A Spatial Analysis: Mapping Rooftop Solar Potential & Solar Equity in Los Angeles Madeline Wander UP 206: Introduction to Geographic Information Systems Final Project Write-Up Professor Leo Estrada Teaching Assistants: Nic Jay Aulston & Peter Capone-Newton March 18, 2011 Wander 2 Introduction With its abundance of sunshine and large amounts of rooftop space, Los Angeles is a prime location to cultivate solar energy. A rooftop solar feed-in tariff (FiT) program would enable Los Angeles residents and business owners to install rooftop solar panels and sell the power generated back to the electrical grid.1 This type of program would not only reduce tenant and owner utility costs, but it would help Los Angeles meet its ambitious renewable energy goals and create an estimated 11,000 new jobs.2 Since the costs of solar module production are increasingly inexpensive, the time is ripe for the City of Los Angeles to adopt such a program. The Program for Environmental & Regional Equity (PERE) at USC—which I work for—is collaborating with the UCLA Luskin Center for Innovation and the Los Angeles Business Council (LABC) to create a vision for a rooftop solar FiT program in Los Angeles. In addition to the clear environmental and economic benefits, we aim to better understand the equity considerations of such a program. Specifically, we will identify rooftop solar capacity of multi-family residential, industrial and commercial buildings in Los Angeles and demonstrate the benefits of such a program to low-income tenants, homeowners and workers. Through a spatial analysis using GIS, this study explores the following research questions: 1 Where does multi-family, industrial and commercial rooftop solar potential exist in Los Angeles, and where is solar potential clustered? What are the income levels and unemployment rates of the neighborhoods surrounding potential solar sites? Where are the photovoltaic (PV) solar installation training programs in relation to potential solar sites? Which potential sites would maximize benefits to low-income tenants, homeowners and workers trained to install rooftop solar? See the UCLA Luskin Center for Innovation’s website: http://luskin.ucla.edu/news/school-public-affairs/uclaresearch-solar-energy-prompts-coalition-campaign-rooftop-solar. 2 This number is based on the UCLA Luskin Center for Innovation’s proposed ten-year 600 Megawatt rooftop solar FiT program for the City of Los Angeles. See the UCLA Luskin Center’s 2010 report, “Bringing Solar Energy to Los Angeles: An Assessment of the Feasibility and Impacts of an In-basin Solar Feed-in Tariff Program,” at http://luskin.ucla.edu/content/bringing-solar-energy-los-angeles. Wander 3 Map Layouts This section describes each map layout, details methodologies used to create each layout, and explains the data analysis and sources behind each layout. Although we are conducting an analysis to understand the equity considerations of a rooftop solar FiT program for the City of Los Angeles, it is important to understand the benefits of such a program at the regional scale as well. Therefore, this report’s spatial analysis is at the county level. Layouts 1 – 5 each represent individual elements of indices (Layouts 6 – 10) identifying the optimal locations for rooftop solar installation that would maximize social equity benefits of this type of program. Layouts 1 & 2: Multi-Family and Industrial/Commercial Solar Opportunities in LA County Layout 1 (left) displays the spatial distribution of the number of multi-family buildings—or “opportunities”—with rooftop solar potential by Census block group. Most multi-family solar potential falls within the City of Los Angeles, with some pockets of solar potential in the south central part of LA County outside of the city boundary. Layout 1 Based on parameters outlined in the UCLA Luskin Center’s 2010 report “Bringing Solar Energy to Los Angeles,” this study only identifies multi-family buildings with the capacity, or “system size,” to hold 50 kilowatts (kW) or more of rooftop solar PV. Indeed, the UCLA Luskin Center claims that Los Angeles has the potential to generate 1,000 megawatts (MW) of solar energy from projects with system sizes of 5-10 kW, 1,500 MW from projects with system sizes of 10-50 kW, and 1,800 MW from projects with system sizes 50- Wander 4 500 kW. However, the 2010 report also states that the most viable program would install solar panels on large commercially owned rooftops that can take advantage of federal tax incentives. Thus, this study only examines multi-family buildings with system sizes of 50 kW or more. To create Layout 1 (previous page), I used data from the UCLA Luskin Center and ESRI. I conducted the following selection by attribute to obtain the multi-family parcels with system sizes of 50 kW or more: Residential Parcels: SolarParcels_20110113; "USE_TYPE" = 'Residential' Multi-Family Parcels: SolarParcels_20110113_Residential; "USE_DESCRI" = 'Five or more apartments‘; "USE_DESCRI" = 'Four Units (Any Combination)‘; "USE_DESCRI" = 'Rooming Houses‘; "USE_DESCRI" = 'Three Units (Any Combination)‘; "USE_DESCRI" = 'Two Units‘ Multi-Family Parcels > 50 KW: SolarParcels_20110113_Residential_MF; "SYS_SIZE" > 49.9 I then spatially joined the multi-family parcels with system sizes of over 50 kW to Census block groups, summing on number of parcels per block group. Last, I classified the data using the Natural Breaks (Jenks) method and divided the data into five classes. The number of parcels per block group ranges from zero to 66. Layout 2 (right) displays the spatial distribution of the number of industrial and commercial buildings with rooftop solar potential by Census block group. Most industrial/commercial solar potential falls outside the City of Los Angeles. Industrial/commercial solar potential appears clustered along corridors in the southeastern part of the LA County. For reasons outlined above, this study only identifies industrial and commercial buildings with system sizes of 50 kW or more. To create Layout 2, I used that same data sources and methodology as in Layout 1, and conducted the following Layout 2 Wander 5 selection by attribute: Commercial and Industrial Parcels – Layer: SolarParcels_20110113; "USE_TYPE" = 'Commercial'; "USE_TYPE" = 'Industrial' Commercial and Industrial Parcels > 50 KW – Layer: SolarParcels_20110113_Ind_Comm; "SYS_SIZE" > 49.9 Again, using the same process and classification method as in Layout 1, I summed the number of industrial/commercial parcels per block group and divided the data into five classes. The number of parcels per block group ranges from zero to 486. Layout 3: Median Household Income by Census Block Group in LA County To identify which areas contain low-income homeowners and tenants who would most benefit from energy savings generated by a rooftop solar FiT program, Layout 3 (left) displays the spatial distribution of median household income by Census block group in LA County. As expected, the lowest income block groups mostly concentrate in the central part of the City of Los Angeles, with lowincome pockets reaching across the southeastern part of LA County. Layout 3 To create Layout 3, I used data from the U.S. Census Bureau American Community Survey (ACS) 2005-2009 and ESRI. I classified the data the using the Quantile method—so each class has the same number of features—and divided the data into four classes. The data ranges from far below poverty level ($21,954 for a family of four in LA County) to above $200,000. Wander 6 Layout 4: Unemployment Rate by Census Block Group in LA County To identify which areas contain high unemployment that would most benefit from jobs created by a rooftop solar FiT program, Layout 4 (right) displays the spatial distribution of unemployment rates by Census block group in LA County. Similar to Layout 3 (previous page), the greatest levels of unemployment fall within the southeastern part of LA County. However, it appears that similar amounts of high unemployment fall inside and outside the City of Los Angeles. Additionally, unemployment rates are slightly more dispersed than the lowest median household incomes, which appear more concentrated. To create Layout 4, I used data from the U.S. Census Bureau ACS 2009 and ESRI. Similar to Layout 3, I classified the data the using the Layout 4 Quantile method and divided the data into four classes. The data ranges from 0% to well above California’s current rate of 12.4 percent. Layout 5: Distance of Each Block Group to Closest PV Solar Installation Training Program To identify areas in which workers have the most access to PV solar installation training programs, Layout 5 (next page) displays the distance of each block group to the closest training program in LA County. Nearly all the training programs are located in the southern part of LA County. Most programs are located outside of the City of Los Angeles, and spread fairly evenly across the southern part of the county. To create Layout 5, I used data from the U.S. Census Bureau ACS 2005-2009 and ESRI. I also gathered locations of the 21 training centers in the LA County Community College system, as well as one additional program (East Los Angeles Skills Center). I then geocoded all the training centers and created an original data layer (PV Solar Installation Training Programs in Los Angeles County). Wander 7 Next, I converted each block group to its centroid point, and used the “Closest Facility” function in Network Analyst to find the distance between each block group and the closest training program. I then assigned each block group polygon its distance to the closest training program, classified the data using the Natural Breaks (Jenks) method, and divided the data into four classes. The data ranges from 0 to 36.1 miles. (Network Analyst was unable to find the distance between the closest training program and 10 [of over 6000] block groups.) Layout 5 Layouts 6 – 7: Solar Equity Indices for Potential Multi-Family and Industrial/Commercial Solar Sites To identify the optimal locations where a rooftop solar FiT program would maximize benefits to low-income households and workers, I used data from Layouts 1 – 5 to create a Solar Equity Index for both multi-family and industrial/commercial opportunities. As Table 1 (next page) shows, I divided the data for each element of the index—number of multi-family or industrial/commercial opportunities, median household income, unemployment rate, and distance to closest training facility—into four categories, and assigned each category a score between zero and three. For number of opportunities per block group, I divided the data into four categories with an even number of block groups within each category. I assigned higher weights to block groups with more opportunities, and lower weights to block groups with fewer opportunities. Wander 8 For median household income, I assigned higher weights to block groups with lower incomes, and vice versa. I divided the data into categories based on the following information: $21,954 = poverty level for a family of four in LA County $24,850 = extremely low income for a family of four in LA County $41,400 = very low income for a family of four in LA County $66,250 = low income for a family of four in LA County For unemployment rate, I assigned higher weights to block groups with higher unemployment rates, and lower weights to block groups with lower unemployment rates. I divided the data so that block groups with unemployment rates closer to the U.S. rate (8.9%) received a lower score than block groups with rates closer to California’s rate (12.4%). Lastly, for distance to closest training facility, I assigned higher weights to block groups with shorter distances to the closest training center, and vice versa. I assigned the highest weight to those block groups within a walkable distance (< 0.05 mile) of a training center, and the next highest weight to those block groups within a bikeable distance (< 3 miles) of a training center. Table 1 below details the weights and categories of each element in the Solar Equity Index. Table 1 – Elements of Social Equity Index Layout 6 (next page) displays the spatial distribution of optimal locations containing multifamily solar sites where a rooftop solar FiT program would maximize benefits to low-income households and workers. Layout 7 (next page) shows a similar spatial distribution of optimal locations containing industrial/commercial solar sites. Locations that maximize benefits have the highest “solar equity potential.” Layouts 6 and 7 (next page) show similar patterns. Most of the block groups with maximum solar equity potential are within the City of Los Angeles, specifically in the central part of the city. Interestingly, for both multi-family and industrial/commercial sites, areas at the southern tip of the city also maximize solar equity potential. Layout 7 displays a few pockets of solar equity potential scattered throughout the southeastern part of LA County. Wander 9 Layout 6 Layout 7 To create Layouts 8 and 9, I used data from Layouts 1 – 5. I classified the data using the Natural Breaks (Jenks) method, divided the data into five classes, and assigned labels ranging from lowest to highest solar equity. Layouts 8 – 9: Hotspot Analysis of Potential Multi-Family and Industrial/Commercial Solar Sites To test a different method of identifying the optimal locations where a rooftop solar FiT program would maximize benefits to low-income households and workers, I conducted a Hotspot Analysis. Using the same index elements as above—number of multi-family or industrial/commercial opportunities, median household income, unemployment rate, and distance to closest training facility—Layouts 8 and 9 display the spatial distribution of locations that would maximize solar equity potential. Layouts 8 and 9 show similar clustering patterns similar to Layouts 6 and 7 above. Not surprisingly, the highest solar equity potential falls within the City of Los Angeles. However, Layouts 8 and 9 show more clustering in the northern part of the city than the Solar Equity Wander 10 Index in Layouts 6 and 7. Overall, Layouts 8 and 9 show a higher concentration of solar equity potential in the southern part of LA County than in the northern part. Layout 8 Layout 9 To create Layouts 8 and 9, I used data from Layouts 1 – 5 and created the following models: Wander 11 I then used the Raster Calculator to add the index element raster layers to create Hotspot Analyses of both multi-family and industrial/commercial solar sites: Lastly, I classified the data using the Equal Interval method, divided the data into four classes, and assigned categorical labels ranging from lowest to highest solar equity. Layouts 10 – 11: Multi-Family and Industrial/Commercial Solar Hotspots – Political Analyses In order to pass a rooftop solar feed-in tariff program for the City of Los Angeles, the policy needs political support. Layouts 10 and 11 display the Los Angeles City Council Districts overlapping the spatial distribution of solar equity (the hotspot analyses from Layouts 8 and 9). Layout 10 Layout 11 Wander 12 By overlaying solar equity potential and PV solar installation training center locations with boundaries of political jurisdictions, we can identify which council districts would benefit the most from a rooftop solar FiT program. Layouts 10 and 11 also contain tables identifying how many multi-family and industrial/commercial opportunities are within each council district. Specifically, Layout 10 shows a concentration of multi-family solar hotspots in Council District (CD) 4 (718 opportunities), CD 8 (687 opportunities), and CD 2 (666 opportunities). Likewise, Layout 11 shows a concentration of industrial/commercial hotspots in CD 5 (207 opportunities), CD 2 and 11 (both with 172 opportunities), and CD 4 (144 opportunities). Since CD 4 (Councilmember Tom LaBonge) and CD 2 (Councilmember Paul Krekorian) contain the largest amounts of multi-family and industrial/commercial opportunities combined, its residents and workers stand to gain much of the benefits from a rooftop solar FiT program in Los Angeles. (Since PERE is doing a case study of Bonnie Brae Village in Westlake—an affordable housing development for formerly homeless seniors with 100% solar capacity already installed— Layouts 10 and 11 display an outline of the Westlake Community Plan Area as well.) To create Layouts 10 and 11, I used data from Layouts 1 – 5, as well as some additional shapefiles from ESRI. I then spatially joined City Council Districts with both multi-family and industrial/commercial parcels with system sizes of over 50 kW, and summed the number of opportunities within each council district. I then overlaid other cities within LA County to identify surrounding areas containing training centers. The classification method and number of categories are the same as in Layouts 8 and 9. Wander 13 Conclusion By identifying clusters of multi-family and industrial/commercial opportunities for rooftop solar installation, median household income and unemployment rate by block group, and distances between block groups and closest PV solar installation training programs in LA County, this study identifies the geographic areas in which a rooftop solar feed-in tariff program would maximize benefits to low-income households and workers. Within LA County, the City of Los Angeles contains the highest concentrations of solar equity hotspots. Specifically, Council Districts 2 and 4 stand to gain the most from a rooftop solar FiT program for Los Angeles. Since this analysis measures the spatial relationships between demographic and metric characteristics of block groups, GIS is an essential piece of this study. Without GIS, I would not have been able to identify the number of rooftop solar opportunities within geographic areas, find the distances between block groups and the closest PV solar training programs, or create the solar equity indices and hotspots. Clearly, this study relies heavily on spatial analysis. Ultimately, due to the clustering of solar equity hotspots in the City of Los Angeles, many residents and workers are poised to gain economic and environmental benefits from the adoption of a rooftop solar feed-in tariff program. This study clearly supports the adoption of such a policy. Wander 14 References J.R. DeShazo and Ryan Matulka. (2010). Bring solar energy to Los Angeles: An assessment of the feasibility and impacts of an in-basin solar feed-in tariff program. UCLA Luskin Center for Innovation, School of Public Affairs.