² The author wishes to thank those who reviewed early versions of this paper: George Burmeister, Ed Holt, David Hoog, Greg Kats, Ken Langer, Alan Miller, Robert Nordhaus, Jon Pietruszkiewicz, Karl Rábago, Roby Roberts, Adam Serchuk, Michael Tennis, Carl Weinberg, and Ryan Wiser. The final draft is the responsibility of the author and does not necessarily reflect the opinions of REPP, the REPP Board of Directors, or the reviewers.
Back to Article

³ Percentages in this paragraph are calculated from U.S. Department of Energy (DOE), Energy Information Administration (EIA), Renewable Energy Annual, 1995 and 1998 (Washington, D.C.). Hydroelectric power has been excluded. Green power is often consid-ered to include power from small hydroelectric facilities, but the EIA does not publish separate data for small hydro.
Back to Article

⁴ Center for Resource Solutions, Summary of Accomplishments of Green-e in 1998 (no date).
Back to Article

⁵ Ibid.
Back to Article

⁶ Litigation delayed New Hampshire’s plan to unbundle its retail market at the beginning of 1999. For current state-by-state information on retail competition, see <http://eia.doe.gov/electricity/ch_str/tab5rev.html>.
Back to Article

⁷ California marketers currently offering products that are certified under the Green-e program are listed in Appendix A.
Back to Article

⁸ Center for Resource Solutions, op. cit. note 4.
Back to Article

⁹ An early study of the California green power market found green premiums ranging from a little more than 2¢ to about 3¢ per kilowatt-hour (kWh) for products generated solely from renewable sources and a proportionately smaller premium for products consisting of a mixture of green and conventional power. Ryan H. Wiser and Steven J. Pickle, Selling Green Power in California: Product, Industry, and Market Trends (Berkeley, CA: Ernest Orlando Lawrence Berkeley National Laboratory, May 1998).
Back to Article

¹⁰ For an analysis of the potential shortfall in green generating capacity, see work paper posted at <http://realliance.org/insurance/>. Demand for green power presumably would be met in the sense that, in the absence of price controls, price would rise until supply and demand were in balance. However, in the absence of some mechanism for bridging the gap between consumers’ short-term commitments and lenders’ long-term concerns, it is not clear that a higher price would by itself elicit much additional supply.
Back to Article

¹¹ For the assumption that the price risk would be borne initially by marketers, see note 18.
Back to Article

¹² DOE and Electric Power Research Institute, Renewable Energy Technology Characterizations, EPRI TR-109468 (Washington, DC: December 1997).
Back to Article

¹³ This differs from the premium assumed for this evaluation, which is 0.05 cents per annual kilowatt-hour of coverage.
Back to Article

¹⁴ See R. Wiser and E. Kahn, Alternative Windpower Ownership Structures: Financing Terms and Project Cost (Berkeley, CA: Lawrence Berkeley Laboratory, Energy and Environment Division, May 1996).
Back to Article

¹⁵ Ibid. The authors found that the cost of wind power could be reduced by as much as 29% if developers could raise capital on the same terms as electric utilities.
Back to Article

¹⁶ An assurance of an adequate flow of income for 10 years can support a loan amortization term longer than 10 years. Loan service after 10 years is of less concern to a lender because part of the loan will have been amortized by then and because the cost of competing conventional power is expected to have increased, at least in nominal terms.
Back to Article

¹⁷ This will not necessarily be true if the market has excess green generating capacity. Until the excess is absorbed by growth in demand, the market price for green power may be too low for even the most efficient new generating facility to be profitable. This situation may now exist in some states where substantial amounts of green generating capacity were built on the basis of power purchase agreements with utilities.
Back to Article

¹⁸ A marketer could also protect itself by purchasing its power under short-term contracts or contracts with market-responsive price clauses. Such measures would not reduce the price risk, but would shift it upstream to the developer, who would therefore be the potential purchaser of the price insurance. For this paper, it is assumed that the marketer purchases power under a long-term contract (or, at least, not at a market-responsive price). On that assumption, the marketer would purchase the insurance.
Back to Article

¹⁹ The projection of green power demand underlying the proposal reviewed in this paper assumed that 14 states would unbundle their retail electric markets by the end of 2005. See DOE, Green Power Consumer Demand: 2000-2005 (n.d.), posted at <http:// www.realliance.org/insurance/>.
Back to Article

²⁰ Some observers have suggested that the failure of California consumers to embrace the conventional and green power choices offered to them is due to the terms under which the California retail market was unbundled. One of those terms required utilities to give their customers a temporary 10% reduction in rates. This provision has raised the bar for a marketer seeking to attract customers from the utilities. It appears that residential consumers have been slow to abandon traditional utility suppliers in Massachusetts and Rhode Island for similar reasons. See source noted in note 4.
Back to Article

²¹ See note 18.
Back to Article

²² See note 9.
Back to Article

²³ The marketer might decide to maintain its retail price at least temporarily. Its failure to meet competition would tend to result in a loss of sales that would leave the marketer with excess power under its power purchase agreements. The marketer generally could sell the excess power in the spot market, but at a lower price. Its loss then would be based on an average of its (unchanged) retail price and its (lower) spot market price.
Back to Article

²⁴ The latter might occur because there was a decline in the price of conventional power: although the wholesale green premium was unchanged, the market price of green power could decline because the premium was being added to a lower base. It would, of course, be possible to create insurance that protected green marketers against the risk of a decline in the wholesale price of green power, and not just against a decline in the green premium component of that price. However, the risk of a decline in the other component — the price of conventional power — is one that is borne by green and conventional marketers alike. Only the risk of a decline in the green premium is unique to green marketers, and only that risk would be insured under the proposal evaluated here.
Back to Article

²⁵ This assumes that the premium would be paid on annual green power sales of 4.8 billion kWh over a 10-year period. For the basis for the latter assumption, see Appendix C.
Back to Article

²⁶ The maximum payout of 1¢ per kWh on annual sales of 4.8 billion kWh would be $48 million per year.
Back to Article

²⁷ The market price of green power in two states may differ by the cost of moving power from one state to the other. Market prices may also differ because of the different characteristics of the power produced by different technologies. In general, the market price will be higher for dispatchable green power — those forms that can supply energy on demand — than for intermittent varieties such as solar or wind. A price difference may also emerge based on consumer preference for one form over another. For example, if consumers prefer solar power to power generated from municipal solid waste, solar power will enjoy a price advantage.
Back to Article

²⁸ The fact that no price index is published for a particular state does not necessarily mean that changes in a published price cannot adequately measure the price risk in that state. Suppose, for example, that an index is published for Massachusetts but not for New Hampshire. The green premium in the two states may be different. However, if there are green power sales between the two states, their green power market prices will be linked, and New Hampshire marketers may consider that changes in the Massachusetts price adequately measures the New Hampshire price risk.
Back to Article

²⁹ The alternative is likely to involve information from actual power purchase agreements. For such information to be used, problems of confidentiality would have to be resolved.
Back to Article

³⁰ DOE, Annual Energy Outlook 1998, Table A-17, p. 122.
Back to Article

³¹ See Gregory Kats and Kenneth Langer, Green Power Finance Initiative: Executive Summary (draft) (August 26, 1998). The authors state that mandated capacity accounts for “the bulk” of the capacity projected to be added between 2000 and 2005 under existing conditions.
Back to Article

³² One implication of this point should be noted. The insurance premium set in those negotiations would be a floor as well as a ceiling. It is possible that insurance companies will consider that there are prospects that would not pay the agreed insurance premium because they are low risk but could profitably be insured at a lower premium for the same reason. The companies should be free to pursue those prospects by offering a reduced premium. However, counting the insurance sold at the lower premium against the companies’ obligations would run the risk of directing more of the insurance to capacity that would have been built without it.
Back to Article

³³ See note 31. For purposes of the calculation of the cost effectiveness of the proposed price insurance in Appendix C, it is assumed that 25% of the projected “build-anyway” capacity (310 megawatts) would take insurance that would count against the insurance companies’ obligations. That appears to be a conservative assumption, but even a significantly higher one — say 30 or 40% — would not affect the basic conclusion.
Back to Article

³⁴ Industry participants in the discussions and companies indicating an interest in offering or buying the insurance are listed in Appendix D.
Back to Article

³⁵ Arizona, Connecticut, Maine, Massachusetts, Nevada, and New Jersey.
Back to Article

³⁶ Under a renewable portfolio standard (RPS), the additional amount paid by each consumer is small — much smaller than the green premium paid by consumers who purchase green power in an unbundled retail market. However, this is because under an RPS, the additional cost of the green power is spread over all of the electricity consumers in the jurisdiction. It is not because the total cost of the green power is lower. Note that proponents of the insurance proposal hope that the companies would find it profitable to continue to offer the insurance after their obligation had been satisfied.
Back to Article

³⁷ The cost per kWh is calculated in Appendix C.
Back to Article