By Lynne Holt, Paul Sotkiewicz, and Sanford Berg 1. April 8, Abstract. I. Background

Transcription

1 NUCLEAR POWER EXPANSION THINKING ABOUT UNCERTAINTY By Lynne Holt, Paul Sotkiewicz, and Sanford Berg 1 April 8, 2010 Abstract Nuclear power is one of many options available to achieve reduced carbon dioxide emissions likely to be mandated by an (as yet) undefined national climate change policy. Investment costs in nuclear power are greater than in any other conventional generating technology. They are irreversible and involve uncertainties during the project s development, construction, and commercial operation. This article extends a real-option value model (Robert Pindyck, 1993) to explain the uncertainties facing prospective nuclear plant developers and applies that model to describe mitigation strategies available for the development, construction, and operation of new nuclear plants. I. Background Prospects for the construction of new nuclear plants are perhaps better now than in the past 30 years given the improved performance and availability of the nuclear fleet over the past decade, the recent volatility of fossil fuel prices, the Obama Administration s support for new nuclear plant construction, and the prospect for Federal climate change policy. However, memories of the accidents at the Three Mile Island (1979) and Chernobyl (1986), large cost over-runs for many units in the current fleet, long construction horizons, a history of poor initial performance following commercial operation, and the unresolved disposition of high-level waste makes the prospect for new nuclear power facilities daunting. Added to this mix are the uncertain final form, assuming one emerges, of Federal climate change policy and the uncertainties regarding the ultimate effectiveness of evolving power technologies. It is easy to see that prospective nuclear plant developers, both merchant and regulated, face formidable challenges regarding their ability to recover the costs of their investments in new nuclear projects. Then there is the challenge of determining the costs of such projects in the first place. An article in this journal shows how difficult it is to estimate the construction costs of new nuclear plants, in large part because of the paucity of recent data. 2 Until recently no new nuclear plant had been ordered in the U.S. since Why have over three decades elapsed since the last new plant was ordered? The answer may reside in the inability of prospective developers to mitigate the perceived uncertainties governing investments in new nuclear plants. To that end, in 1993 Economist Robert Pindyck applied a real-options model to the uncertainties governing development and construction costs for rate-regulated nuclear power plants.

2 His model specifies the option value of waiting for construction cost uncertainties to resolve themselves. 3 Since the development of Pindyck s model, the electricity industry has been restructured in certain parts of the United States with the development of large and liquid wholesale power markets and retail competition in 14 states as an alternative to the traditional rate-regulated paradigm. In particular, restructured wholesale markets introduce uncertainty in commercial operation for nuclear plant operators with respect to prospective future revenue streams that depend on fossil fuel costs, fluctuations in demand, and outcomes in policies related to climate change and renewable energy technologies. This article addresses the role of uncertainty during development, construction and commercial operation phases of a new nuclear plant, and examines policies that address these uncertainties, through the lens of Pindyck s real-options model. In addition, it extends Pindyck s model to account for uncertainties in commercial operation, described in this paper as revenue and operating uncertainty. While Pindyck acknowledges the presence of this type of uncertainty, his analysis focuses on development and construction. Uncertainties associated with new nuclear plant development, construction, and commercial operation may be mitigated, at least to some extent, by federal and state policies which were crafted in large part to revitalize the nuclear industry by creating more favorable conditions for investment. These uncertainties are more pronounced in states with restructured retail electricity markets than in those with rate-regulated generation. Regulators in states with traditional rate regulation may authorize incentives that shift the investment risk for new nuclear plants from developers to ratepayers, an option not available to developers in restructured states. Therefore, the incentives used to mitigate uncertainty have potentially different effects on consumers in states with market-based and rate-regulated generation. Finally, federal and state incentives can never completely eliminate investment uncertainty. Perhaps the most dramatic examples of intractable uncertainty are a presently undefined national carbon policy and long-term fuel volatility. II. The Nature of, and Methods for Mitigating, Uncertainty during Development and Construction The types of uncertainties and associated risks faced by nuclear developers, and the manner in which those risks can be mitigated, will help determine the expected value of a project and whether a developer will choose to go forward immediately with a project or wait until more information is gained. Additional information may help resolve uncertainty and make it profitable for the developer to move forward with the project rather than wait for still more information to be revealed. During project development and construction a developer faces two types of uncertainty, according to Pindyck: (1) technical uncertainty and (2) input cost uncertainty. Technical uncertainty refers to the amount of time, effort, and materials needed to bring the project to commercial operation. Technical uncertainty is present when developers choose new or advanced designs, with which there is little or no previous experience. Pindyck explains that technical uncertainty can be largely controlled or mitigated by the nuclear plant developer. A developer can commence

3 development and construction and gain valuable information regarding the actual amount of time, effort, and materials necessary to bring the project to commercial operation. If it is revealed that the project can be completed within expected parameters and will remain financially viable, the project can proceed to completion and commercial operation. However, if commencing development and construction reveals that the project cannot be brought to commercial operation without expending significantly more time, effort, and resources, the developer may decide to abandon the project. In the presence of technical uncertainty, the expected value of the project increases if the developer is able to commence development and construction of the project, gain information about completion costs, and subsequently abandon the project if it appears not to be financially viable. An example, using Pindyck s typology, shows how the option to abandon the project in the face of technical uncertainty increases the expected value of the project. Assume that a new, commercially operational nuclear power plant has a discounted revenue stream that is known with certainty of $4200 per kilowatt of capacity. If a range of possible costs is used to simplify calculations, it might cost only $2000 per kilowatt to build the new plant with a probability of 0.5, but the cost of construction could be as high $7000 per kilowatt with a probability of 0.5. Prior to commencing development and construction, the net present value of the project per kilowatt of capacity is $4200 (0.5)*$2000 (0.5)*$7000 = -$300, and therefore would not be undertaken at all. However, suppose the developer can commit to spending $2000 per kilowatt and with a probability of 0.5 the project will be completed, and with a probability of 0.5 that the project would require the additional $5000 per kilowatt in outlays which would imply that the project would be abandoned. Under this scenario, the project would be started because it has a positive net present value per kilowatt of -$ (0.5)*$ (0.5) * $0 = +$100, since it would be abandoned (zero outlays) if an additional $5000 per kilowatt were required to bring the project to commercial operation. The example illustrates how the option to abandon the project helps mitigate the technical uncertainty. Input cost uncertainty refers to the uncertainty surrounding the prices paid for inputs such as labor and materials as well as changes to government regulations and policies. This uncertainty occurs whether or not changes to input costs are due to supply-demand fundamentals or to regulatory rules that affect the cost of construction, given that the time, effort, and materials required to bring the project to commercial operation are known. The longer the time between commencing construction and commercial operation, the greater the input cost uncertainties. And unlike technical uncertainty, input cost uncertainty cannot be mitigated or resolved by commencing development and construction: it is pervasive throughout a plant s development and construction phases. Moreover, information regarding input costs is revealed regardless of whether the project is started or not. Consequently, to mitigate input cost uncertainty, a developer is likely to wait for more input cost and regulatory information to be revealed before commencing with development and construction. This strategy stands in contrast to the manner in which a developer can mitigate technical uncertainty. Therefore, the ability to wait for input cost information to be revealed prior to commencing development and construction increases the expected value of the project.

4 Assume a new, commercially operational nuclear power plant has a discounted revenue stream that is known with certainty of $4200 per kilowatt of capacity as in the previous example. It might cost only $2000 per kilowatt to build the new plant with a probability of 0.5, but the cost of construction could be as high as $4800 per kilowatt with a probability of 0.5. Prior to commencing development and construction, the net present value of the project per kilowatt of capacity is $4200 (0.5)*$2000 (0.5)*$4800 = $800, and appears to have positive expected profits. However, suppose the developer can wait one year for the cost of construction information to be revealed. With a probability of 0.5, the project will cost $2000 and the revenue stream will be $4200; otherwise the project will not be undertaken. The option to wait results in a net present value per kilowatt of capacity of (0.5)*($4200 $2000) = $1100 one year from now, which when discounted to the present is still a higher value than that derived from the option to start the project and wait for the costs to be revealed. Therefore, waiting to commence development and construction mitigates input cost uncertainty and creates the incentive to wait for additional input cost information to be revealed prior to commencing construction. III. The Nature and Mitigation of Uncertainty during Commercial Operation Once in commercial operation a new nuclear unit will face uncertainties regarding the stream of future revenues that will allow the developer of the plant to recover its fixed costs including capacity investment costs and a rate of return. In a rate regulated environment the stream of revenues to cover these costs are likely known and designed specifically to cover costs plus a rate of return. Developers in states with retail competition, by contrast, have no such assurance: lacking a defined customer base from which they may recover prudently incurred costs, they face heightened risks in selling power in competitive wholesale markets. In a competitive wholesale market environment, the stream of revenues is derived from transactions in various wholesale energy, capacity, and ancillary service markets. Market revenues will depend on plant availability and performance, the overall supply-demand balance, the cost of fuels of competing technologies such as coal and natural gas, the mix of generation fuel types in the market, and environmental policies that place a price on emissions stemming from prospective Federal climate change policy. A developer has two options for mitigating uncertainty in commercial operation: 1) It can enter into long-term contracts that improve the predictability of revenue streams, much the same as in a rate regulated environment; and 2) it can decide to wait for additional information about operating conditions to be revealed. That information may be gleaned from the evolution of stable wholesale energy and capacity prices over time, or the resolution of major environmental policy uncertainties such as those related to potential Federal climate change policy. Assume the cost of construction is known and is $3500 per kilowatt of capacity. With a probability of 0.5, the realization of operating conditions and environmental policies results in net wholesale market revenues of $2000 per kilowatt of capacity. With a probability of 0.5, the realization of operating conditions and environmental policies results in net wholesale market revenues of $4800 per kilowatt of capacity. The expected value of the nuclear plant is then (0.5)*$ (0.5)*$ $3500 = -$100 per kilowatt of capacity (the value of expected net wholesale market revenues less levelized construction

5 costs). Suppose instead that the developer can either attempt to sign contracts for power sales that ensure wholesale market revenues to be $4800 per kilowatt or wait for the realization of operating conditions and environmental policy that results in revenues of $4800, but that only has a 0.5 probability of success. The option of waiting to do either results in a positive value per kilowatt of capacity for the project of (0.5) ($ $3500) = $650. Consequently, developers can exercise the realoption to wait for greater certainty in commercial operating conditions prior to committing to going forward with the project. IV. Policies for Mitigating Uncertainty Because there are so many uncertainties facing prospective developers of new nuclear plants, federal and state incentives are designed to counter the otherwise logical propensity of developers to take a wait and see attitude and learn from early movers experiences. Most federal subsidies are targeted to improving the predictability of a new nuclear plant s cash flow in some manner. State incentives are often targeted to improving the predictability of regulatory ratemaking, and consequently the cash flow, for the proposed project through all its phases (development, construction, and commercial operation). These incentives are designed not only to stimulate early investment by some developers, but to increase the information available to other developers. Therefore, these incentives effectively reduce the benefits from delaying investment decisions, since information on the regulatory and technical environment is revealed through the actions of first-movers. Technical uncertainty may be mitigated largely through long-term contracts for fuel, supplies, and project management. Input cost uncertainty cannot be so easily hedged to the extent that it relates to changes in government policies that emerge after the project has commenced but before the new nuclear plant supplies electricity. Therefore, it is perhaps no surprise that the federal and state incentives listed below in Table 1 are targeted largely to allaying input cost uncertainty for the development and construction phase of projects. Only federal production tax credits and state ratemaking policies are explicitly designed to mitigate revenue and operating uncertainty. Policy Federal Production Tax Credits Table 1 Federal and State Policies for Mitigating Uncertainty Impact The federal production tax credit, authorized by EPACT 2005, is $18 per MWh for 6,000 MW of new nuclear capacity for the first eight years of operation, with a limit per plant of $125 million per year. This incentive primarily addresses revenue and operating uncertainty by providing a more favorable and predictable revenue stream during commercial operation to hedge against unpredictable price fluctuations, such as those associated with fuel prices and energy demand during the multiple decades of a nuclear plant s operation.

6 Federal Loan Guarantee Program Federal Insurance for project delays Federal Streamlined Licensure Federal Pre-Approved Plant Designs Although the main thrust of the production tax credit is to reduce revenue and operating uncertainty, it also attempts to mitigate technical uncertainty by encouraging potential developers to be pioneers and build the first projects: eligible nuclear plants must be placed in service by January 1, Therefore, production tax credits mitigate both technical and revenue and operating uncertainties associated with a plant s commercial operations. Federal loan guarantees of up to $18.5 billion were authorized by EPACT Advanced nuclear energy facilities are among the projects eligible for loan guarantees of up to eighty percent of a project s construction costs. These guarantees are scheduled to expire on September 30, 2009, unless Congress extends them. 4 The Obama Administration is proposing $54.5 billion triple the amount authorized in This incentive would reduce input cost uncertainty affecting cash flow during the construction phase of new nuclear plants. Because this incentive is targeted to first movers, it also would mitigate technical uncertainty. Access to a loan guarantee should lower a project s overall capital costs, thus increasing the net present value of cash flows for the commercial operations of the plant. The federal government will cover debt service for the first six plants if commercial operation is delayed for reasons involving litigation and failure on the part of the Nuclear Regulatory Commission to meet established schedules. The coverage is limited to $500 million for the first two reactors and $250 million for the next four reactors. This incentive would reduce input cost uncertainty affecting cash flow during the construction phase of new nuclear plants. Because this incentive is targeted to first movers, it also would mitigate technical uncertainty. All applications for licensure of new nuclear plants are reviewed by the NRC for safety, environmental impact, and (previously) for antitrust implications. In 1989, the NRC instituted a streamlined approval process that combines construction and operating licenses based on a review of the same information required of each previously. If it functions as intended, streamlined licensure would mitigate input cost uncertainty because it would shorten the time needed for project development. However, at the same time, streamlined licensure may impose input cost uncertainty for developers because its novelty may make it difficult for them to predict compliance costs. The standardization of designs for new reactors is one way of driving down costs. The NRC issues design certification for new reactors independent of a specific site based on a rule making process. Each certification is valid for 15 years from the date it is issued and may be extended. To date, the NRC has issued four design certifications, the most recent in The availability of pre-approved plant designs may mitigate input cost uncertainties associated with development and construction of new nuclear plants because it should theoretically shorten construction time. However, it may give rise to technical uncertainty because of the prospect that changes to pre-approved designs may subsequently be needed. For example, the NRC took issue with Westinghouse s AP-1000 design because of a concern that the shield building component in the reactor may not

7 Federal Early Site Permits State Rate- Making Principles State Accelerated Cost Recovery Mechanisms for New capacity State Incentives for Project Abandonment adequately protect the reactor from earthquakes, tornadoes, and high winds. Early site permitting is a means of expediting NRC siting approval for new nuclear projects. The NRC approves these sites independent of an application for a specific construction permit. These permits are valid for 10 to 20 years and may be renewed. To date, four early site permits have been issued. Early site permits would reduce input cost uncertainty by shortening the development time for a project. Because nuclear plants are so expensive to build -- $10-12 billion before financing costs 5 developers in rate-regulated states look for assurance that the rules governing the treatment of a project s costs will be predictable. Developers will be particularly concerned if regulators change the way they will treat the return on equity, the depreciable life of a project, or other cost recovery issues after the project is built. To allay those concerns, states such as Iowa, Kansas, and Wisconsin have enacted legislation that allows their state public service commissions to adopt rate-making principles which are binding and applicable to all rate proceedings governing the project. These principles provide predictability to investors and therefore reduce input cost uncertainty to developers (and shift them to ratepayers) during the development and construction phases of the project and reduce revenue and operating uncertainty during the plant s commercial operations. Some states with traditional rate regulation, including Georgia, Florida, Kansas, Louisiana, Michigan, Mississippi, North Carolina, South Carolina, and Virginia, authorize accelerated cost recovery mechanisms for new nuclear plants. Expedited cost recovery mechanisms are intended to lower interest rates on debt and reduce cash flow constraints during the latter part of a plant s construction when costs are highest. Expedited cost recovery mechanisms are generally considered beneficial to a developer since it lowers the risk of default. For ratepayers, the risk is less clear cut. On the one hand, ratepayers may incur costs for the project before the plant ever generates electricity. There is also the possibility that the plant will never be commercially viable and ratepayers will therefore realize no benefits from it. On the other hand, if a developer s long-term debt is reduced because its credit rating is improved or maintained, its cost of capital will likewise be lower. A lower cost of capital may result in more favorable rates in the long-term for customers. Therefore, expedited cost recovery mechanisms would reduce revenue and operating uncertainty as cost recovery is set in advance and is accelerated. If the cost of capital is reduced, input cost uncertainty primarily for development and construction is reduced. Some states with traditional rate regulation, including Florida, Georgia, Iowa, Mississippi, North Carolina, and South Carolina, have enacted laws that authorize developers to abandon work on a nuclear plant and still recover prudently-incurred costs. To the extent that costs are recovered through rates, customers receive none of the benefits of expanded generation but only higher electric bills. Project abandonment provisions reduce technical uncertainty for development and construction of new nuclear plants. Developers can begin construction and then re-evaluate the cost inputs needed

8 to complete the project: if the inputs are no longer determined to be cost-effective, developers can elect to abandon the project with reduced risks for investors but greater risks to ratepayers. Nuclear Waste Insurance The Price-Anderson Act, which took effect in 1957, protects electric utilities for liability associated with nuclear accidents. Over the years the primary insurance pool has increased to over $10 billion and the secondary pool to approximately $8.6 billion. The Energy Policy Act of 2005 extended the Price-Anderson Act to the end of This insurance protection reduces input cost and revenue and operating uncertainties throughout the entire life of a new nuclear project. V. CONCLUSION The goal of the federal and state policies identified in Table 1 is to reduce uncertainties through means other than project delay, so that projects can move forward and uncertainty can be reduced for the next set of developers. Federal and state policies to reduce uncertainty facing prospective developers have been largely directed toward reducing input cost uncertainty during the development and construction phases of a new nuclear project. However, project delays have spill-over implications for a new nuclear plant s cash flow once the plant is commercially viable. With the notable exceptions of the Federal production tax credit and the authorization granted public service commissions in several states with traditional rate regulation to adopt binding rate making principles, most federal and state policies are geared toward reducing uncertainty during project development and construction. A looming uncertainty, with significant implications for revenue and operating uncertainty, is the treatment of carbon emissions at the national level and the long-term volatility of fuel costs. Nuclear generation becomes more attractive in terms of capital costs and construction time relative to other generation sources if a national climate change policy cap and trade or a tax on emissions ---is adopted. Even if the ultimate nature of Federal climate change policy is presently unknown, the ongoing debate over the potential impacts of additional government incentives for new nuclear plant development will continue, particularly in light of the recent bail-outs of large financial institutions and automobile companies. For example, if Congress decides to authorize an expansion of the Federal loan guarantee program for new nuclear plants and some of those projects subsequently default, how much risk will taxpayers, as well as investors, assume? What level of risk will investors and taxpayers be willing to tolerate going forward? The debate itself is a certainty but the responses to the questions generated by it are not. Ultimately, the long-term effectiveness of federal and state mitigation strategies to reinvigorate the construction and operation of the new nuclear plants is the biggest uncertainty of all.

9 1 Dr. Lynne Holt is a Policy Analyst for the Public Utility Research Center (PURC), at the University of Florida. Dr. Paul Sotkiewicz is the Chief Economist at the PJM Interconnection, Dr. Sanford Berg is a Distinguished Service Professor (Economics) and the former Director of PURC at the University of Florida. The ideas expressed here do not necessarily reflect the views of PURC sponsors or of PJM member companies, the PJM Board, or any PJM employees. This article summarizes and extends an earlier paper, which focuses on the impact of federal and state policies to mitigate uncertainties facing nuclear energy. See L. Holt, P. Sotkiewicz, and S. Berg, (When) To Build or Not to Build?: The Role of Uncertainty in Nuclear Power Expansion,, TEXAS J. OIL, GAS, AND ENERGY LAW, Vol. 3, No. 2 (2008), at We gratefully acknowledge the review of and suggestions by Ted Kury, Director of Energy Policy, Public Utility Research Center. 2 So Much Will It Cost to Build a Nuke? ELEC. J., Jan/Feb. 2010, at 1, R. S. Pindyck, Investments of Uncertain Cost, J. FIN. ECON. Vol. 34. No. 1(1993), at President Obama announced on February 16, 2010 loan guarantees totaling $8.3 billion for two new light-water reactors in Georgia, to be developed by Southern company and two partners. 5 R. Smith, U.S. News: Costs Cloud Texas Nuclear Plan, WALL ST. J., Dec. 5, 2009, at A. 2.