installation is to provide a solar energy rooftop potential map. ..... renewable energy tax credit of 30 percent, which expires in 2016, and the North Carolina ...
Reprinted with permission of the American Planning Association. This article originally appeared as the November/December 2010 issue of PAS Memo, the bimonthly online publication of APA’s Planning Advisory Service. PAS is a subscription service providing agencies and organizations with the latest planning resources and customized research assistance; learn more at www.planning.org/pas/about/.
PAS Memo — November/December 2010
Developing a Solar Energy Potential Map By Lyle Leitelt and Todd BenDor Throughout the U.S., municipalities are beginning to look for ways to decrease emissions and carbon output while increasing energy conservation. Part of the shift towards "green cities" is driven by a combination of state incentives and renewable energy portfolio standards that have encouraged the growth of efficient, renewable energy sources in many communities (U.S. DOE 2009a). There are several initiatives and projects that cities have begun to promote, including the installation of solar energy systems throughout the existing urban landscape. Increasingly, local governments are attempting to make solar energy more accessible and affordable for homeowners and businesses (Herbst 2009). One of the most innovative methods in helping communities with solar energy system installation is to provide a solar energy rooftop potential map. Solar mapping is a cost-efficient economic development tool that promotes renewable energy and helps planners satisfy a community's emissions goals. In addition, solar mapping helps planners continue their role as an information source by providing the public with vital property data. A solar map is a web-based, publicly accessible map that provides data on the solar suitability and energy potential of a community's built environment. Like many online planning department mapping portals, the information from solar maps helps the public make vital decisions that affect their property. By providing general solar potential information, the map attempts to encourage owners of suitable property to further research and construct solar panel systems. By constructing solar panel systems, both the property owner and the community receive numerous financial, infrastructure, and environmental benefits. This PAS Memo is intended to guide planners through the process of creating a solar energy potential map through a map-creation case study for the Town of Chapel Hill, North Carolina. By using widely available GIS tools, we created a map depicting the photovoltaic (PV) solar potential for every structure's rooftop. This map allows us to determine suitable areas for solar panels, while estimating energy savings potential, installation costs, long-term monetary savings (including rebates), and potential carbon emissions savings. The steps outlined below may be useful for local governments, utilities, and home/business owners interested in harnessing potential alternative energy sources on their property.
Background Rooftop Solar Power Until the past decade, the U.S. energy community has reluctantly ignored the potential of solar rooftops because of high upfront system construction costs (Shuford et al. 2010). Solar arrays represent a type of distributed power generation, whereby small-scale units generate electricity in lieu of traditional, centralized power plants. There are a number of benefits to solar energy, including increased energy security and independence (including diversification of energy sources), nearelimination of noise pollution and carbon emissions, and low operating and maintenance costs (Mantha 2008). Energy generated by solar panels is usually "net-metered," which enables customers to offset their grid-based energy consumption (from traditional power plants) over a billing period by allowing their electric meters to run backwards when they generate excess electricity. In addition, net metering encourages customer investment in solar energy and increases the value of the generated electricity by allowing the "banking" of excess energy (U.S. DOE 2009b). There is significant potential rooftop-generating capacity across the country. The U.S. Energy Information Administration has estimated that the amount of available single-story commercial roof space capable of supporting photovoltaic electrical power in the U.S. is more than 481 square miles
(Jeppesen 2004). The Energy Foundation concluded that all the capable commercial and residential rooftop space in the country could accommodate up to 710,000 MW of solar electric power, which is about 75 percent of the total electricity-generating capacity in the U.S. today (2005). Despite all the available capacity, solar energy makes up a miniscule amount of our power generation, with distributed solar generation making up a fraction of that. In the U.S., there is currently 1,106 MW of installed PV solar capacity, with only 5 MW existing in North Carolina (US DOE 2009a). The disadvantages of solar photovoltaic rooftop energy stem from the variability of the technology; solar power is entirely dependent upon the sun, while the PV devices lose energy during the power conversion process (CET 2010). Also, solar energy systems are highly dependent upon location and proper placement to optimize power generation. Nevertheless, there are a handful of municipalities that have created solar maps as a means of realizing the benefits of distributed solar energy and the available rooftop energy potential. Municipal Solar Mapping Benefits Within the last three years, solar mapping has become an important tool for communities trying to encourage sustainable and renewable energy alternatives. Starting with San Francisco in 2006, several large cities across the country have created maps to help inform their citizens. The number of large communities with solar maps is growing, as the Solar America Cities partnership has begun to encourage the development of maps within the program's 25 cities (U.S DOE 2009c). According to CH2M Hill, "solar maps radically impact citizens' ability to understand, evaluate and adopt solar energy in their homes and places of business" (McDermott 2008). Solar mapping is a gateway for solar installation in communities, as it creates a "solar portal" — an access point for citizens and business owners to specific information on the solar potential of their land (Herbst 2009).
Table 1: Comparison of Municipal Solar Energy Productions Number of PV Systems
System Annual Capacity (in Production MW) (in MWh)
CO2
Boston
78
2.7
3,240
1,857
N/A
Portland
457
3.2
3,491
1,851
$317,994
Berkeley
626
2.9
4,100
2,300
$538,000
San Francisco
2,043
9.2
13,437
4,036
N/A
San Diego
9,321
76.0
N/A
N/A
N/A
Savings (in tons)
Energy Cost Savings
Data provided from each municipality's solar map as of October 2010.
Municipal solar mapping has proved to be effective in communities that have initiated public, solar information "portals." San Francisco has seen tremendous success since creating the nation's first municipal solar portal. PV installations have grown by 60 percent and have doubled the amount of solar electricity generated (Herbst 2009). The city has even used its own solar mapping platform to install municipal solar systems, including constructing California's largest PV system with 25,000 panels generating about 5 MW (S.F. Public Utilities Commission 2010). The benefits of solar mapping in other cities can be seen alongside San Francisco in Table 1 (above). These solar portals help determine many of the factors involved in the solar energy decision making process, including solar electricity potential, installation costs, availability of rebates, estimated energy saving, and listings of installation contractors (Herbst 2009). Solar portals also benefit communities by increasing access to information, enhancing productivity and investment, helping to achieve state and federal energy goals, improving public-private collaboration, and increasing cost savings (Herbst 2009). When created as a public good, solar mapping provides people access to basic information in an efficient and cost-effective manner. Solar Mapping in Smaller Communities Although solar mapping has taken place in large cities across the U.S., there has been little effort to integrate smaller, less dense communities into the process. Some of this is attributed to the lack of interest or knowledge of solar mapping, but it may also be due to the lack of adequate data to create a meaningful system. In some cases, concerns over privacy, development costs, and staff
qualifications could hinder solar map development (Hyams 2009). Major urban centers are typically looked at in utilizing solar energy because of the ample number of rooftops they provide. However, many urban rooftops are used for HVAC systems, skylights, green roofs, or as outdoor space, which diminish their solar roof potential. In addition, the variation of building heights and the immediacy of buildings to one another creates increased shading and system variability. Having more rooftop square footage does not translate into having better rooftops for solar systems. Much of the solar rooftop potential is dependent upon a community's existing urban structure, because it affects density and available capacity. A New Zealand study found that residential patterns that have a density of 7.25 households per acre (more suburban than urban) have the greatest potential for domestic energy sustainability (Ghosh et al. 2006). Areas where populations are not concentrated in urban areas but rather are spread out over rural areas typically generate more power since more space per person is related to more roof area (Ludwig 2009). Thus, communities with urban forms like Chapel Hill, North Carolina, could have extensive benefits from rooftop solar energy analysis. Solar mapping is a critical tool for smaller communities to understand the quantitative benefits of their rooftop energy potential. Solar Mapping Technology Although most use basic GIS software, several variations of public solar maps and portals have been created using an array of computer software applications. Depending on the data used to create a solar map, some systems have used supplemental applications to create more detailed analyses. The City of Boston used ESRI's spatial analyst extension and flex widget polygon modeler to create their solar potential map (City of Boston 2010). In addition to selecting existing buildings potential, the flex widget allows users to draw polygons over suitable rooftops for solar panels (ESRI ArcGIS Resource Centers 2009). Although the flex widget is a free and useful tool, it does require significant web-based experience and has been reported to cause problems for other users (ESRI ArcGIS Resource Centers 2009). Western cities like San Francisco, Los Angeles, Berkeley, San Diego, Sacramento, and Portland, Oregon, have all utilized CH2M Hill's Solar Automated Feature Extraction (SAFE) technology to create their solar maps. SAFE assesses the solar potential of buildings through a combination of aerial imagery and advanced 3D modeling (Herbst 2009). The method takes into account factors such as roof obstructions like air conditioning units, chimneys, vents, azimuth (the direction of the sun), shadowing from other buildings, and roof slants. SAFE also calculates total roof area, usable roof area for solar panels, the amount of electricity the panels can produce, the electricity cost reduction, and resulting CO2 reduction (Palizzi 2008). CH2M Hill's solar mapping contract with the remaining Solar America Cities partnership communities will promote uniform mapping and easy comparison of solar energy potential across major U.S. cities. However, although SAFE will be wonderful resource for these communities, it can be quite costly, with estimates ranging from $20,000 for a low resolution map for a small city to $200,000 for a high resolution solar map for a large city (McDermott 2008).
Case Study: Chapel Hill, North Carolina Study Area The Town of Chapel Hill, North Carolina, is located in the southeast corner of Orange County in the state's Piedmont region. Chapel Hill is known as the home of the University of North Carolina and is a part of the Research Triangle along with the cities of Durham and Raleigh (Figure 1). The town covers an area of 21.2 square miles and has a population of 51,519 (Town of Chapel Hill 2010a). Figure 1
Chapel Hill, North Carolina
During the past decade, Chapel Hill was designated a Sierra Club "Cool City" by the U.S. Council of Mayors and joined the Cities for Climate Protection Campaign (Town of Chapel Hill 2010b). Both of these programs are intended to reduce the town's fossil-fuel dependency and greenhouse gas emissions. In 2006, Chapel Hill became the first U.S. municipality to pledge a CO2 emissions reduction of 60 percent below 2005 levels by 2050 (Town of Chapel Hill 2010b). To achieve these reductions, the town has implemented several sustainability-focused programs, including an
ordinance within their Land Use Management Ordinance (LUMO) that precedes and provides stronger protection than the North Carolina solar access law. Article 4.6.7.d of Chapel Hill's LUMO includes prohibitions against neighborhood or homeowner's association covenants, or other conditions of sale, that restrict or prohibit the use, installation, or maintenance of solar collection devices (Town of Chapel Hill 2004). By using the solar map that we have created, Chapel Hill can help continue its efforts of achieving sustainability and renewable energy. Solar mapping in North Carolina has been partly driven by Duke Energy's N.C. Solar PV Distributed Generation Program. Duke Energy (2009) plans to install and own solar panels on rooftops across the state, while paying a rental fee to property owners, in what will be one of the nation's earliest and largest demonstrations of distributed generation. A municipal solar map would help Duke Energy and local residents determine which sites can benefit from solar technology installation.
Methodology Evolution from other methods Several models and software applications currently exist to create a solar energy map, most of which were developed either by municipalities or GIS programmers. For this project, the initial model used came from the County of Los Angeles Solar Mapping Portal (Greninger 2009), which based much of their process on the City of San Francisco's solar map. However, when collecting the data needed to implement the Los Angeles model, several data discrepancies occurred. For example, Chapel Hill does not have digital surface model (DSM) data, which provides topographic data of the earth's surface features, such as buildings, roads, vegetation, and natural terrain features (Intermap 2010). Instead, the Chapel Hill model uses data available in most communities rather than inaccessible and expensive DSM data and detailed satellite imagery. Given this and other data discrepancies, the Chapel Hill model became very different from the initial Los Angeles model. In order to understand how the Chapel Hill map was created, a detailed step-bystep explanation is provided. Communities with similar base data can follow this process to construct their own solar potential map. GIS Model and Map Creation To overcome the lack of data on the height of features in the landscape, we created a "pseudo-DSM" model by subtracting digital elevation model data (DEM) from raw LiDAR data obtained from the NC Floodplain Mapping Program. LiDAR is a remote sensing system that is based on the transmission of relatively short-wavelength laser bursts from an airplane to the ground. The amount of light reflected back from the ground to the plane reveals the height of the terrain, yielding an extremely accurate elevation map (Jensen 2000). The raw LiDAR data were converted through a series of steps to a useable raster GIS data format. First, the cell assignment type was set to maximum to get the first returns or highest points (i.e. rooftops and treetops). Second, the cell size was set to create 8 ft2 cells. Cell size needs to be larger than average distance between LiDAR data points to avoid having enormous numbers of cells with no data in them. Cell size is recommended to be several times larger than the average point spacing, but small enough to identify larger data problems that warrant further investigation (Crawford 2008a). We determined that in our case, 8-ft2 cell size achieved this goal. After the LiDAR file conversion, we mosaicked LiDAR raster datasets (which usually arrive in several pieces) together. It was important to use the "maximum mosaic method," which selects the highest values of overlapping maps, as a way of eliminating some the empty value cells. At this point, all of the maximum LiDAR values for Chapel Hill existed on one file. However, there were still missing data cells due to return errors from the unprocessed data, which resulted in imagery with a salt and pepper effect. To solve for these errors, we estimated the value of missing cells by taking the average of each missing cell's neighborhood.1 Despite filling many of the no-data cells that created the salt and pepper effect, some larger void areas remained. These areas were accepted as is since our interpolation effort was not intended to fill major voids (Crawford 2008b). This process concluded the creation of the maximum elevations needed for the model. Since the maximum elevation levels were created, the base elevation layer needed to be applied. Bare digital elevation model (DEM) data2 were acquired from a Department of Transportation GIS system maintained by the State of North Carolina. By subtracting DEM elevations from our LiDAR elevations data, a pseudo-DSM elevation layer (depicting the heights of urban feature such as buildings) was created3 (Figure 2).
Figure 2
DSM-like Elevation Layer Image of UNC-Chapel Hill
This base elevation layer was then used with the area solar radiation tool, one of the solar radiation tools available in the spatial analyst extension of ArcGIS 9.2 that derives incoming solar radiation from a raster surface (see the ESRI ArcGIS 9.2 Desktop Help 2008 for more information). The output raster of the area solar radiation tool is presented in wH/m2/year [watt-hours per square meter per year]. This tool requires that several parameters be set, including the latitude, the time horizon with which to run the mode (we selected an entire year), and the interval for analyzing solar potential (e.g. hourly intervals). This tool took significant time to run (15-plus hours for an area the size of Chapel Hill) and required several trials to find a comparable solar radiation level to the National Renewable Energy Laboratory's Typical Meteorological Year 2 data (TMY2; from 1961-1990) for the surrounding area (4.4 kWh/m2/day for Chapel Hill is considered average; NREL 1994). After running the area solar radiation tool several times, the outputs were not matching the incoming solar radiation reported under the NREL TMY data. According to Mark Greninger, the creator of the County of Los Angeles Solar Map, this is a common issue for the tool (e-mail correspondence). He found the same data discrepancies in Los Angeles and adjusted his figures upward by 28 percent to match the incoming solar radiation as reported by their local NREL TMY2 stations. For Chapel Hill to reach comparable figures in the Raleigh data set, several of the radiation parameters in the area solar radiation tool were adjusted. This included changing the "diffuse proportion" from the default of 0.3 for generally clear sky conditions to 0.2 for very clear sky conditions. Diffuse proportion is the fraction of global normal radiation flux that is diffused (ESRI ArcGIS 9.2 Desktop Help 2008). The "solar radiation transmittivity" was also adjusted from the default of 0.5 for generally clear sky to 0.7 for very clear sky conditions. Solar radiation transmittivity is the fraction of radiation that passes through the atmosphere; it is inversely related to diffuse proportion (ESRI ArcGIS 9.2 Desktop Help 2008). Adjusting these factors finally created the comparable solar figures needed to create the solar
potential map for Chapel Hill. Results from the tool for a sample area are presented in Figure 3. Once we had solar radiation returns for the entire town, we wanted to identify areas of the built environment. Since Chapel Hill did have an existing building footprint shapefile, structures were easily identifiable and the solar radiation returns were extracted for buildings alone. This layer allowed for an easy and efficient way to attribute the data. Figure3
Area solar radiation returns for UNC-Chapel Hill
Analyses Aggregate Solar Analysis With the application of the solar radiation tool, some basic solar statistics for the entire town can be determined. The town has roughly 5,621,224 ft2 of available rooftop, which averages 3.89 kWh/m2/day of solar radiation. The gross rooftop solar radiation potential for the town is 2,735,582 kWh/m2/day or 998,488 MWh/m2/year. It is important to note that these figures represent pure solar radiation and do not take into account conversion or system efficiency when applied to these solar areas. If Chapel Hill were to place solar panels on the entire available roof area, the town could see significant benefits. Solar panel benefits were calculated using the power capacity and dimensions of the SunPower Corporation's 225 Solar Panel. This panel was chosen because it is one of the industry's most efficient, requires minimal rooftop area, and produces high energy per square foot (SRoeCo Solar 2010). In addition, the County of Los Angeles Solar Mapping Portal uses the same panel in their analysis. Some of the SunPower 225 specifications are presented in Table 2.
Table 2: SunPower 225 Solar Panel Key Specifications Solar Panel Key Specifications
Peak Power
225 W
CEC PTC Rating
207.1 W
Warranty
25 year limited power warranty Dimensions
Width
31.42 inches (2.62 feet)
Length
61.39 inches (5.12 feet)
Values from SunPower Corporation (2008)
Based on the available roof area of 5,621,244 ft2, the town could install 374,750 of the SunPower 225 solar panels. The complete system would have a capacity of 77.57 MW of power. To convert power to energy, megawatts (MW) need to be adapted to the average amount of watthours (Wh) in a day (4.93). Since power is not held constant from its source to conversion, a de-rate factor of 77 percent is applied to account for losses. Once all these factors are added, the solar panels have a potential annual energy output of 107,484 MWh. This is almost nine times the power needed for the town's municipal operations, which consumed 12,139 MWh in 2009 (Callaway 2010). According to the EPA (2010), 1,134.88 lbs of CO2 emissions are eliminated for every MWh of new solar energy capacity installed. Using this conversion, the entire roof space of Chapel Hill could produce an annual reduction of nearly 61,000 tons of CO2 emissions. See Table 3 for the town's aggregate solar potential values.
Table 3: Chapel Hill Aggregate Solar Analysis Total Available Roof Area
5,621,244 sq. ft.
Number of Potential Panels
374,750
System Size (in MW)
77.57
System in Wh (4.93h)
382,435,714.3
De-Rate Factor (77%) *
294,475,500
Annual Solar Energy Production (MWh)
107,484
CO2 Reduction Factor (lb/MWh)
1,134.88
CO2 Reduction (lbs/year)
121,980,940
* The DC-to-AC de-rate factor accounts for losses from the DC nameplate power rating and is the product of default loses from numerous system factors (NREL 2010).
Solar Suitability Payback Analysis Once the aggregate solar rooftop area was calculated, actual suitable roof area for solar panels needed to be determined. To find this area, a solar panel cost/efficiency ratio was applied to Chapel Hill by using a similar model created by the County of Los Angeles to calculate solar suitability (email communication with Mark Greninger). Solar suitability was determined by the amount of solar insolation (measure of solar radiation on a specific area) needed to create enough energy to offset the installation costs of a solar panel system. Areas with higher solar insolation are more desirable for solar panels since they would generate energy faster and payback the costs of a solar system sooner. This model bases its solar panel analysis on SunPower 225 Solar Panel, which has a warranty and expected lifespan of 25 years (SunPower Corporation 2008). Thus, solar suitability for Chapel Hill is based on the assumption that solar system costs would be paid off in 25 years. The cost of installing a PV system was calculated by taking the average installation costs of $10,000 per kW, divided by the 225's output of 4.83/kW (Duke Energy 2010). This value was then divided by
the panel size of 1.39 m2 for a total cost of $1,489/m2, which is lowered when applicable tax credits are considered. Only federal and state residential tax credits were considered in this analysis. The federal residential renewable energy tax credit of 30 percent, which expires in 2016, and the North Carolina renewable energy tax credit of 35 percent, which expires in 2015 were applied (Database of State Incentives for Renewables and Efficiency 2009 & 2010). Residents can receive both credits concurrently and save approximately 65 percent of their installation costs, reducing PV system costs to $521/m2. The model considers several other factors that affect suitability. In order to offset the $521/m2, the average cost per kWh was applied. In Chapel Hill, Duke Energy's residential service rate is roughly $0.09 per kWh (Duke Energy 2010). Although it is nearly impossible to predict future energy costs per kWh accurately, it can be assumed that the cost of kWh will not be static over 25 years. To account for any increases in electricity costs, an estimated annual rate of inflation of 3 percent over 25 years was applied to $0.09 per kWh. To determine suitable solar areas in kWh/m2/year with inflation, we can use the following equation, which relates system cost to system and panel efficiency, over a number of years (with increasing power costs; P = principle power cost, r = interest rate, and n = years): [Suitability Threshold] = [System Cost] / P(1+r)n...25 / [System Efficiency] / [Panel Efficiency] / [Daily Conversion] When applying Chapel Hill factors, the equation provides the following results: $521 / .09(1+.03)n...25 / 90% / 18% / 365 days in a year = 2.53 kWh/m2/day Under these conditions, to pay back the rebated PV system cost of $521 per m2, a solar system must generate 150.17 kWh/m2/year when power costs are increasing 3 percent each year. After applying the 90 percent system and 18 percent panel efficiency constraints, the suitable solar insolation threshold is approximately 2.53 kWh/m2/day. Based on the minimum solar radiation of 2.53 kWh/m2/day during a 25-year payback period, the model identified a substantially larger suitable roof top area of 4,516,416 ft2, which could support approximately 301,094 panels with a capacity of 62.33 MW. This capacity could create an annual energy output of 86,358 MWh, reducing emissions by approximately 49,000 tons of CO2. See Table 5 for the town's suitable solar potential figures and Figure 4 for a sample suitable area map.
Table 5: Chapel Hill Suitable Solar Analysis Total Suitable Roof Area
4,516,416 sq. ft.
Number of Panels
301,094
System Size (in MW)
62.33
System in Wh (4.93h)
307,269,846
De-rate factor (77%)
236,597,782
MWh/year
86,358
CO2 Reduction Factor (lb/MWh)
1,134.88
CO2 Reduction (lbs)
98,006,183
Figure4
Suitable solar areas (covered in red points) including energy cost inflation factors around the Polk Place area of the UNC-Chapel Hill campus.
Limitations of the Model This solar mapping procedure is not intended to produce a final, completely accurate solar potential map. Rather, it is intended to give users a general understanding of where solar panels may be suitable in their communities as well as give homeowners and businesses interested in applying solar panels to their property some basic understanding before getting substantially involved in the installation process. These types of maps are only informational tools. Ultimately, homeowners and businesses should contact a solar professional to discuss the feasibility of solar technology for their property. Because our case study solar map is not a detailed analysis of every structure in Chapel Hill, some liberties were taken. Unlike major cities, Chapel Hill does not have access to DSM data and highresolution imagery. Instead, we created a "pseudo-DSM" elevation layer for this analysis. Unfortunately, due to the spacing of LiDAR data and the need to limit large voids of empty data cells, the cell size for this analysis was set at 8 ft2, which is a low resolution compared to similar maps (e.g. 5 ft2 for Los Angeles's map). The large cell size meant that some rooftop attributes were not accurately portrayed. Slopes were grossly generalized and visible only on larger structures and in areas with steep drop-offs. Roof features that are typically recognized using CH2M Hill's SAFE technology were also neglected, which means that existing roof elements were also mistakenly included in the available solar area. Along with the typical roof elements, this analysis also includes structures lacking rooftops, like stadiums and parking structures. In addition, the large cells underestimated the town's available roof area by 3.6 percent; although our analysis determined the town has a rooftop area of around 5.62 million square feet, the town's building footprint layer contains rooftop area of 5.83 million square feet. The cost and age of available data is also important. Collecting LiDAR data typically cost tens of thousands of dollars depending on the flyover area. Therefore, minor updates to solar maps are
intermittent and typically done when an entire set of new LiDAR data is available. Our DEM and LiDAR data were over five years old, meaning that vegetation levels could have significantly changed in some areas. Representatives from Chapel Hill claimed that their building footprint layer was updated weekly, yet there are several recent buildings missing within the layer and therefore neglected in our analysis. It is also important to note that the area solar radiation tool is not necessarily the most accurate. To reach the local NREL TMY2 levels of solar radiation across Chapel Hill, the solar radiation tool was adjusted to show diffuse proportion and transmittivity at levels similar to very clear sky conditions. This adjustment acted as a calibration step to simulate solar levels that were similar to the local NREL TMY2 data published by the Department of Energy. Other solar maps used similar adjustments to calibrate their resulting values. Although the ArcGIS area solar radiation tool does not always provide the most accurate analysis, it is still the most accessible and user-friendly method for creating solar maps. In developing the solar suitability analysis, we made several additional assumptions. The suitability analysis is dependent upon applying a particular PV solar system, in this case the Sunpower Corporation's 225 Solar Panel. This panel is considered highly efficient, yet may cost more than other available panels. Also, the flat rate of $10,000 is on the higher end of average installation costs for residential application and is significantly higher than the average costs for larger commercial systems. When applying tax credits to reduce installation costs, the 65 percent tax rebate in the analysis refers only to residential applications and personal tax credits. There are also restrictions to these rebates; for example, solar energy systems attempting to receive the 35 percent North Carolina tax rebate cannot exceed $10,500 (Database of State Incentives for Renewables and Efficiency 2009). Corporate tax rebates are significantly larger with fewer restrictions, yet these rebates were negated in our analysis in order to focus on the wider residential audience in Chapel Hill. Further work could incorporate these tax factors and uncertainties into future analysis scenarios. Finally, the time to create a solar map is variable and is dependent upon city size, data processing, computing capabilities, and the number of times needed to run the area solar radiation to get comparable returns. It took roughly 60 hours to create the Chapel Hill map, not including the time spent running the solar radiation tool overnight. Creating a solar map is a trial and error process and takes adjustments and patience to create an accurate product.
Action Steps for Planners Before any municipality attempts to create a solar map for its community, some basic data and processing resources are recommended. Most communities should be able to obtain these without much effort or costs. After the data are collected, some straightforward steps can be taken to determine what rooftops are considered suitable for solar panels. z Obtain either Digital Surface Model (DSM) or Digital Earth Model (DEM) and LiDAR data. With
DSM data, little work will be required to create a municipal solar map, whereas with only DEM data, you will need to create a "pseudo-DSM" model by subtracting the DEM data from LiDAR data (see GIS Method and Map Creation Section). z Have access to ArcGIS's Spatial Analyst Extension and the Area Solar Radiation Tool. Use NREL's
Data Manual to find returns comparable to your community. Allow significant time to run this tool, typically overnight. Finding comparable solar levels is a trial and error process. Adjust the "diffuse proportion" or "transmittivity" factors, or apply a percentage factor to achieve desired solar figures. z Must have access to local building footprints or other GIS layers that delineate structures or
other urban features within the community. z Once the solar radiation levels of the communities' rooftops is identified, one can find the roof's
energy potential of suitable locations. Find the specifications of a desired solar panel that would be typical in your community. Apply the costs of installing a PV system and divide that by the panel's power output to find the cost/m2. z Apply any local, state, or federal tax or financial incentives to install a PV system. This will
significantly decrease the cost/m2. z Find the average cost per kWh in your community. Apply an inflation factor to the cost per kWh
that would last over the life of the PV system. In the Chapel Hill example, the SunPower panels
are expected to last at least 25 years. z Once cost and lifespan is determined, apply the system and panel efficiency factors over a year.
This should provide a minimum solar radiation level in kWh/m2/day (see Solar Suitability Payback Analysis Section). z With the minimum solar radiation level, one can use the solar radiation layer in GIS to identify
rooftop areas that are at or above this solar factor. z When all potential areas are identified, CO2 reduction factors can be applied (from the DSIRE
database – see Resources section).
Conclusion Solar mapping can be a helpful tool for planners and municipalities that are trying to encourage communities to adopt solar technology. Creating a solar suitability map for the Town of Chapel Hill has shown that with minimal data and resources, many communities can create their own generalized solar suitability map. The creation of a potential solar map is based on two basic processes. First, we created an elevation layer and applied the area solar radiation tool. Second, we applied several suitability factors to determine which areas are appropriate for rooftop solar energy. The solar map we created will be helpful to the residents and businesses of Chapel Hill who are thinking about applying photovoltaic solar panels on their structures. In terms of familiarity with GIS, an intermediate knowledge of GIS and a few GIS educational courses or a year of GIS work experience is recommended. Experience with the spatial analyst extension is also helpful. Municipal solar maps are a critical tool of information sharing and can be applied in every community, even those with minimal solar energy capabilities. The creation of solar maps usually corresponds with other sustainability efforts like energy independence or carbon reduction programs. By creating these maps, the hope is that businesses and homeowners will become interested in solar energy and will contact a solar installer for a more detailed estimation. Municipal solar maps can be an important tool to help planners disseminate valuable information to their communities. About the Authors Lyle Leitelt is a recent graduate of the Department of City and Regional Planning at the University of North Carolina at Chapel Hill. His master's degree focused on land use planning and design and preservation of the built environment. He is particularly interested in energy issues and its effects on planning. He previously was a long range and transportation intern for the Town of Chapel Hill Planning Department. Todd BenDor is an assistant professor of City and Regional Planning at the University of North Carolina at Chapel Hill. His research and teaching focus on environmental and land use planning. He is particularly interested in using technology and modeling to better understand the impacts of urban growth on sensitive environmental systems. The authors would like to thank those that helped with this study, including Mark Greninger, (Los Angeles Information Office), Amanda Henley (University of North Carolina at Chapel Hill), and Alexander Winn (George Washington University Solar Institute). Notes 1. The following conditional (Con) expression was used in the spatial analyst calculator to fill empty data cells (Crawford 2008b). In this example, the output file is called "output": Con(IsNull([output]), FOCALMEAN ([output] , RECTANGLE , 3 , 3, data), [output]) 2. A digital elevation model (DEM) is a bare-earth model in which features such as buildings, roads, and vegetation canopy are digitally removed from the landscape (Intermap 2010). 3. The expression used in the raster calculator to create the desired elevation layer was similar to the following example: [LiDAR_elevation]-[DEM_elevation] Resources GIS Data Obtaining LiDAR Data: z USGS LiDAR Information: http://lidar.cr.usgs.gov z USGS LiDAR Data: http://lidar.cr.usgs.gov/LIDAR_Viewer/viewer.php
z State, regional, and local governments may provide their own free data; check websites or
contact state offices (emergency management, floodplain programs, departments of transportation, natural resources, etc.)
ArcGIS Help: z Importing Terrain Data Sets into ArcGIS:
http://webhelp.esri.com/arcgisdesktop/9.3/index.cfm?TopicName= Importing_terrain_dataset_source_measurements z Assessing LiDAR Coverage and Sample Density:
http://blogs.esri.com/Dev/blogs/geoprocessing/archive/2008/10/29/LidarSolutions-in-ArcGIS_5F00_part-1_3A00_-Assessing-Lidar-Coverage-and-SampleDensity.aspx z Creating Raster DEMs and DSMs from LiDAR Point Collections:
http://blogs.esri.com/Dev/blogs/geoprocessing/archive/2008/12/15/LidarSolutions-in-ArcGIS_5F00_part2_3A00_-Creating-raster-DEMs-and-DSMs-from-largelidar-point-collections.aspx z Solar Radiation Tool: http://webhelp.esri.com/arcgisdesktop/9.2/index.cfm?
TopicName= An_overview_of_the_Solar_Radiation_tools
Data Resources NREL Solar Radiation and Data Manuel (to compare radiation levels for a community): http://rredc.nrel.gov/solar/pubs/redbook/PDFs/ DSIRE's List of Renewable and Energy Efficiency Government Incentives: www.dsireusa.org/ EPA's eGRIDweb Database converts MWH into pounds of CO2 per region (helps in calculating CO2 reduction): http://cfpub.epa.gov/egridweb/ghg.cfm Municipal Solar Map Websites Berkeley, California: Berkeley Solar Map:http://berkeley.solarmap.org/solarmap_v4.html Boston: Solar Boston: http://gis.cityofboston.gov/solarboston/ Los Angeles: Los Angele County Solar Map: http://lacounty.solarmap.org Portland, Oregon: Oregon Clean Energy Map: http://oregon.cleanenergymap.com/ Sacramento, California: Solar Sacramento: http://smud.solarmap.org/map.html San Diego, California: San Diego Solar Map: http://sd.solarmap.org/solar/ San Francisco: San Francisco Solar Map: http://sf.solarmap.org/ Commercial Mapping Resources Although municipal solar mapping works well as an informational tool for those communities that have them, most municipalities do not provide solar mapping as an amenity. For those still interested in solar mapping benefits, there exists a handful of private solar mapping websites. These sites typically follow similar methodology in how they measure solar potential, but differ in how the information can be accessed. Rather than mapping out entire communities, they focus purely on specific locations customers want analyzed. Commercial solar website portals (for individual buildings): z www.solarrating.ca/ z www.solar.coolerplanet.com z www.sungevity.com z roofray.com z 1bog.org
z greenpowerlabs.com
References Callaway, B. 2010. Town of Chapel Hill greenhouse gasemissionsinventorymunicipaloperations:2005through2009. Retrieved from http://dc.lib.unc.edu/cdm4/item_viewer.php?CISOROOT=/s_papers&CISOPTR=1314. Center for Ecological Technology. 2010. Advantages and disadvantages of photovoltaics. Retrieved from www.cetonline.org/Renewables/PV_pro_con.php. City of Berkeley. 2010. Berkeley Solar Map. Retrieved from http://berkeley.solarmap.org/solarmap_v4.html. City of Boston. 2010. Solar Boston. Retrieved from http://gis.cityofboston.gov/solarboston. City of Portland Bureau of Planning and Sustainability. 2010. Oregon Clean Energy Map. Retrieved from http://oregon.cleanenergymap.com. City of San Diego. 2010. San Diego Solar Map. Retrieved from www.sd.solarmap.org/solar. County of Los Angeles. 2009. Los Angeles County Solar Map. Retrieved from http://lacounty.solarmap.org. Crawford, Clayton. 2008. LIDAR solutions in ArcGIS_part1: assessing lidar coverage and sample density. Retrieved from http://blogs.esri.com/Dev/blogs/geoprocessing/archive/2008/10/29/Lidar-Solutions-inArcGIS_5F00_part-1_3A00_-Assessing-Lidar-Coverage-and-Sample-Density.aspx. Crawford, Clayton. 2008. LIDAR solutions in ArcGIS_part2: Creating raster DEMs and DSMs from large LIDAR point collections. Retrieved from http://blogs.esri.com/Dev/blogs/geoprocessing/archive/2008/12/15/Lidar-Solutions-inArcGIS_5F00_part2_3A00_-Creating-raster-DEMs-and-DSMs-from-large-lidar-pointcollections.aspx. Duke Energy. 2010. Customer generation and solar energy FAQ. Retrieved from www.dukeenergy.com/pdfs/customer-generation-solar-energy-faq.pdf. Duke Energy. 2009. North Carolina solar PV distributed generation program: harnessing the power of the sun. Retrieved from www.duke-energy.com/north-carolina/renewable-energy/nc-solardistributed-generation-program.asp. Database of State Incentives for Renewables and Efficiency. 2009. North Carolinaresidential energy tax credit (personal). Retrieved from www.dsireusa.org/incentives/incentive.cfm? Incentive_Code=US37F&re=1&ee=1. Database of State Incentives for Renewables and Efficiency. 2010. Federal residential renewable energy tax credit. Retrieved from www.dsireusa.org/incentives/incentive.cfm? Incentive_Code=US37F&re=1&ee=1. Energy Foundation. 2005. Energy foundation study finds residential and commercial rooftops could support vast U.S. market for solar power. ESRI ArcGIS Desktop 9.2 Help. 2008. Area solar radiation. Retrieved from http://webhelp.esri.com/arcgisdesktop/9.2/index.cfm?TopicName= Area_Solar_Radiation. ESRI ArcGIS Desktop 9.3 Help. 2009. Importing terrain dataset source measurements. Retrieved from http://webhelp.esri.com/arcgisdesktop/9.3/index.cfm?TopicName= Importing_terrain_dataset_source_measurements. ESRI ArcGIS Resource Centers. 2009. Sample solar radiation analysis and widget. Retrieved from http://resources.esri.com/arcgisserver/apis/flex/index.cfm?fa= codeGalleryDetails&scriptId=16110. Ghosh, S., R. Vale, and B. Vale. 2006. Domestic energy sustainability of different urban residential patterns: a New Zealand approach. International Journal of Sustainable Development, 9(1), 16-37. Greninger, M. 2009. Poster: Developing the Solar Model for the County of Los Angeles Solar Mapping
Portal. Herbst, D. 2009. Solar mapping: demystifying solar potential. Renewable Energy Focus 10(4), 32-35. Hyam, M. 2009. JUCCCE mayor's training program case study: San Francisco solar map. Retrieved from http://energy.sipa.columbia.edu/researchprograms/urbanenergy/documents/ SF% 20solar%20map%20FINAL.pdf. Intermap Technologies. 2010. Digital Elevation Models. Retrieved from www.intermap.com/digital-elevation-models Jensen, J. 2000. Remote sensing of the environment: an earth resource perspective. Upper Saddle River, N.J.: Prentice Hall. Jeppesen, B. 2004. "Rooftop solar power: The solar energy potential of commercial building rooftops in the USA." Refocus 5(4), 32-34. Ludwig, D. 2009. Mapping the solar potential of rooftops: Germany's SUN-AREA research project uses GIS. Retrieved from www.esri.com/news/arcnews/fall09articles/mapping-the-solar.html. Mantha, P. 2008. Presentation: GIS as a tool to evaluate the solar potential of fifty largest public buildings and all public parking lots in Pittsburgh. Retrieved from www.slideshare.net/pmantha/mantha-consolidated-thesis. McDermott, M. 2008. More cities to get their own solar power potential maps. Retrieved from www.treehugger.com/files/2008/10/25-united-states-cities-to-get-solar-powerpotential-maps.php. National Renewable Energy Laboratory. 1994. Solar radiation data manual for flat-plate and concentrating collectors: North Carolina data tables. Retrieved from http://rredc.nrel.gov/solar/pubs/redbook/PDFs/NC.PDF. National Renewable Energy Laboratory. 2010. How to change PVWatts parameters. Retrieved from www.nrel.gov/rredc/pvwatts/changing_parameters.html. Palizzi, A. 2008. Presentation to California Solar Forum: CH2M Hill enterprise management solutions solar mapping web portal using CH2m Hill S.A.F.E. methodology. Retrieved from www.californiasolarcenter.org/pdfs/forum/2008.2.1_SolarForum_TPalizzi-SolarMapch2mhill.pdf. Sacramento Municipal Utility District. 2010. Solar Sacramento. Retrieved from http://smud.solarmap.org/map.html. San Francisco Department of the Environment. 2010. San Francisco Solar Map. Retrieved from http://sf.solarmap.org. San Francisco Public Utilities Commission. 2009. Mayor Newsom signs legislation establishing California's largest solar photovoltaic installation at Sunset Reservoir. May 15. Retrieved from http://sfwater.org/detail.cfm/MC_ID/12/MSC_ID/139/MTO_ID/243/C_ID/4554. Shuford, S., S. Rynne, and J. Mueller. 2010. Planning for a New Energy and Climate Future. PAS Report 558. Chicago: American Planning Association. SRoeCo Solar. 2010. Solar panel comparison. Retrieved from http://sroeco.com/solar/learnsolar/solar-panel-comparison/. SunPower Corporation. 2008. 225 solar panel. Retrieved from http://us.sunpowercorp.com/downloads/product_pdfs/residential/ SunPower_225bk_res_en_lt_w_ra.pdf. Town of Chapel Hill. 2004. Land use management ordinance. Retrieved from http://townhall.townofchapelhill.org/planning/planning_development/ pdfs/lumo_art4.pdf. Town of Chapel Hill. 2010. Fast facts about Chapel Hill. Retrieved from www.ci.chapelhill.nc.us/index.aspx?page=43. Town of Chapel Hill. 2010. Toward a sustainable community. Retrieved from www.townofchapelhill.org/index.aspx?page=235.
U.S. Department of Energy. 2009. 2008 renewable energy data book. Retrieved from http://www1.eere.energy.gov/maps_data/pdfs/eere_databook.pdf. U.S. Department of Energy. 2009. Green power markets: net metering policies. Retrieved from http://apps3.eere.energy.gov/greenpower/markets/netmetering.shtml. U.S. Department of Energy. 2009. Solar America cities: about. Retrieved from www.solaramericacities.energy.gov/about/. U.S. Energy Information Administration. 2010. Glossary. Retrieved from www.eia.doe.gov/glossary/. U.S. Environmental Protection Agency. 2010. eGRID2007 version 1.1 year 2005 GHG annual output emission rates. Retrieved from http://cfpub.epa.gov/egridweb/ghg.cfm.
©Copyright 2010 American Planning Association All Rights Reserved
