TODAY’S STUDY: The Enormous Potential Of Rooftop Solar
Rooftop Solar Photovoltaic Technical Potential in the United States: A Detailed Assessment
Pieter Gagnon, Robert Margolis, Jennifer Melius, Caleb Phillips, and Ryan Elmore, March 2016 (National Renewable Energy Laboratory)
This report quantifies the technical potential of photovoltaic (PV) systems deployed on rooftops in the continental United States, estimating how much energy could be generated by installing PV on all suitable roof area. The results do not exclude systems based on their economic performance, and thus they provide an upper bound on potential deployment rather than a prediction of actual deployment.
Although methods have been developed to estimate rooftop PV technical potential at the individual building level, previous estimates at the regional and national levels have lacked a rigorous foundation in geospatial data and statistical analysis. This report helps fill this gap by providing a detailed data-driven analysis of U.S. (national, state, and ZIP-code level) rooftop PV availability and technical electricity-generation potential. First, we use light detection and ranging (lidar) data, geographic information system (GIS) methods, and PV-generation modeling to calculate the suitability of rooftops for hosting PV in 128 cities nationwide—representing approximately 23% of U.S. buildings—and we provide PV-generation results for a subset of these cities. Second, we extend the insights from this analysis of areas covered by lidar data to the entire continental United States. We develop two statistical models—one for small buildings and one for medium and large buildings—that estimate the total amount of roof area suitable for hosting PV systems, and we simulate the productivity of PV modules on the roof area to arrive at the nationwide technical potential for PV.
Our analysis of the trends in the suitability of rooftops for hosting PV systems reveals important variations in this key driver of rooftop PV technical potential that have not been captured by previous approaches. Figure ES-1 shows the results—from our statistical modeling grounded in lidar data—for the percentage of small buildings that are suitable for PV in each ZIP code in the continental United States. In the figure we can identify regional trends in the suitability of small building rooftops, with high densities of suitable small buildings in California, Florida, and the West South Central census division. Such trends are also critical to estimating PV technical potential at finer resolution, and our report illustrates this with a high-resolution analysis of 11 representative cities.
Figure ES-2 shows the annual energy generation potential from rooftop PV as a percentage of each state’s electricity sales in 2013. The estimates of energy generation are based on the rooftop suitability of small, medium, and large buildings as well as specific roof orientations, local solar resources, PV system performance assumptions, and building footprints.
Figure ES-2 shows that California has the greatest potential to offset electricity use—its rooftop PV could generate 74% of the electricity sold by its utilities in 2013. A cluster of New England states could generate more than 45% because these states’ low per-capita electricity consumption offsets their below-average solar resource. Washington, with the lowest population-weighted solar resource in the continental United States, could still generate 27%. Some states with below-average solar resource (such as Minnesota, Maine, New York, and South Dakota) have similar or even greater potential to offset total sales compared to states with higher-quality resource (such as Arizona and Texas).
The difference between Florida and other South Atlantic states illustrates the interplay between variables that affect technical potential. Florida can offset 47% of its total consumption despite having an average household consumption of 130% of the national average. This is largely explained by significantly below-average electricity consumption outside of the residential sector, which makes the state’s total per-capita electricity sales slightly lower than the national average. In contrast, the other South Atlantic states range from a potential 23% to 35% of electricity offset owing to lower average rooftop suitability (see Figure ES-12 ), slightly lower quality solar resource, and higher per-capita total electricity sales.
Table ES-1 shows our aggregate results.3 The total national technical potential of rooftop PV is 1,118 gigawatts (GW) of installed capacity and 1,432 terawatt-hours (TWh) of annual energy generation. This equates to 39% of total national electric-sector sales, and it is significantly greater than a previous National Renewable Energy Laboratory estimate of 664 GW of installed capacity and 800 TWh of annual energy generation (Denholm and Margolis 2008). The difference can be attributed to increases in module power density, improved estimation of building suitability, higher estimates of the total number of buildings, and improvements in PV performance simulation tools that previously tended to underestimated production.
Although only 26% of the total rooftop area on small buildings (those with a footprint smaller than 5,000 ft2 ) is suitable for PV deployment, the sheer number of buildings in this class gives small buildings the greatest technical potential. Small building rooftops could accommodate 731 GW of PV capacity and generate 926 TWh/year of PV energy, which represents approximately 65% of rooftop PV’s total technical potential. Medium and large buildings have a total installed capacity potential of 386 GW and energy generation potential of 506 TWh/year, which represents approximately 35% of the total technical potential of rooftop PV.
These results are sensitive to assumptions about module performance, which is expected to continue improving over time. For example, this analysis assumed a module efficiency of 16% to represent a mixture of various technology types. If a module efficiency of 20% were assumed instead, which corresponds to current premium systems, each of the technical potential estimates would increase by about 25% above the values stated in this report. Furthermore, our results are only estimates of the potential from existing suitable roof planes, and they do not consider the immense potential of ground-mounted PV. Actual generation from PV in urban areas could exceed these estimates by installing systems on less suitable roof area, mounting PV on canopies over open spaces such as parking lots, or integrating PV into building facades.
Because our results are estimates of technical potential, they do not consider what would be required to use all the potential PV-generated energy. In practice, the integration of a significant quantity of rooftop PV into the national portfolio of generation capacity would require a flexible grid, supporting infrastructure, and a suite of enabling technologies.
Many opportunities exist for expanding on the data, methods, and results from this report. Our base data set can be combined with numerous other data sets—such as insolation data, market data, and end-use electricity consumption data—to provide more precise and actionable information than has been available previously. To that end and to encourage municipalities, utility providers, solar energy researchers, and other PV stakeholders to use the data for their own purposes, we have posted supporting data and documentation of the methods we used to perform our analysis on the NSRDB Data Viewer website. 4 The models developed here also can be applied generally in areas where lidar data may not be available, both for making estimates of rooftop areas for arbitrary geographies and in stochastic simulations where statistically representative buildings are needed.