Panel on Public Affairs Workshop on Electricity from Renewable Sources
David Bodansky
Problems of fossil fuel resource limitations and pollutants, including CO2, continue to motivate the exploration of alternatives, in particular nuclear and renewable energy sources. A major use of these would be for electricity generation.
Electricity production and use
There has been rapid growth in the use of electricity in the US throughout the past century. During the second part, from 1950 through 1992, electricity sales increased almost tenfold, from 33 to 315 GWyr, and the fraction of primary commercial energy consumption devoted to electricity generation rose from 14% to 36% (1). (For units and abbreviations used here, see (2).) In more recent years, since the beginning of the first perceived energy crisis in 1973, electricity growth has slowed, but it has nonetheless outstripped growth in total energy consumption and in population (Table 1)
Table 1. Comparison of increases in US electricity use, 1973-1992, with changes in related parameters.
|
Increase (%) |
The energy input for electricity generation (quad) |
49% |
Electricity Sales |
61% |
Population |
20% |
Gross Domestic Product (Constant $) |
51% |
Total primary energy consumption (quad) |
11% |
Fossil fuels and nuclear power account for most electricity generation. In 1990, the last year for which comprehensive data are readily available, renewable sources collectively provided only 12% of generation (Table 2). Most of this was from hydroelectric, with smaller amounts from biomass and geothermal, and near negligible amounts from wind, solar thermal, and photovoltaics. Except for hydroelectric, most renewable generation is not by utilities but by so-called non-utility generators (NUG). There was little change in total generation between 1990 and 1992, with nuclear generation increasing 7%, hydroelectric dropping 14%, and fossil fuels (collectively) almost unchanged.
Table 2. US electricity generation, by source, for 1990: net generation (gigawatt), fraction from non-utility generators (NUG), and share of total generation
Source |
Generation (net GWyr) |
Percent from NUG |
Percent of Total Gen |
Coal |
181 |
2 |
53 |
Natural Gas |
41 |
26 |
12 |
Petroleum |
14 |
4 |
4 |
Nuclear |
66 |
0 |
19 |
Hydroelectric |
33 |
2 |
9.5 |
Biomass |
4.5 |
95 |
1.3 |
Geothermal |
1.7 |
43 |
0.5 |
Wind |
0.24 |
100 |
0.07 |
Solar Thermal |
0.07 |
100 |
0.02 |
Other |
2.6 |
100 |
0.75 |
Total |
344 |
7 |
100 |
In recent years, there have been projections of a much larger future renewable contribution, including greatly increased electricity generation of photovoltaic, wind, and thermal solar power. For example, a 1990 study spearheaded by the Solar Energy Research Institute (since redesignated as the National Renewable Energy Laboratory) had scenarios that projected future electricity generation from renewable sources (excluding hydroelectric power) in 2030 extending up to a primary energy input of 33 quads, compared to under 1 quad today (3). Other projections vary widely, some considerably less optimistic (4). A great deal depends upon future costs. A target for competitiveness with fossil fuel or nuclear generation is often taken to be about 4 or 5 cents/kWh, although the actual level will depend in part upon future fossil fuel prices and on taxes or emission standards established for fossil fuels.
Motivation for Workshop
In view of the importance of the issues, it is desirable to subject such projections to careful analysis. The projections for nuclear power, made when enthusiasm was high and skepticism relatively muted, can provide a cautionary note as an example of the poor track record of past projections. Thus, a 1972 Atomic Energy Commission report foresaw nuclear growth from about 0.3 quad (of primary energy) in 1970 to over 21 quads in 1990 and to 48 quads in 2000, while actual US nuclear energy consumption in 1990 was only 6 quads and is very unlikely to be significantly higher in 2000.In view of the importance of the issues, it is desirable to subject such projections to careful analysis. The projections for nuclear power, made when enthusiasm was high and skepticism relatively muted, can provide a cautionary note as an example of the poor track record of past projections. Thus, a 1972 Atomic Energy Commission report foresaw nuclear growth from about 0.3 quad (of primary energy) in 1970 to over 21 quads in 1990 and to 48 quads in 2000, while actual US nuclear energy consumption in 1990 was only 6 quads and is very unlikely to be significantly higher in 2000.
The APS Panel on Public Affairs (POPA) has a substantial history of studies of energy issues, including a 1979 study on photovoltaic energy conversion (5). It was concluded by POPA that it would be useful to have an independent, analytic study of electricity generation from renewable sources, although in view of the nature of some of the issues it was not clear that the APS was the proper group to undertake such a study. To assess the appropriateness of an APS study in this area, a workshop was organized by the POPA and held in November 1992 in Washington, D.C. (6) Talks and discussion at the Workshop focused primarily on solar thermal, photovoltaic, wind, and biomass electricity sources (7).
Renewable Sources for Electricity Generation
In solar thermal electricity generation, sunlight is concentrated by large factors and used to heat a fluid to drive a steam or gas turbine. Three geometric configurations are being actively explored: (a) parabolic trough arrays; (b) central receivers (power towers), and (c) parabolic dishes.In solar thermal electricity generation, sunlight is concentrated by large factors and used to heat a fluid to drive a steam or gas turbine. Three geometric configurations are being actively explored: (a) parabolic trough arrays; (b) central receivers (power towers), and (c) parabolic dishes.
In parabolic trough systems, the heated fluid passes through tubes lying on the focal line of long troughs. The troughs track the sun with single-axis rotation. By 1991, nine units with a total capacity of 354 MWe had been installed in California, but the company installing them declared bankruptcy in 1991 due to a lapse in tax advantages and low natural gas prices. The units continue to provide electricity to the California grid, despite the cessation of new construction. They represent the only solar thermal technology that provides commercial power.
In central receiver systems, a large number of individual heliostats, with two-axis tracking, reflect sunlight upon an elevated receiver. A 10 MWe pilot plant facility, Solar One, operated in California from 1982 to 1987. It is scheduled to be replaced by a more advanced unit, Solar Two, at the same site. Pending experience with the latter, construction may be undertaken in the late 1990s on a 100 MWe commercial unit. Technical advances being explored for Solar Two include: stretched membrane reflectors in which two membranes are used, with their curvature determined by the pressure in the gap between them; molten nitrate salt for heat reception, transmission and storage; and air, with suitable particle loading, for heat reception and transmission (with bricks for heat storage). Higher efficiencies are anticipated for an air + gas turbine system than for a molten salt + steam turbine system.
Parabolic dish units can have individual engines at the focus of each dish, or linked receivers driving a common larger engine. Typical units are in the 25 kWe range, and are more suited to remote locations than as suppliers to the electricity grid. Of the three technologies, it is expected that the central receiver will prove to be the most economical, eventually providing electricity at 5 to 7 cents/kWh.
The use of solar photovoltaic power is at present limited by high costs to specialized applications, such as for portable consumer devices, satellites, and remote locations such as lighthouses. World sales of photovoltaic modules in 1992 amounted to only 60 MW of capacity, divided roughly equally between suppliers in the US, Japan, and Europe. Crystalline silicon flat plate units remain the dominant technology, but many alternatives are being explored in the search for low cost, high efficiency, and durability. These include other crystalline materials, thin films, multiple-junction units, and Fresnel lens concentrator units.
Since 1980, the estimated cost of electricity from photovoltaic sources has dropped from about 90 cents/kWh to about 25 to 30 cents/kWh. If this rate of price reduction continues for another 20 or so years, photovoltaic power would become economically competitive. Learning curve methodology suggests that such a price decrease may occur, in conjunction with a large increase in the volume of photovoltaic sales (8).
In principle, wind resources are sufficient to supply a large fraction of the total US electricity demand (9). Aside from cost, possible constraints on the extent to which wind power will contribute arise from the uneven national distribution of wind resources (greatest in the upper midwest), the intermittent nature of the wind, and the possible environmental impacts of large wind energy farms. At present, there is appreciable use of wind in the US only in California, where it provides about 1% of the electricity. World generation by wind was under 0.4 GWyr in 1990, of which about 75% was in California. However, there is increasing interest in other countries, and at present most of the new wind turbines being installed in the US are from Japan or Denmark.
The present cost of electricity from wind is 7-10 cents/kWh. It is projected to drop to 5 cents/kWh by 1995 and to 4 "/kWh by the year 2000. Together with a 1.5 cents/kWh incentive incorporated in the 1992 Energy Policy Act, this would make wind power economically competitive with fossil fuels or nuclear power.
Wind turbines installed in the early 1980s were mostly under 100 kWe in capacity, but new US units are now several hundred kWe and still larger units are being investigated in Europe. There have been significant improvements in recent years in wind turbine materials and blade configurations. This may permit an increase in unit size, while preserving long term durability under rapidly varying stresses. Although land requirements are high, of the order of 600 km2 per gigawatt of average output, most of the land can also be used for grazing or other agricultural applications. There has been increasing recent interest in biomass for electricity generation. Biomass now provides about 4% of total US energy consumption and, in 1990, a little over 1% of electricity generation, mostly produced for internal use of companies in the wood product industries. Any large expansion in the use of biomass for electricity generation is expected to rely on new plantations of dedicated energy crops. It is estimated that 140,000 to 800,000 km^2 of land is available in the US for such plantations.
Harvested crops can produce steam directly or can be converted into liquid or gaseous fuels for use with steam turbines, gas turbines, or, eventually, fuel cells. Gas turbines offer high efficiency, especially in conjunction with combined cycle operation or co-generation. They can be started relatively quickly, making them complementary to intermittent sources. Under favorable assumptions for crop growth rates and turbine efficiencies, about 2,000 km^2 of land would be needed per gigawatt of average output.
The intermittent nature of wind, solar thermal, and photovoltaic sources can create problems if these represent a large fraction of the electricity supply. This is not yet the case in California, where intermittent sources now account for 1 to 2% of the electricity supply. The problems can be ameliorated by complementary renewable sources, such as hydroelectric power and biomass, or through explicit storage. At present, only pumped hydroelectric systems provide large storage capacities. Other storage possibilities include compressed air, thermal storage, batteries, flywheels, and superconducting magnets. Of course, if electricity is used more efficiently, demand is less and supply problems are eased.
The proponents of the various renewable technologies envisage that their costs will become competitive with those of fossil fuels, at 4 to 5 cents/kWh, before 2000 or 2010 for biomass and wind, and somewhat later for solar thermal and photovoltaic generation.
Workshop Conclusions
Overall, the organizing group was impressed by the substantial progress made in the development of renewable technologies, but concluded that it would be inappropriate for the APS to attempt an assessment of the overall prospects of renewable electricity generation, including future costs and market penetration. Many of the issues related to costs and environmental impacts have little physics-related technical component, and therefore do not naturally lend themselves to evaluation by a physics group.Overall, the organizing group was impressed by the substantial progress made in the development of renewable technologies, but concluded that it would be inappropriate for the APS to attempt an assessment of the overall prospects of renewable electricity generation, including future costs and market penetration. Many of the issues related to costs and environmental impacts have little physics-related technical component, and therefore do not naturally lend themselves to evaluation by a physics group.
However, it was concluded that it would be valuable to carry out a somewhat more limited study, emphasizing a critical evaluation of technological aspects of the main generation methods, including their current status and their potential progress, problems, and opportunities. Possible topics include metal fatigue and structural dynamics problems for wind turbines, fuel cells for electricity generation from gasified biomass, photodegradation in amorphous silicon cells, and energy storage methods. A study with a technical emphasis could be of value in identifying possible strengths and weaknesses in proposed technologies, without necessarily attempting to answer broader questions as to the future role of renewable energy.
Acknowledgements. I have drawn heavily upon talks and discussions at the POPA workshop, plus the report prepared subsequently by members of the POPA organizing committee, and I am indebted to these contributors.
Unless otherwise indicated, data on energy consumption are based on Annual Energy Review 1991, Report DOE/EIA-0384 (91) (US Department of Energy, Washington DC, 1992) and Monthly Energy Review, March 1993, Report DOE/EIA-0035(93/03) (US Department of Energy, Washington, DC, 1993).
- Common units are kilowatt-hour (kWh) and gigawatt-year (GWyr), where 1 GWyr = 8.76 x 109 kWh. For renewable energy, it is common to assign a nominal primary energy input based on the energy content of the displaced fossil fuel, presently in the US at 10335 BTU/kWh (33% thermal efficiency). Thus, 1 quad of primary energy corresponds to an output of 11 GWyr, where 1 quad = 10^15 BTU = 1.055 x 10^18 J. Often, generation capacity is specified in units such as megawatts-electric (MWe), to emphasize that the reference is to the electrical output rather than the thermal input.
- The Potential of Renewable Energy, An Interlaboratory White Paper, Report SERI/TP-260-3674 (Solar Energy Research Institute, Golden,1990).
- For example, in the National Energy Strategy (1991), electricity from renewable sources corresponds to only 12 quads of primary energy in 2030, while the scenarios of reference (3) have primary inputs ranging from 13 to 38 quads.
- Principal Conclusions of the APS Study Group on Solar Photovoltaic Energy Conversion, H. Ehrenreich, Chairman (APS, New York, 1979).
- The POPA Committee which organized the Workshop consisted of: David Bodansky (Chair), Henry Ehrenreich, Anthony Fainberg, Daniel Fisher, J.D. Carcia, Pierre Hohenberg (Chair, POPA), Roberta Saxon, and Francis Slakey. Talks were given by: George D. Cody (Exx0n), William Fulkerson (ORNL), Allan R. Hoffman (DOE), Pascal De Laquil (Bechtel), Arthur H. Rosenfeld (Berkeley), Robert A. Stokes (NREL), Carl J. Weinberg (PG&E), and Robert H. Williams (Princeton).
- A brief summary of the Workshop is presented in: Report on POPA Workshop on Electricity Generation from Renewable Sources (1993). Single copies may be obtained by writing (with an enclosed self-addressed mailing label) to Ms. Nancy Passemante, APS, 525 14th Street NW, Suite 1050, Washington, DC 20045.
- G.D. Cody and T. Tiedje, "The potential for utility scale photovoltaic technology in the developed world: 1990-2010," in Energy and the Environment, B. Abeles, A.J. Jacobson, and P. Sheng, editors (World Scientific, Singapore, 1992), pp. 147-215; and POPA workshop talk.
- D.L. Elliott, L.L. Wendell and G.L. Gower, An Assessment of the Available Windy Land Area and Wind Energy Potential in the Contiguous United States, Report PNL-7789/UC-261, (Pacific Northwest Laboratory, Richland, 1961). For winds of Class 4 or higher (mean wind speeds greater than about 5.5 m/sec (12.4 mi/hr) at a height of 10 m), wind turbine hubs at a height of 50 m, "severe" environmental restrictions on the placement of wind turbines, and an overall efficiency of about 19% in converting wind power to electrical power, the authors conclude that wind could generate 250 GWyr per year.
The author is at the Department of Physics, University of Washington, Seattle, Washington 98195.