Federal failure to fulfill spent-fuel obligations creates expensive risks.
Stephen Maloney is a managing consultant in Towers Perrin’s risk and financial services practice. Email him at firstname.lastname@example.org
For more than 50 years, the federal government has failed to manage spent nuclear fuel (SNF) and high-level radioactive waste (HLW), imposing the burdens for this critical function on the private sector. Nuclear plant operators incurred upwards of several hundred million dollars per reactor in uncompensated expense and risk premiums, and potentially face decades of additional costs and risks coping with SNF and HLW.
When federal policies fail, they fail in a big way. The most obvious costs to utilities are in SNF custodial facilities and services that were supposed to be provided by the Department of Energy (DOE) in exchange for the fees paid by utility customers.
But, the more insidious costs are the decades of reimbursed interest and risk premiums, and their ripple effects lowering utility credit ratings and raising capital costs for years to come.
Companies that sold nuclear facilities in the past two decades suffered additional losses at the time of sale, because DOE’s breach put the buyers of those plants at significant risk. No nuclear waste repository was, or is, available. At the time, dry-cask storage technology and safety regulations were in the embryonic stage and success uncertain. Many decommissioning trusts were underfunded. These plants were distress sales.
Plants located in hostile regulatory environments were particularly susceptible to high risk and consequential market devaluations, both direct and indirect. Prospective owners of new nuclear facilities often estimated SNF custodial costs over a plant’s lifetime and facility decommissioning costs and dialed these costs into their bids. In addition to the expected costs, there remained the risk premiums on those expenditures and capital costs. If there were no solution to spent-fuel pool capacity limits, some plants faced the risk of early shutdowns. The owners would lose their entire investments while still facing the responsibility for decommissioning the facility and managing SNF into perpetuity. Sellers took a significant hit on plant valuations on both an expected value and risk basis.
The costs of DOE’s breach continues to this day. Utility customers still pay SNF fees, at the rate of one mill per kilowatt hour, for services still not rendered long after those fees were supposed to end—apparently for nothing. Yucca Mountain effectively is now dead. Before the project was terminated, Yucca Mountain’s cost estimates were fast approaching $100 billion. While such costs might seem little more than a rounding error in the current federal budget, they constitute significant dollars in their own right—and taxpayers are liable for potentially several billion dollars more in damages the courts have, and likely will, award utilities.
Before America begins building a new fleet of nuclear power reactors, federal policy makers should resolve the uncertainty about spent-fuel management once and for all—as they’re legally obliged to do.
More than 25 years ago, Congress passed the Nuclear Waste Policy Act (NWPA) to solve the nuclear waste problem. Among other things, NWPA directed utilities to contract with DOE, which would be responsible for building and operating a waste repository. Contracting utilities were to pay a one-time fee for the electricity generated and sold prior to April 7, 1983, and a continuing fee of one mill/kWh thereafter until the repository was fully funded. DOE was to begin collecting and disposing spent SNF and HLW no later than Jan. 31, 1998.
The federal government quickly became a serial breacher of the standard contracts. Within just a few years, it was clear DOE wouldn’t complete the repository for decades. It also was clear the repository was insufficiently sized to support continued operation of the U.S. fleet of reactors. Faced with spent-fuel pools fast approaching capacity, many plant operators explored other technologies both to mitigate the impending breach and perhaps provide a more effective resolution.
Some were able to gain a few years of operation without violating Nuclear Regulatory Commission (NRC) safety requirements through modifications to the storage racks in their spent-fuel pools. A new industry soon was born that would provide spent-fuel storage in dry casks on plant grounds outside the plant facilities. These casks still needed to pass NRC scrutiny. For quite some time, it wasn’t obvious dry casks would be licensable or available in time to preclude premature shutdowns of plants.
Soon after the 1998 breach, many utilities filed claims against DOE to recover damages. Claims involving an express or implied contract with the United States brought within the six-year statute of limitations are heard by the U.S. Court of Federal Claims (CoFC). While companies in other industries might aggressively protect their interests, utilities typically shy away from litigation, especially when the federal government is involved. Lawyers cost money and a lot of information can be revealed through discovery.
Finally, the remedies available through the CoFC can’t make utilities whole. For example, CoFC is precluded from awarding remote or consequential damages resulting from the federal government’s breach. Yet, a significant portion of costs are in that category. Also, CoFC doesn’t award interest accrued on damages. Not surprisingly, the Justice Department sought to delay trials as long as possible.
Further, CoFC is not too free with federal money. For example, CoFC can’t award damages considered speculative. Anything less than “absolute exactness or mathematical precision” potentially qualifies as speculative, which is problematic since some claims involving plant devaluations are intrinsically inferential in nature; financial engineering methods used in such cases are statistical or stochastic. While mathematically precise in expected results, any statistical or stochastic estimate falls within a band of uncertainty—which fits well in a cost-accounting framework, but which the court might consider speculative by definition.
To date, two classes of claims have been brought: the so-called “cancelled check” claims by plant owners, and “diminution of value” claims by companies who’ve sold their plants and exited the nuclear industry. Cancelled-check cases are straight-forward. Under a cancelled-check case, a company tallies up the money spent on spent-fuel pool re-racking, dry cask storage, and associated operating expenses incurred to mitigate DOE’s breach. Most nuclear companies are pursuing claims of this kind.
Of course, these costs are never done. Since there’s no reason to presume when, if ever, DOE will take custody of SNF, these costs will continue accumulating for 50 years or more. Thus, companies need to routinely bring their claims back to CoFC. The impacts on a utilities’ risk capital aren’t part of a cancelled check claim, and ratepayers and stockholders will continue to foot the risk premium bill for years to come. And, operating plants are worth less today than they would have been because of the SNF perpetual liability.
The second class of claims involves diminution of value. These claims were filed by companies who sold their plants to other nuclear operating companies some 10 years ago. Under a diminution claim, DOE’s breach is considered to have caused a loss in market value of the entire plant at the time of sale in addition to the money required to store SNF.
Utilities have pursued many of both types of cases and some decisions have been rendered. In a case filed by Xcel subsidiary Northern States Power (NSP), CoFC awarded more than $116 million for damages NSP incurred mitigating DOE’s breach at its operating reactors at the Prairie Island and Monticello sites. The company originally sought to recover more than $172 million in SNF mitigation costs. The company’s cost of capital was $54.5 million, or just under one-third of total costs, which the court excluded from the award in its decision. CoFC also made several other lesser adjustments in otherwise granting NSP’s claim for its operating and capital costs.
From the court’s perspective, cost of capital was merely interest expense, which is precluded. From the company’s perspective, cost of capital is tangible and reflects, in part, the risk the company faces with any capital project and the costs capital markets impose for assuming that risk.
NSP’s stockholders also were denied interest on the damages while awaiting the trial and decision. At 12-percent interest compounded over 10 years, NSP’s award would have ballooned to a 2008 value of $435 million. Justice delayed benefited the federal government by more than $300 million in financing, at the expense of NSP and its customers.
Next, a seller’s case. The first diminution case decision involves Boston Edison (now NStar). Approximately 10 years ago, the Commonwealth of Massachusetts enacted legislation restructuring the electric utility industry in response to FERC Order 888. Included in this legislation is the requirement that regulated companies, such as Boston Edison, divest or functionally separate electric generation from transmission and distribution services. Boston Edison elected to sell its power generation, and this process led to a competitive auction of the Pilgrim Station. The winning bidder was Entergy Nuclear, which now owns and operates the plant.
As Pilgrim’s owner, Entergy is pursuing its own claim against the United States. But Boston Edison’s case and the information readily available in CoFC’s decision illustrates the magnitude of costs. When Pilgrim was sold, Boston Edison added funds to the plant’s decommissioning trust. The decommissioning trust is intended to return the site to greenfield status and store SNF until such time as DOE reasonably is expected to assume custody. Boston Edison determined it needed to add $86.2 million more to Pilgrim’s decommissioning trust to pay for additional SNF storage costs arising from DOE’s breach. The utility also sought $2.4 million for additional racks installed in the spent-fuel pool to provide interim storage. Finally, the company claimed it received $36.4 million less for Pilgrim due to price reductions made by Entergy to account for risks associated with the prospect of indefinite storage of SNF not otherwise addressed in the decommissioning trust-fund transfer.
Boston Edison’s claimed damages total $125 million. In a decision rendered Feb. 20, 2008, CoFC awarded Boston Edison $40.03 million in damages plus payment for the cost of the lawsuit. Of this total, CoFC awarded less than half the funds Edison added to the decommissioning trust, and nothing for the additional storage racks. CoFC also concluded Boston Edison failed to establish Pilgrim’s purchase price was diminished by any given amount due to DOE’s breach. Thus, of the $125 million claimed by the utility, CoFC awarded just under one-third, shorting Boston Edison’s stockholders $84.7 million on their damages claim.
Risk Premiums & Valuations
As the NSP and Boston Edison decisions make clear, damages awarded are often a fraction of damages claimed—and damages claimed usually are less than costs actually incurred. Risk is among the more important costs not captured in damage claims and awards. In the NSP decision, for example, cost of capital for SNF modifications was simply barred. In the Boston Edison decision, the company claimed only a portion of the diminution of value suffered due to SNF risk, and even that amount was rejected.
Unfortunately, risk isn’t without cost. The world is fundamentally uncertain and the cost of that uncertainty is reflected in risk premiums. Risky investments, such as bonds, evidence the cost of risk through higher discount rates demanded by capital markets. So-called junk bonds, considered below investment grade in quality and posing greater risk of default, pay comparatively high returns.
Companies owning or purchasing nuclear power plants incur greater risks than companies holding only non-nuclear base-load generation. This nuclear risk premium is some hundreds of basis points, and reflects the operator’s exposure to safety regulations and the intrinsic risks of nuclear technology, which present the potential for extended shutdown or even premature retirement. Nuclear plants with special exposures might experience even higher risk premiums. For example, purchasers of plants sold early in the wave of nuclear sales in the mid-late 1990s were known to have priced the capital investment at discount rates above comparably sized fossil units.
Further, a high discount rate exceeding a company’s cost of capital is common for high-risk capital projects. Hurdle rates used for capital planning purposes are distinctly higher than equity holders’ average rates of return, and much higher than the return on debt. This practice is evident in other high-risk investments where risk premiums commonly reach or exceed 1,000-basis points or more above a company’s weighted average cost of capital.
The Boston Edison decision illustrates how actual costs can exceed damages claimed. CoFC says in its Boston Edison decision that Entergy employed “an IRR [internal rate of return] that exceeded 20 percent.” Such rates weren’t unreasonable because this was a risky venture for Entergy at that time; Boston Edison’s Pilgrim unit was only the second nuclear power plant sold, and the Commonwealth of Massachusetts wasn’t considered particularly supportive of nuclear plant operations.
Market depth, market liquidity, and the corresponding irreversibility in a nuclear transaction also factor into the risk. Unlike other generating technologies, there’s a limited market for nuclear power plants, substantially defined by the ability to transfer the operating license. Only a handful of companies are qualified to run a nuclear facility. The sales transaction is costly, time-consuming and subject to intervention and disruption by local authorities. Thus, once purchased, a buyer likely will be stuck with the deal and will incur sunk costs associated with the transaction. Since the nuclear plant market is neither deep nor very liquid, a company considering a nuclear purchase might try to capture in its IRR or discount rate the market’s view of investment irreversibilities and potential losses. Losses might include the potential for an accident, an extended regulatory shutdown or, worst of all, a scenario requiring premature shutdown and decommissioning of a nuclear unit.
In the Boston Edison decision, the court reports Entergy’s analysts were concerned about the potential for premature shutdown, decommissioning risk, and market price risk, while speculating on importance and ranking. Incredibly, CoFC doesn’t appear to recognize that most of these risks can be substantially hedged away or have nothing to do with SNF.
For example, market price risk would apply to any generator of Pilgrim’s size at that location. Market price risk can be, and routinely is, hedged through power agreements, and offsetting financial and physical hedges. In any case, market risk doesn’t present the potential for catastrophic losses of the kind associated with a premature shutdown.
Nuclear capacity-factor risk is dominated by operational events and regulatory risk, and there are ways to manage and offset this risk. Again, short of a premature shutdown, the effects on investment returns are far from catastrophic. Since Boston Edison provided a decommissioning fund when the plant was sold and decommissioning costs appeared to be trending downward, decommissioning risk was mostly hedged, as well.
The only remaining component that couldn’t be hedged away because of its magnitude was the risk of premature shutdown and loss of the entire investment. Standard computational methods exist for estimating the valuation sensitivity to this premium. Today those methods would include embedded options, and a formal treatment of market irreversibilities.
For purposes of illustration, however, the more familiar net present value (NPV) calculation of the kind often used 10 years ago to value assets can be used to test the sensitivity to three potential nuclear risk premiums: 250, 500, and 750 basis points (100 basis = 1 percent). The court reports Entergy assumed that the plant would operate through 2012, subject to risk of premature shutdown encapsulated in the higher IRR used. Lacking any other information concerning Entergy’s analysis, one makes the simplifying assumption that the plant generates a uniform sequence of after-tax cash streams over its life. A simple NPV calculation of these three cases indicates the nuclear risk premium, including SNF uncertainty, decreases the actual purchase price by between 11 percent and 44 percent, with a midpoint of about 24 percent from what the plant would otherwise have been worth on a NPV basis. For an $80 million acquisition, these risk premiums imposed on the entire investment translate to a value diminution ranging from nearly $9 million to more than $32 million. The mid-point is $19.3 million or roughly 24 percent decrease on the $80 million spent.
These results are scalable. For a $200 million acquisition under similar simplifying assumptions, a 20-percent IRR increases the value of the claim by between $22 million and $80.3 million with a midpoint of $48.3 million. A $400 million acquisition ranges from $88 million to $160.3 million with a midpoint of $96.3 million.
The SNF component of the nuclear risk premium can be inferred from these results using a variety of standard financial engineering methods. For some plants faced with an imminent shutdown, the SNF component accounts for risks in the range of 90 percent of the total.
Again, these results are based on simplifying assumptions (i.e., uniform net after-tax cash streams), and might not reflect other aspects of the transaction structure (i.e., any commodities or assets with salvage value, or that can be sold easily in a liquid market). By comparison, in the Boston Edison case, Edison’s expert testified that the SNF premium was $38.7 million, based on market experience that places SNF’s impact at Pilgrim in the range of 500 basis points to 750 basis points, depending on deal structure.
Clouding Nuclear’s Future
As claims made by NSP and Boston Edison in their court cases attest, risk has real costs. There are also consequential effects. Risk impacts the risk capital a company must set and cuts into liquidity. Unhedged and uninsured risks are borne by the available risk capital. If a loss is realized, thereby overwhelming the firm’s available liquidity and risk capital, the firm fails.
As seen with market and capacity factor risks, such risks typically are hedged to some degree using physical and financial instruments. Decommissioning trust funds ensure adequate monies are available after plant retirement, but the balance between risk and available risk capital impacts the value of a company’s securities. Rating agencies are taking a growing interest in the risk capital levels at energy firms. The balance between exposure level and available risk capital can be an important factor in the credit rating. In some cases, a company might find its rating agency imputing debt to cover potential risk exposures insufficiently hedged or unduly exposed to counterparty or systemic risk.
These standards are impacting balance sheets at a difficult time for the industry. Over the past year or so, utility stock prices have declined generally. Kilowatt-hour deliveries are down as much as 8 percent with industrial sales often much worse. Some companies face significant pension shortfalls due to market declines. Overall, cash flow and liquidity management are critical functions at utilities now. For some, this means deep cost cuts and slashed dividends. Companies face significant challenges maintaining risk capital and ensuring liquidity.
Until such time as rate increases are forthcoming or revenue increases, many companies will remain barely investment grade. And debt imputations will cut into available risk capital. As dry casks continue to accumulate on plant sites, nuclear companies face continued exposure to SNF-related risk, effectively into perpetuity. Will circumstances change in the future, limiting their ability to add casks needed to store used fuel and thereby continue operating? Will additional challenges to liquidity arise as a result?
Some rating agencies think so. Moody’s and Standard & Poor’s routinely comment on the continued failure to resolve the SNF issue. As one recent S&P report observes, “Effective nuclear waste disposal and storage are crucial aspects in solidifying nuclear plants as viable resource alternatives on a large scale.” For those companies contemplating nuclear investments, S&P promised to closely monitor developments “to determine if, and when ratings should change to reflect the challenges ahead.”
With dry cask storage capacity capable of supporting up to 60 years of operation, those contemplating building the next generation of nuclear plants assume DOE won’t provide custodial and disposal services, perhaps a prudent assumption given the forecasted federal deficits. Until the federal government resolves the SNF management issue, however, the industry will continue facing risks that translate directly into lost shareholder value and higher customer rates. As S&P observes, “It is contradictory for the federal government to provide incentives for new nuclear generation on one hand, [while] not delivering on its commitment to provide long-term storage on the other.”