limitations. For instance, there are likely discrepancies
between the elevation data used in this analysis and actual landfill topography. Moreover, site
visits were not conducted in the current study and this may have caused inaccuracies in the
digitization of available area at landfills. Finally, the current work is a purely technical study and
neglects to account for many economic or policy factors that might make the development of
given technologies for specific sites more or less viable.
The present study is dependent on two large assumptions. First of all, dish-Stirling
systems have not been reported to be installed on landfills. This study assumes these systems
could physically be installed on such sites. Furthermore, because of a lack of data on the
technology, several attributes used for PV geomembrane systems were taken from general PV
thin film characteristics.
Providing an estimate for the solar potential of closed landfills presents challenges that
stem from the uncertainties in modeling solar radiation and converting this figure to energy
output. Esri’s Solar Analyst, used in this study to estimate solar radiation, does not empirically

85

measure incoming solar radiation and therefore is only an estimate of the power source. For
instance, Solar Analyst does not consider reflective radiation when calculating global radiation.
While this type of radiation is very small compared to direct and diffuse irradiance, incoming
radiation available for PV geomembrane cells could in theory be slightly higher than values
estimated by Solar Analyst.
Solar Analyst also does not model cloud cover directly, rather it uses radiation parameters
such as transmittivity and diffuse proportion to estimate the average fraction of radiation passing
through the atmosphere or irradiance that is diffuse. In many cases, these parameters are
sufficient to account for cloud cover in solar radiation estimations over multiple days (Fu and
Rich 1999, 29). The defaults of 0.3 and 0.5 were used for diffuse proportion and transmittivity
parameters, respectively, in all analyzed landfills; these parameters account for ‘generally clear
skies’ (Esri 2012). Since the cloud cover of all landfills accounted for in this study do not
necessarily fall into the category of ‘generally clear skies’, results from the present work do not
fully account for cloud cover of the analyzed sites.
To address the concern of cloud cover effects on solar radiation, NREL (2012) DNI data,
which accounts for the phenomenon through remote censored data, was compared to the average
kWh/m
2
/day from Solar Analyst estimates. This assessment found that Solar Analyst results
were typically smaller, with an average of 3.7 kWh/m
2
/day, compared to values between 2.2 and
8.8 kWh/m
2
/day found in NREL data (2012). Figures 1 and 6 illustrate NREL DNI data used for
this comparison. It can then be projected that, while Solar Analyst does not directly account for
variation in overcast conditions, radiation predicted in the current study is lower that from
empirically measured data.

86

There is a lack of accurate data regarding landfills in the United States and this is another
challenge faced in the current study. As addressed in the previous section there is currently no
official, accurate, and comprehensive database of landfill data necessary to conduct a more
thorough inventory of these facilities.
GIS files used to represent landfill areas in this study were converted between vector and
raster formats multiple times in the present study to work around limitations of the data types.
Outlines of landfills were digitized in vector format, which were then used to clip raster elevation
data. The geoprocessing necessary to find areas leftover for PV geomembrane systems after
locations were assigned to dish-Stirling facilities, as shown in Figure 3, converted raster to vector
data. This vector data was then used to clip raster elevation data. Raster and vector
transformations were necessary three times in the present study. The intrinsic differences
between these two data types cause another source of error for this study since the two data types
cannot overlap perfectly.
On average, this inaccuracy should not affect the results of the current study in any
particular direction. However, if the raster representation of an area is larger or smaller than the
area digitized in vector format, converting between the data types may increase or decrease the
area analyzed for solar potential accordingly. Results, in watt hours, derived from this area
would be skewed lower than predicted if the raster representation of the area was smaller than
the digitized landfill size. This inaccuracy, however, is extremely difficult to quantify and was
therefore not accounted for in the present study.
Modeling the terrain of a landfill using the NED is another source of uncertainty in this
process. The NED data used to obtain slope and aspect information represents the most current
elevation data from the USGS. However, unless a landfill was already closed for five years at

87

the time of the NED creation, the location’s terrain will change significantly from accepting
additional waste due to the settlement and grading of that waste. This would have a direct
impact on the usable land available at the site.
No site visits were conducted in the present study. Therefore, some presumptions were
necessary to digitize usable area at each site surveyed. It is possible that structures, vegetation,
and other obstacles visible in the imagery have since been removed or additional obstacles to
solar power systems have been created.
An economic analysis was not conducted for the present study. While dish-Stirling
systems are more efficient than PV geomembrane technology, it is possible that the latter could
be a better financial investment after relevant economic factors are considered. Furthermore,
siting solar power facilities on landfills, as proposed in the current work, could present financial
challenges unaddressed in the present study’s analysis.
The present study assumes that dish-Stirling systems can be installed on top of landfills.
However, an example of such an installation could not be found. A detailed analysis is therefore
necessary to determine if dish-Stirling systems have this capability. Such an analysis is
described in the following section.
This study assumes that attributes for PV geomembrane technologies are similar to other
second generation PV technologies. EPA and NREL studies found that ballasted thin film PV
systems were the most effective PV technology for landfills, but these studies did not test for PV
geomembrane (Salasovich and Mosey 2011a). It was assumed that PV geomembrane would be
economically comparable to ballasted thin film panels. Likewise, efficiency factors for PV
geomembane were not readily available. Therefore this study also assumes that PV

88

geomembrane has an efficiency factor similar to other thin film PV technologies. These
assumptions were necessary because PV geomembrane technology is relatively new and little
data are available on the subject.
5.4 Future Work
Since this study is the first to estimate the potential of landfills for hosting dish-Stirling
and PV geomembrane technologies, several areas of future work are presented. A technical
study is necessary to yield information on combination dish-Stirling, PV geomembrane, and LFG
to energy facilities. A further study regarding the ability of landfills to host dish-Stirling systems
is needed; more data on PV geomembrane’s use on landfills are also required. Studies that
measure effects of overcast conditions on the analyzed landfill sites are also an area for future
research.
Additional areas for future work can be cited. Because there is a lack of studies
analyzing solar energy potential that report power production per acre, studies that give this point
of reference are needed. Furthermore, a financial analysis of the proposed facilities would
answer questions concerning the economic feasibility of these systems. Also, a site specific
study is necessary for a practical installation of solar capturing technologies on a landfill.
Finally, as solar energy facilities sited on landfills are constructed, empirical data must be
collected from these locations to calibrate the model proposed in this study.
The present work shows the theoretical potential of combining PV geomembrane, dish-
Stirling, and LFG to energy systems. A combination of dish-Stirling and LFG to energy systems
deserves particular attention, since the two technologies could share a Stirling engine. Although
PV geomembrane and dish-Stirling systems could share transmission lines and related

89

equipment, these savings would likely be relatively small compared to two technologies
powering the same thermal engine. Because such a Stirling engine has not yet been engineered,
there is a future need to explore cost benefit calculations for these sites after these shared systems
are established. LFG to energy facilities would both increase ROI for these facilities and provide
a secondary power source that can be utilized when solar energy is low to match the timing of
electrical demand. This analysis would further refine landfill solar inventory modeling for future
studies.
Technical research is required to explore the feasibility of placing dish-Stirling systems
on landfills and to acquire additional information regarding the use of PV geomembrane on such
sites. Since dish-Stirling systems have not been installed on landfills in the past, it is unknown if
such an installation is feasible. Additionally, installation requirements used in the present study
should be compared in detail to empirical data from existing PV geomembrane installations on
landfills. Because such specifics were unavailable, this comparison was not possible in the
current work.
Economic studies such as cost-benefit analyses are necessary for a comprehensive
analysis of solar potential on closed landfills. A landfill must produce adequate electricity to
cover initial capital costs and ongoing operation and maintenance costs for a site to be
economically viable. The price that local utilities are willing to pay for electricity from these
plants determines how much revenue the location makes for a given amount of electricity
produced. Additionally, government policies affect the economic viability through grants,
government aid, limitations on grid-connected facilities, and other political variables that differ
from place to place. These issues were outside the scope of the present study but must be
addressed nonetheless by solar power developers.

90

As discussed in the previous section, radiation parameters used in Solar Analyst for this
study’s irradiation estimations were calculated under ‘generally clear skies’. In reality, each site
would have unique values for transmittivity and diffuse proportion depending on local climate.
If landfills presented in this study were to be analyzed more thoroughly, it would be necessary to
find radiation parameters that optimize Solar Analyst to account for local weather patterns.
There are few inventory studies estimating solar energy potential that report power
production per acre. Of all studies reviewed in the literature review of the present work, Carrión
et al. (2008) was the only report that provided enough information to compare results on an acre
to acre basis. Therefore, it was difficult producing a point of reference to relate this study’s
results to other works. It would be beneficial to any inventory study measuring solar potential to
compare energy produced per acre to past works; hence, a study comparing results of several
analyses in this way is needed to establish a good basis for comparison.
The methodology presented in this study represents a rough estimate of potential
locations and productivity available on closed landfills. A more detailed study is needed if any
of these sites were to be developed. A collaboration of landfill owners, engineers, experts in
dish-Stirling and PV geomembrane solar technology, and related professionals is necessary to
provide enough data to justify an actual installation at these sites.
Lastly, as solar facilities are installed on landfills in practice, the assumptions of the
current work will be tested. Solar electricity is a relatively new technology and siting these
systems on landfills is still an emerging concept with few real world examples. Because of this,
the present work was forced to make several assumptions. As more solar power is harvested on
landfills, the proposed model must evolve to take advantage of knowledge accumulated from the
successes and failures of these facilities related to specific site characteristics. With this

91

knowledge, inventory studies of such sites around the world using GIS modeling could be
accomplished with greater precision.

92

References
Abbasi, S., and N. Abbasi. "The Likely Adverse Environmental Impacts of Renewable Energy
Sources." Applied Energy, 4th ser., 65, no. 1 (April 2000): 121 - 44.
ArcGIS Desktop, Release 10, Esri, Redlands, California, United States
Arnette, Andrew N., N., and Christopher W. Zobel. "Spatial Analysis of Renewable Energy
Potential in the Greater Southern Appalachian Mountains." Renewable Energy 36, no. 11
(May 2011): 2785 - 798. http://www.sciencedirect.com.
Aydin, Nazliy Y. "GIS-Based Site Selection Approach for Wind and Solar Energy Systems: A
Case Study from Western Turkey." Thesis, Middle East Technical University, 2009.
http://etd.lib.metu.edu.
Badran, Qmar, Emad Abdulhadi, and Hussien Sarhan. "Solar Sites Assessment in Jordan Using
Fuzzy Logic." Applied Solar Energy 44, no. 3 (2008): 211 - 17.
http://www.sciencedirect.com.
Bravo, Javier D., Xavier G. Casals, and Irene P. Pascua. "GIS Approach to the Definition of
Capacity and Generation Ceilings of Renewable Energy Technologies." Energy Policy 35
(June 11, 2007): 4879 - 92. http://www.sciencedirect.com.
Byrne, John, and Lado Kurdgelashvili. "Solar Electricity." SAGE: Encyclopedia of Geography:
Barney Warf: 9781412956970. Accessed October 21, 2012. http://www.sagepub.com.
California. California Energy Commission. "CEC CalRecycle Database." CEC CalRecycle
Database. Accessed 2012.
http://www.calrecycle.ca.gov/SWFacilities/TechServices/EmergingTech/ CECData.xls.
California. Office of Planning and Research. Renewable Energy in California. Accessed January
21, 2013. http://www.opr.ca.gov.
Carrión, J., A. Espinestrella, F. Aznardols, and A. Ridao. "The Electricity Production Capacity of
Photovoltaic Power Plants and the Selection of Solar Energy Sites in Andalusia (Spain)."
Renewable Energy 33, no. 4 (April 2008): 545 - 52. doi:10.1016/j.renene.2007.05.04.
Demirbas, A. "Global Renewable Energy Projections." Energy Sources, Part B: Economics,
Planning, and Policy 4, no. 2 (December 09, 2011): 212 - 24.
http://dx.doi.org/10.1080/15567240701620499.
Dixon, Christopher John, and Bridget Leach. Sampling Methods for Geographical Research.
Norwich: Geo Abstracts, 1977.
Dubayah, Ralph, and Paul M. Rich. "Topographic Solar Radiation Models for GIS."
International Journal of Geographical Information Systems 9, no. 4 (1995): 405 - 19.
doi:10.1080/02693799508902046.

93

European Union. European Commission. European Research on Concentrated Solar Thermal
Energy. By Philippe Schild. European Communities, 2004. http://ec.europa.eu.
Esri. "ArcGIS Help 10.1 – Area Solar Radiation (Spatial Analyst)." ArcGIS Help 10.1. Accessed
December 14, 2012. http://resources.arcgis.com.
Federal Emergency Management Agency. "U.S. Renewable Green Energy Electric Transmission
ArcGIS Shapefile Map Layer." Accessed May 07, 2012.
http://www.mapcruzin.com/renewable-energy-us-electric-transmission-shapefiles.htm.
Fu, Pinde, and Paul Rich. "Design and Implementation of the Solar Analyst: An ArcView
Extension for Modeling Solar Radiation at Landscape Scales." In Proceedings of the
Nineteenth Annual ESRI User Conference. Proceedings of Nineteenth Annual ESRI User
Conference. 1999. http://proceedings.esri.com.
Gastli, Adel, and Yassine, Charabi. "Solar Electricity Prospects in Oman Using GIS-based Solar
Radiation Maps." Renewable and Sustainable Energy Reviews, 2009, 790 - 7.
doi:10.1016/j.rser.2009.08.018.
Germany. Federal Ministry for the Environment, Nature Conservation and Nuclear Safety.
Renewable Energies: Innovations for the Future. By Sigmar Gabriel. Berlin, 2006.
http://www.ppre.uni-oldenburg.de.
Green, Martin A. "Thin film Solar Cells: Review of Materials, Technologies and Commercial
Status." Journal of Materials Science: Materials in Electronics 18, no. S1 (April 3, 2007):
15-19. doi:10.1007/s10854-007-9177-9. http://copilot.caltech.edu.
Green, Martin A., Keith Emery, Yoshihiro Hishikawa, and Wilhelm Warta. "Solar Cell
Efficiency Tables (version 37)." Progress in Photovoltaics: Research and Applications
19, no. 1 (December 2010): 84-92. doi:10.1002/pip.1088.
Hetrick, W. A., Paul M. Rich, Fairley J. Barnes, and Stuart B. Weiss. "GIS-Based Solar
Radiation Flux Models." American Society for Photogrammetry and Remote Sensing
Technical Papers 3, GIS. Photogrammetry and Modeling (1993): 132 - 43.
http://www.creeksidescience.com.
Hofierka, Jaroslav, and Marcel Súri. "The Solar Radiation Model for Open Source GIS:
Implementation and Applications." In Proceedings of the Open Source GIS - GRASS
Users Conference 2002. Proceedings of Open Source GIS - GRASS Users Conference
2002, Trento, Italy. 2002. http://skagit.meas.ncsu.edu.
Huang, Shengli, and Pinde Fu. "Modeling Small Areas Is a Big Challenge." ArcUSer, Spring
2009, 28 - 31. http://www.esri.com.
"Installation Manual for PVL on Membrane: A Building Integrated Photovoltaic Flat Roof
Solution Using TPO and EPDM Membrane Roofing Systems." Uni-Solar. January 15,
2010. http://www.uni-solar.com.

94

International. International Atomic Energy Agency and United Nations Department of Economic
and Social Affairs. Energy Indicators for Sustainable Development: Country Studies on
Brazil, Cuba, Lithuania Mexico, Russian Federation, Slovakia and Thailand. 2007.
http://www.un.org.
Jason, Janke R. "Multicriteria GIS Modeling of Wind and Solar Farms in Colorado." Renewable
Energy 35 (2010): 2228 - 34. http://www.sciencedirect.com.
Kumar, Lalit, Andrew K. Skidmore, and Edmund Knowles. "Modeling Topographic Variation in
Solar Radiation in a GIS Environment." International Journal of Geographical
Information Science 11, no. 5 (1997): 475 - 97. doi:10.1080/136588197242266.
Lacey, Stephen. "2011 Outlook for Renewable Energy." Power Engineering, ISSN 0032-5961,
01/2011, Volume 115, Issue 1, P. 34 115, no. 1 (January 2011): 34 - 35.
http://www.proquest.com.
Lee, Amy H.I., Hsing Hung Chen, and He-Yau Kang. "A Model to Analyze Strategic Products
for Photovoltaic Silicon Thin film Solar Cell Power Industry." Renewable and
Sustainable Energy Reviews, February 2011, 1271 - 83. doi:10.1016/j.rser.2010.10.008.
Lehman, Ronald G. Concentrated Solar Thermal Facilities: A GIS Approach for Land Planning.
Thesis, University of Southern California, 2011. Ann Arbor, MI: ProQuest LLC., 2011.
http://www.proquest.com.
Lisell, Lars, Ian Metzger, and Jesse Dean. "SolOpt: A Novel Approach to Solar Rooftop
Optimization." ASHRAE Transactions 117, no. 2 (July 2011): 236 - 43.
http://go.galegroup.com.
Malik, Virginia A., Steven L. Lerner, and Donald L. MacLean. "Electricity, Methane and Liquid
Carbon Dioxide Production from Landfill Gas." Gas Seperation and Purification 1, no. 2
(December 1987): 77 - 83. http://www.sciencedirect.com.
Microsoft Excel, 2010, Microsoft, Redmond, Washington, United States
National Pipeline Mapping System. "NPMS Public Map Viewer." NPMS Public Map Viewer.
Accessed 2012. https://www.npms.phmsa.dot.gov.
O’Connell, Kim A. "Sorry, Landfill Closed." Waste Age 32, no. 7 (January 2001): 46 - 50.
http://www.proquest.com.
O'Leary, Philip, and Patrick Walsh. "Landfill Closure and Long-Term Care." Waste Age 34, no.
1 (January 2003): 44 - 7. http://www.proquest.com.
Pan, Tze-Chin, Jehng-Jung Kao, and Chih-Po Wong. "Effective Solar Radiation Based Benefit
and Cost Analyses for Solar Water Heater Development in Taiwan." Renewable and
Sustainable Energy Reviews 16, no. 4 (May 2012): 1874 - 82.
http://www.sciencedirect.com.

95

Patel, Mukund R. Wind and Solar Power Systems. Boca Raton: CRC Press, 1999.
http://books.google.com.
Pons, Xavier, and Miquel Ninyerola. "Mapping a Topographic Global Solar Radiation Model
Implemented in a GIS and Refined with Ground Data." International Journal of
Climatology 28, no. 13 (November 15, 2008): 1821 - 34. doi:10.1002/joc.1676.
Rasmussen, Sarah. "Producing Electricity from Methane in Iowa." BioCycle 46, no. 11
(November 2005): 37 - 9. http://www.cabdirect.org.
Rylatt, M., S. Gadsden, and K. Lomas. "GIS-based Decision Support for Solar Energy Planning
in Urban Environments." Computers, Environment and Urban Systems 25, no. 6
(November 2001): 579 - 603. doi:10.1016/S0198-9715(00)00032-6.
Salim, Mariam G. "Selection of Groundwater Sites in Egypt, Using Geographic Information
Systems, for Desalination by Solar Energy in Order to Reduce Greenhouse Gases."
Journal of Advanced Research 3, no. 1 (2012): 11 - 19. doi:10.1016/j.jare.2011.02.008.
SCS Engineers. Comparative Analysis of Landfill Gas Utilization Technologies. Reston, VA:
SCS Engineers, 1995. http://www.nrbp.org.
Suflita, Joseph M., Charles P. Gerba, Robert K. Ham, Anna C. Palmisano, William L. Rathje,
and Joseph A. Robinson. "The World's Largest Landfill: A Multidisciplinary
Investigation." Environment, Science, and Technology 26, no. 8 (1992): 1486 - 95.
doi:001 3-936W92~0926-1486$03.00
Tadock, James E. A GIS Analysis on Possible Photovoltaic Cell Use for Energy Reduction
During Peak Hours in Huntington, West Virginia. N.p.: MS Thesis, Marshall University,
Huntington, 2009. Web. 29 July 2012. http://www.marshall.edu.
Themelis, Nickolas J., and Priscilla A. Ulloa. "Methane Generation in Landfills." Renewable
Energy 32 (2007): 1243 - 57. doi:10.1016/j.renene.2006.04.020.
Tsatsarelis, T., A. Karagiannidis, and G. Perkoulidis. "Technologies of landfill gas management
and utilization." Protection and Restoration of the Environment VIII, Chania (2006): 3 -
7.
United States. Department of Energy. Energy Efficiency and Renewable Energy. 2008 Solar
Technologies Market Report. By Selya Price and Robert Margolis. January 2010.
http://www1.eere.energy.gov.
United States. Energy Information Administration. Annual Energy Outlook 2012 with
Projections to 2035. 2012. http://www.eia.gov.
United States. Energy Information Administration. "Residential Average Monthly Bill by Census
Division and State, Table 5.a." Electric Sales, Revenue, and Average Price. September
26, 2012. http://www.eia.gov.

96

United States. Energy Information Administration. State Electricity Profiles 2010. January 2012.
http://www.eia.gov.
United States. Environmental Protection Agency and National Renewable Energy Laboratory.
Screening Sites for Solar PV Potential. Accessed October 21, 2012. http://www.epa.gov.
United States. Environmental Protection Agency. Criteria for Solid Waste Disposal Facilities: A
Guide for Owners/Operators. 1993. http://www.epa.gov.
United States. Environmental Protection Agency. "EGRID 2012, Year 2009 Summary Tables."
Clean Energy. Accessed January 12, 2013.
http://www.epa.gov/cleanenergy/documents/egridzips/eGRID2012V1_0_year09_Summa
ryTables.pdf.
United States. Environmental Protection Agency. "Landfill Methane Outreach Program
Database." Accessed 2012. http://epa.gov/lmop/documents/xls/lmopdata.xls.
United States. Environmental Protection Agency. Solid Waste and Emergency Response. Solar
Power Installations on Closed Landfills Technical and Regulatory Considerations. By
Gabriel Sampson. 2009. http://clu-in.org.
United States. Environmental Protection Agency. Title 40: Protection of Environment; Part 258 -
Criteria for Municipal Solid Waste Landfills. Accessed March 12, 2012.
http://ecfr.gpoaccess.gov.
United States. National Renewable Energy Laboratory. "Lower 48 and Hawaii DNI 10km
Resolution 1998 to 2009." Solar Data. September 2012. http://www.nrel.gov.
United States. National Renewable Energy Laboratory. Energy Efficiency and Renewable
Energy. By Doug Dahle. July 14, 2008. http://www.blm.gov.
United States. National Renewable Energy Laboratory. Feasibility Study of Economics and
Performance of Solar Photovoltaics in Nitro, West Virginia. By Lars Lisell and Gail
Mosey. 2010. http://www.nrel.gov.
United States. National Renewable Energy Laboratory. Feasibility Study of Economics and
Performance of Solar Photovoltaics in the Commonwealth of Puerto Rico. By James
Salasovich and Gail Mosey. 2011. http://www.nrel.gov.
United States. National Renewable Energy Laboratory. Feasibility Study of Economics and
Performance of Solar Photovoltaics at Massachusetts Military Reservation. By Byron
Stafford, Robi Robichaud, and Gail Mosey. 2011. http://www.nrel.gov.
United States. National Renewable Energy Laboratory. Feasibility Study of Economics and
Performance of Solar Photovoltaics at the Refuse Hideaway Landfill in Middleton,
Wisconsin. By James Salasovich and Gail Mosey. 2011. http://www.nrel.gov.

97

United States. National Renewable Energy Laboratory. Feasibility Study of Economics and
Performance of Solar Photovoltaics at Johnson County. By James Salasovich and Gail
Mosey. 2012. http://www.nrel.gov.
United States. National Renewable Energy Laboratory. Modeling Photovoltaic and
Concentrating Solar Power Trough Performance, Cost, and Financing with the Solar
Advisor Model: Preprint. By Nate Blair, Mark Mehos, and Craig Christensen. Golden,
CO: National Renewable Energy Laboratory, 2008. http://www.nrel.gov.
United States. National Renewable Energy Laboratory. Photovoltaics on Landfills in Puerto
Rico. By James Salasovich and Gail Mosey. 2011. http://www.nrel.gov.
United States. National Renewable Energy Laboratory. Solar-augment Potential of U.S. Fossil-
fired Power Plants. By Craig S. Turchi, Nicholas Langle, Robin Bedilion, and Cara
Libby. Golden, CO: National Renewable Energy Laboratory, 2011. http://www.nrel.gov.
United States Geological Survey "USGS TNM 2.0 Viewer." USGS TNM 2.0 Viewer. Accessed
2012. http://viewer.nationalmap.gov/viewer/.
United States. Federal Emergency Management Agency. "U.S. Renewable Green Energy
Electric Transmission ArcGIS Shapefile Map Layer." Accessed 2012.
http://www.mapcruzin.com.
Van Hoesen, John, and Steven Letendre. "Evaluating Potential Renewable Energy Resources in
Poultney, Vermont: A GIS-based Approach to Supporting Rural Community Energy
Planning." Renewable Energy 35, no. 9 (2010): 2114 - 22.
doi:10.1016/j.renene.2010.01.018.
Waldo, Beek, and Rene Janseen. Hybrid Nanocomposites for Nanotechnology: Electronic,
Optical, Magnetic and Biomedical Applications. New York: Springer, 2009.
http://www.springerlink.com
Yang, Fan. Large Area A-Si/μc-Si Thin Film Solar Cells, Solar Cells - Thin film Technologies.
InTech, 2011. doi:10.5772/22642.
Zweibel, Ken, James Mason, and Vasilis Fthenakis. "A Solar Grand Plan." Scientific American
298, no. 1 (January 01, 2008): 64 - 73. http://people.bu.edu.

98

Appendix A - Landfills Passing Prescreen Requirements
CEC List
Name
EPA List
Name
Landfill City Landfill County Coordinates

Twin Bridges
LF
Anderson Shasta 40.495803, -122.202519
Contra Costa
SLF (aka
Pittsburg or
GBF LF)
Central Contra
Costa SLF
Antioch Contra Costa 37.9875, -121.845
Azusa LF
Azusa Land
Reclamation
Company, Inc.
Azusa Los Angeles 34.119639, -117.927593
China Grade
SLF
China Grade
SLF
Bakersfield Kern 35.425, -118.929
Beale AFB LF
Beale Air
Force Base
SLF
Beale Air Force
Base
Yuba 39.072978, -121.392578
Lamb Canyon
DS
Lamb Canyon
Disposal Site
Beaumont Riverside 33.88389, -116.99722
Big Bear RDS
Big Bear
Refuse
Disposal Site
Big Bear City San Bernardino 34.305549, -116.819583
Boron SLF Boron SLF Boron Kern 34.9903, -117.647
Buttonwillow
SLF
Buttonwillow
SLF
Buttonwillow Kern 35.4121, -119.46678
Clover Flat LF
Clover Flat
Landfill
Calistoga Napa 38.584, -122.534
Chiquita
Canyon
Chiquita
Canyon SLF
Castaic Los Angeles 34.434664, -118.645356

99

CEC List
Name
EPA List
Name
Landfill City Landfill County Coordinates
Colton LF
Colton
Sanitary
Landfill
Colton San Bernardino 34.04553, -117.345575
Fink Rd LF Fink Road LF Crows Landing Stanislaus 37.3882, -121.136
Edom Hill DS
Edom Hill
Disposal Site
Desert Hot Springs Riverside 33.88196, -116.438735
San Marcos
LF
San Marcos
LF
Escondido San Diego 33.090004, -117.197451
Echo Gold
Goldstone
Deep Space
Comm
Complex
Fort Irwin (Mil Res) San Bernardino 35.304479, -116.798329
Tri-Cities LF
Tri-Cities
Landfill
Fremont Alameda 37.49277, -121.99229
Chateau
Fresno LF
Chateau
Fresno LF
Fresno Fresno 36.687607, -119.945266
Orange Ave.
Orange
Avenue
Disposal Inc.
Fresno Fresno 36.687211, -119.761645
McCourtney
Rd LF
McCourtney
LF
Grass Valley Nevada 39.1726, -121.112
Hanford LF Hanford SLF Hanford Kings 36.297902, -119.598116
Hesperia RDS
Hesperia
Refuse
Disposal Site
Hesperia San Bernardino 34.34728, -117.3483
Highgrove LF
Highgrove
SLF
Highgrove Riverside 34.006708,-117.282228
Exeter DS
Exeter
Disposal Site
Lindsay Tulare 36.228947, -119.151619
Vasco Road
LF
Vasco Road
SLF
Livermore Alameda 37.753182, -121.722447
Harney Lane
LF
Harney Lane
SLF
Lodi San Joaquin 38.0994, -121.1364
Austin Rd. LF
Austin Road
Landfill
Manteca San Joaquin 37.879122, -121.191308
Yuba Sutter
Disposal Area
LF (YSDA)
Yuba-Sutter
Disposal Area
Marysville Yuba 39.17018, -121.550807
Newby Island
Newby Island
SLF Phases I,
II, & III
Milpitas Santa Clara 37.459837, -121.943829

100

CEC List
Name
EPA List
Name
Landfill City Landfill County Coordinates
Badlands DS
Badlands
Disposal Site
Moreno Valley Riverside 33.9535, -117.118
Kirby Canyon
LF
Kirby Canyon
Recycling &
Disposal
Facility
Morgan Hill Santa Clara 37.18507, -121.67109
Oasis DS
Oasis
Disposal Site
Oasis Riverside 33.439, -116.081
Milliken Milliken SLF Ontario San Bernardino 34.0365, -117.558
Oro Grande
Oro Grande
LF
Oro Grande San Bernardino 34.634903, -117.306705
Lewis Rd. LF
Lewis Road
SLF
Pajaro Monterey 36.880753, -121.699169
Redding SLF
(Benton)
City of
Redding/
Benton LF
Redding Shasta 40.571425, -122.411256
San Timoteo
SWDS
San Timoteo
Sanitary
Landfill
Redlands San Bernardino 34.01283, -117.21477
Sacramento
City LF
Sacramento
City LF
Sacramento Sacramento 38.58736, -121.45592
Crazy Horse
LF
Crazy Horse
Landfill
Salinas Monterey 36.80365, -121.618273
Miramar
SWLF
West Miramar
SLF
San Diego San Diego 32.856, -117.162
Guadalupe
SLF
Guadalupe
Sanitary
Landfill
San Jose Santa Clara 37.2114, -121.901
Cold Canyon
Cold Canyon
LF Solid
Waste
Disposal Site
San Luis Obispo San Luis Obispo 35.1873, -120.596
City of Santa
Maria LF
City of Santa
Maria Refuse
Disposal Site
Santa Maria Santa Barbara 34.950187, -120.377115
French Camp
LF
French Camp
Landfill
Stockton San Joaquin 37.916, -121.295

101

CEC List
Name
EPA List
Name
Landfill City Landfill County Coordinates
Bradley Ave
East & West
Bradley
Landfill
Sun Valley Los Angeles 34.240171, -118.384513
Lopez Canyon
LF
Lopez Canyon
SLF
Sylmar Los Angeles 34.293849, -118.392602
Corral Hollow
Corral Hollow
LF
Tracy San Joaquin 37.67, -121.457
Twentynine
Palms DS
Twentynine
Palms
Disposal Site
Twentynine Palms San Bernardino 34.1192, -115.965
City of Ukiah
SWDS
City of Ukiah
Solid Waste
Disposal Site
Ukiah Mendocino 39.169731, -123.165457
Buena Vista
DS
Buena Vista
Disposal Site
Watsonville Santa Cruz 36.91738, -121.81142
BKK West
Covina DS
BKK Landfill-
Phases I & II
West Covina Los Angeles 34.037973, -117.902573
Puente Hills
LF
Puente Hills
LF
Whittier Los Angeles 34.0203, -118.006
Teapot Dome
DS
Teapot Dome
Disposal Site
Woodville Tulare 36.0211, -119.106
Forward LF
Forward Inc.
Landfill
San Joaquin 37.874599, -121.188254

102
Appendix B - Solar Calculations
Landfill System
Raster
Cell
Count
Raw Area (m
2
)
Final
Area
(Acres)
Mean Wh/m
2

Raw Annual Wh for
Landfill
Annual AC Produced
(Wh)
Annual DC
Produced
(Wh)
Total Annual Energy
from Landfill (Wh)
Total Annual
Energy from
Landfill
(MWh)
BKK
PV 6,963 902,417.96 167.24 1,446,888.00 979,273,288,057.32 107,720,061,686.31 82,944,447,498.46
BKK
CSP 852 110,420.81 20.46 1,057,375.00 87,567,153,202.36 21,891,788,300.59 16,856,676,991.45 99,801,124,489.91 99,801.12
Clover Flat
PV 695 90,073.31 16.69 1,364,949.00 92,209,109,442.57 10,143,002,038.68 7,810,111,569.79 7,810,111,569.79 7,810.11
Cold
Canyon PV 1,936 250,909.26 46.50 1,395,041.00 262,521,527,730.32 28,877,368,050.33 22,235,573,398.76
Cold
Canyon CSP 1,229 159,280.72 29.52 1,019,117.00 121,744,269,290.97 30,436,067,322.74 23,435,771,838.51 45,671,345,237.27 45,671.35
Central
Contra
Costa CSP 3,271 423,927.78 78.57 946,137.90 300,820,606,194.68 75,205,151,548.67 57,907,966,692.48 57,907,966,692.48 57,907.97
Echo Gold
CSP 417 54,043.99 10.02 1,480,695.00 60,016,997,253.11 15,004,249,313.28 11,553,271,971.22 11,553,271,971.22 11,553.27
Exeter
CSP 1,817 235,486.63 43.64 999,151.70 176,465,153,113.70 44,116,288,278.43 33,969,541,974.39 33,969,541,974.39 33,969.54
Forward
CSP 4,712 610,684.11 113.18 956,180.40 437,943,129,332.06 109,485,782,333.02 84,304,052,396.42 84,304,052,396.42 84,304.05
Guadalupe
PV 964 124,936.22 23.15 1,430,941.00 134,082,271,790.75 14,749,049,896.98 11,356,768,420.68
Guadalupe
CSP 1,373 177,943.39 32.98 992,714.80 132,485,281,311.72 33,121,320,327.93 25,503,416,652.51 36,860,185,073.18 36,860.19
Hanford
CSP 3,590 465,270.79 86.23 995,551.20 347,400,666,323.44 86,850,166,580.86 66,874,628,267.26 66,874,628,267.26 66,874.63
Lewis Rd
PV 432 55,988.02 10.38 1,339,951.00 56,265,899,002.79 6,189,248,890.31 4,765,721,645.54 4,765,721,645.54 4,765.72
Milliken
CSP 7,060 914,989.34 169.57 1,066,866.00 732,128,265,626.84 183,032,066,406.71 140,934,691,133.17 140,934,691,133.17 140,934.69
Miramar
PV 3,089 400,340.24 74.19 1,450,874.00 435,632,432,079.52 47,919,567,528.75 36,898,066,997.14
Miramar
CSP 12,941 1,677,178.06 310.83 1,070,763.00 1,346,895,157,082.20 336,723,789,270.55 259,277,317,738.32 296,175,384,735.46 296,175.38
Oasis
CSP 1,098 142,302.88 26.37 1,033,737.00 110,327,810,490.97 27,581,952,622.74 21,238,103,519.51 21,238,103,519.51 21,238.10
Orange
Ave. CSP 1,353 175,351.36 32.50 988,864.20 130,049,009,645.12 32,512,252,411.28 25,034,434,356.69 25,034,434,356.69 25,034.43
Redding
PV 930 120,529.76 22.34 1,278,695.00 115,590,598,891.65 12,714,965,878.08 9,790,523,726.12
Redding
CSP 1,682 217,990.38 40.40 915,046.60 149,603,516,338.77 37,400,879,084.69 28,798,676,895.21 38,589,200,621.34 38,589.20
Twentynine
Palms CSP 1,749 226,673.71 42.01 1,088,693.00 185,083,557,436.25 46,270,889,359.06 35,628,584,806.48 35,628,584,806.48 35,628.58

103

Appendix C - Maps of Analyzed Landfills

104

105

106

107

108

109

110

111

112

113

114

Abstract (if available)

Abstract Solar radiation is a promising source of renewable energy because it is abundant and the technologies to harvest it are quickly improving. An ongoing challenge is to find suitable and effective areas to implement solar energy technologies without causing ecological harm. In this regard, one type of land use that has been largely overlooked for siting solar technologies is closed or soon to be closed landfills. By utilizing Geographic Information System (GIS) based solar modeling, this study takes an inventory of solar generation potential for such sites in the State of California. The study takes account of various site characteristics in relation to the siting needs of photovoltaic (PV) geomembrane and dish-Stirling technologies (e.g., size, topography, closing date, solar insolation, presence of landfill gas recovery projects, and proximity to transmission grids and roads). ❧ This work reaches three principal conclusions. First, with an estimated annual solar electricity generation potential of 3.7 million megawatt hours (MWh), closed or soon to be closed landfill sites could provide an amount of power significantly larger than California’s current solar electric generation. Secondly, the possibility of combining PV geomembrane, dish-Stirling, and landfill gas (LFG) to energy technologies at particular sites deserves further investigation. Lastly, there are many necessary assumptions, challenges, and limitations when conducting inventory studies of solar potential for specific sites, including the difficulty in finding accurate data regarding the location and attributes of potential landfills to be analyzed in the study. Furthermore, solar modeling necessarily simplifies a complex phenomenon, namely incoming solar radiation. Lastly, site visits, while necessary for validating details of the site, are largely impractical for a large scale study.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Calculating solar photovoltaic potential on residential rooftops in Kailua Kona, Hawaii

PDF

Modeling the spatio-temporal variability of solar radiation on buildings: a case study of Lewis Hall

PDF

Concentrated solar thermal facilities: A GIS approach for land planning

PDF

Site suitability analysis: small-scale fixed axis ground mounted photovoltaic power plants in Fresno, CA

PDF

Suitability analysis for wave energy farms off the coast of Southern California: an integrated site selection methodology

PDF

Concentrated solar thermal facilities: A GIS approach for land planning [map]

PDF

Identification and analysis of future land-use conflict in Mecklenburg County, North Carolina

PDF

A model for placement of modular pump storage hydroelectricity systems

PDF

Onshore wind power systems (ONSWPS): a GIS-based tool for preliminary site-suitability analysis

PDF

Site suitability analysis for implementing tidal energy technology in southern California

PDF

Improving the communication of water allocation decisions using interactive maps

PDF

Wrong way driving in San Antonio, Texas: a transportation route study using Network Analyst

PDF

Being green and the common citizen: developing alternative methods of renewable energy investment through solar power and efficient building design

PDF

Community gardens for social capital: a site suitability analysis in Akron, Ohio

PDF

Semi-automated visualization of spatial information in unstructured text

PDF

A spatial narrative of alternative fueled vehicles in California: a GIS story map

PDF

Historical temperature trends in Los Angeles County, California

PDF

The geography of voter power in the U.S. electoral college from 1900-2012

PDF

Modeling burn probability: a Maxent approach to estimating California's wildfire potential

PDF

Implementing spatial thinking with Web GIS in the non-profit sector: a case study of ArcGIS Online in the Pacific Symphony

Asset Metadata

Creator Munsell, Devon R. (author)

Core Title Closed landfills to solar energy power plants: estimating the solar potential of closed landfills in California

School College of Letters, Arts and Sciences

Degree Master of Science

Degree Program Geographic Information Science and Technology

Publication Date 02/17/2013

Defense Date 01/24/2013

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag ArcMap,California,dish-stirling,GIS,landfills,OAI-PMH Harvest,photovoltaic geomembrane,renewable energy,solar analyst,solar power

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Vos, Robert O. (committee chair), Ruddell, Darren M. (committee member), Swift, Jennifer N. (committee member)

Creator Email dmunsell@usc.edu,munsellde@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-220265

Unique identifier UC11292846

Identifier usctheses-c3-220265 (legacy record id)

Legacy Identifier etd-MunsellDev-1437.pdf

Dmrecord 220265

Document Type Thesis

Rights Munsell, Devon R.

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA