Table of Contents *

1. Introduction and Summary *

2. Nuclear Generation Projections *

3. Decommissioning Funding *

4. Spent Nuclear Fuel Transportation and Long-Term Storage Costs *

5. Conclusion *

6. References *

Table 2.1 Retired Nuclear Generating Units With Capacity 40 MW and Larger *

Table 2.2 Key Input Assumptions to Nuclear Retirement Analysis *

Table 2.3 Summary of Nuclear Retirements in the Reference, Low and High Cases *

Projections: *

Table 4.1 Funding Scenarios for High Level Nuclear Waste Disposal Program *

Table A.1 Nuclear Units Identified at Risk in the Reference Case A-*

Table A.2 Nuclear Units Identified at Risk in the Low Case A-*

Table A.3 Nuclear Units Identified at Risk in the High Case A-*

Table B.1 Decommissioning Funding Status by Company B-*

Table B.2 Decommissioning Funding Status By Generating Unit B-*

DOE 1996 Fee Adequacy Assessment With Real Interest at 2% C-*

DOE With Synapse Reference Nuclear Projection C-*

DOE With Synapse Low Nuclear Projection C-*

Independent Cost Assessment C-*

Independent Cost Assessment With Synapse Reference Nuclear Projection C-*

Independent Cost Assessment With Synapse Low Nuclear Projection C-*

1. Introduction and Summary
1. Introduction

This paper summarizes the results of an integrated three-part project dealing with: (1) the prospects for early nuclear power plant closure, (2) the potential unfunded liability for decommissioning, and (3) the potential unfunded liability for spent nuclear fuel transportation and storage. This is all considered in the context of electric industry dergulation.

When utilities and regulators determine funding amounts for decommissioning it is typically assumed without question that the nuclear generator will continue to produce electricity until the end of its 40-year operating license. Similarly, for determining funding adequacy for the high-level waste disposal program, the Department of Energy typically assumes that all reactors will run to (or nearly to) the end of their operating licenses. This faith in the longevity of existing nuclear power plants is unfounded, inconsistent with nuclear experience to date, and can lead to imprudent and inefficient decision-making.

 

2. Nuclear Scenarios

Based upon our comparison of nuclear unit operating costs with projected market prices for electricity, we have developed three projections of nuclear unit retirements. The High, Reference, and Low Nuclear Generation cases have 20, 34, and 90 nuclear units retiring prior to the end of their operating licenses. In each case, the average shutdown for the units closed "prematurely" is about 15 years prior to the end of the license.

 

3. Decommissioning Funding Shortfalls

We estimate that for the fleet of currently operating nuclear power plants, the investor-owned utilities’ portion of the unfunded liability for decommissioning is about $24 billion, at year end 1997 in 1997 dollars. If all of the units operate to the end of their licenses and the decommissioning cost estimates turn out to be correct, then the full amount needed for decommissioning could be collected during the units’ operating period. However, with early retirements, we estimate the unfunded decommissioning liability at the time of closure, summed for all of the units projected to be retired early, to total $4.1 billion, $7.1 billion, and $15.3 billion in 1997 dollars for the High, Reference, and Low Nuclear Generation cases, respectively.

 

4. Spent Fuel Funding Shortfalls
   We find that the prospect of plant retirements reducing the revenue stream to fund the disposal of spent nuclear fuel suggests that the current one mill per kWh fee collected by DOE should be increased. Even more significantly, the DOE’s cost estimate for implementing the spent fuel disposal program appears to be out-of-date and optimistic. If a recent independent cost assessment putting the total program cost roughly 50% above DOE’s estimate is correct, then the fee (in nominal dollars) may have to be increased to something in the range of 2.6 mills per kWh (for the EIA generation projection) to 4.5 mills per kWh (with the Synapse low case nuclear generation projection.
   
   These required increases, however, could result in additional nuclear retirements. For example, with the Synapse Reference Case for nuclear retirements and the independent cost estimate for the waste disposal program, we find that the fee increase from one to 2.9 mills per kWh can be expected to result in the early retirement of an additional ten nuclear units. This would, in turn, cause a need to further increase the fee. It appears that the high level waste disposal program may not be able to simultaneously satisfy both of its principles. This potential conflict should be recognized, and avoided to the extent possible by timely adjustments to the fee. To the extent that required prospective fee increases are not feasible due to the feedback effect upon nuclear generation, other funding approaches, such as retroactive assessments upon the generators of the waste may be necessary.

 

2. Nuclear Generation Projections
1. Recent Nuclear Generator Retirements

One need only consider the list of units that have shutdown before the end of their operating licenses to realize that "premature" closure – retirement prior to the expiration of the operating license – is more than a remote possibility. Table 2.1 lists the nuclear power plants in the US that have been retired, or for which shutdown has been announced.

Table 2.1 Retired Nuclear Generating Units With Capacity 40 MW and Larger

Plant	State	Capacity (MWe)	Year Closed	Approximate Age at Retirement (years)
Hallam	Nebraska	75	1964	2
Pathfinder	South Dakota	66	1967	3
Fermi 1	Michigan	61	1972	9
Indian Point 1	New York	265	1974	12
Peach Bottom 1	Pennsylvania	40	1974	8
Humboldt Bay	California	65	1976	14
Dresden 1	Illinois	200	1978	19
Three Mile Island 2	Pennsylvania	926	1979	1
Shippingport	Pennsylvania	72	1982	25
La Crosse	Wisconsin	48	1987	19
Rancho Seco	California	918	1989	15
Shoreham	New York	820	1989	0
Fort St. Vrain	Colorado	330	1989	10
San Onofre 1	California	436	1992	25
Yankee Rowe	Massachusetts	175	1992	31
Trojan	Oregon	918	1992	18
Haddam Neck	Connecticut	582	1996	29
Big Rock Point	Michigan	72	1996	34
Maine Yankee	Maine	840	1997	25
Zion 1 and 2	Illinois	2080	1998	25
Millstone 1	Connecticut	660	1998	28
Oyster Creek	New Jersey	650	announced	29

 

The reasons given for nuclear plant retirement decisions generally include poor forward operating economics, and recently electric industry deregulation has been noted as an increasingly important factor. For example, Geoffrey Rothwell’s assessment of the decisions to close the Yankee Rowe and Trojan plants concludes that the "plants were closed after their owners determined that the Net Present Value (NPV) of continued operations was negative or nearly negative" (Rothwell, 1997).

In the case of Maine Yankee, "economic pressure" was identified as a root cause of problems that eventually led to the unit’s retirement:

Economic pressure to be a low-cost energy producer has limited available resources to address corrective actions and some plant improvement upgrades. Management has effectively prioritized available resources, but financial pressures have caused the postponement of some needed program improvements and actions. (NRC 1996, page 71)

Maine Yankee’s Cultural Assessment Team reported that

The current economic and political environment is considered precarious, and Maine Yankee’s survival is seen to be based on a formula of low cost and high production. There is an associated fear among many employees that highlighting any negative issue could endanger the plant’s continued operation. . . . At Maine Yankee, the Team found an organization struggling with forces requiring unprecedented change. These include evolving performance standards as well as deregulation within the electric utility industry. (Bradford et al., 1996, pages 8-9)

Electric industry restructuring is, in general, magnifying pressure to cut costs. To the extent that cost cutting in operating budgets can be done without creating other problems this could work to the advantage of existing nuclear power plants. On the other hand, cutting current operating costs may be counter productive to the extent that cost cutting leads to declines in performance or to additional costs in the longer term. In either case, electric market deregulation is creating an environment where it is increasingly difficult to continue operating uneconomic plants. While some subsidies to nuclear plant operation have been provided for in "transition" plans, the pressure to mitigate stranded generation costs by closing uneconomic nuclear plants is considerable. At the same time, some companies may be waiting until stranded cost recovery is assured before deciding to retire uneconomic nuclear units.

 

2. Deregulation and Projections of Nuclear Generator Retirements

The recent experience with nuclear plant early retirement announcements raises the question of how this trend might or might be expected to continue into the future. Recent analyses of this question have found that a significant portion of the nuclear fleet is at risk of shutting down on the basis of poor operating economics.

For example, Geoffrey Rothwell (1998) analyzes the economics of the nuclear fleet using econometric estimates to simulate costs in a probabilistic comparison with electricity market prices. He concludes that "if costs are not reduced, there are approximately two dozen units at risk of early retirement before 2006, when nuclear power unit operating licenses begin to expire" (page 12).

Jim Riccio’s analysis for Public Citizen (1998) took a more straightforward approach. Riccio compared average nuclear fuel and O&M costs for the 1994 to 1996 period with the estimated cost of replacement power in the region as estimated by the Nuclear Regulatory Commission (1992). This comparison identified 42 nuclear units that are not competitive.

Moody’s Investor Services (1995) examined nuclear operating costs, and concluded that there are "at least 10 nuclear plants (out of 109 in the U.S.) that might be closed in the event of deregulation" (page 7). In November 1996 Moody’s reported that

the bond ratings of 24 nuclear operating electric utilities have been downgraded, some more than once. Only 8 IOUs [investor-owned utilities] with nuclear investments have been upgraded. . . The average bond rating for electric utilities with significant exposure to nuclear investments is Baa1 versus A2 for those with no nuclear investments and an industry-wide average of A3.

The Interstate Natural Gas Association of America released a report in May, 1998, on the "Need for Natural Gas Increases with More Nuclear Plants Shut Down." INGAA concluded that 34 of 72 US nuclear reactor sites are vulnerable to shutdown because their annual production costs are higher than projected market prices. These sites represent 34% of the nuclear generating capacity in the U.S., or 37,859 MW.

A recent annual survey of utility CEOs and managers (WIEG, 1998) found that only 42 percent of the respondents believe that "nuclear plants can compete in a price conscious market" while less than half (49 percent) believe that "most nuclear plants will remain in operation through their initial license term" – down from 67 percent in 1997. Virtually twice as many respondents (39 percent) as last year (20 percent) believe that a "large number of nuclear plants will be shut down in the next five years."

In this context of emerging competition in electricity markets and changing perceptions of the ongoing role of nuclear power generation, we set out to analyze the economics of continued plant operation.

 

3. Method and Assumptions for Nuclear Plant Retirement Analysis

The prospects for retirement of the nuclear fleet depend primarily upon the operating economics. For the most part, it is reasonable to assume that nuclear units with operating costs above the market value of their electricity will be shut down when subjected to competitive pressure. Here, we’ve constructed a framework for simulating the unit owners’ decision-making on a forward-looking basis. The basic decision-rule is that the expected present value of the costs of operating the unit must be less than the expected present value of the energy produced. Where this is not the case, the unit is assumed to be retired. Projections of present value cost and revenues are done for each unit in each year of the study.

For nuclear operating costs, we calculated averages for the 6 year period from 1992 through 1997. For our Reference case, these recent period averages were simply projected into the future with no change (in real dollars) except for a modest decline in capacity factor during the last five years of a unit’s license period. High and Low cases were developed as variations from the reference case, as indicated in Table 2.2.

 

 

Table 2.2 Key Input Assumptions to Nuclear Retirement Analysis

Variable	Reference Case	Low Nuclear Generation Case (high nuclear costs and low market prices)	High Nuclear Generation Case (low nuclear costs and high market prices)
Nuclear Capacity Factor	6 year average, with annual decline at 1% in last 5 years of license	6 year average, declining at 0.25% annually, plus annual decline at 2% in last 5 years of license	6 year average, increasing at 0.25% annually, no adjustment for nearing end of license
Nuclear Fuel, O&M and Capital Additions Costs	6 year average cost per kWh, escalating at the general inflation rate	annual escalation at 0.5 percent real	annual decline at 0.5 percent real
Near-term Electricity Value (1996)	1996 regional average of reported marginal energy costs plus $5/MWh for capacity value	7 percent less than the reference case	7 percent greater than the reference case
Long-term Electricity Value (2005 and beyond)	EIA’s projected market prices by region (based largely upon the cost of new combined cycle generation with gas)	15 percent less than the reference case in 2020 (based upon EIA’s analysis with lower natural gas prices)	13 percent greater than the reference case in 2020 (based upon EIA’s analysis with higher natural gas prices)

 

For the value of generation from the nuclear generators, we used system marginal cost data for 1996 and projections of market prices for electricity by region produced by the EIA using its National Energy Modeling System (Beamon, 1997). These assumptions, for the Reference, High, and Low cases, are summarized in Table 2.2. The EIA forecast of market prices includes 2 to 3 mills/kWh for "general and administration" costs, and so 3 mills/kWh of G&A costs were included in the nuclear costs for this analysis. G&A includes labor related benefits and taxes that are typically higher for nuclear plants than for other generating facilities.

 

4. Results of Nuclear Plant Retirement Analysis
   We find that many existing nuclear units are uneconomical to continue operating. In the reference case, 34 units are found to be uneconomical to operate. Most of these would be retired as soon as they are subjected to competitive pressure. This points to an interesting economic implication of the timing of electric industry restructuring. There appear to be a number of units that are uneconomic over the full period of their remaining lives beginning in 1998, but that would be economic over shorter, later periods due to projected increases in market prices over time. If competition comes slowly or these units are protected from competitive pressures for several years, then their owner/operators may well choose to keep them open despite their uneconomic status.
   
   In the low case, we find that most of the existing fleet of nuclear units is uneconomic to operate, and should be closed. In this case, the extent to which individual nuclear units will be retired early will be moderated by a price feedback. That is, as the first wave of nuclear units are retired, the electricity markets will tighten and the value of capacity and energy will rise. This will make the remaining units relatively more attractive to operate.
   
   In the high case, with very optimistic assumptions for nuclear plant costs and performance, we still find 20 nuclear units to be uneconomic to operate.
   
   The projected operating nuclear capacity for the Reference, Low, and High cases is plotted in Figure 2.1. All three cases have total nuclear capacity in the US declining from about 100 GW today to zero by 2037, but the drop off in the early years is much steeper in the low case.
   
   The number of nuclear of units retired early, the "lost" operating years, and the average size of the retired units are listed for our three cases in Table 2.3. In the Reference case, a total of 508 unit-years of operation are "lost" as a result of the 34 early generating unit retirements.
   
   Limitations and caveats are identified at the conclusion of this paper, in the section on "further research." It should also be noted that the results reported here understate the extent of early retirements in that feedback from nuclear waste program fees is not incorporated in the analysis. We have conducted some initial exploration of this feedback effect – increases in the level of the high-level waste disposal fee leading to additional retirements leading to additional fee increases – and describe those results in Section 4.3 of this report.
   
   It should also be noted here that the ranges used here for the low and high case input assumptions are rather tight. That is, experience for particular variables could easily fall outside of the range of projections incorporated here. It was decided that for this analysis, the range depicted in Figure 2.1 would serve well as a sufficiently wide range for analyzing decommissioning and spent fuel funding issues.

The particular nuclear units identified as at risk in the Reference, Low, and High cases are listed in the tables in Appendix A.

 

Table 2.3 Summary of Nuclear Retirements in the Reference, Low and High Cases

Case	Units Retired Prior to License Expiration Date	Total Unit Years of Operation Lost	Total Giga-Watt Years of Operation Lost	Average Size of Retired Nuclear Units (MW)
Reference Case	34	508	479	943
Low Nuclear Generation Case	90	1338	1386	1036
High Nuclear Generation Case	20	304	283	931

 

3. Decommissioning Funding
1. Background on Nuclear Plant Decommissioning

All nuclear power plants must eventually be decommissioned. The decommissioning process includes draining the plant’s fluid systems; decontaminating pipes, equipment, and structural materials that have become radioactive; and, either immediately or after some delay period, dismantling the reactor and surrounding structures and shipping the radioactive waste to a low-level waste burial facility.

The "irradiated" or "spent" nuclear fuel accumulated at the site during the plant’s years of operation must be removed from the spent fuel pool prior to decommissioning the facility. The cost of transporting and storing the fuel is typically not considered a part of decommissioning. In this report we will address spent nuclear fuel costs separately (in Section 4).

The Nuclear Regulatory Commission has required that commercial nuclear power plants collect funds for their eventual decommissioning, and set the funds aside in external trust funds. The primary reason for this requirement is to ensure that funds will be available for decommissioning the plants after they have retired. When a nuclear plant is retired, it no longer generates a revenue stream, and there is little incentive for its owner to spend money to clean it up. A plant owner that is under financial stress (perhaps related to the unanticipated, "premature" retirement of its nuclear asset) might not have the resources to responsibly decommission its facility. With external funding the assurance of the availability of funds for the safe and timely decommissioning of retired nuclear plants is improved.

There is also an equity argument for collecting funds for decommissioning during the operating life of a facility. That is, the customers who benefit from the electrical power produced by the nuclear plant should be responsible for paying its clean up cost. The equity considerations are complex, however. For example, for those nuclear plants that cost several billion dollars to construct it is not accurate to say that the customers bearing the brunt of those construction costs are "benefiting" as a result of receiving the excessively high priced electricity from the facility. Nonetheless, there is much merit to the concept that we should provide resources to ensure the safe dismantlement of today’s nuclear plants during their operating lives, rather than leaving this responsibility to future generations.

With decommissioning funding typically based upon the license period of individual nuclear units, it is nearly certain that in the event of a shutdown prior to the end of the operating license there will be a funding shortfall. The extent of the shortfall depends upon when in its license period the unit closes, the pattern of funding, and the interest accumulated on the decommissioning fund. There have been funding shortfalls for each of the nuclear units that has been closed to-date, and this is likely to be the case for many currently operating units that are shut down in the future.

Whether and to what extent companies are allowed to recover decommissioning fund shortfalls in the event of early retirement will depend upon the institutional arrangements. Some regulatory commissions and/or legislatures may require customers to pay such charges in non-bypassable electricity distribution fees while others may require the plant owners to fund the shortfall. The principle that only costs for generators that are "used and useful" should be charged in regulated rates is long-established. On the other hand, regulators have shown a willingness to make exceptions in the case of unfunded decommissioning costs, charging such costs to customers in regulated rates even after the facility is closed.

 

2. Current Decommissioning Funding Status

In Table B.1 of Appendix B, we list the estimated decommissioning cost and the amount in the external decommissioning funds, for all of the investor-owned utilities with large amounts of nuclear entitlements. The total for the 51 IOUs listed amounts to an estimated cost of $38.8 billion (in 1997 dollars) and a fund balance of $15.1 billion. Counted in this way, the current level of the unfunded decommissioning liability amounts to $23.6 billion, or about 61 percent of the total estimated decommissioning cost in today’s dollars. The decommissioning information relied upon here is based primarily upon data reported by utilities in their 10Ks, supplemented by information from regulatory proceedings in particular instances.

In Table B.2 of Appendix B the decommissioning funding information is presented by unit. For this table, some calculations were necessary. For utilities that own more than one unit, the costs were allocated based upon capacity ratings in cases where we did not have unit-specific data. For units with public power minority owners the figures were scaled up from the IOU data in order to reflect the full generating unit. Note that units owned entirely by public power entities are not included in the table since data on decommissioning funding were not available.

The figures in Table B.2 include 8 nuclear units that have already been shutdown (most of which still need to be dismantled). For these units the total estimated decommissioning cost is $3.6 billion (in 1997 dollars) while the collected funds amount to only $1.3 billion (at year-end 1997). While state regulators have sometimes allowed utilities to recover in regulated rates the decommissioning cost shortfalls for nuclear units retired early, the magnitude of the shortfall for these 8 units (about $2.3 billion) illustrates the need for timely and adequate funding of decommissioning. In a less regulated environment, it will be increasingly difficult for utilities to recover decommissioning funding shortfalls for generators that are not operating. There are also strong equity and efficiency arguments for not requiring customers to pay for such costs in regulated rates.

Table 3.1 Nuclear Plant Decommissioning Funding Summary

	Total Estimated Decommissioning Cost (billions of 1997$)	Total Decommissioning Fund Balance (billions of 1997$)	Total Decommissioning Funding Shortfall (billions of 1997$)
Data as of Year-End 1997:
51 Investor-Owned Utilities (listed in Table B.1)	38.8	15.1	23.6
102 IOU Nuclear Units (listed in Table B.2)	42.5	15.6	26.9
8 Closed Units	3.6	1.3	2.3
94 Operating IOU Nuclear Units	38.9	14.3	24.6
Projections:
Synapse Reference Case	38.9	31.8	7.1
Synapse Low Nuclear Case	38.9	23.6	15.3
Synapse High Nuclear Case	38.9	34.8	4.1

 

3. Estimated Decommissioning Funding Shortfalls
   
   With annual funding amounts collected in electricity prices charged to customers, and placed in external trust funds, the current total decommissioning funding shortfall can be expected to decline over time. Still, if funding levels are based upon the operating license periods and nuclear generators are retired prior to the end of their operating licenses, the funding available for decommissioning at the time of plant closure will be inadequate.
   
   For each of the three Synapse nuclear generation scenarios we projected the extent of the total shortfall for the nuclear industry in the US. For these estimates, it was assumed: (1) that the current decommissioning cost estimates are accurate; (2) that annual funding contributions are set in order to achieve the target levels in the year of operating license expiration; and (3) that any interest on the fund balances is exactly matched by escalation in the decommissioning costs.
   
   Given these assumptions, and the nuclear scenarios described in Section 2, we estimate that the total unfunded decommissioning liability for units retired before the end of their operating licenses would amount to $4.1 billion, $7.1 billion, and $15.3 billion, for the high, reference, and low scenarios, respectively. These figures are in 1997 dollars, and do not include the units operated by public power entities or the units already closed.
   
   The current set of decommissioning cost estimates is, of course, subject to considerable uncertainty. The rapid rate of escalation in the estimates over the past two decades suggests that further escalation is a distinct possibility, and that the unfunded liability could be much greater than the figures reported here.

 

4. Spent Nuclear Fuel Transportation and Long-Term Storage Costs
1. High Level Radioactive Waste Policy in the U.S.

The prospect of nuclear plant retirements has implications for spent fuel disposal as well. Our nation’s policy for spent nuclear fuel disposal is based upon two potentially conflicting ideas. First, the costs of disposal are to be fully paid for by the owners and generators of spent nuclear fuel through a fee paid to the DOE for nuclear kWh generated and sold. At the same time, the DOE is precluded from changing the fee retroactively. That is, the DOE can raise the fee that it charges per kWh of future generation from nuclear power plants, but it cannot go back to nuclear electricity generated in prior years if the program revenues are found to be inadequate to cover program costs. This policy results in a classic dilemma when confronted with numerous nuclear plant retirements.

The DOE periodically checks whether the one mill per kWh fee (along with some one-time payments) will be adequate to cover the costs of the disposal program. In estimating the revenue side of its program, the DOE typically assumes that all existing nuclear plants will operate (and pay the DOE one mill per kWh) to the end of their operating licenses, unless they have specifically announced plans to close early. DOE’s last assessment of the fee adequacy (done in October, 1996) found the one mill fee to be adequate. The DOE is expected to produce a new fee adequacy study soon.

 

Table 4.1 Funding Scenarios for High Level Nuclear Waste Disposal Program

Scenario	Revenues	Costs of Waste Disposal Program	Shortfall in 2071 (billions of 1997 $)	Necessary Fee to Cover Costs
1. DOE 1996 Fee Adequacy Assessment With Real Interest at 2%	Nuclear generation from EIA 1994 projection, adjusted to remove cancelled TVA units.	DOE’s cost estimate of from September 1995 TSLCC Report.	$1.9 billion	1.1 mills/kWh
2. DOE with Synapse Reference Nuclear Projection	Early retirement of 34 additional nuclear units (decreasing DOE’s generation forecast by about 10%)	DOE projected program costs decreased by 2.8% recognizing 5.6% lower total nuclear generation with half of costs fixed.	$3.8 billion	1.2 mills/kWh
3. DOE with Synapse Low Nuclear Projection	Early retirement of 90 additional nuclear units (decreasing DOE’s generation forecast by about 57%)	DOE projected program costs decreased by 16.4% recognizing 32.8% lower total nuclear generation with half of costs fixed.	$6.7 billion	1.5 mills/kWh
4. Independent Cost Assessment	Nuclear generation from EIA 1994 projection, adjusted to remove cancelled TVA units.	Cost estimate from PIC’s Independent Cost Assessment.	$45.9 billion	2.6 mills/kWh
5. Independent Cost Assessment with Synapse Reference Nuclear Projection	Early retirement of 34 additional nuclear units (decreasing DOE’s generation forecast by about 10%)	Independently projected program costs decreased by 2.8% recognizing 5.6% lower total nuclear generation with half of costs fixed.	$46.5 billion	2.9 mills/kWh
6. Independent Cost Assessment with Synapse Low Nuclear Projection	Early retirement of 90 additional nuclear units (decreasing DOE’s generation forecast by about 57%)	Independently projected program costs decreased by 16.4% recognizing 32.8% lower total nuclear generation with half of costs fixed.	$43.0 billion	4.5 mills/kWh

 

2. Nuclear Waste Funding Scenarios and Shortfalls

In Table 4.1, we summarize six scenarios for the spent fuel disposal program cash flow. In the first case, we take the analysis of the DOE’s latest fee adequacy report with one modification: the assumed real interest rate is reduced to 2.0 percent. DOE’s report presents results for a range of interest rate assumptions, but appears to favor a 2.8 percent rate based on a DRI forecast. We believe that 2.8 percent is optimistic for a risk-free return, and that a figure of 2.0 is preferable for waste fund planning purposes. Ibbotson (1998) for example reports a long-term inflation adjusted rate for government bonds of 2.1 percent.

This first scenario shows a resulting fund shortfall of $1.9 billion (in 1997$) in the year 2071, which can be avoided by increasing the fee slightly – to 1.1 mills per kWh. This case essentially forms the basis for the DOE’s belief that the funding level need not be increased.

In the second scenario for the nuclear waste program, we incorporate the Synapse reference case for nuclear unit retirements. This results in a forecast of future nuclear generation (and hence revenue from the fee) that is about 10 percent lower than that assumed by the DOE. Because nearly half of the total nuclear generation from our country’s fleet of nuclear units is behind us, however, the total nuclear generation (and hence the approximate total amount of nuclear waste) is reduced by only 5.6 percent. For the cost side of the program, we assume here that the disposal program costs are half fixed (unchanging with the amount of waste generated) and half variable (scaling proportionally with the amount of waste generated). The specific nature of how the program costs change with differing quantities of waste generated, transported, and stored, over different time periods is an important topic for detailed engineering analysis which remains to be undertaken. Note, however, that the decrease in program costs is likely to be much lower than the decrease in revenues, as a result of the structure of the program funding mechanism and the fact that we are at or near the mid-point in cumulative electricity production from our nation’s nuclear plants. The result for the reference case scenario is a projected funding shortfall of $3.8 billion (in 1997$) in the year 2071, at the conclusion of the spent fuel program. A relatively minor adjustment to the one mill per kWh fee – to 1.2 mills per kWh – is enough to offset the shortfall, if the adjustment is made in the next few years.

A third scenario, with Synapse’s Low case projection of nuclear generation, shows a funding shortfall of $6.7 billion (in 1997$ in the year 2071). This can be avoided by a fee increase to 1.5 mills per kWh. Here, the adjustment to the program costs amounts to 16.4 percent, based upon the same half fixed, half variable assumption used in the prior case.

There are many reasons to believe that the current official estimates of program costs are understated. A recent Independent Cost Assessment prepared for the Nevada Agency for Nuclear Projects (PIC, 1998) found that program costs are likely to be roughly 50 percent higher than assumed by the DOE. In our fourth nuclear waste program scenario, we substitute this cost estimate for the DOE’s, and find an expected shortfall of $45.9 billion (in 1997$ in 2071). This huge funding shortfall, a gross violation of the principle that the costs of the program are to be recovered from the generators of the waste in the fee charged to nuclear generation, can be avoided by increasing the fee to 2.6 mills per kWh.

In scenarios 5 and 6, we combine the Synapse nuclear plant retirement projections with the independent cost estimates for the spent fuel program. The results are funding shortfalls similar to that of scenario 4, but the necessary fee increases are larger, owing to the decreases in nuclear generation. In cases 5 and 6, the fee must be raised to 2.9 and 4.5 mills per kWh, respectively.

 

3. Feedback Effect of Fee Increases Upon Nuclear Retirements
   As the spent fuel disposal fee is increased to internalize the costs of nuclear waste, there is an important and troubling feedback effect upon fee adequacy. A higher fee will tend to cause additional nuclear unit retirements, which in turn will lead to a need to increase the fee. It is quite possible that in some scenarios this reinforcing feedback could result in a situation where increasing the fee is counterproductive. This prospect should be avoided, by making necessary adjustments to the fee in a timely manner, as the need becomes apparent. Delays in implementing fee increases could make it impossible to satisfy the "full cost recovery" principle for program funding, without implementing retroactive assessments.
   
   As an initial investigation of this feedback effect, we analyzed some variations on Scenario 4. In Scenario 4 we have the Synapse Reference case projection of nuclear generation and PIC’s independent cost estimate for the waste disposal program. As reported in Table 4.1, in order for the fee revenues to cover the cost of the program in this scenario, the one mill per kWh fee would need to be raised to 2.9 mills starting in 1999. When we put this increased fee level into our model for projecting nuclear retirements, we find that an additional 10 nuclear units should be retired early. The total early retirements in this scenario amount to 44 units, 671 unit-years of operation, or 654 giga-watt years. This, of course, in turn has implications for the level of the fee. With less nuclear generation the fee would have to be raised further in order to cover the full cost of the waste disposal program, possibly resulting in additional nuclear retirements.
   
   It is interesting also to consider the implications of delaying the fee increase. As the timing of the increases is deferred, the magnitude of the required fee goes up since there is less prospective nuclear generation to apply the fee to. We estimate that if the fee increase were delayed for five years, then instead of increasing to 2.9 mills per kWh, the fee would need to be increased to 4.0 mills (starting in January of 2004). If the fee increase is delayed by ten years, to 2009, then the level of the required fee is 6.0 mills per kWh.
5. Conclusion
1. Limitations of the Analysis and Areas for Further Research

The analysis described here depends upon numerous simplifying assumptions. Future research should address key issues including refinement of the nuclear cost and performance projections, the incorporation of feedbacks into the methodology, and the exploration of policy options for internalizing costs and eliminating subsidies to both fossil and nuclear generation.

The development of nuclear generation projections depends critically upon trends in nuclear plant operating costs and performance with age, particularly during the final years of a unit’s operating license. The analysis presented above considers a fairly wide range for these parameters, and not surprisingly comes to a wide range for the resulting high and low scenarios. Some uncertainty is inevitable, but further research could help to indicate likely trends, perhaps narrowing the range of input assumptions. Also, the experience toward the end of a unit’s license period deserves additional attention. It may be increasingly difficult to maintain a staff of skilled and motivated employees at a nuclear facility that is scheduled to be closed and in the context of an industry that is in decline.

The range and volatility of electricity prices in regional markets is a worthy topic for analysis. In the study described above we considered a range of projected market prices, but we did not examine the role of price volatility. On the one hand, nuclear power offers an advantage relative to fossil fuels by diversifying exposure to oil and gas price volatility. On the other hand, nuclear power has experienced its own operating cost escalation in the past. Moreover, a period of several years with market prices below the longer term average could lead to a near term increase in nuclear unit retirements.

Plant and company specific considerations, including plans for major equipment replacement, should also be addressed in future research projects. For facilities facing major equipment investments such as steam generator replacement, the economics of continued operation might be unfavorable, particularly in a deregulated market. In this analysis we have taken a broad view of the nuclear industry using data for cost and performance in the recent past. The result is that we present view of things that is smooth in the sense that we do not recognize the implications of large specific repairs required in particular future years. Also, the results presented here, while broadly reasonable for the industry as a whole, are not intended to be accurate for individual units.

The role of potential nuclear plant license extensions should be examined. Some units may be granted approval to operate for an additional period beyond the expiration of their current operating licenses. If license extension becomes common, then funding shortfalls projected here could be substantially decreased.

Tightening environmental regulations for fossil-fueled power generation should be considered. Additional costs could be imposed upon fossil fuel generators to reduce emissions of sulfur dioxide, nitrogen oxides, particulates, mercury (and other toxics), and carbon dioxide. Such additional regulations would tend to improve the economics of nuclear power plant operation. In most regions, however, in the longer term the marginal resource type determining the market price for wholesale electricity is expected to be natural gas fired combined-cycle generation. Since these gas units are relatively quite clean, the impact of tighter emissions standards will be felt mainly by existing, inframarginal fossil fuel generators.

Feedback of nuclear retirements upon market prices for electricity could be an important phenomenon, particularly in regions and scenarios where several units are retired early. The closure of baseload generators would tend to increase market prices, making additional early retirements less economically attractive. This is a complex phenomenon that can be readily modeled with a computerized system dispatch simulation model. Such modeling would tend to be for particular regions, and is beyond the scope of the analysis conducted here.

For nuclear plant decommissioning cost estimates and fund balances, the data presented here has good coverage of investor-owned utilities. Additional research could usefully be aimed at determining the status of decommissioning funding for public owners of nuclear power plants.

The nuclear waste fee results are sensitive to the relationship between total nuclear waste volume and DOE spent fuel program costs. Here we made a simple scaling assumption: that one half of the costs of the high level waste program are fixed (do not vary with the amount of waste) and half are variable (scaling proportionately with the amount of waste). This is a relationship that could be usefully explored by detailed engineering analysis of the nuclear waste program cost components.

Proposals for interim storage of nuclear waste and litigation about responsibility for costs incurred as a result of delays in the DOE development of high-level waste storage facility both play an important role in our overall nuclear waste policy. These considerations, however, are beyond the scope of the present analysis.

Financial assumptions (i.e., the inflation rate and real interest rate) have a great influence upon the economics of the spent fuel disposal program. While we are comfortable with the assumptions employed in this analysis, further research into the appropriate financial assumptions to use in this analysis would certainly be useful. The cost streams for the high level waste program extend well into the next century, and the economics are quite sensitive to the financial assumptions.

Consideration should be given to the role of Price-Anderson liability limits and nuclear insurance in a competitive electricity market. If the market is to function efficiently, then subsidies such as the limit on liability in the event of a nuclear accident should be eliminated. It may be possible for the market to internalize these costs in the form of insurance premiums.

There is also a crucial feedback between nuclear waste disposal fees and the number of units that are uneconomic to operate. As the fees are increased (due, in part, to early nuclear plant retirements) the economics deteriorate for continued operation of the remaining plants, perhaps leading to additional early retirements. Our initial research suggests that this feedback loop could be strong enough in some scenarios to lead to the early shutdown of a substantial amount of nuclear generating capacity. In a situation where increasing disposal fees cause additional early retirements there would be tremendous political pressure to break the loop by violating one of the policy principles of the nuclear waste disposal program – either charging fees retroactively upon prior nuclear generation or obtaining funds from general tax revenues.

 

2. Conclusions and Policy Implications
   Perhaps, the most important set of considerations for immediate attention are those that are within the control of regulators and policy-makers. As the electric utility industry is deregulated, we will have to decide whether and to what extent specific generating technologies should be subsidized. As a general principle of competition, the owners of nuclear power plants should be required to bear their full costs, including accident risk, nuclear waste disposal, and the costs of dismantling the plants. The public policy implications are far-reaching, particularly through time.
   
   It appears, for example, that the one mill per kWh fee for spent fuel transportation and long-term storage may be inadequate to fund the program. At the same time, legislation has been proposed that would cap the fee at one mill. This could lead to a violation of the principle of "full cost recovery" of the Nuclear Waste Policy Act, "under which all costs related to the waste disposal services will be paid for by the owners and generators of SNF and civilian and defense high-level radioactive waste." If we wait to increase the fee, then we may find that at some future time an even higher increase is unavoidably necessary to cover the costs of the program but that the higher fee will lead to additional plant retirement decisions, undermining the economics of the program further. Indeed, it is likely that we are already beginning to face such a situation.
   
   Similarly, the current understanding of projected nuclear decommissioning costs poses a policy dilemma in a deregulated electricity market. Recovery of "catch-up" amounts through wires charges and other "stranded cost" charges imposes a special burden on electric customers in geographic areas historically served by nuclear power plants, potentially distorting market signals and dampening economic activity in those areas. While equity considerations may justify such charges for past consumption of nuclear energy, both fairness and efficiency militate against such subsidies for future consumption of nuclear power.
   
   Consequently, the advent of competition and deregulation will inevitably force policy-makers and regulators to face the "bottom-line" question that has troubled nuclear power from the earliest stages of its commercial development: who pays how much and for how long? The answer to this question will no doubt be controversial, but is nonetheless essential and unavoidable. While this paper does not provide the answer, we hope that it offers some useful insights into the nature and the magnitude of the problem.
6. References