Kevin F. Forbes & Ernest M. Zampelli* (Forbes@cua.edu)
In an economically rational society, policies would be adopted that would seek to attain the society’s environmental goals at the least possible cost. For example, to minimize the cost of reducing carbon emissions in the power sector emissions from all carbon sources should be cut such that marginal abatement costs are equalized across all carbon sources within the power sector. Under a cap-and-trade scheme, there is reason to believe that this condition would be realized if firms are profit maximizers and the carbon market is perfectly competitive. Under these assumptions firms would find it profitable to reduce carbon emissions until marginal abatement costs equals the price of carbon. With a uniform carbon price within the power sector, marginal abatement costs would therefore be equalized across generating units both within and across firms, thus minimizing total abatement costs given the targeted reduction in emissions.
Unfortunately, cap-and-trade legislation is currently stalled in the United States Senate. This actually suits some individuals as quite fine. These individuals advocate instead for policies that would simply increase the supply of renewable energy such as wind energy. For example, Bjorn Lomborg, the author of The Skeptical Environmentalist and Cool It argues that it is a waste of time trying to actually cut carbon emissions and that society instead should focus instead on making solar and wind energy competitive with electricity from fossil fuels.[i] We do not mean to suggest that this view is representative of the environmental movement. Many environmental advocacy organizations strongly support cap-and-trade legislation. The Sierra Club has called the carbon emission allocation system “pivotal in determining the success of global warming legislation.” The National Resources Defense Council has stated that a “cap on carbon pollution is the single most important policy that our nation can adopt to move toward a clean energy economy” and that cap-and-trade is a “proven strategy for achieving a desired level of emissions reductions.” Whatever the motivations of Mr. Lomborg, when faced with politically difficult and complicated legislation, solely promoting the simple iconic symbol of a windmill as a solution to our problems has appeal. As in many areas of public policy, simplistic popular solutions may not be the most effective ones.
In this paper, we assess the economics of a policy approach that attempts to reduce carbon emission solely by stimulating the supply of green energy. In section 2 of the paper we present an economic model that addresses whether more supplies of green power can lead to a first best solution in terms of emissions. In section 3, we present data on the cost of wind energy, wind energy being the fastest growing form of green power. In section 4, we briefly discuss the challenge that wind integration creates for electricity market balancing. In section 5, we econometrically assess the effect of wind energy on the dispatch of both coal and natural gas. Evidence is presented that increases in wind energy disproportionally displace natural gas instead of coal. The effect of this is to mute wind’s potential contribution to reducing carbon given that the emissions from natural gas are less than half those of coal.
2. Can Additional Supplies of Green Energy Lead to a First-Best Optimum?
In a first-best optimum, the externalities associated with the production of electricity would be internalized such that the marginal social cost of electricity equals the price. Can additional supplies of green energy alone attain this solution? In general, the answer is no.[ii] To see this, assume there are two sources of power, Dirty Power (DP) and Green Power (GP). For simplicity, assume that the marginal social costs for GP equals its marginal private costs, i.e., there are no external costs associated with the production of GP. In contrast, assume that the marginal social costs of DP are considerably larger than its marginal private costs. That is, society bears a cost, pollution, in the production of DP that the producers of DP do not have to pay for. Assume further that the marginal social costs of GP (MSCGP) is less than the marginal social costs of DP (MSCDP) but that the marginal private costs of DP (MPCDP) are less than marginal private costs of GP (MPCGP). This situation is represented graphically in Figure 1. Under Laissez-Faire, the supply curve would be SLF which is simply the horizontal sum to the GP and DP supply curves (observe however, that SLF will coincide with SDPLF when price is less than or equal to P0).
The market equilibrium under Laissez-Faire is represented in Figure 2. The equilibrium price and quantity are PLF and QLF, respectively. Observe that dirty and green power output equal QDP,LF and QGP,LF, respectively where QDP,LF + QGP,LF equal QLF.
The first best optimum price and quantity are represented in Figure 3 as P* and Q*, respectively. At this point, the marginal social cost of producing electricity equals the price that consumers are willing to pay. Observe that in this case, the externality is so large that the optimum quantity of dirty power is zero, i.e., only green power is produced at the optimum (this is a simplifying assumption so as to make it easier to represent the optimum). Observe that Q* < QLF. A subsidy induced rightward shift in the supply of green power would pivot the total supply curve to the right (the supply of dirty power would not shift since suppliers of dirty power are assumed ineligible for the subsidy). The equilibrium price would decline while total output, which already exceeds Q*, would rise instead of decline. Because of the lower price, QDP would fall. But the decline in dirty power would be less than in increase in green power. The general result therefore is that subsidies to green power should not be expected to lead to a first-best optimum.
3. Wind Energy : Not So Cheap in Terms of Private Costs
When the cost of electricity transmission is ignored, the United States is believed to have more than 8,000 GW of wind resources that the industry estimates can be developed at a cost of less than or equal to approximately $80 per MWh.[iii] When transmission costs are factored in, only about 600 GW of resources could be available at a delivered price less than $100 per MWh (Figure 4).[iv] To put this $100 price in perspective, the monthly average price among the three default load aggregation points (LAPs) in the California ISO was $49.15/MWh in January 2010.[v] The price “gap” between the current price of electricity and the projected supply cost of wind may be of some relevance given that the California Public Utilities Commission and the California Energy Commission have recommended a 33 percent renewable energy requirement as a key strategy to reduce greenhouse gases with cap-and-trade only accorded a relatively minor role. The plan, if implemented, would increase the wind energy capacity available to the California independent system operator by almost 500 percent by 2020 with wind energy capacity accounting for approximately 18 percent of installed nameplate capacity.[vi]
4. Wind Energy: A Challenge to System Operations
The stability of a power grid requires that the amount of power generation in a balancing authority area match exactly, on a near-instantaneous basis, the system load, net of losses and interchange with other balancing authority areas. In this context, wind energy can represent a challenge to operations because production levels are largely uncontrollable with upward dispatch not being possible. There is also evidence in the case of ERCOT that wind energy production levels are difficult to forecast and that the forecast errors have adverse operational impacts.[vii] There is also evidence that the root-mean-squared-errors in the day-ahead wind forecasts in two out of the four electricity control areas in Germany are about three times the errors in the load forecasts.[viii] The wind forecasting errors in the Transpower system in Germany (formerly known as E.ON Netz) have been shown to have implications for its unscheduled power flows with other systems.[ix] Unscheduled electricity flows can be a significant challenge to system operations since generation needs to be dispatched to offset their effect on system frequency. This may be a tolerable situation when neighboring systems have low wind energy penetration levels. It would be arguably far less tolerable if and when wind achieves 20 percent integration everywhere.
5. Wind Energy: Largely Displacing Natural Gas and thus Not So Green
According to data reported by EIA, the carbon emissions per kWh are about 57 percent lower from a natural gas combined cycle generating unit than when the electricity is produced using coal.[x] The contribution of wind energy to lower carbon emissions is therefore maximized when wind displaces coal.
To examine the impact of higher wind penetration levels on the dispatch of coal vs. gas, electricity production data was obtained from ERCOT for the period 13 June 2009 through 31 January 2010. The following equation was estimated:
Logistick = Ck + ΣδkHi + αkDaWindShri + βkSchedLoadi (1)
Logistick,i is the logistical transformation of the generation produced using fuel k in period i (in MWh) relative to period i’s scheduled load, k = gas, coal
Hi is a vector of binary variables representing the hour of the day (exclusive of hour 1 so as to avoid the “dummy variable” trap).
DaWindShri is the hour 12 day-ahead forecasted level of wind energy for market period i relative to the scheduled level of load for that period.
SchedLoadi is the scheduled level of load in period i.
The regression results are presented in Table 1. To economize on space, the coefficients on the binary variables are not reported but are available on request.
All of the reported coefficients are highly statistically significant as evidenced by their t-statistics. The coefficient on scheduled load is positive in case of gas and negative in case of coal indicating that the share of load accounted for by gas rises as load rises while the share of load accounted for by coal rises as load declines. This is consistent with the fact that coal is less flexible than natural gas in terms of dispatch. With respect to the coefficient on DaWindShr, the coefficients are negative for both fuels indicating that increases in forecasted wind reduce the expected dispatch of both fuels. While both coefficients are negative, the coefficient on DaWindShr for gas is about 6.5 times the estimated coefficient for coal. The difference between the two coefficients is statistically significant. This indicates that increases in forecasted wind energy reduce the dispatch of both fuels but that the reduction of electricity from natural gas is far greater than the reduction of electricity from coal. This occurs because system operators base the dispatch of non-wind generation on private marginal costs with priority being given to generating units with the lowest private marginal costs. An increase in wind energy will therefore reduce expected generation from higher private marginal cost generating units (one exception would be the output from gas turbines which are dispatched to help balance supply with demand). While natural gas combined cycle plant are more efficient in power conversion than coal plants, the reality is that the short run private marginal costs of generating a MWh from natural gas is generally more than the short run private marginal costs of generating a MWh of electricity from coal. Thus, in the absence of cap-and-trade legislation, an increase in wind energy has the unintended consequence of largely displacing the cleanest fossil fuel.
The findings in Table 1 are consistent with a 2008 report by United States Department of Energy.[xi] One implication of these findings is that increases in wind penetration will have only a modest effect on carbon emissions. Specifically, the 2008 DOE study projects that carbon emissions from the electricity sector in 2030 will be substantially above the target level even if the 20 percent wind penetration level is attained (Figure 5).
The integration of wind energy into the power grid is perceived as an important metric of action to reduce carbon emissions. However, as we have seen, simply having more green energy on the power grid will not yield a first best solution. To reduce carbon at the least cost to society, firms need incentives to reduce emissions from the most carbon intensive fuels. It is not a socially desirable outcome to have carbon reduction undermine the reliability of the power grid. We need a policy that does more than promote iconic symbols at the expense of promoting proven policies that use markets to cost effectively achieve the important societal goal of reducing carbon emissions.
We thank Michael A. Forbes of MIT for writing the software that performed the approximately 5,500 downloads of the ERCOT hourly wind forecast data and the programs used to process the data. We also thank the staff of ERCOT for providing us with access to their data series and for answering our many questions. Any errors are the full responsibility of the authors.
Forbes, K., Stampini, M. and Zampelli E. (2010). Do Higher Wind Power Penetration Levels Pose a Challenge to Electric Power Security? Evidence from the ERCOT Power Grid in Texas, unpublished manuscript.
Forbes, K., Stampini, M. and Zampelli E. (2009). Wind Power Variability, Wind Power Forecasting Errors, and the Cost of Balancing Power: Evidence from the E.ON Transmission System in Central Germany, unpublished manuscript.
California ISO (2010) Market Performance Report January 2010, February 25, 2010 http://www.caiso.com/2424/2424d03b3f610.html
Energy Information Administration (2010). Electric power annual. http://www.eia.doe.gov/cneaf/electricity/epa/epa.pdf
Department of Energy, Office of Energy Efficiency and Renewable Energy (2008). 20% Wind Energy by 2030:Increasing Wind Energy's Contribution to U.S. Electricity Supply. Available on the Internet at http://www1.eere.energy.gov/windandhydro/pdfs/41869.pdf
Gold, Russell (2010) Natural Gas tilts at Windmills in Power Feud, Wall Street Journal March 2 2010 http://online.wsj.com/article/SB10001424052748704188104575083982637451248.html
Hawkins, David , CAISO’s Plan for Integration of Renewable Resources, July 20, 2008 NARUC Meeting http://www.narucmeetings.org/Presentations/Hawkins%20-%20CAISO%20Renewables%20presentation.pdf
Lomborg, B. (2009). Don’t Waste Time Cutting Emissions, New York Times, April 25 2009. http://www.nytimes.com/2009/04/25/opinion/25lomborg.html
* Authors are with The Center for the Study of Energy and Environmental Stewardship, Department of Business and Economics, The Catholic University of America, Washington, D.C.
[i] Lomborg, B. (2009). Don’t Waste Time Cutting Emissions, New York Times, April 25 2009. http://www.nytimes.com/2009/04/25/opinion/25lomborg.html
[ii] The answer would be yes if the demand for electricity is perfectly inelastic.
[iii] Department of Energy, 2008, p. 8.
[iv]Department of Energy, 2008, p. 9.
[v] California ISO (2010), p.6
[vi] Based on the values reported by Hawkins(2008)
[vii] See Forbes, Stampini, and Zampelli (2010)
[ix] See Forbes, Stampini, and Zampelli (2009)
[x] Based on EIA (2010), Tables Table 5.4 and Table A3.
[xi] Department of Energy (2008), p. 12. This report finds that 20 percent wind integration could “ displace about 50% of electric utility natural gas consumption and 18% of coal consumption by 2030.” This finding of wind displacing natural gas is also consistent with the observations of Gold(2010).