17
This report is available at no cost from the
National Renewable Energy Laboratory (NREL)
at www.nrel.gov/publications
5 Conclusions
Table 8 and Table 9 summarize the U.S. utility-scale PV and CSP land-use requirements
evaluated in this report. Average total land-use requirements are 3.6 acres/GWh/yr for PV and
3.5 acres/GWh/yr for CSP. Average direct-area requirements are 3.1 acres/GWh/yr for PV and
2.7 acres/GWh/yr for CSP. On a capacity basis, the total-area capacity-weighted average for all
solar power plants is 8.9 acres/MWac, with 22% of plants within 8 and 10 acres/MWac. For
direct land-use requirements, the capacity-weighted average is 7.3 acre/MWac, with 40% of
power plants within 6 and 8 acres/MWac. Solar land-use estimates from the literature generally
fall within these ranges. Within the broad technology categories of PV and CSP, land-use metrics
are also impacted by specific technology choices, such as cell efficiency, tracking method, and
inclusion of thermal energy storage, and are a function of the solar resource available at
each site.
Although our results stem from an empirically based effort to estimate solar land use, several
caveats are warranted. Some solar-technology categories have relatively small samples sizes,
which must be considered when interpreting the robustness of reported results. Over 26 GWac of
PV and CSP are under development as of February 2013 (SEIA 2013), and the results reported in
this study must be understood in light of a rapidly growing installed base. Additionally, various
data sources were used when gathering information about solar projects. Although we tried to
obtain the highest-quality sources (project applications and regulatory documents, referred to as
“official documents” in this report), we collected official documents for only 20% of all projects
evaluated. Other data sources are expected to have higher levels of uncertainty (although how
much higher is unclear), which could contribute to the observed variability in results. With the
exception of a few CSP projects, we collected reported capacity of power plants but not annual
generation. The generation-based land-use results are expected to have higher levels of
uncertainty because annual generation is simulated. Although generation-based results provide a
more consistent approach when comparing land-use requirements across technologies, capacity-
based results are useful for estimating land area and costs for new projects because power plants
are often rated in terms of capacity. Finally, owing to the rapid evolution of solar technologies as
well as land-use practices and regulations, the results reported here reflect past performance and
not necessarily future trends.
We analyze elements that affect the area of solar impact, but we recognize that the duration of
use and impact on land quality are also important when considering land use impacts. Future
analyses could include evaluating the quality of land impacts, assessing both the initial state of
the land impacted and the final states across a variety of factors, including soil quality and
overall ecosystem quality. Finally, larger sample sizes and additional data elements would
improve the robustness of the conclusions and enable a more thorough investigation of the
impacts of additional factors, such as tilt angle, azimuth, PV module technology, CSP solar
multiple, and storage technologies.