Title:Renewable Electricity Generation
Hardcover Dimensions:6" x 9"
- 80 Hardcover pages
- Buy the Book
Download PDF Despite widespread political support and large direct and indirect subsidies from both the federal and state governments, renewable electricity—wind and solar power, in particular—produces only 3.6 percent of US power generation. This small market share suggests inherent limitations that can be overcome only at very high cost. This first in a three-part Outlook series discusses these limitations. [Read part 2] [Read part 3]
Key points in this Outlook:
- Political and policy support for renewable electricity—wind and solar power, in particular—is long-standing, substantial, and bipartisan.
- Despite this support, wind and solar power have exhibited poor outcomes in terms of market competitiveness with power generation using such conventional fuels as coal and natural gas.
- This weak performance results from three inherent characteristics of wind and solar power: their unconcentrated energy content, siting limitations and high transmission costs, and low availability and intermittency.
- These alternative energy sources will not go mainstream so long as barriers such as high associated costs for land, capital, transmission, and backup capacity persist.
Renewable electricity—wind and solar power, in particular—receives very large direct and indirect subsidies from the federal and state governments. This policy support is far larger than that enjoyed by such conventional electric generation technologies as coal, natural gas, nuclear fuels, or hydroelectric facilities. Moreover, a majority of states have mandated some form of guaranteed market shares for renewable electricity. This political support for renewable power is substantial, broad-based, bipartisan, and long-standing.
Nonetheless, renewable electricity—particularly, wind and solar power—has very high costs, and this is likely to remain true for the foreseeable future. As a result, these sources have achieved only small market shares. Renewable electricity generation from all nonhydroelectric sources was only 3.6 percent of total US generation in 2010. The Energy Information Administration estimated in 2007 that the proportion in 2030 would be that very same 3.6 percent but this year has increased that projection to 11 percent.
It is not clear what changes in important parameters have yielded that increase in the projected market share over the course of only a few years. As I will discuss in detail in this series of three Outlook essays, no sound rationale, whether economic or technological, can explain this change in the official wisdom. To the contrary, both economic and technological factors strongly suggest that wind and solar power will remain uncompetitive; heavily dependent on subsidies, both direct and indirect; and small relative to the electricity market as a whole.
Energy policies in the United States for decades have pursued energy sources defined in various ways as alternative, unconventional, independent, renewable, and clean in an effort to replace such conventional fuels as oil, coal, and natural gas. These long-standing efforts have, without exception, yielded poor outcomes, in a nutshell because they must swim against the tide of market forces. That is why the only reliable outcome has been one disappointment after another, and there are powerful reasons to predict that the same will prove true with respect to the current enthusiasm for renewable electricity. In this Outlook, I will look at the inherent limitations of wind and solar power generation—that is, the reasons they are likely to remain uncompetitive without very large subsidies and other policy support, or even with such support.
Policy preferences for renewable electricity at both the federal and state levels are substantial, as direct and indirect financial subsidies and other forms of support.1 The relative magnitudes of the federal subsidies for various forms of electricity, as estimated by the Energy Information Administration, are instructive.2 As previously mentioned, nonhydroelectric renewable power generation for 2010 was 3.6 percent of all generation, but it received 53.5 percent of all federal financial support for the electric power sector. Wind power, providing 2.3 percent of generation, received 42 percent of such support. This combination of substantial policy support and meager market competitiveness suggests the presence of important impediments to the growth of renewable power.
The technical literature reveals several such problems that have not received widespread attention in the popular discussion.
Unconcentrated Energy Content
The energy content of wind flows and sunlight, which varies depending upon air speed and sunlight intensity, is far less concentrated than that of the energy contained in fossil or nuclear fuels.3 To compensate for this unconcentrated nature of renewable energy sources, the facility operator or power utility must invest large amounts in land and materials to make renewable generation technically practical for generating nontrivial amounts of electricity. A wind farm would require 500 windmills of 2 megawatts (MW) each to provide a theoretical generation capacity of 1,000 MW. Because the wind turbines must be spaced apart to avoid wake effects (wind interference among the turbines), a 1,000 MW wind farm would require on the order of 48,000–64,000 acres (or 75–100 square miles) of land. With an assumed capacity factor for a typical wind farm of 35 percent, reliable wind capacity of 1,000 MW would require an amount of land (perhaps at different locations) on the order of three times that rough estimate.4 In contrast, a 1,000 MW gas-fired plant requires about ten to fifteen acres; conventional coal, natural gas, and nuclear plants have capacity factors of 85–90 percent.
"The combination of substantial policy support and meager market competitiveness suggests the presence of important impediments to the growth of renewable power." The same general problem afflicts solar power. The energy content of sunlight, crudely, is about 150–400 watts per square meter depending on location, of which about 20–30 percent is convertible to electricity depending on the particular technology. Accordingly, even in theory, a square meter of solar energy–receiving capacity is enough to power roughly one 100-watt light bulb, putting aside such issues as sunlight intensity. This problem of land requirements for solar thermal facilities is of sufficient importance that most analyses assume a maximum generation capacity of 50–100 MW, which, conservatively, would require approximately 1,250 acres, or two square miles.
In short, transformation of the unconcentrated energy content of wind and sunlight into a form usable for modern applications requires massive capital investment in land and wind turbines and solar-receiving equipment. This means that energy from renewable sources, relative to that from conventional forms, by its very nature is limited and expensive.
Siting Limitations and Transmission Costs
Conventional power-generation plants can be sited, in principle, almost anywhere, and such fuels as coal and natural gas can be transported to the generation facilities. This means that investment-planning decisions can optimize transmission investment costs along with the other numerous factors that constrain and shape generation investment choices: land costs, environmental factors, reliability issues, transmission line losses, and the like. Wind and solar sites, on the other hand, must be placed where the wind blows and the sun shines with sufficient intensity and duration. (Photovoltaic installations, suitable for small applications, face the transmission problem either not at all or to a far smaller degree than solar thermal plants but still are constrained by the intensity of sunlight.) Because appropriate sites are limited, with the most useful (lowest cost) ones exploited first, the cost of exploiting such sites must rise as more sites are used. As a result, even if wind and solar technologies exhibit important scale economies in terms of capacity and/or generation costs, scale economies may not characterize a broader cost calculation including the cost of finding and using particular sites.
In other words, scale economies may not be present at the industry level even if they are present at the project (or even at the turbine or parabolic dish) level. This reality is consistent with data on capacity factors for 1998–2009 published recently by the Energy Information Administration.5 The capacity factors for nonhydroelectric renewables declined almost monotonically from 57 percent to 33.8 percent, suggesting that as renewables capacity has expanded, developers have been forced onto increasingly unfavorable sites.
Because conventional power-generation investments can optimize transmission costs and other reliability factors more easily than is the case for wind and solar capacity, it would be surprising if such costs were not higher for the latter. This is exacerbated by the physical realities that wind conditions are strongest in open plains regions, while solar generation generally requires regions with strong sunlight and, for thermal solar plants, sizable open areas. For the United States, the best wind capacity sites are in a region stretching from the northern plains down through Texas, and the best thermal solar sites are in the Southwest. The United States simply lacks significant east-west high-voltage interconnection transmission capacity to transport wind and solar power to the coasts. One national study of this problem notes that “wind development will require substantial additions to the nation’s transmission infrastructure . . . due to the locational dependence of wind resources [and] the relatively low capacity factor of wind plants.”6
Some analyses of these transmission costs are available. One survey of forty transmission studies for wind projects conducted during 2001 through 2008 finds a median transmission cost of $15 per megawatt hour (mWh).7 The survey was limited to studies of transmission requirements for multiple new wind plants with a combined capacity greater than 300 MW. A California Public Utilities Commission analysis concludes that implementation of a 20 percent renewable electricity standard (or requirement) for the state by 2020 would require four new major transmission lines at a cost of about $4 billion, while a 33 percent standard would require seven new lines at a cost of $12 billion. For that 33 percent requirement, the assumptions in the study suggest transmission costs of about $6.39 per mWh, a figure that is implausibly low.
"Energy from renewable sources, relative to that from conventional forms, by its very nature is limited and expensive." A study done for the National Renewable Energy Laboratory examined the transmission requirements and attendant costs for four alternative wind capacity scenarios for the Eastern Interconnection (which spans the continental United States east of the Rocky Mountains, minus Texas, plus parts of southeastern Canada). This study reports a cost of wind “integration” of about $5 per mWh, but other data in the study suggest transmission costs of about $17 per mWh, a figure roughly comparable to the $15 median reported in the survey noted above.8
The Energy Information Administration has published a comprehensive comparison of various cost categories across generation types. The data show that conventional generation—coal and natural gas combined cycle—has transmission costs of about $3.60 per mWh, less than half those of wind generation ($8.40) and about a third those of thermal solar generation ($10.40).9 These projections for transmission costs are consistent with the hypothesis that wind and solar power are highly constrained in terms of capacity factors and sites, and so impose higher transmission costs than is the case for conventional generation.
Low Availability and Intermittency
Electric energy cannot be stored in batteries in large amounts at low cost because of technological limitations; indirect storage in the form of water in dams is the only economic option. This means that the production and consumption of electricity in a given power network must be balanced constantly to prevent blackouts and, more generally, to preserve system reliability. Because unexpected surges in demand and outages of generating equipment can occur, backup generation capacity must be maintained. Such backup capacity is termed the “operating reserve” for the given network.
This operating reserve has two components. The first is the “spinning reserve.” These generators are already connected to the network, and their output can be increased by raising the torque applied to the turbines. The typical system requirement is that spinning reserves be 50 percent or more of total operating reserves. The second component of operating reserves is the supplemental reserve, which comprises generation capacity that can be brought online within five to ten minutes and electric power that can be obtained quickly by borrowing from other networks or by withholding power being distributed to other networks. Additional reserve capacity often is provided by generators that require up to an hour to come online; this backup capacity is not included in measures of the operating reserve for a system because of the length of time required for availability.
Electric supply systems respond to growing demands (“load”) over the course of a day or year by increasing output from the lowest-cost generating units first and then calling upon successively more expensive units as electric loads grow toward the daily (or seasonal) peak. Because of the uncertainties caused by the unreliability of wind and sunlight, most electric generation capacity fueled by renewable energy sources cannot be assumed to be available on demand when their use is expected to be most economic. Accordingly, this capacity cannot be scheduled (or “dispatched”). Instead, it requires backup generation capacity to preserve system reliability.
Therefore, the cost of that needed backup capacity becomes a crucial parameter usually not mentioned in public discussions of wind and solar power. One study, using figures from the California Independent System Operator, projects that California’s renewable generation capacity will increase between 2009 and 2020 to about 17.7 gigawatts (GW) for a 20 percent renewable requirement and about 22.4 GW for a 33 percent requirement.10 The projected needs for backup capacity (of varying types) are, respectively, 0.8 GW (or 4.5 percent) and 4.8 GW (or 21 percent).
What would that cost? US wind and solar generation capacity in 2009 was about 34,000 MW. If we assume, conservatively, that this renewable capacity has required investment in backup capacity of about 3 percent (rather than 4.5 percent), that requirement would be about 1,000 MW. Cost estimates published by the EIA suggest that this backup capacity has imposed fixed capital and operations and maintenance costs of about $1.7 billion, variable operating costs of approximately $2.00–$4.50 per mWh, and total costs per mWh of about $368.11
"The cost of needed backup capacity is a crucial parameter usually not mentioned in public discussions about wind and solar power." That rough estimate may be biased downward. Because state renewables requirements require system operators to use renewable power when it is available, conventional generation must be cycled—that is, turned on and off—in coordination with renewable power availability. This cycling means less efficient operation for both coal-fired and gas combined-cycle backup generation, particularly for coal. A recent study of the attendant emissions effects for Colorado and Texas found that requirements for the use of wind power impose significant operating and capital costs because of these backup generation cycling needs and actually exacerbate air pollution problems.12
The EIA estimates wind (onshore) and solar costs in 2016 at about $149 and $257–396 per mWh, respectively; if we add the rough estimate for backup costs, the total is about $517 for wind and $625–764 for solar generation.13 The EIA estimates for gas- or coal-fired generation are about $80–110 per mWh. Accordingly, the projected cost of renewable power in 2016, including the cost of backup capacity, is at least five times higher than that for conventional electricity. At the same time, outages of wind capacity because of weak wind conditions are much more likely to be correlated geographically than outages of conventional plants, and the same is true for solar electric generation because of the geographic concentrations of thermal solar sites and photovoltaic systems.
The higher cost of electricity generated with renewable energy sources is only one side of the competitiveness question; the other is the value of that generation, as not all electricity is created equal. In particular, power produced at periods of peak demand is more valuable than off-peak generation. In this context, wind generation, in particular, is problematic because, in general, winds tend to blow at night and in the winter, which corresponds inversely to peak energy demand during daylight hours and in the summer.
The most striking economic characteristic of wind and solar power is a small market share stubbornly resistant to strong political support and very substantial direct and indirect subsidies. This general problem of poor competitiveness is the result of factors intrinsic to renewable electricity: unconcentrated energy content, siting limitations and additional resulting transmission costs, and poor reliability and the attendant costs of backup capacity. These problems are inherent in the technologies and can be overcome only at considerable expense.
Despite these excess costs, political support for wind and solar power remains strong. No state has formally abandoned or weakened its renewable electricity requirements, and federal policies to promote renewable technology in electricity production remain in place. Can renewable electricity prove itself worthy of this support? The five central rationales commonly offered in support of these state and federal policies will be the focus of my next Outlook.
Benjamin Zycher ([email protected]) is a visiting scholar at AEI. This first in a series of three Outlooks is based on his new book, Renewable Electricity Generation: Economic Analysis and Outlook (AEI Press, 2011). The author thanks Kenneth Green for his suggestions on an earlier draft.
1. These will be summarized in the third part of this Outlook series. For a detailed list of such policies, see the Database of State Incentives for Renewables and Efficiency, www.dsireusa.org (accessed November 8, 2011).
2. See tables ES4 and ES5 in US Energy Information Administration, Direct Federal Financial Interventions and Subsidies in Energy in Fiscal Year 2010 (Washington, DC: US Department of Energy, July 2011), www.eia.gov/analysis/requests/subsidy/pdf/subsidy.pdf (accessed November 8, 2011).
3. The energy content of different fuels varies greatly. Per unit of fuel—tons of coal, millions of cubic feet of natural gas, miles per hour of wind speeds, or hours of sunlight—this variation can be thought of usefully as the degree of concentration of the energy content of a particular energy source.
4. The capacity factor for a generation facility (or technology) is its actual production over a given time period divided by its theoretical maximum production over that time period. For standard assumptions on capacity factors for the various generation technologies, see Energy Information Administration, “2016 Levelized Cost of New Generation Resources from the Annual Energy Outlook 2010,” n.d., www.eia.doe.gov/oiaf/aeo/pdf/2016levelized_costs_aeo2010.pdf (accessed November 8, 2011). The assumed capacity factor for onshore wind generation in that analysis is 34.4 percent.
5. See table 5.2 in Energy Information Administration, Electric Power Annual 2009, revised January 4, 2011, www.eia.doe.gov/cneaf/electricity/epa/epat5p2.html (accessed November 23, 2011).
6. Andrew Mills, Ryan Wiser, and Kevin Porter, The Cost of Transmission for Wind Energy: A Review of Transmission Planning Studies (Berkeley, CA: Lawrence Berkeley National Laboratory, February 2009, p. vii), http://eetd.lbl.gov/EA/EMP/reports/lbnl-1471e.pdf (accessed November 8, 2011).
7. Ibid, 6–8.
8. See tables 1–4 and figure 3, EnerNex Corporation, Eastern Wind Integration and Transmission Study, NREL/SR-550-47086 (Golden, CO: National Renewable Energy Laboratory, January 2010), www.nrel.gov/wind/systemsintegration/pdfs/2010/ewits_executive_summary.pdf (accessed November 29, 2011).
9. Energy Information Administration, “2016 Levelized Cost of New Generation Resources from the Annual Energy Outlook 2010,” n.d., www.eia.doe.gov/oiaf/aeo/pdf/2016levelized_costs_aeo2010.pdf (accessed November 8, 2011).
10. KEMA Inc., Research Evaluation of Wind Generation, Solar Generation, and Storage Impact on the California Grid, prepared for the California Energy Commission, CEC-500-2010-010, June 2010, 1, www.ovcr.ucla.edu/uploads/file/CA%20Energy%20Commission_PIER%20Final%20Project%20Report_June%202010.pdf (accessed November 29, 2011).
11. See table 8.2 in EIA, “Electricity Market Module,” April 2010, www.eia.gov/oiaf/aeo/assumption/pdf/electricity.pdf#page=3 (accessed November 8, 2011).
12. Bentek Energy, How Less became More: Wind, Power and Unintended Consequences in the Colorado Energy Market, prepared for the Independent Petroleum Association of Mountain States, April 16, 2010, 25–33, www.wind-watch.org/documents/wp-content/uploads/BENTEK-How-Less-Became-More.pdf (accessed November 29, 2011).
13. Energy Information Administration, “2016 Levelized Cost of New Generation Resources from the Annual Energy Outlook 2010,” n.d., www.eia.doe.gov/oiaf/aeo/pdf/2016levelized_costs_aeo2010.pdf (accessed November 8, 2011).