National Academies Press: OpenBook
« Previous: Chapter 5: Disposal of Plutonium Without Irradiation
Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

6
Comparing the Options

In this chapter we compare the reactor-related options for disposition of weapons plutonium (WPu) using the criteria developed in Chapter 3. The array of options treated comprises: currently operating light-water reactors; currently operating heavy-water reactors; currently operating liquid-metal reactors; evolutionary light-water reactors; advanced light-water reactors; advanced liquid-metal reactors; modular high-temperature gas-cooled reactors; molten-salt reactors; particle-bed reactors; accelerator-based conversion systems; and immobilization with fission-product wastes. The emphasis in these comparisons is on the options that we and the parent committee have concluded offer the greatest promise of reducing the security hazards associated with surplus WPu over the next 30 years or so—the use of currently operating reactor types or evolutionary adaptations of them to incorporate WPu into spent fuel on a once-though basis, and vitrification of WPu with defense high-level wastes (HLW).

We begin by comparing the options on criteria related to security, turning then to economics and environment, safety, and health (ES&H). Many of the reactor-option characteristics that influence these evaluations depend to some degree on details of core design, fuel composition, and refueling modes and schedules that may vary considerably within a given reactor type. For example, given a pressurized-water reactor (PWR) with a core loaded two-thirds with low-enriched uranium (LEU) and one-third with mixed-oxide (MOX) fuel, the quantity and quality of plutonium in the spent fuel will depend in detail on the initial percentages of uranium-235 (U-235) and the various plutonium isotopes in the fresh fuel, on the burnup, and on the neutron-energy spectrum, which can

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

vary with core and fuel design, presence of burnable poisons, and so on (see Chapter 2).

The calculations used to determine spent fuel characteristics subject to these variations are fairly complicated and are performed with large, standardized computer codes and neutronics databases. We have not performed such calculations ourselves for this study, but have relied instead on the calculations performed for similar purposes over the past few years by reactor manufacturers and national laboratories. The quantitative comparisons in this chapter are based largely on the set of such calculations summarized in Table 6-1, which includes a range of reactor types-and, within types, ranges of fuel characteristics and operating modes-sufficient to illustrate the key dependencies and variations.1

The bulk of the comparison of options in this chapter focuses on options for disposition of U.S. WPu. This was inevitable, given the amount of relevant information available in the United States. Given the importance of disposition of plutonium in Russia as well, however, several paragraphs at the end of each section in this chapter are devoted to a preliminary discussion of how the options in Russia compare by the same criteria. The panel believes that additional study rigorously comparing plutonium disposition options in Russia against the criteria outlined in this chapter is urgently needed.

SECURITY COMPARISONS

As noted in Chapter 3, the primary motivation of the U.S. government in preparing to carry out disposition of excess WPu is to minimize the risks to national and international security posed by the existence of this material. Thus, options must meet this objective to be worthy of further consideration.

The panel was not charged with examining many of the issues related to security that are described in the report of our parent committee (NAS 1994), such as the interaction of WPu disposition with the future efforts to reduce nuclear arms and stem their spread. The panel takes note, however, of a number of important considerations outlined in the committee's 1994 report:

1  

The presentation of numbers in this table to three- or four-digit precision should not be taken either as indicative of the actual accuracy of the calculations, which is generally less, or as suggesting that small differences in these values are important, which generally they are not: the largely illusory precision in Table 6-1 is maintained simply to facilitate consistency checks and to permit associating the values in the table with particular calculations in the literature. Nor should it be assumed that the presence in the table of reactors designed by particular manufacturers constitutes a preference by the panel for these manufacturers' designs in comparison to other manufacturers' designs of the same general type; unless otherwise noted, the purpose of this specificity is simply to associate the tabulated parameters with the particular design, fuel type, and operating mode for which they were calculated.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-1 Plutonium Disposition Characteristics of Some Representative Reactor Types

Description

Burnup (MWd/ kgHM)

Metal Inventory (MT)

Metal Input (MT/yr)

Pu Input (kg/yr)

Pu Output (kg/yr)

Δ Pu (kg/ GWe-yr)

Pu/HM in Spent Fuel(%)

CLWR: 3,800-MWt/1,200-MWe PWR @ CF=0.75

100% LEU (3.8%)

42.0

99.2

24.8

0

253

+281

1.0

33% MOX (4.0%)

42.0

99.2

24.8

330

387

+63

1.6

100% MOX (4.0%)

42.0

99.2

24.8

996

636

-400

2.6

ELWR: 3,817-MWt/1,256-MWe PWR @ CF=0.75

100% MOX (6.8%)

42.2

99.2

24.8

1,672

1,215

-485

4.9

ELWR: 3,926-MWt/1,300-MWe ABWR @ CF=0.75

100% LEU (2.6%)

27 5

155.2

39.2

0

338

+347

0.9

100% MOX (3.0%)

37.1

155.2

29.0

867

562

-313

1.9

ALWR: 3,880-MWt/1,220 MWe APWRs (2 × 1,940-MWt/610-MWe) @ CF=0.75

100% MOX (5.5%)

40.0

133.8

26.6

1,462

1,038

-463

3.9

100% MOX (4.0%)

50.0

98.0

21.2

850

421

-469

2.0

CHWR: 5,664-MWt/1,538-MWe CANDUs (2 × 2,832-MWt/769-MWe) @ CF=0.80

100% natural U

8.3

234.7

199.4

0

768

+624

0.4

100% MOX (1.2%)

9.7

232.5

170.6

2,124

1,405

-584

0.8

100% MOX(2.1%)

17.1

226.0

96.8

2,000

1,333

-542

1.4

MHTGR: 3,600-MWt/1,717-MWe MHTGRs (6 × 600-MWt/286-MWe) @ CF=0.75

100% PuO/PuO2

595.5

4.7

1.7

1,656

589

-829

36

CLMR: 1,470-MWt/560-MWe BN-600 LMFBR @ CF=0.75

100% MOX (15.6%)

60.9

 

6.6

1,032

1,070

+90

16.2

ALMR: 4,200-MWt/1,440-MWe ALMR @ CF=0.75

100% U/Pu (10.6%)

75.2

113.0

15.3

1,626

1,781

+144

11.6

ABBREVIATIONS:

ABWR = advanced boiling-water reactor.

ALMR = advanced liquid-metal reactor.

ALWR = advanced light-water reactor.

APWR = advanced pressurized-water reactor.

CANDU = Canadian deuterium-uranium (reactor).

CF = capacity factor.

CHWR = current heavy-water reactor.

CLMR = current liquid-metal reactor.

CLWR = current light-water reactor.

ELWR = evolutionary light-water reactor.

GWe = gigawatt-electric.

HM = heavy metal.

kg = kilogram.

LEU = low-enriched uranium (figure in parentheses is U-235 enrichment).

LMFBR - liquid-metal fast-breeder reactor.

MHTGR = modular high-temperature gas reactor.

MOX = mixed-oxide fuel (figure in parentheses is Pu percentage of heavy metal).

MT = metric ton.

MWd = megawatt-day.

MWe = megawatt-electric.

MWt = megawatt-thermal.

PWR = pressurized-water reactor.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

NOTES to Table 6-1:

CLWR: These cases calculated for a nominal current PWR by Battelle Pacific Northwest Laboratories (Pritchard 1993); LEU case shows current practice for comparison.

ELWR: First case is ABB-Combustion Engineering System-80+ evolutionary PWR at 3.817 MWt (ABB-CE 1993); this case shows a plutonium loading higher than practical in current PWRs, making it possible for a single large reactor to load 50 tons of WPu in a 30-year operating lifetime. Second case is General Electric's ABWR, actually an evolutionary reactor in our terminology (GE 1993). First row parameters for use of LEU without MOX in this reactor scaled from American Physical Society study values for a nominal boiling-water reactor (APS 1978).

ALWR: Two Westinghouse PDR-600 APWRs (Westinghouse 1993). The second row illustrates a case using an advanced annular fuel design to achieve very high burnup and a 50-percent plutonium destruction fraction starting from 4-percent WPu MOX fuel; this destruction fraction is considerably higher than those attained with more typical LWR parameters.

CHWR: Two CANDU reactors of the Bruce Station type, as analyzed for conventional natural uranium fuel and “reference" and "advanced" MOX fuels by Atomic Energy of Canada, Ltd. (AECL 1994). CANDUs are given an advantage in capacity factor—an assumed 0.80 in contrast to the 0.75 assumed for the other reactor types listed—because unlike the others they do not need to shut down to refuel.

MHTGR: Six General Atomics 600-MWt gas-turbine MHTGR modules, as analyzed for 100%-percent PuO/PuO2 fuel by General Atomics (GA 1994). Plutonium weight fraction in spent fuel in this case refers to weight fraction of plutonium in plutonium plus higher actinides plus fission products.

CLMR: The CLMR is the BN-600 LMFBR operating in Russia, the largest of the world's three commercially operating LMRs. Its parameters using 100 percent MOX were calculated at Battelle Pacific Northwest Laboratories (Pritchard 1993); whether 100 percent MOX loading in this reactor is actually feasible is in doubt, however (see Chapter 4).

ALMR: Five GE 840-MWt ALMR modules, as analyzed for metallic plutonium-uranium-zirconium (Pu-U-Zr) fuel under the "spent fuel alternative" by General Electric (GE 1993). Figures given here for burnup (75.2 MWd/kgHM) and initial plutonium content in metallic fuel (10.6 percent) relate to heavy-metal content only (not including the zirconium in the ternary Pu-U-Zr fuel) and are averages for driver and blanket fuel assemblies. (Initial driver fuel plutonium content is 20.3 percent WPu in heavy metal.)

  • Current arms reductions agreements do not require the dismantlement of the nuclear weapons involved; nor do these agreements place any controls on the fissile materials these weapons contain. Dismantlement of these weapons and disposition of the resulting fissile material would significantly contribute to building confidence in the "irreversibility" of nuclear arms reductions, a goal enunciated by President Clinton and Russian President Yeltsin at their summit in January 1994. For disposition of WPu to contribute to this goal in the near term would require a clearly enunciated plan for disposition, with an expeditious beginning of implementation.

  • The foundation of international nuclear nonproliferation efforts is the Non-Proliferation Treaty. This treaty is based on a bargain between the nuclear-weapon states and the non-nuclear-weapon states, which included the nuclear-weapon states' commitment to negotiate in good faith toward nuclear disarmament. A clear commitment to disposition of excess WPu, with an early start on implementation, would contribute

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

to confidence that the United States and Russia were making good on this commitment.

  • The current transformations in the former Soviet Union create nuclear risks of three general kinds:

    • "breakup," meaning the emergence of multiple nuclear-armed states where previously there was only one;

    •  "breakdown," meaning erosion of government control over nuclear weapons and materials within a particular state; and

    • "breakout," meaning repudiation of arms-reduction agreements and pledges, and reconstruction of a larger nuclear arsenal.

It would be desirable not to prolong these risks as they apply to excess WPu any longer than necessary.

The panel notes that all of these considerations point to:

  1. timing as an absolutely critical part of minimizing security risks; and

  2. the importance of disposition of excess WPu not only in the United States, but in Russia as well.

These points are emphasized throughout the remainder of this section, in comparing the security impacts of different reactor options. We begin our comparisons of these security impacts with some broad generalizations, related to the various issues mentioned above and the potential threats considered in Chapter 3, that have been important to our process of narrowing the range of options given serious consideration. We then focus on the timing of different options, followed by discussions of the accessibility of the excess WPu during the course of the various disposition options and the degree of difficulty of recovering the excess WPu when they are complete.

General Considerations

Some general conclusions about the security dimensions of alternative approaches to disposition of WPu follow directly from consideration of the character of the threats likely to be of greatest concern (see Table 3-1). In particular:

  1. Risks of Storage. Prolonged storage of excess WPu in readily weapons-usable forms would mean a continuing risk of breakout, as well as of theft from the storage site. In addition, extended storage of large quantities of excess fissile materials, particularly in the form of weapon components, could undermine the arms reduction and nonproliferation regimes (the severity of this problem depending in part on the specific arrangements for custody of the materials in question). Thus, in judging the attractiveness of disposition options, we give heavy weight to (1) minimizing the time before processing of WPu can begin and (2) minimizing the subsequent time lag before disposition has succeeded in

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

reducing the accessibility of the last excess WPu (e.g., when the last WPu has been loaded into a reactor or a vitrifier). The timing for each disposition option is dependent on three factors: its technical readiness or uncertainty, the speed with which public and institutional approval (including relevant funding) could be gained, and the time required to implement it once developed and approved.

  1. Risks of Handling. Nearly all disposition options other than indefinite storage as pits require processing and usually transportation of plutonium, in ways that could increase access to the material and complicate accounting for it, thus increasing the potential for diversion and theft. The biggest risks of these kinds involve the steps before the WPu has been either irradiated in a reactor or mixed with radioactive wastes. In order to ensure that the overall process reduces net security risks, an agreed and stringent standard of security and accounting must be maintained throughout the disposition process, approximating as closely as practicable the security and accounting applied to intact nuclear weapons. The parent committee called this the "stored weapons standard." Hence, choices among long-term disposition options comparable in terms of timing should be weighted in favor of those that minimize:

    • any processing steps with high accessibility and low accountability;

    • the number of transport steps and the risks involved in each; and

    • the number of sites at which plutonium is handled and the risks at each site.

  1. Risks of Recovery. A third key security criterion for judging disposition options is the risk of recovery of the plutonium after disposition—by the state from whose weapons the WPu came (either covertly or overtly), or by potential proliferators (acquiring the material by covert theft, overt theft in the event of a loss of national authority, or overt forcible theft). Options that left the excess WPu substantially more accessible for weapons use than the global stock of plutonium in civilian spent fuel would mean that this material would continue to pose a unique safeguards problem indefinitely. Conversely, as long as the large stocks of plutonium in civilian spent fuel exist and continue to accumulate, options that made the excess WPu much less accessible than these larger stocks (for example, by eliminating it entirely or nearly so) would provide little additional security benefit, unless the same were done with the much larger stock of civilian plutonium. These considerations lead naturally to the "spent fuel standard" enunciated earlier. In any case, none of the disposition options that could plausibly be completed in less than 50 years would destroy more than

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

70-80 percent of the excess WPu—and whether the amount of WPu that remains in spent fuel (or vitrified waste) is 10 or 50 tons makes little difference to the overall security picture when the total stock of plutonium in spent fuel by that time will amount to over 1,500 tons.

  1. Indirect Impacts. The goal of long-term disposition of WPu should be not only to ensure that the plutonium from dismantled weapons is not reused in weapons, but also to avoid substantially increasing security risks from other fissile materials. Thus, policy-makers must be attentive to possible indirect effects that the choice of disposition options might have on the proliferation risks posed by other fissile materials in the world, in addition to its direct effects on the surplus weapon material. Disposition options that entail use of MOX fuel in, and/or fuel reprocessing for, civilian power reactors could potentially encourage the expanded use of these approaches in ways that increase the vulnerability of reactor plutonium (RPu) to diversion for nuclear weaponry (including acquisition of weapons by countries that possess no WPu). Conversely, it is possible that development of MOX fuel or reprocessing approaches for WPu disposition would lead to improvements in the safeguardability of these technologies (or an increase in society's determination to safeguard them), with beneficial results in current or future civilian nuclear-energy programs of which these technologies are a part. Policy-makers examining disposition options will have to take into account these possible indirect impacts of the options on the management of civilian plutonium, and how they fit with broader national fuel-cycle policies. For either reactor or vitrification options, if the United States wishes to maintain a policy of generally discouraging fuel cycles involving the use of separated plutonium, or if it wishes to make support for such cycles contingent on stringent safeguards and security measures, it will need to make a clear statement of how its choice fits within that broader context. It is important to note, in this context, that since the WPu is already separated, choice of a reactor option would not necessarily reopen the contentious question of reprocessing in the United States.

Timing

The issue of timing, which as we have argued above is an important aspect of security, has a number of dimensions: (1) the length of time until a disposition scheme can begin receiving and processing plutonium, (2) the length of time thereafter until the total quantity of surplus plutonium has entered the process, (3) the length of time from the start of operations until all of the plutonium has reached its final dispositioned state, and (4) the lengths of time that the plu

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

tonium spends in its various intermediate forms and locations during the disposition campaign (especially, of course, the most vulnerable forms and locations).

The first of these characteristic times—the length of time until disposition operations can commence—is important from the standpoint of sending an early "signal" that the commitment to remove the plutonium from the military inventory is really being carried out, as well as being a key element in the timing of the whole disposition process. The determinants of this critical length of time include: (1) the time needed to make a decision about how to proceed; (2) any time required for research, development, and design of elements of the scheme before they can be constructed at the needed scale; (3) the time needed to obtain the requisite permits and licenses, including time for any analyses required as input to the permitting and licensing processes; and (4) the time needed for construction and startup testing of any of the necessary facilities that do not already exist.

Some of these time periods can and should overlap: research and development on a variety of options can proceed in parallel with a process of decision about which option to select for deployment; construction on some elements of a disposition scheme can be underway while other elements are still under development; and licensing does not necessarily need to be complete before construction begins. (Overlap of construction with licensing can be the cause of costly mid-construction design changes, however, as experience with commercial nuclear reactors has demonstrated from time to time.) Estimation of the time requirements for the various steps that must precede commencement of disposition operations-and of the degree of overlap that can or will occur in these steps-is difficult, and the results will necessarily be approximate.

We present in Table 6-2 our estimates of the minimum plausible time requirements for the steps preceding commencement of disposition operations for a representative array of reactor-related options in the United States. The times specified in Table 6-2 for U.S. options are based on the assumption that (1) a U.S. government decision to proceed with a particular option is made early in 1996, and (2) the WPu disposition mission is given high national priority with corresponding resources-as the panel recommends be done.2

2  

We do not think that a decision to proceed with any of the advanced-reactor options could or should be made so quickly, but we chose to assume the same decision date for all the options in making these estimates in order to be able to compare, on an equal footing, the time requirements once a decision is made. The estimates given in Table 6-2 are in reasonably close agreement with those developed by the Fission Working Group Review Committee in the 1992-1993 Department of Energy Plutonium Disposition Study (Omberg and Walter 1993, Figure 5.1-8). We and the Fission Working Group Review Committee are more pessimistic about this timing than were the contractor studies commissioned in Phase I of DOE's Plutonium Disposition Study, where the contracts required (unrealistically in most cases, we think) that proposals be presented for completing the plutonium disposition mission by 2018.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-2 Minimum Plausible Time to Commencement of Disposition Operations

Option

Steps and Time Requirements

Vitrification with defense HLWa

Research on remaining process and repository issues: 1996-1998.

 

Fabrication and installation at Savannah River of plutonium metal-to-oxide conversion facilities 1998-2001

 

Fabrication, installation, and testing at Savannah River of suitable melters and other process equipment for adding plutonium to vitrification operations: 1999-2004.

 

PLUTONIUM CONVERSION TO OXIDE AND VITRIFICATION OPERATIONS COMMENCE 2005

CLWR one-third MOX. FMEFb

Contract for completion of MOX fabrication plant, do completion work, test and license plant: 1996-2000.

 

Choose LWRs to be used and negotiate arrangements, obtain needed permits and licenses: 1996-1999

 

PLUTONIUM CONVERSION TO OXIDE AT MOX PLANT COMMENCES 2001, MOX FUEL LOADING IN REACTORS COMMENCES 2002

CLWR full-MOX, FMEFc

Contract for completion of MOX fabrication plant, do completion work, test and license plant: 1996-2000

 

Choose LWRs to be used and negotiate arrangements, make any needed modifications, obtain needed permits and licenses: 1996-2001.

 

PLUTONIUM CONVERSION TO OXIDE AT MOX PLANT COMMENCES 2001, MOX FUEL LOADING IN REACTORS COMMENCES 2003

CLWR one-third or full-MOX, new MOX plantd

Contract for construction of MOX fabrication plant, site, construct, test, and license plant: 1996-2002.

 

Choose LWRs to be used and negotiate arrangements, obtain needed permits and licenses: 1996-1999/2001

 

PLUTONIUM CONVERSION TO OXIDE AT MOX PLANT COMMENCES 2003, MOX FUEL LOADING IN REACTORS COMMENCES 2004

ELWR full-MOX, FMEFe

Contract for completion of MOX fabrication plant, do completion work, test and license plant 1996-2000

 

Select site and contractors for ELWR, construct, test, and license plant: 1996-2004.

 

PLUTONIUM CONVERSION TO OXIDE AT MOX PLANT COMMENCES 2001, MOX FUEL LOADING IN REACTORS COMMENCES 2005

ALWR full-MOX, new MOX plantf

Contract for construction of MOX fabrication plant, site, construct, test and license plant: 1996-2002

 

Complete development and design work on ALWR, select site and contractors, construct, test, and license plant: 1996-2007.

 

PLUTONIUM CONVERSION TO OXIDE AT MOX PLANT COMMENCES 2003, MOX FUEL LOADING IN REACTORS COMMENCES 2008

MHTGR or ALMRg

Complete development and design work on reactors and fuel cycles, select site and contractors for co-located fuel fabrication facility and reactors, construct, test, and license: 1996-2012.

 

PLUTONIUM CONVERSION TO OXIDE COMMENCES 2010, MOX FUEL LOADING IN REACTORS COMMENCES 2013

MSR, PBR, or ABCh

Extensive development and design work needed for these concepts likely to add 5-10 years to time scales for MHTGR and ALMR.

 

FUEL LOADING IN REACTORS COMMENCES BETWEEN 2018 AND 2023

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

NOTES to Table 6-2: All estimates assume a U.S. government decision at the beginning of calendar 1996 to proceed with the option. Elements affecting total time are entered in boldface. Estimates are highly uncertain, for reasons discussed in the text.

a Assumes addition of WPu at 1-2 percent by weight to glass logs incorporating 20 percent by weight defense HLW, as now scheduled to begin production (without WPu) at Savannah River in the late 1990s.

b Current Light-Water Reactor with 1/3 MOX cores; fuel produced in Fuel Materials Examination Facility (FMEF) at Hanford, Washington.

c Same as previous case, except that it is assumed that CLWRs can be licensed to use full-MOX cores. This reduces the number of reactor sites that need to be chosen, agreed, and permitted, perhaps shortening the time required for these steps, but may increase the time needed for reactor licensing for MOX use and related analysis and testing. That time could increase from four to six years, as indicated, without changing the overall start dates from those applying in the previous case. Options involving completing existing reactors such as the WNP facilities are considered to be included in the CLWR options, as provision of the needed MOX capacity would in all probability still be the rate-limiting step.

d Same as previous two cases except fuel produced in a new MOX plant, which adds at least two years to this pacing element.

e Same as second case except uses newly constructed evolutionary LWRs instead of existing ones, with time estimated for final design, siting, construction, testing, and licensing at nine years. Reactor rather than MOX plant is now pacing element; use of new MOX plant rather than FMEF would delay commencement of conversion to oxide by two years but would not affect start of MOX loading in reactors.

f Same as previous case except uses advanced rather than evolutionary LWRs, assumed to require an additional three years of design and development time. We assume that on the resulting longer overall time scale it will be decided to employ newer MOX fabrication technology in a new plant, rather than updating FMEF. Use of FMEF would permit starting conversion to oxide two years sooner but would not affect start of MOX loading in reactors.

g Modular High-Temperature Gas-cooled Reactor or Advanced Liquid Metal Reactor. We rate these two reactor types as comparable in the amount of remaining development and design effort needed before they could be brought into operation using plutonium fuel. The indicated operation date of 2013 is based on assuming a serious national commitment to achieving this, and thus is optimistic. The associated fuel fabrication facility assumed to be completable and licensable at least three years sooner than reactor, to permit accumulation of fuel for first full core by the time reactor is otherwise ready for operation.

h Molten-Salt Reactor, Pebble-Bed Reactor, or Accelerator-Based Convertor.

We emphasize that the dates shown in Table 6-2 reflect what we think is possible, which is not necessarily the same thing as what is likely. In recent years the Department of Energy (DOE) has not had great success in carrying large projects of this kind through to completion; success in this case will require intense, high-level commitment and oversight. In the current institutional environment in the United States, moreover, delaying or halting large nuclear projects is substantially easier than carrying them out on schedule.

There is little doubt that the large-scale processing of plutonium required for virtually any disposition option will engender controversy and that any option will face opponents. At each stage of the lengthy political and regulatory process that will be required to implement any of these options, there are likely to be efforts to block or delay the process, through lobbying the relevant legislatures and regulatory bodies and legal actions in the courts. If successful, such

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

actions could delay implementation of disposition options for years. (Licensing issues could be particularly important in governing how rapidly options can be implemented and are summarized in the appendix to this chapter.) In short, the time estimates in Table 6-2 are optimistic: while it is extremely unlikely that the disposition options considered here could be carried out more rapidly, it is very possible, even likely, that they will ultimately be carried out more slowly.

Timing of plutonium disposition in Russia is even more uncertain than timing in the United States. The most critical limiting factor on timing in Russia is likely to be the availability of the financial resources needed to carry out the job (see discussion of economics below). As in the United States, making use of existing facilities to the extent possible would speed the process.

In the Russian context, if appropriate support and resources were available, the timing of the vitrification option might be more favorable than in the U.S. context, as Russian vitrification of HLW is already ongoing, new melters are regularly replacing older ones (so that a critically safe plutonium disposition melter could be inserted into the process without undue delay), and regulatory obstacles may be less substantial. To incorporate substantial quantities of plutonium safely, however, would probably necessitate switching from phosphate to borosilicate glasses, which could impose significant delays.

In the case of the light-water reactor (LWR) options, Russia, like the United States, does not have a production-scale operational MOX fabrication facility, but has a partially completed one; Russia's facility is farther from completion, but would have a larger capacity than the Fuel Materials Examination Facility (FMEF). Russia does have two pilot-scale MOX fabrication lines, and therefore might be in a position to carry out initial demonstrations of MOX use in LWRs more rapidly than could the United States. A new MOX facility could probably be built from scratch in Russia on a time scale comparable to that in the United States, assuming that the job had adequate resources and governmental priority—but even more than in the United States, that is a very large assumption, given the many other urgent problems the Russian government must address. (Foreign participation, as has been suggested by some European MOX manufacturers, could potentially help keep such an effort on track.)

As indicated in Chapter 4, Russian VVER-1000 reactors are similar to U.S. LWRs in their adaptability to MOX burning. Regulatory issues in Russia are difficult to predict, but could potentially impose smaller delays, because a legal process allowing multiple opportunities for opponents of a project to intervene does not yet exist to the same degree as it does in the United States. In short, the timing arguments as between the two main options discussed in this report— vitrification and use as MOX in LWRs—do not appear to be radically different in the Russian case than they are in the U.S. case.

Russia's one operating liquid-metal reactor (LMR), the 560-megawatt-electric (MWe) BN-600, would require 50 years' operation to process 50 tons of WPu on a once-though basis (see Table 6-1), even assuming that it could be

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

modified to use 100-percent MOX fuel. (As noted in Chapter 4, Russian Ministry of Atomic Energy (MINATOM) officials have expressed the opinion that such modification is not practical for this reactor.) The reactor has already been operating for about 15 years, moreover, which means that its remaining operating lifetime may be less than half of the 50 years needed to disposition 50 tons of WPu. As noted in Chapter 4, the record of the BN-600 and its smaller predecessor, the BN-350, raises doubts about its reliability if not its safety. In sum, its suitability for the disposition mission is questionable, and under the most optimistic of assumptions the timing with which it could perform the mission would still be unattractive.

The timing implications of choosing more advanced reactor types than current LWRs and current LMRs are similar in Russia to those in the United States: that is, the use of advanced reactors not yet ready for immediate construction would delay accomplishing the mission considerably. As noted in Chapter 4, Russian officials have argued that because Russia has tested plutonium in LMRs and has not yet explored plutonium use in LWRs, the option of completing several large LMRs and using the excess WPu in them could be accomplished more quickly than the LWR option. But LWRs have the considerable advantage of existing already, and the extensive experience of LWR MOX use in other countries would allow an LWR MOX option to be implemented rapidly in Russia, just as in the United States. The panel believes that storing the WPu until new LMRs can be financed and constructed in Russia would needlessly delay achievement of the spent fuel standard in the protection of this material, substantially extending the period of high security risks—direct and indirect—associated with storage of WPu in forms readily usable in weapons. Proposals from some other Russian analysts to save the excess WPu until other types of advanced reactors became available—such as high-temperature gas-cooled reactors (HTGRs) or lead-cooled liquid metal reactors—would create even larger delays.

In the case of use of reactors in other countries for disposition of either U.S. or Russian WPu, the timing might differ in some respects, but it is difficult to say whether the overall result would be a faster process, a slower one, or no change:

  • If U.S. plutonium were to be loaded into European reactors already licensed to use one-third MOX cores, this would circumvent the time requirement for licensing U.S. reactors for such fuel; but it would add a time requirement for negotiating the terms of such an arrangement. The pacing element in the use of existing reactors for WPu disposition might still be the licensing and construction or upgrading of MOX fuel fabrication facilities rather than reactor licensing, unless this fabrication took place in existing or planned European plants. Depending on the size of the European MOX fabrication capacity in relation to civilian

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

MOX fabrication demand in this period, this might require the negotiation of arrangements and terms for the displacement, in European MOX fabrication plants, of the recycled civilian plutonium they would otherwise be using.3 This approach also would require overcoming political, environmental, and safeguards objections to transoceanic international shipment of U.S. WPu. If the MOX fabrication did not take place in the existing or planned European plants, then the timing would be governed by the requirements of fabricating it in the United States—hence would not differ from that shown in Table 6-2 for the one-third MOX and full-MOX current LWR options in this country—or the requirements of siting, constructing, and building new MOX plants in Europe, which even in the absence of any controversies could not occur much more rapidly than the optimistic U.S. schedule assumed in Table 6-2.4

If U.S. WPu were to be loaded into CANDU (Canadian deuterium-uranium) reactors in Canada, the MOX fuel fabrication for these reactors would probably take place in the United States. Thus this pacing element of the timing would not change from the CLWR options shown in Table 6-2, and the overall timing would be the same—assuming that Canadian licensing of MOX use in CANDUs, and the negotiation of the terms under which this would be done, would require about the same amount of time as the corresponding negotiations and licensing activities in the United States.

Concerning the possibility of loading Russian WPu into European or Canadian reactors, the timing would hinge both on the time needed to reach agreement on terms and arrangements for the transfer of Russian weapon material to other countries and on the time needed to accomplish the MOX fabrication step. The latter, if done in Europe, would be subject to the same constraints and possibilities described above in connection with making MOX from U.S. WPu in Europe; if done in the United States or Russia, it would be subject to the timing constraints on deploying new MOX capacity in those countries.

In summary, the most plausible way in which an international variant might offer timing better than that of the U.S. current light-water-reactor scenarios

3  

If MOX fabrication using civilian plutonium had to be postponed, that would mean either that the separated civilian plutonium previously scheduled for MOX fabrication would accumulate at reprocessing plants (meaning that accelerated shrinkage of the stock of separated military plutonium would be achieved only at the cost of creating an equivalent addition to the stock of separated civilian plutonium) or that scheduled civilian reprocessing operations would have to be deferred (increasing the European stock of—and hence the need for storage capacity for—civilian spent fuel).

4  

Controversies are more likely to delay construction and licensing of a new MOX plant in the United States than to delay such an effort in Europe, so the chances of actually meeting the schedule indicated in Table 6-2 might well be better in Europe.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

depicted in Table 6-2 would be through the use of existing or planned European capacity for MOX fabrication, avoiding the time-consuming completion or new construction of MOX fabrication in the United States or Russia. Whether this approach would really save time would depend on whether the needed international agreements on transfer and custody of the WPu, together with any needed technical and economic arrangements for displacing the civilian plutonium scheduled to be fabricated in these MOX plants, could actually be concluded in less time than it would take to complete new MOX fabrication capacity in the United States or Russia.

Table 6-3 presents working scenarios for the timing of disposition following commencement of operations for those of the options in Table 6-2 that could begin operating before 2010, assuming a high national priority for this mission. (Later-starting options are of less interest, for reasons adduced above, and any scenarios we might devise for the phasing of their introduction would be even more speculative than those we offer for the nearer-term possibilities.) Our scenarios encompass a range of assumptions about the number and scale of disposition facilities, representing a variety of possible choices concerning the trade-offs among cost, start date, and completion date of the campaign. Table 6-4 shows, for these scenarios, the key dates and the post-2000 integrated inventories of WPu in the two most intrinsically vulnerable forms relevant to these disposition schemes: (1) metal in pits and (2) plutonium oxide as powder or MOX, before loading into reactors or HLW-bearing logs.5

The main conclusions we draw from these considerations of the interaction of timing with security can be summarized as follows:

  1. Timing of all the WPu disposition options is highly uncertain, in large part because of the substantial controversies likely to attend any WPu disposition option in the United States. Estimates of the time required to accomplish major nuclear projects in the United States have historically tended to be optimistic. Thus, judgments about the absolute value of start dates and completion dates, and even about the relative timing of different options, must be considered somewhat tentative. These estimates are based on qualitative judgments, not on detailed design studies or detailed analysis of the political and regulatory processes required to implement these options.

5  

We have assumed that the dismantling of surplus nuclear weapons containing the nominal 50 tons of plutonium has been completed by the end of the year 2000, and we begin the calculation of the integrated inventories at the beginning of the year 2001. Since none of the disposition options we are considering would begin the processing of pits before 2001, their integrated inventories of plutonium in pits will be identical up to that point and therefore irrelevant to comparisons. Conversion of pits to oxide is assumed to occur in connection with, and on the same time scale as, the MOX fuel fabrication or vitrification process.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-3 Timing of Disposition Following Commencement of Operation

U.S. Disposition of U.S. Weapons Plutonium

Vitrification with defense high-level waste (HLW): 1.3 weight percent plutonium, eight years. The indicated figures, suggested by preliminary analysis at Savannah River, should be considered illustrative. If a loading of 1.3 percent WPu in borosilicate glass is feasible with respect to control of criticality in glass manufacture and with respect to repository criticality, and if vitrification operations incorporating WPu commence in 2005 (Table 6-2), then addition of WPu at this loading to 2,200 logs scheduled for manufacture at the Savannah River site between 2005 and 2013 would account for the 50 tons of WPu.

CLWR, one-third MOX FMEF: six reactors, 25 years. We assume, for the United States, use of 3,800-MWt LWRs @ CF = 0.75 and burnup of 42 MWd/kg with 4.0 percent plutonium in one-third MOX cores, implying 8.3 tons MOX and 330 kg plutonium per reactor-year (kgPu/RY); FMEF MOX capacity is 50 tons of heavy metal per year (MTHM/yr), thus feeds six such reactors in steady state. First MOX batch requires 33 tons MOX per reactor, hence 200 tons MOX for six reactors. At 50 MTHM/yr, FMEF makes a one-third core MOX batch for one reactor every eight months. Assuming four months' lag between completion of a batch at FMEF and completion of loading of that batch at a reactor, start of FMEF operation on January 1, 2001, implies completion of loading a one-third MOX core at the first reactor on January 1, 2002. FMEF has finished fabricating 50 tons of WPu into MOX on December 31, 2025, and the last one-third MOX reload into a reactor takes place on May 1, 2026.

CLWR, full MOX, FMEF: two reactors, 25 years. Same reactor parameters as above but with full MOX enable two reactors to load the 50 tons of WPu in 25 years; first cores contain 100 tons of MOX, so first reactor can start May 1, 2003, after January 1, 2001, start of MOX plant operation and second reactor can start May 1, 2005; as in the first case, the last of the 50 tons of WPu will be transformed into MOX in 2025 and loaded into a reactor May 1, 2026.

CLWR, full MOX FMEF: two reactors, 15 years. Same reactor parameters as above but with 6.8 percent plutonium in heavy metal. Start dates are as in previous case, but shorter campaign duration resulting from higher plutonium loading leads to last of WPu being loaded in 2016.

CLWR, new MOX plant: 12 reactors at one-third MOX or 4 reactors at full MOX, 12.5 years. Same reactor parameters but twice as many reactors as preceding two cases enable loading all of the 50 tons of WPu in the 12.5 years following MOX plant startup if a new MOX plant with capacity of 100 MTHM/yr is employed (this option trades increased cost and dispersion of disposition operations for shortened campaign duration); if new MOX plant commences operation in 2003 (Table 6-2), the last of the 50 tons of WPu will be transformed into MOX in 2015 and loaded into reactors in 2016.

ELWR, full MOX, FMEF: one reactor, 30 years. If licensing and political considerations lead to decision to build a new reactor at a U.S. government site, the fastest starting of the ELWR options examined in DOE's Plutonium Disposition Study would be an ABB-Combustion Engineering (ABB-CE) System-80+ or a General Electric (GE) ABWR, and minimizing plutonium transport would call for building it at the Hanford site (assuming FMEF provides the MOX fuel fabrication). Minimizing government investment favors building only one reactor and processing the 50 tons of plutonium in its 30-year lifetime (e.g., at 6.75 percent plutonium and 42.2 MWd/kg for an ABB-CE System-80+ with 25 MTHM/yr MOX fabrication, or 3.3 percent plutonium and 21.5 MWd/kg for a (GE ABWR with 50 MTHM/yr MOX fabrication). MOX fabrication timing could be the same as for preceding FMEF cases, but reactor would not be ready for loading fuel until 2005 (Table 6-2) and last of WPu MOX would not be loaded until about 2030.

ALWR, full MOX, new MOX plant: four 610-MWe reactors, 17 years. If it were decided to use the WPu mission in order to demonstrate a more advanced LWR technology than the evolutionary reactors considered in the previous case, one might (for example) build four of the Westinghouse PDR-600 ALWRs and use them in conjunction with a 50-MTHM/yr MOX plant making 5.5 percent enriched fuel, which at a burnup of 40 MWd/kgHM would process the plutonium in about 17 years of operation. The MOX plant could commence operation in 2003 and finish in 2020, and the reactors would load MOX fuel starting in 2008, would load the last of it in 2021.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

International Variations

Russian CLWR, one-third MOX. In Russia, use of the four 3,130-MWt VVER-1000 PWRs suitable for MOX operation at the same burnup, plutonium enrichment, and capacity factor assumed here for the U.S. LWRs (loading 272 kgPu/RY in one-third MOX cores) would require almost 50 years to load 50 tons of WPu (probably longer than the life expectancy of these reactors); if four currently unfinished VVERs can be completed by 2000, then the total of eight could finish loading the 50 tons WPu by about 2025. If some of the older VVERs-those already operating-had to be retired at, say, age 30, the campaign's duration would increase correspondingly.

Russian CLWR, full MOX. If use of full-MOX cores with 4.0 percent WPu and burnup to 42 MWd/kgHM is feasible in VVER-1000 PWRs, then four of these could load 3.28 tons of WPu per year, and the nominal 50 tons of surplus Russian WPu could be loaded in about 15 years (given a MOX fuel fabrication capacity of 100 MTHM/yr).

Russian CLMR, full MOX. If the one operational Russian LMR, the 560-MWe BN-600, could be modified to use a full-MOX core, it would need 50 years to load the nominal 50 tons of surplus Russian WPu.

CANDU reactors, 100-percent MOX cores, FMEF. At a burnup of 9.7 MWd/kgHM, and CF = 0.80, a 2,832-MWt CANDU reactor loads about 85 MTHM/yr, containing 1,062 kgPu/yr at 1.2 percent, and two such reactors load 170 MTHM/yr containing 2,124 kgPu/yr. If the fuel is fabricated at the U.S. FMEF at a capacity of 170 MTHM/yr (which Atomic Energy of Canada, Limited, analyses indicate is practical given the simplicity of CANDU fuel), and the FMEF begins operation at this level in 2001, then the first CANDU could be loaded with a full-MOX core by the beginning of 2002 and the second by the beginning of 2003, and the last of the 50 tons of WPu would be fabricated into fuel in 2024 and loaded into the reactors in 2025.

  1. The panel's best estimate of the minimum time requirements for accomplishing various disposition options indicates that, of all the disposition options considered in this report, the spent fuel options utilizing power reactors of currently operating types would have the earliest plausible start dates for conversion of pits to oxides and for the most important vulnerability-reducing step (loading into reactors in the case of reactor options, loading into melters for vitrification). The difference in start date between this option and the vitrification option, however, is smaller than the uncertainties in estimates of the start dates for these options.

  2. Within the category of reactor options using reactors of currently operating types, the specific options with the earliest plausible start dates are (a) those that use existing or planned European MOX fabrication capacity (assuming rapid conclusion of any needed international agreements and arrangements for displacing civilian plutonium from these MOX plants) and (b) those that rely on completing the existing FMEF at Hanford for MOX fabrication.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-4 Key Dates and Integrated Inventories for Disposition Scenarios

 

Plutonium Conversion

Plutonium Loading:a

Integrated Inventoryb (ton-yr Pu) as:

Option

Start

End

Start

End

Pits

Oxidesc

VITRF

2005

2013

2005

2013

425

13

CLWRa

2001

2025

2002

2026

625

33

CLWRb

2001

2025

2002

2026

625

38

CLWRc

2003

2015

2004

2016

413

25

CLWRd

2003

2015

2004

2016

413

30

CANDU

2001

2024

2002

2025

600

33d

ELWR

2001

2030

2005

2032

625

79

ALWR

2003

2019

2008

2021

525

66

a Loading means loading into a reactor or a melter.

b Integrated inventories are computed from the year 2001 (see text).

c Oxides refer here to plutonium oxides before loading into a reactor or melter.

d Could be smaller, depending on how the continuous refueling capability of the CANDU is used.

Key to options and integrated-inventory calculations:

VITRF = Vitrification with defense HLW in the Defense Waste Processing Facility at Savannah River, with illustrative parameters given in Table 6-3. Plutonium is converted from metal to oxide and incorporated into glass at 5.6 tons/yr for 9 years. Pit integrated inventory is 50 tons × 4 yr (2001-2004) plus 1/2 × 50 tons × 9 yr (2005-2013) = 425 ton-yr. We assume an oxide stock equal to three months' input to melter will be maintained at the site, hence oxide integrated inventory is 5.6 tons / 4 × 9 yr (2005-2013) = 13 ton-yr.

CLWRa = Six current 3,800-MWt LWRs with one-third MOX, FMEF MOX fabrication. Pit integrated inventory is 1/2 × 50 tons × 25 yr (2001-2025) = 625 ton-yr of metal. Assuming that MOX fabrication plant runs 1 yr accumulating 50 MTHM before loading first 33-MTHM one-third MOX core and keeping 4-month (= 17 MTHM) minimum reserve, MOX integrated inventory as heavy metal is (1 yr × 25 tons) + (23.67 yr × 33 tons) + (0.33 yr × 25 tons) + (0.33 yr × 33 tons)= 825 MTHM-yr, × 0.04 MTPu/MTHM = 33 MTPu-yr.

CLWRb = two current 3,800-MWt LWRs with full MOX, FMEF MOX fabrication. Pit integrated inventory as in CLWRa. Assuming that MOX fabrication plant runs 2.33 yr accumulating 117 MTHM before loading first 100-MTHM full-MOX core and keeping 4-month (= 17 MTHM) minimum reserve, MOX integrated inventory as heavy metal is (2.33 yr × 58.5 tons) + (2 yr × 67 tons) + (20.33 yr × 33 tons) + (0.33 yr × 33 tons) = 959 MTHM-yr, × 0.04 MTPu/MTHM = 38 MTPu-yr.

CLWRc = Twelve current 3,800-MWt LWRs with one-third MOX, new MOX fabrication plant. Pit integrated inventory is 50 tons × 2 yr (2001-2002) + 1/2 × 50 tons × 12.5 years (20032015) = 413 ton-yrs. Assuming that MOX fabrication plant runs two-thirds of a year accumulating 66 MTHM before loading first 33-MTHM one-third MOX core and keeping 4-month (= 33 MTHM) minimum reserve, MOX integrated inventory as heavy metal is (2/3 yr × 33 tons) + (11.67 yr × 50 tons) + (0.33 yr × 42 tons) + (0.33 yr × 17 tons) = 620 MTHM-yr, × 0.04 MTPu/MTHM = 25 MTPu-yr.

CLWRd = Four current 3,800-MWt LWRs with full MOX, new MOX fabrication plant. Pit integrated inventory is as in CLWRc. Assuming that MOX fabrication plant runs 1.33 years accumulating 133 MTHM before loading first 100-MTHM full-MOX core and keeping 4-month (= 33 MTHM) minimum reserve, MOX integrated inventory as heavy metal is (1.33 yr × 66 tons) + (3 yr × 83 tons) + (8 yr × 33 tons) + (0.33 yr × 42 tons) + (0.33 yr × 17 tons) = 755 MTHM-yr, × 0.04 MTPu/MTHM = 30 MTPu-yr.

CANDU = Two 769-MWe CANDU with full MOX at 1.2 percent WPu, FMEF MOX fabrication. (According to Atomic Energy of Canada, Limited, relative simplicity of CANDU fuel makes it possible for FMEF to meet the 170-MTHM/yr fabrication requirement for this scenario.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

Rate of plutonium processing into fuel is 2,124 kg plutonium per year, so the fabrication operation runs for 24 years from 2001 and the integrated inventory in pits is 1/2 × 50 tons Pu × 24 years = 600 ton-yr. If the average working inventory of fabricated MOX fuel before loading into the reactors is one half core's worth per reactor, hence one full core's worth altogether, the associated integrated inventory of plutonium in oxide is 116 MTHM × 0.012 MT WPu per MTHM × 24 yr = 33 ton-yr.

ELWR = One evolutionary 3,900-MWt LWR with full MOX (GE ABWR at 21.5 MWd/kgHM and 3.3 percent plutonium), FMEF MOX fabrication at 50 MTHM/yr. Assuming the MOX plant is operated as soon as possible (in order to begin the disposition process as soon as possible, even though the reactor will not be ready to receive the fuel as rapidly as the MOX plant can produce it), the pit integrated inventory will be 625 MTHM-yr as in the other FMEF cases. The oxide inventory ramps from zero at the beginning of 2001 to 200 MTHM at the end of 2004 just before fuel loading into the reactor begins in 2005, so the contribution to the integrated inventory from this period is 1/2 × 200 tons × 4 yr = 400 MTHM-yr; then the inventory drops to about 50 MTHM as the first full core is loaded, and for the next 26 years this 50 MTHM is a base or reserve above which the MOX plant's annual output ramps an additional inventory from zero just after a reload to 50 MTHM just before a reload; thus the average inventory during this 26 years is 75 MTHM in oxide and the contribution to the integrated inventory is 1,950 MTHM-yr; at the end of one more year the last 50 tons are loaded, so this year contributes another 50 MTHM-yr, and we have altogether an integrated inventory of 400 + 1,950 + 50 = 2,400 MTHM-yr, which at 0.033 MTPu/MTHM is 79 MTPu-yr.

ALWR = Four advanced 610-MWe LWRs with full MOX, new MOX fabrication plant at 53 MTHM/yr output. The integrated inventory in pits is 50 tons × 2 yr (2001-2002) + 1/2 × 50 MTHM × 17 yr (2003-2019) = 525 MTHM-yr. The oxide inventory accumulates at 53 MTHM per year from the beginning of 2003 to the end of 2007 just before the reactors are ready to receive fuel, contributing 1/2 × 5 yr × 265 MTHM = 663 MTHM-yr to the oxide integrated inventory; from 2008 through 2019 the working inventory ramps from 0 to 53 tons each year, averaging 26.5 MTHM, and allowing for a 4-month reserve of 18 MTHM makes the integrated-inventory contribution for this period 45 MTHM × 12 yr = 540 MTHM-yr; the total integrated inventory of heavy metal in oxide is then 663 + 540 = 1,203 MTHM, which at 0.055 MTPu/MTHM is 66 ton-yr plutonium.

  1. Use of full-MOX as opposed to one-third MOX cores in the current LWR options, to the extent that it is practical, would reduce the number of reactors needed for the disposition campaign, but it would not significantly affect the start dates (unless demonstrating full-MOX capability, or attaining it through modifications, added more than the two years assumed here to the reactor timetable); and it would only shorten the duration of the campaign, once started, if MOX fuel fabrication capacity were expanded beyond what has been assumed here. Integrated inventories are not very sensitive to the choice of full-MOX versus one-third MOX cores (although of course the geographic dispersion of the inventories is likely to be sensitive to this choice).

  2. Construction of a new MOX fabrication plant with twice the capacity of FMEF (or expansion of FMEF to such higher capacity, which we assume would take a comparable length of time) would delay start dates for the current LWR options but could accelerate completion dates by about a decade (given the use of the doubled number of reactors that a doubled MOX capacity would permit). This choice would also reduce

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

the post-2000 integrated inventory of pits by some 30 percent (the faster drawdown more than compensating for the later start date).

  1. Use of commercial CANDU reactors with full-core MOX produced at FMEF would provide the possibility of carrying out disposition of all of the nominal 50 tons of WPu in two reactors over a period of about 25 years, with start dates, completion dates, and post-2000 pit integrated inventory about the same as those of the CLWR options (assuming that the bilateral negotiations needed to arrange this option could be completed reasonably quickly).

  2. Construction of a new evolutionary LWR for WPu disposition, as might be done if political difficulties preclude using an already operating or now partly completed LWR, would also permit loading all of the 50 tons of WPu into a single reactor during its lifetime, at the cost of modestly delaying the start date for plutonium loading, delaying by about five years the completion of the campaign, and significantly increasing the pre-load oxide integrated inventory.

  3. Use of advanced LWRs would entail the latest start date for plutonium loading of any of the options besides other advanced reactors, and it would not offer compensating advantages over current- and evolutionary-LWR cases in completion date or integrated inventories.

  4. The panel's best estimate of the earliest plausible start date for the vitrification option is a few years later than our estimate of the earliest plausible start date of the near-term reactor options, although as noted earlier this difference is smaller than the large uncertainty in estimating when either of these classes of options could actually begin. The vitrification option could have a completion date earlier than any of the reactor options (excluding implausibly high MOX fabrication capacities). Under the parameters postulated here, it would also have a post-2000 pit integrated inventory equal to the best of the reactor options, and a pre-load oxide integrated inventory better than any of them.

  5. Advanced reactors other than ALWRs—including MHTGRs, ALMRs, MSRs, PBRs, and ABCs—would have such a severe disadvantage in start dates, completion dates, and post-2000 pit integrated inventories that we rate them as unacceptable in their timing for the near-term WPu-disposition mission; choosing one of these options would needlessly prolong the hazards of storing excess WPu in readily weapon-usable form.

  6. The timing for use of analogous options for disposition of Russian plutonium is not likely to be significantly more favorable than that summarized in the preceding numbered subparagraphs and in Table 6-4, and could easily be worse.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

For all reactor options, there is an important trade-off to be made between the number of reactors and sites involved and the completion time of the disposition campaign. Using WPu in a large number of reactors at several sites would substantially reduce the amount of time required to complete disposition, with the associated security benefits described at the outset of this section. In an extreme case, for example, if plutonium fuel could be produced for all the reactors in the United States, the excess WPu could be completely loaded into reactors within a few months, rather than a few decades, after the operation began. On the other hand, approaches involving many sites would also have important disadvantages: (1) there would be more plutonium transportation steps (probably the point in the disposition process when the material is most vulnerable to forcible theft, as described below), and more sites at which the plutonium would have to be secured and accounted for; (2) a larger MOX fabrication capacity would be needed to fabricate more plutonium into fuel in a shorter time (possibly requiring more time to build), and this capacity would then be idled (if it were located in the United States) as soon as the WPu disposition campaign was complete; (3) political and licensing issues would have to be overcome at a larger number of locations; and (4) symbolically, if the United States chooses to continue to generally oppose the use of fuel cycles involving separated plutonium, involving a larger fraction of the U.S. nuclear industry in the use of plutonium would tend to counteract this message.

Ultimately, choices balancing these advantages and disadvantages of increasing the number of sites will have to be based on educated judgment; the panel is not aware of any defensible means by which these advantages and disadvantages can be quantified and rigorously compared. While it is true that thousands of assembled nuclear weapons are transported each year in the United States and the former Soviet Union, and that plutonium shipments in Europe are commonplace, it is nevertheless the panel's judgment that the advantages of limiting plutonium disposition to one or two sites outweigh the timing disadvantages of doing so-particularly as, in the United States, Russia, and some other countries, sites exist with several reactors at a single location, offering the possibility of limiting the time required for plutonium disposition while simultaneously keeping the number of separate sites to a minimum. This judgment is generally reflected in the assumptions used in Table 6-3.

Other Indices, Barriers, and Threat-Barrier Interactions

Some of the security-related characteristics of plutonium in different forms are summarized in Table 6-5. These characteristics have been used, together with the definitions and criteria presented in Chapter 3 (see “Specific Security Concerns and Threat Characteristics" and "A Matrix Scheme for Characterizing Options"), to arrive at the set of characterizations presented in Table 6-6 of in

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-5 Some Security-Related Characteristics of Plutonium in Different Forms

 

 

 

 

 

 

Gamma Dose Rate(rem/hr):

Form of Plutonium

Mass per Item (kg)

Max Item Dim (cm)

Pu per Item (kg)

Pu Conc (kg/kg)

At Surface

At 1 meter

Note

Intact pit (WPu metal)

ca 4

ca 15

ca 4

1

0.85

0.005

1

RPu metal sphere, δ-phase

6

9

6

1

17

0.03

1

PuO2 powder, WPua

(powder @ I g/cm3)

0.88

1

0.009

2

PuO2 powder, RPub

(powder @ I g/cm3)

0.88

20

0.2

2

MOX fuel pellet, WPu

0.006

1

3 × 10-4

0.05

0.05

1 × 10-6

3

MOX fuel pellet, RPu

0.006

1

3 × 10-4

0.05

1

2 × 10-5

3

LWR MOX fuel rod, WPu

2.5

410

0.1

0.04

0.03

1.4 × 10-4

4

LWR MOX fuel rod, RPu

2.5

410

0.1

0.04

0.7

3 × 10-3

4

LWR MOX fuel assembly, WPu

658

410

25

0.038

0.03

4 × 10-3

5

LWR MOX fuel assembly, RPu

658

410

25

0.038

0.7

0.08

5

MHTGR WPu fuel block

100

80

0.8

0.008

0.5

0.02

6

Irradiated LWR MOX fuel assembly, WPu

04 MWd/kgHM,

2 yr

658

410

23

0.035

38,000

4,500

7

 

10 yr

658

410

23

0.035

180

22

7

 

30 yr

658

410

23

0.035

79

9

7

 

100 yr

658

410

23

0.035

16

2

7

40 MWd/kgHM,

10 yr

658

410

18

0.027

18,000

2,200

7

 

30 yr

658

410

18

0.027

7,900

940

7

 

100 yr

658

410

18

0.027

1,600

190

7

50 MWd/kgHM,

10 yr

658

410

9

0.014

23,000

2,800

7

 

30 yr

658

410

9

0.014

10,000

1,200

7

 

100 yr

658

410

9

0.014

2,000

240

7

MHTGR WPu fuel block irradiated to 580 MWd/kgHM

2 yr

100

80

0.2

0.002

6,600

660

8

10 yr

100

80

0.2

0.002

1,800

180

8

30 yr

100

80

0.2

0.002

1,000

100

8

100 yr

100

80

0.2

0.002

200

20

8

Borosilicate glass log with WPu and HLW

small, 1.3% Pu, 20% HLW

250

50

3

0.013

not calculated

9

large, 1.3% Pu, 20% HLW

2,200

300

22

0.013

5,200

900

9

same, + 10 years

2,200

300

22

0.013

4,200

720

9

same, + 30 years

2,200

300

22

0.013

2,600

450

9

same, + 100 years

2,200

300

22

0.013

520

90

9

ABBREVIATIONS:

Max = maximum

Dim = dimension

Conc = concentration

aWPu assumed to contain 0.2 weight percent Am-241 (from initial 0.4 percent Pu-241, aged 14 years)

bRPu, reactor plutonium, assumed to contain 4 weight percent Am-241 (from initial 9 percent Pu-241, aged 12 years).

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

NOTES to Table 6-5:

1. If the rate at which energy is released in gamma rays of a given energy in a sphere or cylinder (or spherical or cylindrical shell) of material is Da (J/kg × h), and if the radius of this material (or, in the case of a shell, the thickness), R (cm), is large compared to the mean length for gamma-energy absorption in the material, L = 1 / [(µPu) × (density of Pu)], where µpu (cm2 /g) is the mass energy-absorption coefficient of plutonium metal for the relevant gamma-ray energy, then it is easy to show that the rate of energy absorption in the material at its outer surface is 0.5 × Dd and the contact dose rate in tissue at the surface is Ds = 0.5 × Da × (µtPu), where µt is the mass energy absorption coefficient of tissue. Now, for WPu containing 0.2 weight percent 430-yr Am-241, which emits a 0.06 million electron volt (MeV) gamma ray in 36 percent of its decays and which dominates the total gamma emission in this plutonium, the gamma-energy release rate is

Da  = (0.002 gAm-241/gPu) × [(6 × 1023 nuclei Am-241)/(241 g Am-241)]

× (0.693 disintegration/nucleus) / [(430 yr) × (8760 h/yr)]

× (0.36 × 0.06 MeV/dis) × (1.6 × 10-13 J/MeV) × (1,000 g/kg)

= 3.2 J/kg-h.

Now, with µ, the mass energy-absorption coefficient of tissue at 0.06 MeV, taken to be equal to that of water at 0.032 cm2/g, and with µPu, the mass energy-absorption coefficient of plutonium at 0.06 MeV, scaled from that of uranium (= 5.78 cm2/g) by 5.78 × (94/92) 2 = 6.03 cm2/g, we have L = 1 / [(6.03 cm2/g) × (19.6 g/cm3)] = 0.0085 cm, so the assumption that this is small compared to the radius or shell thickness is clearly satisfied, and we have

Ds = 0.5 × Da × µtPu = 0.5 × (3.2 J/kg × h) × 0.032/6.03 = 8.5 × 10-3 J/kg-h.

In the S1 units for absorbed dose, grays =joules/kg, this is 8.5 × 10-3 grays/h, which in traditional dose units is 0.85 rads/h; since sieverts = grays × QF and rem = rads × QF, where QF, the quality factor, is unity for gamma rays, this is equivalent to 8.5 × 10-3 sieverts/h or 0.85 rem/h. (This is the dose rate at the surface of tissue in contact with the plutonium sphere; the average dose rate in a body of finite dimension-say, 30 cm thick-in contact with the material would be somewhat lower by virtue of attenuation both by absorption and geometry.) By spherical symmetry, and neglecting the very small attenuation in air over distances of the order of a meter, the dose at the surface of tissue 1 m from the surface of a sphere of the indicated composition would be Dr = Ds × (R/r)2, with R the radius of the surface in meters and r = 1 + R. Thus with R = 0.08 m we have

Dr= 8.5 × 10-3 Sv/h × (0.08/1.08)2 =4.7 × 10-3 Sv/h = 0.005 rem/h.

In the case of a sphere of reactor plutonium, the concentration of the gamma-emitting Am-241 is, for our assumed compositions, 20 times higher, so the surface dose rate is 20 × 8.5 × 10-3 Sv/h = 0.17 Sv/h = 17 rem/h, and for a sphere of radius 4.5 cm the dose rate 1 m from the sphere's surface is

Dr= 0.17 Sv/h × (0.045/1.045)2 = 3.2 × 10-4 Sv/h = 0.032 rem/h.

2. Calculated as in note 1, for spherical powder mass containing 4 kg of plutonium; PuO2 mass energy-absorption coefficient taken to be 0.882 × 6.03 + 0.118 × 0.032 = 5.32 cm2/g, where 0.032 is the mass energy-absorption coefficient for oxygen.

3. Calculated as in note 2, for MOX fuel pellet with diameter 0.8 cm and length 1 cm, density 10.5 gm/cm3, 5.5 percent plutonium in heavy metal, and 0.2 percent Am-241 in plutonium.

4. Calculated as in note 3, with active rod length having 365 pellets, and with dose at 1 m from surface, now scaling according to cylindrical rather than spherical geometry, given by Dr = Ds × R/r, where R is surface radius (here 0.004 m) and r = 1 + R. Rod contains 2.1 kg MOX surrounded with 0.4 kg stainless steel cladding. Effect of 0.5 mm cladding layer on dose rate from 0.06-MeV gamma ray

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

estimated as exp[-(0.96 cm2/g) × (8 g/cm3) × (0.05 cm)] = 0.68, where 0.96 cm2/g is mass energy-absorption coefficient for 0.06 gamma MeV in iron and 8 g/cm3 is density of iron.

5. Calculated as in note 4, with Westinghouse PDR-600 17 × 17 fuel assembly containing 264 fuel rods plus control-rod channels (each assembly contains 461 kgHM, 524 kg MOX, 623 kg MOX with cladding, 658 kg assembly mass including spacers, springs, etc.). Assembly is square, 21.4 cm on a side, approximated for dose-at-distance calculation as a cylinder with r = 4 × 21.4 cm / 2 × pi = 13.6 cm.

6. Based on General Atomics figure of 700 mrem/h surface dose rate on an unirradiated fuel block 36 cm across, containing 0.8 kg WPu with 0.3 percent Am-241 (GA 1993, p. 6-26), scaled to 500 mrem/h for 0.2 percent Am-241, and scaled to dose rate at I m via approximations referenced in note 1, above.

7. Westinghouse PDR-600 fuel assembly, as described in note 5; dose rate at 1 m after two years for 0.4 MWd/kgHM from Westinghouse (1993); surface dose rate derived from this figure via Dr = Ds × (R/r) (see note 4) with R = 13.6 cm, r = 113.6 cm. Dose rates at 1 m after 10, 30, and 100 years for 40-MWd/kgHM and 50-MWd/kgHM irradiations scaled from detailed calculations performed at the Idaho National Engineering Laboratory (INEL) (Schnitzler 1993), using the ORIGEN isotope generation and depletion code and the QAD point-kernel shielding code with a complete model of this fuel assembly, assuming 3.2 percent U-235 fuel irradiated to 33 MWd/kgHM. We assume difference in gamma doses between MOX and LEU spent fuel is small, since both produce about 3 Ci Cs-137 per MWd, and the 0.66-MeV gamma from 85 percent of Cs-137 decays is the dominant source of gamma rays from spent fuel between 5 and 100 years after discharge. INEL results for 33 MWd/kgHM and 3 feet from surface of fuel assembly are 20 Sv/h at 10 years, 8.4 Sv/h at 30 years, and 1.7 Sv/h at 100 years. Scaling factor for radius 1 m instead of 3 feet is (13.6 + 91.4) / (13.6 + 100) = 0.924. Scaling factors for higher irradiations are 40/33 = 1.21 and 50/33 = 1.51. Surface doses scaled from doses at 1 m using Dr = Ds × (R/r) with R = 13.6 cm, r = 113.6 cm. Dose rates for 0.4-MWd/kgHM irradiation at 10, 30, and 100 years scaled linearly with irradiation from 40-MWd/kgHM results. To check the approximate method of notes I and 4, which we use elsewhere in this table when no detailed calculations are available, we here use the approximate method to recalculate, for comparison with the indicated "exact" calculation, the dose to be expected from LEU fuel irradiated to 33 MWd/kgHM. We assume that: (a) the gamma dose from spent fuel is dominated between 5 and 100 years after discharge by the 0.66-MeV gamma ray emitted in 85 percent of decays of 30.17-year half-life Cs-137; and (b) the quantity of Cs-137 in spent fuel at discharge is 3 Ci/MWd. For the fuel assembly described in note 5, the quantity of Cs-137 per assembly after irradiation to 33 MWd/kgHM is (3 Ci/MWd) × (33 MWd/kgllM) × (461 kgHM) = 45.6 × 103 Ci; after 10 years of cooling time the Cs-137 inventory is (45.6 × 103 Ci) × exp(-0.693 × 10/30.17) = 36.3 × 103 Ci. The energy release rate in the assembly is

(36.3 × 103 Ci) × (3.7 × 1010 dis/Ci × sec) × (3,600 sec/h) × (0.85x0.66 MeV/dis)

× (1.6 × 10-13 J/MeV)/ (658 kg) = 659 J/kg × h.

The mass energy absorption coefficient for the whole assembly is estimated by considering the assembly to be a homogeneous mixture of 507 kg U02 (allowing for 3.3 percent burnup) with µ for 0.66MeV gammas = 0.0718 cm2/gm), 134 kg steel with µ for 0.66-MeV gammas = 0.0282, 15 kg fission products (3.3 percent of the initial 461 kgHM) with µ for 0.66-MeV gammas taken as that of tin at 0.032, and 2 kg oxygen (associated with the fission products) with µ for 0.66-MeV gammas = 0.0292; hence

µtot = (507 × 0.0718 + 134 × 0.0282 + 15 × 0.032 + 2 × 0.0292) / 658 = 0.0619 cm2/g.

The mass energy absorption coefficient for water for 0.66-MeV gammas is 0.033 cm2/g, so the dose at the surface is Ds = 0.5 × (659 J/kg × h) × (0.033/0.0619) = 176 J/kg-h, i.e., 176 Sv/h or 17,600 rem/h. The dose at a distance of I m from the fuel assembly surface is

Dr = Ds × (R/r) = (176 Sv/h) × (13.6/113.6)= 21.1 Sv/h.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

This is within 15 percent of the "exact" calculation's result of 18.4 Sv/h for the same conditions, and thus confirms the usefulness of the approximate approach. Note, finally, that the diminution of plutonium over time in spent fuel because of the decay of 14.4-year Pu-241 (12-15 percent of total plutonium in high burnup LWR MOX) to 430-year Am-241 is not reflected in the plutonium column here, because Am-241 is itself weapon-usable (with critical-mass properties similar to Pu-242; see Mark 1993).

8. Doses at I m for times 2, 10, and 30 years are from GA (1994). Surface dose rates calculated from these by the approximation described in note 6. Doses at 100 years calculated from those at 30 years assuming a 30-year half-life of the gamma dose in this time period.

9. The large log has a glass volume of 625 liters, a glass density of 2.7 kg/liter, a glass radius of 29 cm, and a steel canister thickness of 0.9 cm. The glass content amounts to just under 1,700 kg and the steel jacket has a mass of 450 kg. Given that Savannah River wastes to be contained in the glass comprise 2.6 × 108 Ci of Cs-137 and Sr-90, and that the Cs-137 whose 0.66-MeV gamma will dominate the gamma dose constitutes about half of this (Berkhout et al. 1993, p. 184 and note 53), and given the planned campaign of 6,105 logs to contain all of the HLW at that site, the concentration of Cs-137 in the glass will be 1.3 × 108 Ci /(6,105 logs × 625 liters/log × 2.7 kg/liter) = 12.6 Ci/kg. The energy release rate in glass from this cesium's 0.66-MeV gamma is then

Da = 12.6 Ci/kg × 3.7 × 1010 dis/sec × Ci × (0.85 × 0.66 MeV)/dis

× 3,600 sec/h × 1.6 × 10-13 J/MeV

= 150 J/hr-kg.

The mass energy-absorption coefficient for the 0.66 MeV gamma in glass with 1.3 percent plutonium is about 0.787 × 0.0293 + 0.20 × 0.033 + 0.013 × 0.078 = 0.031, where 0.0293, 0.033, and 0.078 cm2/g are the mass energy-absorption coefficients for 0.66-MeV gamma radiation in SiO2, tin, and plutonium, respectively (taking tin as representative of fission products). The energy-absorption length is (0.031 cm2/g × 2.7 gm/cm3)-1 = 11.9 cm, which is about 40 percent of the radius of our log, so the approximate formula that gives the surface dose rate as Ds = 0.5 × Da × (µtglass) is not very good but perhaps is tolerable. It gives Ds = 0.5 × 150 J/kg-h × 0.033/0.031 = 80 Grays/h = 8,000 rem/h. The effect of the 0.9 cm steel container wall can be approximated by

exp[-(0.028 cm2/g) × (7.8 g/cm3) × (0.9 cm)] = 0.82,

where 0.028 cm2/g is the mass energy-absorption coefficient for iron at 0.66 MeV, giving a surface dose rate outside the container of 0.82 × 80 Sv/h = 66 Sv/h = 6,600 rem/h. Assuming the typical log is not produced until another 10 years after the date of the inventory figure with which we started, the dose rate will be down by another exp[-(10 × 0.693/30.2)] = 0.795, giving 5,200 rem/h. To get the dose rate at a radius of r = 1 + R = 1.3 m, notice that if this radius is considered small compared to the 3-m cylinder length, the approximation for that circumstance would give D (r = 1.3m) = 5,200 × (30/130) = 1,200 rem/h. In the limit of a spherical source, we would get D (r = 1.3m) = 5,200 × (3/1.3)2 = 280 rem/h. The first approximation is closer to right, so we estimate 900 rem/h. Doses at subsequent times figured based on the 30-year half-life of Cs-137. No figures are presented for the small log in this table because the energy-absorption length is not small compared to the radius of the log and the approximate approach used here is therefore not valid.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

trinsic barriers to weapons use associated with the main plutonium forms encountered in reactor-related disposition options.

Of particular interest in Table 6-6 are the conclusions about the intrinsic barriers associated with the final forms of the plutonium for the spent fuel and vitrification options—that is, MOX spent fuel assemblies and WPu oxide in glass logs with HLW, respectively. As indicated in the table, the MOX option has a modest advantage in terms of the isotopic barrier, whereas the vitrification option has a modest advantage in terms of the mass and bulk of the objects that a thief or divertor would need to remove and transport. The radiologic barrier is higher for the spent fuel than for the vitrified log, but it is high enough for the latter that we do not give much weight to the difference. It is our judgment that the chemical barrier is roughly comparable for the two options, in terms of the complexity of and technological sophistication needed for the steps required to separate plutonium from the two final waste forms. Overall, we rate both of these final forms as meeting the "spent fuel standard."

Table 6-7 presents characterizations of the implementation-dependent barriers and overall vulnerability to different threats for the vitrification option and for a version of the current light-water-reactor option in which two 1,300-MWe LWRs using full-MOX cores load the nominal 50 tons of WPu over a period of 25 years (option CLWRb in Table 6-4). The implementation-dependent barrier evaluations follow the prescriptions set forth in Chapter 3 under "Specific Security Concerns and Threat Characteristics." The vulnerability evaluations depend, as indicated in Chapter 3, on the interaction of the intrinsic and implementation-dependent barriers with the characteristics of the different classes of threat. 6 Key features of these interactions are:

  • Overt Diversion. Vulnerability to overt diversion will depend mainly on the intrinsic barriers, insofar as most of the implementation-dependent barriers would be irrelevant to the United States or Russia were one of these states to contemplate overtly reversing the disarmament process to reincorporate the WPu into its arsenal. The one exception to the irrelevance of implementation-dependent barriers would be the institutional barriers in cases where the plutonium had actually been placed, at some point in the disposition process, under international control and removed from the territory of the original possessor state.

6  

Recall from Chapter 3 that "overt diversion" refers to the situation in which a country chooses to reuse the (previously surplus) WPu in its possession for weapons purposes, without attempting to conceal this activity; "covert diversion" refers to the same situation except that the country attempts to conceal what it is doing; "theft" refers to acquisition of the material by unauthorized entities other than the initial possessor state, and might be attempted by force ("forcible theft"), by stealth ("covert theft"), or openly but without the need for the force ("overt theft"), as might occur following a breakdown of national authority. Intrinsic and implementation-dependent characteristics relevant to assessing vulnerability of an option to overt theft are an obvious combination of the characteristics germane for the cases of forcible and covert theft, so we give no further separate attention to overt theft here.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-6 Intrinsic-Barrier Characterization for Plutonium Forms Encountered in Reactor-Related Disposition Options

Form of Plutonium

Kg of Material per Kg of Plutonium

Qualitative Characterization of Other Intrinsic Barriers (0 = lowest, 4 = highest)

Overall Intrinsic-Barrier Characterization (0 to 4)

 

 

Isotopic

Chemical

Radiologic

Mass/Bulk

 

Nuclear-weapon pit

1

1

0

2

1

1

WPu oxide powder

1.1

1

1

2

0

1

MOX (WPu/U) powder

20

1

2

1

0

2-

MOX fuel rod

25

1

2

1

2

2

MOX fuel assembly

25

1

2

1

3

2+

MOX spent fuel assemblya

35

2

4

4

3

4

WPu oxide in HLW log

50

1

4

4

4

4

NOTE: The basis for this characterization scheme is elaborated in Chapter 3 under "Specific Security Concerns and Threat Characteristics."

a 40 MWd/kgHM.

In any other circumstances, the vulnerability to overt diversion will be essentially proportional to the physical ease of restoring the plutonium to weapon usability, or, in other words, inversely proportional to the magnitude of the intrinsic barriers: an overt divertor presumably would choose, if plutonium at several stages of disposition were available, to divert that which was closest to weapon-usable form; and a decision whether to divert the "dispositioned" plutonium at all, as opposed to obtaining additional WPu from other facilities from scratch, would likewise be related to the costs of overcoming the intrinsic barriers in comparison to the costs of from-scratch production. With respect to this class of threats and also with respect to covert diversion, the isotopic barrier may have greater importance in relation to the other intrinsic barriers than is the case with respect to theft, in that this characteristic probably is more likely to influence the thinking of the United States or Russia about whether to divert than it is to influence the thinking of countries or subnational groups contemplating theft as a means to acquire or augment a much less sophisticated arsenal.

  • Covert Diversion. Vulnerability to covert diversion will depend not only on the factors just described that pose barriers to processing and using the material for weapons, but also on the intrinsic and implementation-dependent barriers to diverting it without detection. Thus, the

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

dangers of covert diversion are reduced when the material is in forms that are bulky and heavy, when it is held at one or a small number of well-guarded sites, and when it is rigorously monitored and accounted for at each handling and processing step, under international oversight.

  • Forcible Theft. In contrast to the diversion threats, the threats of theft will tend to be on behalf of potential bomb-makers with less sophisticated requirements and less sophisticated materials-processing and weapon fabrication capabilities. Accordingly, in the case of theft the chemical and radiologic barriers will be more important, in the intrinsic-barrier category, and the isotopic barrier less important than in the case of diversion; the dilution and mass and bulk barriers will be somewhat important—less so, however, than for covert diversion and theft—and the implementation-dependent barriers (location and exposure, containment, and institutional) will be extremely important. Transport is probably the hardest operation to protect against forcible theft, so transport links involving material of low intrinsic barriers have the highest vulnerability against this class of threat.

  • Covert Theft. The relevance of intrinsic barriers to assessing the vulnerability to covert theft is much the same as in the case of forcible theft, except that the relevance of dilution and mass and bulk is higher in the covert case. With covert theft, one is almost inevitably dealing with an "insider" component to the threat, and the greatest vulnerabilities tend to occur where material of low intrinsic barriers and high portability flows through processes where maintaining accurate material accounting is most difficult-this means, above all, the processing step in which plutonium metal is converted to oxide (which occurs in both the MOX/spent fuel and vitrification options) and the further processing steps (MOX/spent fuel option only) of blending the mixed oxides and fabricating the fuel pellets.

As is clear from a glance at Table 6-7, the vitrification option has some advantage in overall vulnerability, compared to the MOX/spent fuel option, insofar as vitrification entails fewer processing steps and probably fewer transport steps (unless the MOX fabrication capacity and MOX-burning reactors are co-located) in which vulnerabilities to covert diversion and to theft are highest.7 On the other hand, the vulnerability of vitrification to both overt and covert diversion threats involving the final plutonium form is higher than the corresponding vulnerability of the MOX/spent fuel option, because the isotopic barrier is modestly lower for the vitrification option. Taking into account the modest advantage of the MOX options with respect to security of the final plutonium form

7  

This conclusion is compatible with the opinions of a number of safeguards experts consulted by the panel, who said that they considered MOX fuel fabrication plants more difficult to safeguard than a vitrification plant would be.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

and the modest advantage of the vitrification option with respect to security of the operations preceding final disposition, we characterize the two classes of options as comparable overall in terms of the barriers and vulnerabilities relating to security. To put it another way, the number of different and incommensurable variables, and the uncertainties in assessing each of them, make it impossible to determine whether either of these options offers superior overall security compared to the other.

The other options from Table 6-4, including those that employ CANDU reactors and evolutionary and advanced LWRs, would have substantially the same ratings in the framework of Table 6-7 as those shown there for the current LWRs with full-MOX cores. The differences in this framework would relate mainly to cases in which multiple reactors would not be required (amounting to 1 unit on the location and exposure part of the implementation-dependent barrier characterization for the irradiation step) or in which MOX fuel fabrication was co-located with the reactor (thus omitting the line characterizing vulnerabilities of fresh fuel transport).8 Except for the obvious advantages of cases that minimize the number of reactors used or eliminate transport steps by co-location, then, the main security and vulnerability differences among the variants of the LWR- and CANDU-based MOX options are those associated with timing and the related question of integrated inventories of pits and oxides, as summarized in Table 6-4.

More advanced reactor options, such as the MHTGR and ALMR, would also not differ greatly, in their ratings in the framework of Table 6-7, from the LWR case shown, assuming these advanced reactor types were used on a once-through basis. The MHTGR could gain 1 unit in the isotopic barrier rating when operated at very high burnup, and the co-location of fuel fabrication facilities with the reactor(s)—as would be possible with either of these reactors but also possible, in principle, with LWRs where new reactors or new fabrication capacity were sited to achieve this—would provide some gain in security; but we regard these potential improvements as insufficient to offset the large liabilities with respect to timing suffered by all the advanced reactor types, as discussed in the preceding section. This conclusion is reinforced, of course, by the circumstance that the current-reactor options and the vitrification option already achieve, at their end-points, a standard of security comparable to that of the large quantities of spent reactor fuel that already exist, so that there would be little overall security gain from using an advanced reactor type to push the WPu to a higher-than-spent fuel standard unless the same processing were contemplated for all the civilian spent fuel as well.

8  

There are some additional differences between CANDUs and LWRs, for example, in the size of the fuel bundles (smaller for CANDU) and the height of the radiologic barrier (lower for typical CANDU bumups than for typical LWR bumups), but these differences would not change our assessment of the overall vulnerabilities from the ratings shown for LWRs in Table 6-7.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-7 Barriers and Vulnerability for Selected Disposition Options

 

 

 

Implementation-Dependent Barriers (0 = low, 4 = high)

Vulnerability with respect to the threat of:

Option and Step

Duration (yr)

Overall Intrinsic Barriers (0-4)

Loc/ Exp

Containment

Institutional

Overt Diversion

Covert Diversion

Forcible Theft

Covert Theft

Vitrification with Defense High-Level Wastesa

Pit storage

13

1

3

3

3 4

high

med

med

med

Pit transport

9

1

0

2

3

high

high

high

med

Oxide production

9

1

2

2

2-4

high

high

med

high

Oxide storage

9

1

3

2

2-4

high

med

med

med

Log production

9

1→4

2

2

2-4

high

high

med

high

Log storage

15

4

3

2

2

low

low

low

low

Log transport

5

4

0

2

2

low

low

low

low

Logs in repository

indef

4

4

3

2

low

low

low

low

Current LWRs with Full-MOX Coresb

Pit storage

25

1

3

3

3 4

high

med

med

med

Pit transport

25

1

0

2

3

high

high

high

med

Oxide production

25

1

2

2

2-4

high

high

med

high

Oxide storage

25

1

3

2

2-4

high

med

med

med

MOX production

25

1→2

2

2

2-4

high

med

high

high

MOX storage

25

2

3

2

2-4

high

med

med

med

Fuel assembly production

25

2→2+

2

2

2-4

med

med

med

high

Fuel assembly storage

25

2+

3

2

2-4

med

med

med

low

Fuel assembly transport

25

2+

0

1

2

med

med

high

low

Fuel storage at reactor

25

2+

2

2

2-3

med

med

med

low

Irradiation in reactor

28

2+→4

1

4

2-3

low

low

low

low

Spent fuel at reactor

33

4

3

2

2-3

low

low

low

low

Spent fuel transport

18

4

0

2

2

low

low

low

low

Spent fuel in repository

indef

4

4

3

2

low

low

low

low

NOTES: Duration (years) extends from the time that operations in a given step begin to the time that operations of this type cease. Overall intrinsic barriers (0 lowest to 4 highest) are taken from the last column of Table 6-6. Implementation-dependent barriers (0 lowest to 4 highest)—in the categories location/exposure, containment, and institutional are rated on the basis described in Chapter 3 (see section "Specific Security Concerns and Threat Characteristics"). Vulnerability assessments are based on the interaction of the barriers with the different types of threats, as discussed in Chapter 3.

aVitrification: It is assumed that the repository is ready in 2015 and that the process of shipment of the logs to, and their emplacement in, the repository extends over a period of five years.

bCurrent LWRs with Full-MOX Cores: It is assumed that there is only one MOX fabrication plant. Repository becomes available in 2015. Spent fuel assumed stored at reactor for five years cooling if repository is available.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

In general, the same types of threats and considerations would apply in the Russian case as in the U.S. case, but concerns about the possibility of covert theft during processing and forcible theft during transport would appear to be even more significant. Thus the premium on minimizing processing steps involving high accessibility and low accountability, and on minimizing transport steps, would be even more substantial than in the U.S. case. As noted in the U.S. case, however, this does not appear to be a strong discriminant between the various options considered in this report: while it appears that the overall in-process vulnerability of the vitrification option would be somewhat lower than the LWR MOX option, the attractiveness of vitrified material for weapons purposes would be somewhat higher, making the two options roughly comparable in overall risk.

The panel concurs with the full committee that U.S.-Russian co-operation to improve material protection, control, and accounting in Russia should be an urgent priority accorded a significantly higher level of funding than is currently the case, and notes the specific recommendations made by the parent committee in this regard (NAS 1994, pp. 133-136). Specifically for the case of disposition of excess WPu, the panel concurs with the committee's recommendation that "The United States and Russia should begin discussions with the aim of agreeing that whatever disposition options are chosen, an agreed, stringent standard of accounting, monitoring, and security will be maintained throughout the process-coming as close as practicable to meeting the standard of security and accounting applied to intact nuclear weapons" (NAS 1994, p. 227). Without assurance that such strict standards of security and accounting will be maintained, there is a very real risk that disposition of WPu, with the substantial plutonium processing and possibly transport it entails, could add to the risk of proliferation rather than reducing it.

The panel notes that unlike in the U.S. case, disposition of excess WPu in Russia will take place in a context of a nuclear economy involving substantial civilian plutonium separation and government plans for a large-scale plutonium economy, which also poses proliferation risks. It is unlikely to be possible to "delink" the WPu issue completely from the civilian plutonium issue in Russia. For example, in the U.S. case, a MOX plant that might be constructed or completed for the WPu disposition mission would presumably be closed when that mission was completed (unless U.S. fuel-cycle policy changed in the interim). In the Russian case, by contrast, such a facility would presumably continue to operate, fabricating reprocessed civilian plutonium into fuel, and thereby serving as a key element in the next stage of the plutonium economy in Russia. U.S. policy-makers, when making decisions on steps that could influence Russian plutonium disposition choices (such as provision of assistance for particular options, for example) will have to take into consideration these linkages to the civilian plutonium economy and their relation to U.S. fuel-cycle policy and nonproliferation goals.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

ECONOMIC COMPARISONS

As noted in Chapter 3, economic considerations are less important than security in reaching a conclusion about the relative attractiveness of alternative options for WPu disposition. But a study of costs is nonetheless worthwhile, both to assist in ranking options that are not distinguishable on security grounds and to facilitate planning for the investments that will be required for one option or another. In what follows, we apply the economic assessment methods and assumptions introduced in Chapter 3 to illuminate the comparative costs of the main disposition options.

We consider, in turn, the costs of incorporating WPu into reactor fuel, the costs of using such fuel in currently operating reactors, the costs of using it in reactors now partly completed (which could be completed for the plutonium disposition mission), the costs of building new reactors of evolutionary and advanced types and using them to process WPu-bearing fuel, and the costs of vitrifying the plutonium with defense HLW. In these comparisons, all of the final cost estimates are stated in 1992 dollars.

The economic analyses presented here relate mainly to prospective disposition operations in the United States. One case is considered in which U.S. plutonium would be used in Canadian reactors. A subsection at the end discusses how the economic considerations would differ for the disposition of Russian WPu.

Weapons Plutonium Versus Uranium as Power Reactor Fuel

Considerable discussion and controversy has surrounded the question of whether the use of surplus WPu as fuel for power reactors would save money or cost money, compared to the costs of generating the same amounts of electricity by the means that would otherwise be used—for example, in the case of currently operating reactors, using the same reactors but with U-235 as the primary fissile material. In this section, we seek to clarify this issue by (1) identifying the component costs that bear on the comparative economics of plutonium-based and uranium-based fuels, (2) indicating and explaining the ranges of values for these component costs that have appeared in the literature, (3) offering our own estimates of the values that are most appropriate to the case at hand (i.e., to investigation of the economics of the disposition of WPu over the next 20-40 years), and (4) using these estimates to obtain a plausible range for the total costs (or savings) associated with disposition of WPu through its use as fuel in power reactors.

We take, as the starting point for our treatment, the proposition that surplus WPu would be made available to the power-generating entity at no charge, in the form of plutonium metal. That is, the costs invested in the production of the plutonium in the course of the weapons programs of the United States and the

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

former Soviet Union are to be treated as "sunk" costs, but any costs entailed in converting the plutonium from the metallic form (in which it emerges from surplus weapons) to oxide or other forms (required for fueling some types of reactors) are to be charged to electricity generation.9 Our assumption that the plutonium metal is free, of course, means that the resulting estimates of the costs of the use of WPu for electricity generation will not be relevant to the case of recycle of plutonium from spent reactor fuel; this would be more expensive because of the costs of reprocessing spent fuel to extract the plutonium.

In order to compare the costs of electricity generation with WPu-based versus uranium-based fuels, it is not necessary to estimate all of the generation costs; only those costs that are related to the character of the fuel need be estimated and compared, per kilowatt-hour (kWh) and, in the case of the plutonium-based fuel, per kilogram of WPu. From these figures it will be possible to calculate the incremental cost (or savings) per kilowatt-hour associated with using WPu for electricity generation, as well as the incremental cost (or savings) associated with the disposition in this way of 50 tons of WPu (or 100 tons, or any other quantity). Where these calculations require the choice of a value for the real cost of money, we shall use the figure of 7 percent per year (r = 0.07) specified by the U.S. Office of Management and Budget for the evaluation of government projects that produce effects in the private sector or have any interaction with it (since clearly electricity generation does this).10 Because we are concerned with the costs of plutonium disposition to society rather than the costs to any particular entity, the question of whether particular fuel-related costs are incurred by the federal government or by another entity will concern us only insofar as it affects the magnitude of these costs (as, for example, through the circumstance that a private entity may have to pay for property taxes and insurance while a government entity does not).

We first consider, for specificity, a comparison between the use of conventional LEU oxide fuel in a PWR and the use of MOX fuel of comparable reactivity in the same reactor with the same burnup—i.e., same irradiation in thermal megawatt-days (MWd) per metric ton of heavy metal (MTHM, based on the combined mass of uranium and plutonium in fresh fuel). Since a kilogram of heavy metal (kgHM) in fuel, whether LEU or MOX, then generates the same amount of electricity, it becomes both germane and convenient to compare the costs of the two fuels on a per-kgHM basis. We take up the generalization to different fuel parameters and different reactor types subsequently.

We take the "baseline" LEU fuel to have an initial U-235 enrichment of 4.4 percent and assume it is irradiated on a three-year fuel cycle, at 75-percent ca-

9  

For purposes of comparative assessment it is necessary to account for the costs of any conversions of the WPu from metallic form, because different disposition schemes require different conversions (and some require none at all). See "Issues and Criteria in Economic Evaluation of Alternatives" in Chapter 3.

10  

See Chapter 3 and OMB (1992).

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

pacity factor, to 40,000 MWd/MTHM—typical commercial practice in a modern PWR. We assume that the MOX is made by mixing WPu oxide with depleted uranium oxide (the uranium which contains 0.25 percent U-235) to an initial concentration of 4.8 percent WPu in heavy metal; this initial WPu concentration gives the same end-of-cycle reactivity (i.e., at 40,000 MWd/MTHM) as for the LEU fuel, arguably the most appropriate basis for comparison.11 To compare the costs of these equivalently energy-generating fuels, we then need to estimate, on a comparable basis:

For the LEU fuel, the costs of

  • finding, mining, and milling the uranium ore;

  • converting the resulting uranium-ore concentrate ("yellowcake," typically 60-85 percent U308) to uranium hexafluoride (UF6);

  • enriching the UF6 to the indicated U-235 concentration,

  • converting the enriched UF6, to UO2;

  • fabricating the UO2 into fuel pellets, fuel rods, and fuel assemblies;

  • all the storage and transport steps associated with this chain, including delivery of the fuel assemblies to the reactor and their storage there until they are loaded into the core; and

  • the costs associated with ultimate disposition of the spent fuel after it leaves the nuclear reactor.12

For the MOX fuel, the costs of

  • conversion of WPu metal to PuO2;

  • acquisition of depleted uranium and its conversion, if necessary, to UO2;

  • mixing the oxides and fabricating the MOX into fuel pellets, fuel rods, and fuel assemblies;

  • all the storage and transport steps associated with this chain, including delivery of the fuel assemblies to the reactor and their storage there until they are loaded into the core; and

  • the costs associated with ultimate disposition of the spent fuel after it leaves the nuclear reactor.

11  

LEU and MOX fuels have somewhat different rates of change of reactivity with burnup. The closest match in nuclear performance is obtained by matching the end-of-life reactivities (see, e.g., Battelle 1993).

12  

It is appropriate to include this category explicitly both because these costs conceivably could differ, in some cases, between plutonium-based and uranium-based fuels, and also to remind ourselves, even where these costs do not differ, that a proper comparison must either include them for both fuel types or exclude them for both. (Comparisons in the literature contain some inconsistencies in this regard.) By convention, the costs of interim storage of spent fuel at the reactor, before its removal for ultimate disposition, are considered part of reactor costs rather than as part of fuel costs.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

Many studies of these matters have omitted some of these cost categories, above all the costs of conversion of plutonium metal to oxide and the costs of storage and transport. Since these costs will be quite different for LEU and MOX fuels (because of the much larger radiological and security hazards of the latter), and since they also may differ significantly among different schemes that use MOX (e.g., because of different types and distances of transport), their omission can significantly distort comparisons.

We now turn to an activity-by-activity examination of the costs associated with the use of LEU and MOX fuels, beginning with LEU.

LEU: Composition of the Total Costs

The total costs per kilogram of heavy metal in fresh fuel13 consist of the sum of terms obtained by multiplying, for each step in the fuel production chain, the unit cost for the step (dollars per unit of activity in the step) times the number of units of activity in the step associated with one kilogram of heavy metal (kgHM) in fresh fuel. Thus we have, for LEU

LEU fuel cost ($/kgHM) = {unit cost of uranium acquisition [$/(kgU acquired)] × quantity of U acquired to make 1 kgHM in fresh fuel [kgU/kgHM]} + {unit cost of uranium conversion [$/(kgU converted)] × quantity of U converted to make I kgHM in fresh fuel [kgU/kgHM]} + {unit cost of uranium enrichment [$/(separative work unit; SWU)] × quantity of U enrichment to make I kgHM in fresh fuel [SWU/kgHM] } + {unit cost of LEU fuel fabrication [$/kgHM]} + {unit cost of ultimate disposition [$/kgHM]}.

The cost of acquisition of uranium, as represented by the selling price of uranium-ore concentrate (typically reported as dollars per pound of U308 or as dollars per kilogram of contained uranium), includes the costs of finding, mining, and milling the uranium ore. The cost of conversion of the UF6 to UO2 following enrichment is, by convention, considered to be part of the cost of fuel fabrication. Costs of transport and storage of the uranium between and at the steps up to the fuel's arrival at the reactor are small compared to the other costs and are assumed to be included with them (e.g., cost of shipping to the conversion plant is included with the uranium acquisition cost, and cost of storage at the conversion plant is included with the cost of conversion). The cost of storing

13  

It is customary to normalize all nuclear fuel-cycle costs to the quantity of heavy metal present in the fuel at the point it is loaded into the reactor. Thus, for example, the costs of spent fuel transport and waste management, per kilogram of heavy metal (kgHM), relate to the amount of fresh fuel that contained one kilogram of combined uranium and plutonium metal, not to the slightly larger amount of fuel that would contain 1 kgHM after some of the uranium and plutonium have been consumed by fission (see, e.g., OECD 1992).

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

fresh fuel at the reactor prior to loading it into the core is likewise small and is customarily treated as part of the reactor capital and operating charges rather than as part of fuel costs.

Determination of the activity quantities at each step requires attention to the material balance and separative work requirement for uranium enrichment, as well as to the fractional losses of uranium in other steps. The relevant enrichment relations are summarized in “Uranium and Separative Work Requirement for Enrichment" on p. 288. For our reference case values of 4.4 percent U-235 enrichment and 0.25 percent U-235 content in the enrichment tails, the uranium feed requirement is 8.83 kgU input per kgU in enriched product, and the separative work requirement is 6.59 SWU per kgU in enriched product. We follow the recent OECD fuel-cycle study (OECD 1992) in estimating other uranium losses in processing to be 0.5 percent in conversion to UF 6, 1.0 percent in fuel fabrication, and negligible at other steps. Accounting for these losses, the enrichment requirement per kilogram of heavy metal in fresh fuel is

SW [SWU/kgHM] = 6.59 [SWU/(kgU enriched product)] × 1.01 [(kgU in enriched product)/(kgHM in fresh fuel)]  = 6.66 SWU/kgHM,

and the uranium feed requirement per kilogram of heavy metal in fresh fuel is

F [kgU/kgHM] = 8.83 [(kgU into enrichment)/(kgU in enriched product)] × 1.005 [(kgU into conversion)/(kgU into enrichment)] × 1.01 [(kgU in enriched product)/(kgHM in fresh fuel)]  = 8.96 [(kgU into conversion)/kgHM].

It follows that the cost of LEU fuel under our reference case conditions can be written as

LEU cost [$/kgHM] = 8.96 ´ (UCOST + CNVSN) + 6.66 × ENRCH + FABRN + DSPSL,

where UCOST is the cost of uranium per kilogram uranium delivered to the conversion process, CNVSN is the cost of conversion per kilogram uranium delivered to the conversion process, ENRCH is the cost per separative work unit, FABRN is the cost of fabrication per kilogram of heavy metal in the fresh fuel, and DSPSL is the cost of spent fuel storage and disposition per kilogram of heavy metal in the fresh fuel.

The range of values of the first four of these unit costs (that is, excluding disposal costs) from current experience and recent studies is shown in Table 6-8. Taking the lowest values for all of the unit costs in the table would yield a total cost of LEU fuel, not including costs of spent fuel storage and disposal, of $813/kgHM; the corresponding figure based on taking the highest values in the table is $2,167/kgHM. This is a large difference. In what follows, we examine the individual costs to try to narrow the range that can be considered plausible

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-8 LEU Unit Costs from Current Experience and Recent Studies

Figures have been converted to 1992 U.S. dollars and rounded to the nearest dollar.

Source

UCOST ($/kgU)

CNVSN ($/kgU)

ENRCH ($/SWU)

FABRN ($/kgHM)

Mid-1993 spot-market pricea

17

3

66

NA

Mid-1993 long-term contract pricea

25-43

4-6

87-107

194-291

ERI (1993) range for year 2000b

35-55

6

92-114

194c

USDOE (1993a) Plutonium Disposition Studyd

65

10

125

260

OECD (1992) study, rangee

41-92

6-11

82-133

205-358

OECD (1992) study, reference case

71.7

8

113

282

Berkhout et al. (1992) best estimate

40

7

100

200

This report

55±20

9±1

95±15

200±30

NOTE: NA indicates not applicable.

a USDOE (1993c).

b Energy Resources International, "Nuclear Fuel Cycle Supply and Price Report," as reported in Nuclear Fuel, June 9, 1993 (ERI 1993). The Office of the Assistant Secretary for Nuclear Energy (USDOE 1993c) provided this reference in response to a request from the National Academy of Sciences' Committee on International Security and Arms Control for that Office's judgment of the likely future prices of these commodities and services.

c ERI (1993) gives a figure of $200/kgU (1993 dollars) for PWR fuel and $300/kgU (1993 dollars) for BWR fuel. For this report we use the former figure since the other parameters of our calculations relate to PWRs.

d Assumptions specified by the U.S. Department of Energy for its 1992-1993 Plutonium Disposition Study (USDOE 1993a), apparently based on Delene and Hudson (1993). DOE now concedes that the indicated values for the costs of uranium and its conversion and enrichment are "high compared to today's market and sensitivity studies" (USDOE 1993c, p. 4).

e The values presented in the Organization for Economic Co-Operation and Development study (OECD 1992) represent levelized costs during a presumed period of operation extending from the year 2007 to 2035.

for the circumstances of interest here, namely, disposition of WPu in the period extending roughly from the year 2000 to 2030.

LEU: Uranium Acquisition

Estimates of world uranium reserves and resources versus expected selling price, when compared with plausible demands to 2015 or even 2030, give little reason to expect substantial price increases during this period above the range of current long-term contract prices. 14 That was true even before the advent of

14  

For recent estimates of world uranium resources see, e.g., OECD (1990). For comparisons with growth rates of demand, see ERI (1993). It is worth noting that DOE's own 1993 "Cost Estimate Guidelines for Advanced Nuclear Power Technologies" (Delene and Hudson 1993) call for assuming no real escalation of uranium-ore costs above $25 per pound of U3O8 ($65/kgU) over the 30-year operating lifetimes of plants that would begin operation between 2000 and 2010. The spotmarket price of U3O8 in early 1995, as this report went to review, was about $10 per pound of U3O8 ($26/kgU).

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

agreements that, if implemented, will bring onto the civilian-power uranium market substantial quantities of LEU obtained by blending down surplus highly enriched uranium (HEU) from dismantled nuclear weapons. The Organization for Economic Co-operation and Development study (OECD 1992) estimated the rate of increase of the real cost of uranium acquisition in the period of interest at 1.2 percent per year, with a range from -0.8 percent per year to 2.1 percent per year. We assume, perhaps a bit conservatively, a rate of growth of real uranium costs after the year 2000 of 0-2 percent per year; if we take the year 2015 as the reference point for an economic evaluation relevant to the disposition of WPu, and if we escalate the Energy Resources International (ERI 1993) low estimate for the year 2000 at 0 percent per year and their high estimate for that year at 2 percent per year, our range of values for the year 2015 becomes $35-$75/kgU (1992 dollars).

LEU: Conversion

The figures for conversion are neither controversial nor very important in determining the total costs of LEU fuel. We take the range of plausible values for the midpoint of our period of interest to be $8-$10/kgU.

LEU: Enrichment

As noted in a number of the recent studies (see, e.g., OECD 1992, USDOE 1993c), there exists a worldwide surplus of uranium enrichment capacity and no foreseeable prospect for its disappearance. The evolution of enrichment technologies in the directions already evident, moreover, will tend to lower the costs per separative work unit (OECD 1992). In these circumstances, we think it is difficult to justify an upper-limit estimate for the 2015 enrichment higher than the upper end of the range of long-term contract prices in 1993 ($107/WS), and that the lower end of the range must be at least at the OECD low estimate for the period 2007-2035 ($82/SWU); we therefore take the range to be $80-$110/SWU (1992 dollars).

LEU: Fabrication

Much of the difference in estimates of the costs of LEU fuel fabrication per kilogram of heavy metal appears to be due to the difference between PWR fuel and boiling-water reactor (BWR) fuel. We see little reason to diverge, in our comparison based on the case of a PWR, from the ERI (1993) estimate of about $200/kgHM for PWR fuel, (which the Nuclear Fuel analysis (ERI 1993) predicts to remain unchanged to the end of their forecast period of 2005). The technology of UO2 fuel fabrication is mature and its occupational health and environmental impacts are small, so it is difficult to see what could drive up its cost

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

other than, conceivably, a temporary shortage of capacity that would quickly be remedied. We take the range to be $200/kgHM ± 15 percent, hence $170-$230/kgHM (1992 dollars).

LEU: Spent Fuel Storage and Waste Disposal

As noted earlier, it is customary to count costs of spent fuel storage at the reactor as part of reactor costs while including an estimate of the costs of ultimate disposal of the spent fuel in the fuel costs. Since the exact means of ultimate disposal has not been decided, this item poses some difficulty. In U.S. practice, electric utilities pay DOE a "waste-disposal fee" of $0.001 per electrical kilowatt-hour (kWh) generated. For the parameters of our example, this is equivalent to about $260/kgHM, evaluated as of delivery to the reactor.15 This figure, it must be emphasized, reflects not real-world experience about the ultimate cost of disposing of spent fuel as waste but, rather, DOE's rough estimate of how much money needs to be collected at the time of generation in order to provide a waste-management fund that will be adequate to the task at the future time when disposal actually takes place. The uncertainty in such an estimate is necessarily quite large; we round the figure to $300/kgHM and assign a judgmental 70-percent confidence interval of a multiplicative factor of two up and down, giving a range of $150-$600/kgHM. This range is roughly consistent with the $200-$1,000/kgHM range of disposal costs cited in the OECD (1992) study.

15  

We assume that electric utilities pay the $0.001/kWh waste-disposal assessment annually based on the year's electricity generation. To translate this operating cost into an equivalent contribution to the cost of fuel per kilogram of heavy metal, one must take into account both the number of electrical kilowatt-hours generated from I kgHM (which for our nominal PWR and 40,000-MWd/kgHM burnup is 40 MWd/kgHM × 0.316 MW-electric/MW-thermal × 1,000 kW/MW × 24 hr/d = 303,000 kWh) and the way in which electric utilities calculate carrying charges on nuclear fuel. Without accounting for carrying charges, the waste-disposal assessment would be equivalent to $0.001/kWh × 303,000 kWh/kgHM = $303/kgHM. Carrying charges on fuel are customarily calculated for a period one year longer than the fuel's residence time in the core, hence four years for a three-year residence time. Thus, in the translation between acquisition cost for the fuel per kilogram of heavy metal and the equivalent per-kilowatt-hour cost, the acquisition cost is multiplied by the annualized capital recovery factor—here 0.07 × 1.074/(1.074 - 1) = 0.295/yr for 0.07/yr real cost of money and four-year accounting lifetime. So the fuel's acquisition cost is translated into four annual payments, each equal to 0.295 of the purchase cost, and since each of these payments can be associated with one-fourth of the fuel's lifetime electricity output, the carrying-charge factor by which the per-kilowatt-hour cost exceeds the ratio of the purchase cost to the total electrical output is 0.295/0.250 = 1.18. Thus, for the assumed circumstances, the purchase-cost equivalent of $0.001/kWh is not $303/kgl M but $303/kgHM / 1.18 = $257/kgHM.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

URANIUM AND SEPARATIVE WORK REQUIREMENTS FOR ENRICHMENT

Performing a material balance for enrichment shows that the quantity of uranium "feed" (F) needed per kilogram of enriched uranium "product" is given by

F[(kgU input)/(kgU in enriched output)] = (Xp - Xt) / (Xf - Xt),

where X is the U-235 fraction in the enriched product, Xt is the U-235 fraction in the waste ("tails"), and Xf is the U-235 fraction in the feed material. If the feed material is natural uranium, Xf = 0.0072. The separative work requirement per kilogram of uranium in the enriched product can be shown to be given by

SW [(separative work units)/(kgU in enriched output)

= V(Xp) + (F - 1) × V(Xt)- F × V(Xf),

where V(X), the dimensionless "separation potential," is

(2X-1)In[X/ (1-X)].

The U-235 fraction left in the tails, Xt, influences both the separative work requirement (which decreases as Xt increases) and the uranium feed requirement (which increases as Xt, increases). Its value is chosen to minimize the total cost of producing a kilogram of enriched uranium, and so depends on the relative costs of enrichment work and uranium feed. Under recent and current conditions, Xt has generally been chosen to be 0.0025 or 0.0030. We follow several

Although it often has been assumed that ultimate disposition costs would be about the same for MOX fuel as for LEU fuel, there is at least some basis for supposing that they might be higher. As noted in a recent study by analysts at the German nuclear research center in Karlsruhe (Kessler et al. 1992), spent MOX fuel has a higher thermal power, decaying more slowly, compared to LEU fuel irradiated to the same exposure; the German analysts estimate that this would reduce by about a factor of two the number of fuel elements that could be packaged in each of the containers envisioned for use in the German waste repository, and would increase by a similar factor the repository floor-space requirements for MOX fuel. Also, spent MOX fuel from a PWR would contain 23 percent residual plutonium versus about 1 percent in typical spent LEU fuel, and the associated criticality considerations may likewise dictate a lower packing density in the repository.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

recent studies in assuming a value of 0.0025, although at current low prices of uranium a modest savings—about $30/kgHM for 4.4 percent enriched fuel—would be associated with using X, = 0.0030 instead. The requirements for uranium feed and separative work per kilogram of uranium in the enriched product, for tails fractions of 0.0025 and 0.0030 and various enrichment levels, are shown in the following table.

 

Enrichment

Uranium Feed (kgU input per kgU in enriched output)

 

Separative Work (SWU per kgU in enriched output)

 

(% U-235)

X, = 0.0025

X, = 0.0030

X, = 0.0025

X, = 0.0030

 

 

3.0

5.85

6.43

3.77

3.38

 

 

3.3

6.49

7.14

4.36

3.92

 

 

3.7

7.34

8.10

5.16

4.66

 

 

4.0

7.98

8.81

5.77

5.22

 

 

4.4

8.83

9.76

6.59

5.98

 

 

5.0

10.11

11.19

7.84

7.13

 

 

5.5

11.17

12.38

8.90

8.10

NOTES: The substantial simplification of ignoring the 0.006 percent U-234 in natural uranium produces only an insignificant error in the foregoing calculations.

The separative work unit (SWU), although called "work," is a linear combination of the quantities of material flowing into and out of the enrichment process and thus, by convention, has units of kilograms. The number of SWUs needed is a nonlinear function of the product enrichment desired, the enrichment of the starting material, and the enrichment of the discarded "tails."

Because canister design and repository layout are still far from settled, and because there might be relatively inexpensive ways to accommodate the higher decay power and greater reactivity of MOX fuel compared to LEU, we cannot say much more than that the costs of ultimate disposition of MOX are likely to fall in a range extending from about the same as those for LEU to about twice as much. If the best estimate were 50 percent greater, the $300/kgHM figure for LEU given above would imply a penalty associated with MOX of an additional $150/kgHM. Because the uncertainties about ultimate disposition costs are so large, however, we choose not to add a figure of such questionable validity to the better-defined numbers for other elements of fuel costs. Instead we will present figures for LEU and MOX fuel costs less the costs of ultimate disposal, underlining here that if there is a difference between MOX and LEU in the omitted disposal costs it will be in favor of LEU.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×
LEU: Reactor Panel Best Estimate of Total Fuel Costs Less Disposal

The preceding considerations have led us to the following plausible ranges of unit costs relevant to the determination of total fuel costs less costs of ultimate disposal, evaluated in 1992 dollars for operations in the period 2000-2030: UCOST = $55 ± $20, CNVSN = $9 ± $1, ENRCH = $95 ± $15, FABRN = $200 ± $30. Using the uranium requirements and enrichment quantities that correspond to 4.4-percent enriched fuel and enrichment tails of 0.0025 percent U-235, with losses of 0.5 percent in conversion and 1.0 percent in fabrication, then gives a central estimate of $1,406/kgHM. If it is assumed, for convenience, that the indicated uncertainties represent standard deviations of independent random variables, then the variance of the sum is the sum of the variances, and the standard deviation is the square root of this sum, or $207/kgHM. Since the precision in these figures is clearly illusory, we round off the range to $1,400 ± $200 per kgHM.

MOX: Composition of the Total Costs

The equation for the total costs per kilogram of heavy metal in fresh MOX fuel is, in analogy to the one given above for LEU,

MOX fuel cost ($/kgHM) ={unit cost of uranium acquisition [$/(kgU acquired)] × quantity of U acquired to make I kgHM in fresh fuel [kgU/kgHM]} + {unit cost of uranium conversion to UO2 [$/(kgU converted)] × quantity of U converted to make I kgHM in fresh fuel [kgU/kgHM]} + {unit cost of plutonium conversion to PuO2 [$/(gPu converted)] × quantity of Pu converted to make 1 kgHM in fresh fuel [gPu/kgHM]} + {unit cost of MOX fuel fabrication [$/kgHM]} + {unit cost of ultimate disposition [$/kgHM]},

plus any costs of plutonium storage and transport within the fuel cycle that are not already included in the preceding figures.

The concentration of WPu in our reference-case MOX fuel is 4.8 percent by weight, hence 48 g/kgHM, so the uranium requirement is 0.95 kgU/kgHM. If uranium losses in conversion and fabrication totaled 1.5 percent, as assumed for LEU fuel, the requirement would be about 0.97 kgU/kgHM. The OECD study (1992) assumed the same percentages for plutonium losses in conversion and fabrication as for uranium. This seems to us to be an overestimate, in light of the extra incentives for tight material control over plutonium that arise from its extra ES&H and security dangers; but since changing this estimate would affect the economic conclusions by an amount very small compared to other uncertainties in the calculation, we do not take the trouble to develop an alternative figure. Thus the plutonium requirement is 48 × 1.015 = 49 gPu/kgHM.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×
MOX: Uranium Acquisition and Conversion

We assume this uranium will be in the form of depleted uranium (0.25 percent U-235), which is and will remain in abundant supply in the nuclear-energy industry. The combined cost of acquiring depleted uranium and converting it to UO2 can hardly be more than $10/kgHM, and any error in this estimate cannot be important since it will assuredly be small compared to uncertainties in the other components of MOX fuel costs.

MOX: Conversion of Plutonium Metal to PuO2

As noted above, it is appropriate to assign to the MOX fuel costs a contribution for conversion of plutonium metal to oxide, because some alternative schemes for disposition of the WPu—against which the MOX fuel option must be compared—would not incur this cost.

A 1992 study by DOE (USDOE 1992) estimated that conversion of 100 tons of Russian WPu to oxide over a period of 10 years, in Russia or France, would entail a facility construction cost of $150 million and operating costs of $10 million per year. If we assume that the costs of such an operation in the United States would be 50 percent higher, this would mean capital costs of $225 million and operating costs of $15 million per year (assumed to be 1992 dollars). For a conversion rate of 10 MT/yr for 10 years, and with r = 0.07, these figures translate on a levelized-annualized basis to a conversion cost of $4.70/gPu. If we apply the widely used rule of thumb that the construction costs and operating costs for such facilities increase as the 0.6 power of output (meaning unit costs will decrease as the 0.4 power of output), the unit cost for converting 50 tons of plutonium to oxide in 10 years would be 32 percent higher, or $6.20/gPu.

A 1993 study of the possibility of vitrifying the WPu in radioactive-waste-bearing glass (McKibben et al. 1993) estimated the upstream costs (previtrification costs) of a campaign treating 50 tons of WPu as $400 million. Conversion to oxide is the most substantial of the "upstream" steps. If we assume it accounts for 75 percent of the indicated cost, or $300 million, and that two-thirds of this is the facility cost and one-third represents 10 years of operating costs at 5 tons of plutonium per year, then with r = 0.07 the costs on a levelized-annualized basis would be $7.70/gPu.

Based on the preceding two calculations, and allowing some extra leeway for the approximateness of the estimates on which the calculations are based, we shall take as our estimate of the costs of conversion of plutonium metal to oxide as $7 ± $2/Pu (1992 dollars, in U.S. facilities, based on a campaign treating a total of 50 tons of WPu). 16

16  

Two more recent examinations of the cost of pit conversion (which became available after the panel's economic calculations were largely complete) presented cost estimates that were in one

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×
MOX: Fuel Fabrication

The cost of fabrication of MOX fuel is by far the largest component of the cost of MOX fuel based on cost-free plutonium metal from surplus weapons, and it is also the component for which the widest range of values can be found in the literature. The breadth of this range is attributable in part to the early stage of commercialization of MOX fabrication technology (whereby only a few plants have operated, and these are of low capacity compared to those envisioned for a mature industry) and in part to the difference in the scale of operation between the scale appropriate to a major commitment to MOX use in civilian power generation and the scale needed for a limited campaign for the disposition of 50 or 100 tons of WPu. The range of cost estimates is also due in part to differences in comprehensiveness, accounting conventions, and conservatism—as discussed in general terms in Chapter 3 (see section "Issues and Criteria in Economic Evaluation of Alternatives")—employed in the development of the estimates.

Table 6-9 summarizes the results of our attempt to put on a somewhat more consistent basis a variety of the most recent (1992 and 1993) estimates of MOX fabrication costs found in the literature. The figures shown in the table have all been converted to 1992 dollars and reflect, insofar as has been possible given the information provided in the references, a common set of assumptions about contingency factors, real cost of money, construction time and interest during construction, treatment of dismantling and disposal costs, and other accounting conventions. (See the notes to Table 6-9 for details and "Issues and Criteria in Economic Evaluation of Alternatives" in Chapter 3 for background.) The range of unit costs shown in the table can be further reduced by adjusting for scale: if we apply the 0.6 power law mentioned above for the scaling of cost with output to determine, for each estimate in the table, the corresponding equivalent unit cost at 100 MTHM/yr output, the central values of the estimates for the 10 cases would be (in order of the cases in Table 6-9): $1,094, $1,438, $1,272, $1,486, $1,413, $1,340, $1,500, $1,574, $1,757, and $2,240.

The first of these figures is the GE estimate prepared for the U.S. Department of Energy's Plutonium Disposition Study (PDS); comparison with the other estimates suggests that it is unrealistically low. The third figure, prepared

   

case higher, and in the other case lower, than the panel's estimate, suggesting that the general range of the panel's estimate is correct. A study prepared by the Lawrence Livermore National Laboratory in support of DOE's plutonium disposition effort estimated the capital and operations costs of facilities designed to process 50 tons of plutonium in the form of pits to oxide suitable for fabrication into fuel in 20 years at $360 million and $23 million per year (in the cheapest case) to $664 million and $73 million per year (in the most expensive case) (Walter 1994). By contrast, an examination of the problem performed at Los Alamos, focusing on a somewhat different technique for carrying out the conversion, estimated that the capital and operations costs of a facility capable of processing nearly twice as many pits per year would be $100 million and $10 million per year (Toevs and Trapp 1994). (Both estimates appear to use 1994 dollars.)

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

for the PDS by Westinghouse, must similarly be regarded as low when one recalls that, unlike all the other estimates, it is said to include plutonium metal-to-oxide conversion. The last and highest figure in the list relates to a small European plant of outdated design and cannot be considered very relevant for our purposes. With the omission of these "outliers," the average of the remaining seven estimates is $1,500/kgHM. Inspection of the contributing factors to this total (see the notes to Table 6-9) shows that the difference between paying and not paying property taxes and insurance on the fuel fabrication facility amounts to about $150/kgHM. Hence we take, as our central estimates of fuel fabrication costs in new facilities at a scale of 100 MTHM/yr, figures of $1,425/kgHM if property taxes and insurance are not paid and $1,575/kgHM if these are paid. Our judgmental 70-percent confidence intervals extend ±$300/kgHM from these figures.

Before such estimates can be used in an economic assessment of a program for the disposition of WPu, however, they must be further adjusted to account more fully and consistently for "preoperational costs" such as research and development and safety and environmental analysis and licensing. These preoperational costs are often excluded or understated in cost estimates for a commercial operation (typically on the assumption that they are already "sunk" costs, or because it is assumed that they have been or will be covered by government); but, in assessing the costs to society of a WPu disposition program in a situation in which MOX fuel would not otherwise be being fabricated, the preoperational costs must be assigned to the disposition program.

In the contractor studies of MOX fuel-based disposition schemes performed in the U.S. Department of Energy's PDS, the estimates of preoperational costs varied quite widely. Some of this variation appears to be due to differing conventions by which various types of preoperational costs were lumped with or separated from capital costs. As best we can tell, fuel fabrication plant preoperational costs in the contractor-analyzed cases closest in scale to our 100-MTHM/yr reference level, were $39 million (GE 1993, p. 10.5), $45 million (ABB-CE 1993, p. VI-14), and $68 million (Westinghouse 1993, p. 3-3); as percentages of the sum of direct plus indirect construction costs for the indicated fuel fabrication plants, these figures translate to 11.6, 10.0, and 17.5 percent, respectively. The Technical Review Committee of the PDS criticized all of the contractor studies for understating preoperational costs (USDOE 1993a, p. SC6-5). We therefore estimate the preoperational costs for a 100-MTHM/yr fuel fabrication plant as a 20-percent addition to overnight construction costs (direct + indirect + contingency), hence about $120 ± $30 million for a 100-MTHM/yr operation.17 Assuming these preoperational costs are distributed over a nine-year

17  

As can be seen from the notes to Table 6-9, a fuel fabrication plant that produces 100 MTHM/yr for $1,425-$1,575/kgHM (with and without taxes and insurance) would have overnight construction costs of about $600 million, on which a 20-percent addition makes $120 million.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

S-curve prior to operation, the IDC (interest during construction) factor is 1.41 and the contribution to capital investment at start of operation is 1.41 × $120 ± $30 million = $170 ± $40 million. Using the usual fixed charge rates (FCRs) this translates to $140 ± $30/kgHM without property taxes and insurance and $170 ± $40/kgHM with property taxes and insurance.

The foregoing figures for MOX fabrication costs do not include metal-to-oxide conversion, nor do they include uranium feed. For our reference 4.8-percent plutonium fuel, and based on a conversion rate corresponding roughly to 100 MTHM/yr at 5 percent plutonium, the metal-to-oxide costs would add $7 ± $2/gPu × 49 gPu/kgHM = $343 ± $98/kgHM. The uranium costs are just $10/kgHM, so the addition for these two items is about $350 ± $100/kgHM. Thus we have altogether, for the costs of making MOX fuel at 4.8 percent plutonium and 100 MTHM/yr in new facilities including preoperational costs and plutonium metal-to-oxide conversion:

fabrication plant costs without preoperational-cost increment

$1,425 ± $300/kgHM without paying property tax and insurance

$1,575 ± $300/kgHM if property tax and insurance are paid

preoperational-cost increment

$140 ± $30/kgHM without paying property tax and insurance

$170 ± $40/kgHM if property tax and insurance are paid

cost of metal-to-oxide conversion and uranium acquisition

$350 ± $100/kgHM.

These figures combine to totals of $1,915 ± $318/kgHM without property tax and insurance and $2,095 ± $319/kgHM with property tax and insurance (noting that the ranges combine as the square root of the sum of the squares); we round these to $1,900 ± $300/kgHM and $2,100 ± $300/kgHM, respectively.

In addition to the possibility of building MOX fuel fabrication facilities from scratch, there exists in the United States the possibility of completing the SAFLINE MOX fabrication line that now stands unfinished at the FMEF on DOE's Hanford reservation in Washington state. USDOE (1993b) has estimated the cost of completing the SAFLINE MOX fabrication line to a capacity of 50-MTHM/yr average output, meeting modern health and safety standards, at $75 million, based on somewhat dated proprietary studies that we have not been able to examine. A more recent estimate commissioned in connection with the Isaiah Project from the individual who was responsible for the most recent work on the FMEF facility (Dahl 1993) gives a figure of “less than $150 million," explicitly including metal-to-oxide conversion. Operating costs were given as $1,500/kgHM in the 1988 study cited by DOE (which, assuming these were 1988 dollars, would be $1,740/kgHM in 1992 dollars); this would seem very high if it did not include metal-to-oxide conversion, so we assume that the DOE

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-9 Partially Adjusted Estimates from Literature on MOX Fuel-Fabrication Costs

1992 dollars per kilogram of heavy metal—uranium + plutonium—in MOX fuel, excluding, except where noted, costs of conversion of Pu metal to oxide and "preoperational costs" [e.g., research and development, safety and environmental assessment and licensing, and plant startup and testing]. Extra significant figures are to assist in checking calculations; the precision is illusory.

Cost and Source of Estimate

Basis

$845 to $907 (GE 1993)a

174-MTHM/yr average output U.S. plant operating for 30 years, without metal-to-oxide conversion.

$1,026 to $1,153 (USDOE 1992)b

200-MTHM/yr average output commercial plant in France, operating for 30 years, without metal-to-oxide conversion.

$1,041 to $1,122 (Westinghouse 1993)c

200-MTHM/yr capacity U.S. plant, 150-MTHM/yr average out-put, 30-year operation, with metal-to-oxide conversion.

$1,126 ± $307 (OECD 1992)d

200- to 400-MTHM/yr output commercial plant in Europe, 30-year operation, without metal-to-oxide conversion.

$1,341 to $1,484 (ABB-CE 1993)e

100-MTHM/yr average output U.S. plant, 30-year operation, without metal-to-oxide conversion.

$1,427 to $1,576 (USDOE 1993c)f

100-MTHM/yr capacity U.S. plant, 30-year operation, without metal-to-oxide conversion.

$1,500 ± $300 (USDOE 1993c)g

Range of reported estimates for costs of MOX fabrication in European plants in the 100-MTHM/yr size class.

$1,650 ± $170 (Nuclear Fuel 1992)h

Estimated range for commercial-scale European plants (taken to be 100-125 MTHM/yr).

$1,890 (Kessler et al. 1992)i

Estimate for German plant operating at 120 MTHM/yr.

$3,900 (Nuclear Fuel 1993)j

MOX fabrication fees reportedly being charged by Siemens to German utilities in 1990-1991, at output below 20 MTHM/yr plant.

NOTES: The figures shown are based on a starting point of PuO2 powder delivered to the fuel fabrication plant; costs of conversion of plutonium metal to PuO2 powder and costs of transport of the PuO2 to the fabrication plant (if any is required) are not included in these figures, nor is any charge for the WPu itself, or for the uranium feed and its conversion to UO2 and transport to the fabrication plant, or for cost of ultimate disposal of the spent fuel (or of HLW derived from it). These costs are considered in adjacent subsections of the report. Estimates of construction costs and operating costs for MOX fuel fabrication plants from the indicated references have been put on a consistent basis by the Reactor Panel using the following conventions: contingency factor = 25 percent of direct plus indirect construction costs (i.e., construction-cost estimates that did not include a contingency were multiplied by 1.25, and those that included a different contingency fraction, c, were multiplied by 1.25 / [1+c] ); annual real cost of money r = 0.07; interest during construction (IDC) computed at r = 0.07 on the basis of S-curves for cumulative construction investment over a construction period of six years unless otherwise specified, giving IDC of 0.27; dismantling and disposal (D&D) costs for facilities assumed to require availability at shutdown of a fund equal to 10 percent of the initial capital investment (including interest during construction) (USDOE 1988b, p. 43) accounted for an annuity taken from operating costs and invested at real interest rate i = 3 percent per year (see "Issues and Criteria in Economic Evaluation of Alternatives: in Chapter 3), with magnitude equal to this sum times 0.03 / [(1.03) n - 1)], where n is the operating

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

lifetime; levelized annual capital charges figured as direct plus indirect construction charges, multiplied as indicated by appropriate contingency and IDC multipliers, then multiplied by fixed charge rate FCR = r (1+r)n / [(l+r)n - 1], where n is the total operating lifetime or the duration of the plutonium disposition campaign (if facilities not expected to have any subsequent use). If a private entity were building the facility, the r value might be higher and the FCR would have to be increased by a further 0.02/yr or so to cover property taxes and insurance.

a GE (1993, p. 10.10) estimates the direct construction cost of a BWR fuel fabrication plant with an average output of 174 MTHM/yr (enough for six 1,300-MWe reactors using 29 MTHM/yr each at 37,100-MWd/MTHM average burnup) at $240.2 million, and application of General Electric's (GE's) indirect cost factor of 0.4 gives direct plus indirect costs of $336.2 million. With contingency, overnight costs become 1.25 × $336.2 = $420.4 million, and with interest during construction the total capital cost is 1.27 × $420.4 million = $533.9 million. The levelized-annualized capital costs (r = 0.07, 30 years) are $43.0-$53.7 million/yr or $247-$309/kgHM. GE (1993, p. 10.13) gives the operation and maintenance (O&M) costs for this plant as $103 million/yr, and with D&D annuity equal to $53.4 million × 0.03 / [(1.03)30 - 1] = $1.1 million/yr, the total O&M costs are $104.1 million/yr or $598/kgHM and the total fuel fabrication costs are $845-$907/kgHM. These figures do not include plutonium metal-to-oxide conversion (GE 1993, p. 3.16).

b Indicated cost estimate based on construction of a new MOX fabrication plant in France at a cost of $1,000 million for a plant with average output 200 MTHM/yr, which would run for 10 years at 5 percent WPu to process 100 MT WPu, then run another 20 years in commercial operation using reactor plutonium. We assume the $1,000 million includes contingency but not IDC and D&D. Then IDC at r = 0.07 and 6-year construction is 1.27 giving initial capital cost = $1,270 million. FCR = capital recovery factor (CRF) is 0.0806 (no increment for property taxes or insurance is considered), so levelized-annualized cost is $102.4 million and capital contribution to MOX fuel cost is $512/kg. Operating cost is quoted as $100 million/yr, to which D&D charge adds $127 million × 0.04 / [(1.07)30 - 1] = $2.7 million, so total operating cost is $102.7 million/yr / 200,000 kgHM/yr = $514/kgHM, and the fuel cost is $512/kgHM + $514/kgHM = $1,026/kgHM. Addition of 0.02/yr contribution to FCR, corresponding to property taxes and insurance that would be typical for a private operator in the U.S. context, would increase the total by $127/kgHM to $1,153/kgHM.

c The Westinghouse contractor study for the 1992-1993 Department of Energy Plutonium Disposition Study (PDS) gave an estimate of direct plus indirect construction costs for a plant with nameplate capacity of 200 MTHM/yr and average output of 150 MTHM/yr as $385.4 million (Westinghouse 1993, p. 3-12), a figure said to include the capacity to convert plutonium metal to oxide (Westinghouse 1993, p. 1.4-1). With 25 percent contingency and 27 percent IDC (six years, r = 0.07), the initial capital cost becomes $611.8 million; for a 30-year plant life, FCR = 0.0806/yr without property taxes and insurance and 0.1006 if these are included, so the annualized capital charges are $49.3-$61.5 million, or at 150 MTHM/yr $329-$410/kgHM. Westinghouse (1993, p. 3-22) gives operating costs for this plant as $105.5 million/yr, and the D&D annuity for a 30-year operating period would add $61.2 million × 0.03 / [(1.03)30 - 1] = $1.3 million/yr, so the total is $106.8 million/yr or $712/kgHM, and the sum of capital and operating charges is $1,041-$1,122/kgHM.

d The central figure is the OECD study's "reference case" value of $1,100/kgHM in 1991 dollars, explained in the report's Annex 6 and applicable during OECD's assumed operating period of 2007-2035, and here converted to 1992 dollars as $1,100 × 1/0.977 = $1,126. The figure is based on the assumption that MOX fuel fabrication will cost four times the report's $275/kgHM figure for LEU fuel fabrication. The sensitivity studies cited in the report yielded a range of $800-$1,400/kgHM (1991 dollars) surrounding the "reference" value, hence ±$300 ´ 1/0.977 = ±$307.

e The ABB-Combustion Engineering estimate of direct plus indirect construction costs for a 100-MTHM/yr average output fuel fabrication plant in connection with DOE's PDS is $450 million (ABB-CE 1993, p. VI-14, interpreted as the sum of direct plus indirect charges but without contingency in USDOE [1993a, pp. SC6-10 and A-13]). With contingency, the overnight cost is 1.25 × $450 million = $562.5 million, and the total capital costs with IDC are $562.5 million × 1.27 = $714.4 million. At r = 0.07 and with a 30-year operating life, the FCR is 0.0806 without property taxes and insurance and 0.1006 with these, so the levelized annual capital charges are

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

$57.6-$71.9 million/yr, and at 100 MTHM/yr these translate to $576-$719/kgHM. ABB-CE estimates the operating costs of this plant to be $75 million/yr (ABB-CE 1993, p. VI-16) and the D&D annuity adds $71.4 million ´ 0.03 / [(1.03)30 - 1] = $1.5 million/yr, so total operating costs are $76.5 million/yr, or $765/kgHM. The total MOX fabrication costs are then $1,341-$1,484/kgHM; in the ABB-CE analysis these do not include plutonium metal-to-oxide conversion (ABB-CE 1993, p. 111-82).

f USDOE (1993c) quotes an overnight cost of $440 million (1992 dollars) for a 100-MTHM/yr capacity plant and annual operating costs of $62 million (from the Nuclear Energy Cost Data Base maintained at the Oak Ridge National Laboratory) including D&D annuity. With IDC (r = 0.07, six years) the capital cost is $558.8 million, and the levelized annual capital costs (r = 0.07, 30 years) are $45.0-$56.2 million/yr (with and without property taxes and insurance). We assume this plant operates at 75-percent capacity factor, yielding an average output of 75,000 kgHM/yr, hence capital charges costs of $600-$749/kgHM and operating costs of $827/kgHM, for total costs of $1,427-$1,576/kgHM. We assume that, following conventional practice in LWR fuel-cycle analysis, the input to this fuel fabrication plant was assumed to be plutonium dioxide, so that metal-to-oxide conversion costs are not included.

g USDOE 1993c quotes a range of estimates that have been made for European plants in the 100 MTHM/yr class as extending from $1,200-$1,800/kgHM, hence $1,500 ± $300.

h Quoted in Berkhout et al. (1993). The range given was $1,300-$1,600/kg of MOX, which translates to $1,480-$1,820/kgHM.

i Based on Kessler et al. (1992, p. 25) where MOX fabrication costs "are estimated at about DM 3,000/kg at present in a fabrication plant with a throughput of 120 t/yr and adequate plant utilization." At $0.63/1 DM, 3,000 DM equals $1,890 in U.S. dollars.

j The figure is based on a successful Siemens lawsuit against the state of Hesse for damages in the form of lost revenues in a state-ordered shutdown of that firm's 25-MTHM/yr MOX fabrication plant in Hanau. Lost revenues were estimated at DM 550,000 per day for an output of 103 kg/d of MOX measured as oxide, hence 6,070 DM/kgHM, which at $0.63/1 DM is about $3,800/kgHM in 1991 U.S. dollars or $3,900/kgHM in 1992 U.S. dollars.

capital-cost as well as operating-cost estimates for the FMEF MOX fabrication line do include this conversion. The Dahl (1993) study did not estimate operating costs.

In constructing our own estimate of the cost of MOX fuel from the FMEF facility, we use a range of $100-$150 million in completion costs, which we take to be overnight costs; with an assumed four-year construction time, hence IDC multiplier of 1.19, the initial capital investment would be $119-$179 million. Since this facility is already on a federal site, we use only the lower fixed charge rate corresponding to no property tax or insurance charges, giving levelized-annualized capital charges of $9.6-$14.4 million/yr for a 30-year operating life, or $192-$288/kgHM at 50 MTHM/yr.

Concerning operating costs at FMEF, scaling from a central estimate of $800/kgHM for the operating cost of a new plant producing 100 MTHM/yr (see notes to Table 6-9) would give $1,060/kgHM for a 50-MTHM/yr plant, and adding $226/kgHM for the operating-cost component of metal-to-oxide conversion18 gives roughly $1,290/kgHM. Because per-unit-output operating costs at

18  

Based on $7/gPu total at a scale corresponding to 100 MTHM/yr and 49 gPu/kgHM, increased by 20.4 = 1.32, as derived earlier in this section, assumed split 50/50 between capital charges and operating costs.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

FMEF are expected to be higher than at a new plant, we take this to be the lower limit of a range with the DOE estimate of $1,740/kgHM at the upper end. Assuming dismantling and disposal (D&D) costs require $40 million at the end of plant life, the 30-year annuity payment required to produce this sum if invested at 3 percent per year real interest rate is $0.8 million/yr or $17/kgHM, so our operation and maintenance (O&M) costs are $1,310-$1,760/kgHM. The capital plus operating costs for the FMEF MOX fabrication line, not including preoperational costs, would then be $1,500-$2,050/kgHM.

We suppose that part of the preoperational costs already have been spent in the case of the FMEF MOX fabrication line, and we consider these sunk. Some additional preoperational costs probably are embedded in the DOE and Dahl estimates of the completion costs for the facility. We assume the incremental preoperational costs would be $20-$30 million (20 percent of $100-$150 million); if these are incurred over a six-year period prior to operation (greater than the assumed four-year construction period), they will add 1.27 × $20-$30 million = $25-$38 million, which translates to $40-$60/kgHM. The total costs of MOX fuel fabrication at FMEF, at a scale of 50 MTHM/yr, including metal-to-oxide conversion, are thus $1,530-$2,110/kgHM, which we round to $1,800 ± $300/kgHM.

MOX Incremental Costs of Plutonium Storage and Transport

In the case of LEU, costs of storage and transport at and between the various steps of fuel preparation are included in the cost figures that were given. In the case of MOX fuel the storage and transport costs can be considerably higher (because of the extra ES&H and security hazards posed by the plutonium), and not all of them have been included in the estimates presented above. (It may be presumed that storage at the plutonium conversion plant and MOX fuel fabrication plant is included in the estimates of the costs of these activities, but transport costs and any extra costs associated with the storage of MOX fuel after it leaves the fabrication plant but before it is loaded into the reactor core are not included.) The OECD study (1992) cites plutonium storage costs of $1-$2/gPu per year and plutonium transport costs of $500-$900/kgPu for transport within the European community (i.e., excluding long-distance transport by ship). We assume that the incremental storage costs for a MOX fuel operation in the United States will be between $0 and $2/gPu and that the incremental transport costs will be between $0.5 and $1.5/gPu, so that the total incremental costs for storage and transport are $2 ± $1/gPu, hence about $100 ± $50/kgHM for 49 gPu/kgHM.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×
MOX: Cost of Ultimate Disposal

As indicated in the corresponding discussion under LEU, ultimate disposition costs for MOX fuels may range from the same as to perhaps twice as high as those for LEU fuels, but because of the large uncertainties associated with these costs for either fuel we present our main comparisons without them.

MOX: Estimated Total Fuel Costs, Less Disposal

A campaign to process 50 tons of WPu into MOX fuel with our reference-case plutonium content of 4.8 percent would entail fabricating about 1,000 MTHM of MOX fuel, which could be done in 20 years at an average fabrication rate of 50 MTHM/yr or in 10 years at an average rate of 100 MTHM/yr. If the intent were to use this fuel in one or two reactors with full-MOX cores (as might be desirable to minimize transport), the 50-MTHM/yr option would be adequate. If the aim were to complete the campaign as quickly as possible (at the cost of somewhat higher capital investment in fabrication capacity and the engagement of a larger number of reactors), the 100-MTHM/yr option would be appropriate. This second option would of course also be adequate to handle, over a period of 20 years, a case in which the campaign doubled in size from 50-100 tons of WPu because of either deeper arsenal reductions or acquisition of Russian plutonium for disposition in the United States. A fuel fabrication rate as low as 25 MTHM/yr, applied over a period of 30 years, would suffice to process 50 tons of WPu if plutonium loading were increased to 6.7 percent by weight.

The range of scales of MOX operations likely to be considered for a WPu disposition campaign therefore extends from about 25 MTHM/yr to about 100 MTHM/yr. For our reference-case MOX fuel, designed to deliver 40,000 MWd/MTHM with end-of-life reactivity equal to that of 4.4-percent enriched LEU after the same irradiation, we estimate the fuel-cycle costs at these scales of operation in the period from 2000 to 2030 to be as shown in Table 6-10, excluding repository fees. (Scaling from 100-MTHM/yr operations in a new plant and from 50-MTHM/yr operations at FMEF has been carried out assuming total costs in a plant of a given type scale with the 0.6 power of output, so that unit costs decline with the 0.4 power of output.)

Net Fuel-Cycle Costs of Using WPu in PWRs

For the reference case we have been considering—substitution of WPu in MOX for 4.4-percent U-235 (LEU) PWR fuel achieving an irradiation of 40,000 MWd/MTHM—we found above that the relevant levelized-annualized fuel cost (less the cost of ultimate waste disposal) for LEU fuel in the period 2000-2030 is $1,400 ± $200/kgHM. In Table 6-11, this figure is combined with the MOX

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-10 Summary of MOX Fuel-Cycle Costs (in 1992 U.S. dollars)

 

New Fuel Fabrication Plant

 

No Tax and Insurance

With Tax and Insurance

Fuel Fabrication at FMEF

MOX Production Costs

100 MTHM/yr

1,900±300

2,100±300

NAa

50 MTHM/yr

2,500±400

2,800±400

1,800±300

25 MTHM/yr

3,300±500

3,700±500

2,400±400

Incremental Plutonium Storage and Transport

100±50

100±50

100±50

Totals

100 MTHM/yr

2,000±300

2,200±300

NAa

50 MTHM/yr

2,600±400

2,900±400

1,900±300

25 MTHM/yr

3,400±500

3,800±500

2,500±400

NOTES: MOX production costs include fabrication with preoperational costs, plutonium metal-to-oxide conversion, and depleted uranium acquisition and conversion. Costs are in 1992 dollars/kgHM and apply to the small-scale MOX operations that would be associated with a program restricted to disposition of WPu. They do not apply to a large-scale commercial MOX program, in which economies of scale could be gained from higher outputs and preoperational costs would be distributed over much higher production. Uncertainty ranges are judgmental 70-percent confidence intervals (corresponding approximately to one standard deviation of a normally distributed random variable).

a NA indicates not applicable. To our knowledge, expanding the FMEF MOX fabrication capability to 100 MTHM/yr has not been studied.

fuel-cycle-cost estimates from Table 6-10 to give a set of values for the expected excess of MOX fuel-cycle costs over those for LEU, expressed in dollars per kilogram of heavy metal, per electrical kilowatt-hour, and per gram of WPu.

The differential costs in Table 6-11 can be converted readily into annual costs and into net discounted present values of such cost streams for a full plutonium disposition campaign. For example, at a MOX fabrication rate of 50 MTHM/yr and the indicated plutonium loading of 4.8 percent in heavy metal, a 50-tons plutonium campaign would require 21 years of MOX fuel fabrication and corresponding MOX-based electricity generation, at an excess cost of $91.4 ± $27.4 million/yr (including fuel carrying charges) if the MOX is fabricated in a new plant that pays property taxes and insurance, $73.1 ± $27.4 million/yr if it is fabricated in a new plant that does not pay taxes and insurance, and $29.5 ± $20.6 million/yr if it is fabricated at FMEF. The discounted net present values of these cost streams at the start of operation, with real cost of money at 7 percent per year, are $991 ± $297 million, $792 ± $297 million, and $329 ± $231 million, respectively. (A campaign with this timing and plutonium loading could be accomplished using two 1,250-MWe PWRs with full-MOX cores or six such

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-11 Fuel-Cycle Cost Differentials for MOX Versus LEU Fuel (in 1992 U.S. dollars)

 

MOX from New Fabrication Plant

 

No Tax and Insurance

With Tax and Insurance

MOX Fabrication at FMEF

Cost excess of MOX over LEU fuel in $/kgHM if MOX is fabricated at a scale of

100 MTHM/yr

600+350

800±350

NAa

50 MTHM/yr

1,200±450

1,500±450

500±350

25 MTHM/yr

2,000±550

2,400±550

1,100±450

Equivalent differential in electricity generation cost in $/MWhb

100 MTHM/yr

2.3±1.3

3.0±1.3

NAa

50 MTHM/yr

4.6±1.7

5.7±1.7

1.9±1.3

25 MTHM/yr

7.6±2.1

9.1±2.1

4.2±1.7

Equivalent cost per gram of WPuc

100 MTHM/yr

15±9

20±9

NAa

50 MTHM/yr

30±11

37±11

12±9

25 MTHM/yr

50±14

60±14

27±11

NOTES: Figures apply to fuel irradiated to an average of 40,000 MWd/MTHM, with initial loadings of 4.8 percent WPu in heavy metal (MOX case) and 4.4 percent U-235 in heavy metal (LEU case).

a NA indicates not applicable. Expanding the FMEF MOX fabrication capability to 100 MTHM/yr has not been studied.

b Thermal-to-electric conversion efficiency = 0.33, fuel accounting life five years with real cost of money r = 0.07.

c Equal to cost per kilogram of heavy metal times carrying-charge factor on five-year fuel cycle, divided by 49 gPu/kgHM. Rounded to nearest dollar.

PWRs with one-third MOX cores, and could perhaps begin, if the fuel were fabricated at FMEF, shortly after the year 2000; see Table 6-2.)

Variations in Fissile Content and Burnup

The estimates derived from calculations along the foregoing lines depend on the assumptions chosen about the fissile content and burnup of fuel. At lower fissile content, and burnup, the economic comparison becomes even less favorable to MOX fuel; at higher fissile content and burnup, it becomes more favorable to MOX. The reason for this is that two of the main components of LEU fuel cost per kilogram of heavy metal—uranium-acquisition and enrichment costs—increase sharply as the fissile content rises, while the cost of MOX fuel fabrication per kilogram of heavy metal, which is by far the largest component

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

Table 6-12 MOX-LEU Fuel-Cycle Cost Comparison Versus Enrichment and Burnup (in 1992 U.S. dollars)

 

30,000 MWd/MTHM

40,000 MWd/MTHM

50,000 MWd/MTHM

 

3.3% U-235 in LEU

3.6% WPu in MOX

4.4% U-235 in LEU

4.8% WPu in MOX

5.5% U-235 in LEU

6.0% WPu in MOX

Fuel cost, $/kgHM

1,040

2,120

1,400

2,200

1,780

2,280

Fuel cost, $/MWh

5.2

10.5

5.2

8.2

5.3

6.8

MOX penalty, $/kgHM

— 1,080 —

— 800 —

— 500 —

MOX penalty, $/MWh

—5.3 —

—3.0 —

—1.5—

NOTE: Costs relate to PWRs fed by new MOX fuel fabrication plants paying property taxes and insurance and operating at 100 MT IM/yr.

of MOX fuel cost in a situation where the plutonium metal is obtained free of cost, goes up only a little or not at all with plutonium loading.19

The sensitivity of the cost comparison to fissile content is illustrated quantitatively in Table 6-12, where it has been assumed that the only component of MOX fuel cost, from a new fuel-fabrication plant, that varies with plutonium content is the cost of conversion of plutonium metal to oxide (taken to be $7/gPu at 100 MTHM and 5 percent plutonium by weight in heavy metal). It is apparent that, if the MOX made from WPu were replacing 5.5-percent enriched LEU, the central-estimate cost penalty would be cut almost in half compared to the 4.4-percent enriched LEU substitution, and the penalty per kilowatt-hour of electricity generated would fall by more than twofold. It cannot be assumed, however, that further reductions in the WPu/MOX cost penalty would be obtainable by going to still higher plutonium loadings in PWR MOX, because burnups above 50,000 MWd/MTHM are not available in LWRs with current fuel designs. (This limitation might change in the future, but then fuel fabrication costs probably also would change, invalidating the present calculations.)

Costs of Reactor Modifications and Licensing

The preceding estimates do not take account of the costs of any modifications to existing LWRs in order to enable them to utilize 100-percent MOX cores, nor do they include costs that would be associated with safety analysis

19  

There may be some component of the fuel fabrication cost that depends on plutonium content rather than just on total amount of heavy metal in fabricated fuel, but we have not been able to find an estimate of such an effect either in the literature or in conversations with individuals experienced in the design and operation of MOX fuel fabrication plants.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

and licensing of MOX use in existing LWRs whether at one-third or full-MOX operation.

If it were decided to use only one-third MOX cores, or to use full-MOX cores in one or more of the three operating U.S. LWRs that were designed for full-MOX operation (the three ABB-Combustion Engineering System-80 reactors at the Palo Verde station in Arizona), no costs would be incurred for reactor modification; but there would still be costs for MOX-related safety analyses and licensing. A study of completing, for the purpose of WPu disposition, the WNP-1 1,250-MWe PWR in Washington state estimated the "safety, licensing, and permitting" costs of doing this at $94 million (SAIC 1993a); what fraction of this total is related to MOX use per se, and hence applicable to the case of licensing an already operating reactor for this purpose, is not clear. In the Plutonium Disposition Study of the Department of Energy (USDOE 1993a), vendor estimates of "preoperational costs" (a category that includes research and development and plant testing as well as safety analysis and licensing) for advanced LWRs ranged from $93 to $244 million; and DOE's Technical Review Committee complained that the vendors had underestimated the safety analysis and licensing costs. We assume here that the safety and licensing costs for MOX operation in existing U.S. LWRs would fall in the range of $100 ± $50 million.

We have seen no detailed estimates for the cost of modifying an operating U.S. LWR to be able to use a full-MOX core, if that were required. 20 Our own very rough estimate is that, in a case where substantial modifications to the control systems were in fact necessary, such modifications might cost $100 ± $50 million in labor and equipment, while requiring a shutdown of the reactor for perhaps a year. The lost electricity revenues from a year's shutdown of a 1,200-MWe reactor that had been operating at 75-percent capacity factor would amount to about $390 million if the busbar value of the electricity is $0.05/kWh (see Chapter 3), or $390 ± $130 million allowing for a ±33-percent uncertainty range on the product of shutdown time and electricity value.

Net Economic Costs of Using WPu in Currently Operating PWRs

Consider a reference case in which two currently operating PWRs in the 1,200-MWe class are to load the nominal 50 tons of surplus U.S. WPu over a period of 21 years, using 100-percent MOX cores with 48 kgPu/MTHM and average burnup of 40,000 MWd/MTHM. If the reactors in question need no modification to use 100 percent MOX and if safety analysis and licensing costs

20  

Analyses performed by vendors in the second phase of DOE's Plutonium Disposition Study. which became available late in the panel's deliberations, suggest—contrary to previous assumptions—that several existing LWR types besides the System-80 could in fact use 100-percent MOX cores safely without undergoing significant modification. Further analysis and review will be required before this conclusion can be considered firm. See also "Environment, Safety, and Health" below.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-13 Net Economic Impact of WPu Disposition in Currently Operating LWRs (in millions of 1992 U.S. dollars)

 

No Reactor Modifications Required

Reactor Modifications Required

Fuel From:

FMEF

New Plant no Tax

New Plant with Tax

FMEF

New Plant no Tax

New Plant with Tax

MOX fuel costs

329±231

792±297

991±297

329±231

792±297

991±297

Costs of reactor

Licensing

127±64

127±64

127±64

127±64

127±64

127±64

Modification

0

0

0

1,000±300

1,000±300

1,000±300

Rounded Totals

456±240

919±304

1,118±304

1,456±384

1,919±427

2,118±427

NOTE: Figures are discounted present value, as of start of operation (shortly after the year 2000), of the net incremental costs of using WPu-MOX instead of LEU in two 1,200-MWe PWRs to disposition 50 tons of WPu in 21 years of reactor operation (assuming 48 gPu/kgHM as loaded-49 gPu/kgHM manufactured—and average burnup of 40,000 MWd/kgHM).

of $100 ± $50 million in 1992 dollars are assumed to accumulate in the usual Scurve during a period of six years prior to the start of MOX operation at the reactors, the contribution to the discounted net present value of incremental MOX costs at the time of startup, for real cost of money at 7 percent per year, would be 1.27 × ($100 ± $50 million) = $127 ± $64 million.

If both reactors need substantial modifications to be able to use 100-percent MOX cores, the estimated additional costs are 2 × ($100 ± $50 million) in labor and equipment plus 2 × ($390 ± $130 million) in lost electricity revenues during shutdown, or about $1,000 ± $300 million in total modification costs. Table 6-13 combines these estimates with the above-derived net discounted present value of the extra fuel-cycle costs as of start of operations, giving a range of central estimates of the net economic impact of using currently operating PWRs for WPu disposition extending from $450 to $2,100 million net discounted present value at start of operations, depending on whether FMEF or a new MOX fuel fabrication plant is used and on whether substantial modifications to permit 100-percent MOX use are required or not.

Variations in Reactor Type

The economics of using WPu in reactors would be more attractive for reactors that require high levels of enrichment and derive high burnups from it. In a liquid-metal fast-breeder reactor (LMFBR), for example, fuel to be used in the core would typically be enriched to 20 percent fissile plutonium or 30 percent U-23521 and would achieve a burnup of 100,000-150,000 MWd/MTHM. The uranium requirement for 30-percent enriched fuel is 63.3 kgU/kgHM and the

21  

Plutonium is substantially more reactive in a fast-neutron spectrum than is U-235.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

separative work requirement is 64.2 SWU/kgHM (assuming 0.25 percent U-235 in the tails). The Nuclear Energy Cost Data Base (USDOE 1988b) reference value for the fabrication cost of plutonium-based LMFBR core fuel is $2,700/kgHM (in 1992 dollars), compared to $500/kgHM for fabricating enriched-uranium LMFBR fuel. Thus, with the usual assumption of 0.5 percent losses in conversion and 1 percent losses in fabrication, and with uranium acquisition and conversion costs of ($55 + $9)/kgU and enrichment costs of $95/SWU, the cost of enriched uranium fuel for an LMFBR core (without ultimate disposal costs) would be

1.015 × 63.3 kgU/kgHM × $64/kgU + 1.01 × 64.2 SWU/kgHM × $95/SWU + $500/kgHM = $10,772/kgHM.

On the MOX side, allowing $9/gPu for conversion to oxide and incremental storage and transport costs, as before, and $7/kgHM for acquisition and conversion costs for 0.7 of depleted uranium per kilogram of heavy metal, leads to a corresponding MOX fuel cost of

$7/kgHM + 1.015 × 200 gPu/kgHM × $9/gPu + $2,700/kgHM  = $4,534/kgHM.

Thus a cost advantage of about $6,000/kgHM is predicted for the use of plutonium in a MOX-fueled LMFBR, given cost-free WPu as the raw material (translating to $0.0074/kWh on a levelized basis, assuming r = 0.07, a four-year fuel cycle, irradiation of 100,000 MWd/MTHM, and thermal-to-electric conversion efficiency of 0.40). A similar result could be expected from a comparison of MHTGRs using 100-percent plutonium fuel against the use of 94-percent enriched uranium fuel in these reactors.

It should not be assumed, however, that the fuel-cost benefits of using free WPu in these advanced reactor types, compared to using enriched uranium in them, mean that there is a large economic benefit to be derived from choosing such reactors as the disposition option. The overall economic consequences of such a choice would depend also on the construction costs of such reactors, on their total fuel costs compared to those of LWRs, and on the time delay (and consequent charges for plutonium storage) before they could be deployed. These matters are discussed further in the section “Building New Reactors for Plutonium Disposition" below.

A last alternative reactor case for which the analysis of the economic effect of the use of WPu is especially interesting is that of the CANDU reactor, of which a substantial number are in commercial operation. These reactors normally use unenriched uranium-oxide fuel. A study by the manufacturer of the use of CANDU reactors for WPu disposition (AECL 1994) indicates that two such reactors at Canada's Bruce station could irradiate 50 tons of WPu in 24 years using a current-technology MOX fuel containing 1.2 weight percent plutonium (burnup 9.7 MWd/kgHM) and that four reactors could irradiate 100 tons

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

of WPu in 25 years using an advanced MOX fuel containing 2.1 weight percent plutonium (burnup 17.1 MWd/kgHM). Based on a rather detailed study of FMEF at Hanford, Washington, analysts at Atomic Energy of Canada, Limited, (AECL) conclude that this facility could be modified to fabricate the needed quantities of MOX CANDU fuel for an overnight cost of $118 million, and that the operating costs would be about $64 million per year for the reference MOX fuel and about 20 percent more for the advanced fuel (which, however, would fuel four reactors rather than two and, thus generate twice as much electricity), assuming plutonium metal-to-oxide conversion has been performed (and paid for) elsewhere.22

For AECL's assumed four-year construction time for modifications to the FMEF, the IDC factor would be 1.18 and the initial investment at startup would be 1.19 × $118 million = $140 million. The corresponding levelized constant dollar capital charges at r = 0.07 and a 24-year operating lifetime would be $12 million/yr (no insurance and taxes), which when added to the O&M costs of $64 million/yr gives fuel costs of $76 million/yr. (Divided by the 170.6-ton annual output of fuel, this gives $446/kgHM. Adding metal-to-oxide conversion costs of $7/gPu and incremental plutonium storage and transport costs of $2/gPu would add another $9/gPu × 2.12 MgPu/yr = $19 million/yr or $9/gPu × 12 gPu/kgHM = $108/kgHM.) Fueling costs for the standard natural uranium feed for these reactors are estimated by AECL at $16.3 million/yr for 199.4 MTHM/yr (burnup 8.3 MWd/kgHM), hence $82/kgHM. The costs of reactor-facility modification and reactor licensing to permit use of plutonium fuel are estimated by AECL at $37 million, which under the same assumptions as used above would translate to a levelized constant-dollar cost of $3.8 million/yr. The incremental cost of plutonium disposition under this option can be calculated, therefore, as $12 million/yr + $64 million/yr + $19 million/yr + $4 million/yr $16 million/yr = $83 million/yr for 24 years. The equivalent discounted present value at start of plutonium disposition operations at the reactor is $952 million.

Completing Existing LWRs

Within the option of using U.S. LWRs to process WPu into spent fuel, one of the potentially attractive variants is the possibility of completing, for this purpose, one or more existing partially completed reactors that exist in the United States, such as those of the Washington Public Power Supply System (WPPSS) or the Tennessee Valley Authority (TVA). Two of the WPPSS Nuclear Project reactors-WNP-1 and WNP-3-are of particular interest. The WNP-I facility is

22  

The assumption in the AECL, study, on which its estimates of FMEF completion costs and operating costs are based, that plutonium metal-to-oxide conversion is performed elsewhere, differs from the assumption of FMEF studies (cited above) that this conversion would be performed at the FMEF. The AECL analysts agreed that FMEF could be adapted to perform the conversion step, but they did not estimate the costs of this.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

a 1,250-MWe Babcock and Wilcox PWR located on DOE's Hanford reservation and is 63 percent complete. The WNP-3 facility is a 1,240-MWe Combustion Engineering System-80 PWR located at Satsop (west of Olympia) and is 75 percent complete. WPPSS has been keeping these reactors in mothballs for a number of years, but has recently decided to stop maintaining them and to sell the parts; thus to keep this option open would require government action in the near term.

The attractions of this variant include, above all, the possibility of processing the entire 50 tons of U.S. surplus WPu at a single reactor site (since either reactor could be completed with the capability to utilize a full-MOX core and could be expected to be operable for at least 30 years after startup) and the minimization of shipping of MOX fuel (since WNP-1 is on the Hanford reservation where a MOX fuel fabrication capability is already partly in place, and WNP-3 is less than 200 miles from that site). In addition, the circumstances of these two reactors lend themselves to government acquisition of the facilities, if it is deemed preferable to perform the plutonium disposition mission at a government site; and because the reactors are not now operating, equipping them with the capacity to use full-MOX cores can be accomplished without the loss of revenue that would be associated with shutting down one or more existing LWRs for modifications.

If only one of the two reactors is to be used for the plutonium disposition mission, the choice between them will involve a trade-off between (1) ease of completion and licensing for use with a full-MOX core, in which WNP-3 has an advantage because it is a Combustion Engineering System-80 reactor (designed from the outset to be able to use a full-MOX core), and (2) reducing MOX transport, in which WNP-1 has an advantage because it is located at the Hanford site where the MOX fuel is likely to be fabricated. The advantage of using both reactors for the plutonium disposition mission would be the capacity to load the quantity of MOX containing 50 tons of WPu in half the time that would be required if only one reactor were used. (Another two-reactor variant would be to use WNP-1 together with the WNP-2, which is an operating 1,100-MWe BWR located at the Hanford site. This might or might not entail shutting down WNP-2 temporarily for modifications to permit it to operate with a full-MOX core-see Chapter 4.)

The possibility of using the WNP-1 and WNP-3 reactors for this purpose has been intensively investigated and promoted by a consortium of companies calling this effort the Isaiah Project.23 In connection with this joint venture, the costs of completing the two reactors and operating them in a plutonium disposition mode have been explored. Under the Isaiah Project proposal, the consor-

23  

The consortium comprises Science Applications International Corporation (SAIC), Newport News Industrial Corporation, and Battelle Memorial Institute. The Isaiah Project is described in SAIC (1993a. 1993b).

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

tium would complete the WNP-1 and WNP-3 reactors and operate them as government-owned, contractor-operated facilities that would sell steam to WPPSS, which would retain ownership of the turbine generator facilities at the two reactors and would sell the electricity to the Bonneville Power Administration (BPA).

The U.S. government would pay all fuel-cycle and other operating costs (including, e.g., the costs of conversion of WPu to metal and of MOX fuel fabrication, and the costs of spent fuel disposal). A part of the revenue stream from the operation would provide the return to the private investors in the consortium; additional money borrowed against the rest of the anticipated revenue stream would be used to reduce by $2 billion the debt incurred by BPA in connection with the initial construction of these reactors, and to provide an additional $2 billion to DOE for use in other antiproliferation programs, perhaps including WPu disposition in the former Soviet Union.

We do not address here the merits of the particular institutional and financing arrangements proposed by the Isaiah Project, confining ourselves instead to summarizing and commenting on the Project's analysis of the costs and revenues associated with WPu disposition using the WNP-1 and WNP-3 plants.

The Project estimates that the completion of the two reactors would cost $3.3 billion (1993 dollars)-$1.7 billion for WNP-I and $1.6 billion for WNP-3—including costs of safety analysis, licensing, and permitting for full-MOX operation, but not including the turbine generator portion of the plants, which under the Isaiah proposal would be financed separately. The indicated figures are overnight costs, including both direct and indirect cost components plus a contingency of 10 percent. They are based on extensive recent analyses by the WPPSS and the reactor vendors and architect engineers originally responsible for these plants (WPPSS 1992a, 1992b). In the same analyses, the costs of completing the turbine generator part of the plants were estimated as $240 million for WNP-I and $100 million for WNP-3. The proponents argue that the credibility of these estimates is reinforced by their status as "capped-cost" figures, meaning that the contractors have offered to complete the plants at these costs, absorbing any overruns themselves.

To put these estimates into our own framework for calculating initial capital costs as the basis of a levelized-annualized capital charges calculation, we need to convert the figures to 1992 dollars and add a correction for IDC. The conversion to 1992 dollars consists of a multiplicative factor of 1/1.03 = 0.97 (see Table 3-6). Interest during construction, given the Isaiah Project's 1993 estimate of reactor startup in 1999 if work had been initiated at the beginning of 1994, would contribute a multiplicative factor of about 1.25 at a real cost of money of

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

7 percent per year (see Table 3-7).24 Thus the 1992 dollar completion costs of the two plants as of a 1999 startup date would be

WNP-1: ($1.7 billion + $0.24 billion) × 0.97 × 1.25 = $2.35 billion

WNP-3: ($1.6 billion + $0.10 billion) × 0.97 × 1.25 = $2.06 billion.

These are not the total construction investments as of startup but the incremental investments needed to finish partly completed plants. Thus it is not appropriate, strictly speaking, to apply the rule of thumb from the Chapter 3 section "Principles and Pitfalls in Cost Comparisons" that 10 percent of the initial investment will be needed at the end of a facility's life to cover its dismantling and disposal (D&D) costs. If, however, we use instead the figure recommended in DOE's Nuclear Energy Cost Data Base (USDOE 1988b) of $145 million (1987 dollars) for an 1,100-MWe plant, converting to 1992 dollars and scaling up for a 1,250-MWe plant would give an end-of-life D&D fund of $200 million, which is about what the 10-percent rule of thumb would give if applied to the completion costs. For a nominal 30-year reactor lifetime and real cost of money of 7 percent per year, the effective increment on initial capital investment is $200 million divided by (1.07)30, or $26 million. With this addition, the effective initial capital investments to complete the two plants would become, respectively, $2.61 and $2.32 billion.

Let us now estimate the costs and revenues of using one of these plants to process the entire nominal 50 tons of WPu into spent fuel in an operating lifetime of 30 years; this corresponds (see Table 6-1) to plutonium loading in fresh fuel of 6.8 percent of total heavy metal by weight, capacity factor of 75 percent, average burnup of about 42 MWd (thermal)/kgHM, and fuel fabrication of 25 MTHM/yr. If MOX fuel production costs are $2400 ± $400/kgHM (Table 6-10, for 25-MTHM/yr throughput) and if the incremental costs of plutonium storage and transport are a relatively low $1/gPu (in light of minimal transport between fuel fabrication and reactor), hence $70/kgHM, and with costs of $10/kgHM for uranium acquisition and conversion, the fuel-cycle costs less disposal would be about $2,500 ± $400/kgHM. If carrying charges on the fuel are based on a five-year fuel accounting lifetime at a real cost of money of 7 percent per year, then at 42.2 MWd/kgHM and a thermal-to-electric conversion efficiency of 0.33, these fuel-cycle costs amount to $0.0076-$0.0106/kWh, and the addition of the statutory $0.001/kWh for waste-disposal costs makes this range $0.0086-$0.0116/kWh. The nonfuel operation and maintenance costs for a large PWR are in the range of $0.012-$0.016/kWh according to DOE's Nuclear Energy Cost Data Base (USDOE 1988b, with conversion of the reference's 1987 dollars to 1992 dollars).

24  

Shifting the projected time from 1994 to 1996, and projected completion to 2001, would have only a modest impact on the projected cost.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

For a nominal plant lifetime of 30 years and a real cost of money of 7 percent per year, the levelized-annualized capital charges on the completion costs of $2.3-$2.6 billion would be

  • $2.3-$2.6 billion × $0.0806/yr = $185-$210 million/yr for an entity that did not pay property taxes and insurance, and

  • $2.3-$2.6 billion × $0.1006/yr = $231-$261 million/yr for an entity that paid property taxes and insurance amounting to 0.02/yr of the initial capital investment.

The corresponding range of capital charges per kilowatt-hour at 75-percent capacity factor would be $0.023-$0.032/kWh.

These figures do not include the cost of acquiring one of the WNP reactors from its current owners. The Isaiah Project allowed a cost of $1 billion per reactor for this purpose. (WPPSS had invested some $5 billion in the two reactors, but, inasmuch as they are now slated to be sold for scrap, the owners might be quite content with $1 billion per reactor, possibly less; see Lange and Hanson 1993.) Financing this additional $1 billion at 7 percent real cost of money over 30 years would add costs of $80.6-$100.6 million/yr (depending on whether a charge for property taxes and insurance is assessed against this sum), hence $0.0098-$0.0123/kWh.

In summary, our estimate of the cost of using one of the WNP reactors for disposition of 50 tons of WPu over a 30-year period, if expressed as a cost per kilowatt-hour of electricity generated in the process is:

MOX fuel-cycle costs, including waste disposal

$0.0101 ± $0.0015/kWh

nonfuel operation and maintenance

$0.0140 ± $0.0020/kWh

capital charges on completion costs

$0.0275 ± $0.0045/kWh

capital charges on $1 billion acquisition cost

$0.0111 ± $0.0012/kWh

 

----------------------------

Total (range is square root of sum of squares)

$0.063 ± $0.006/kWh

The range of $0.063 ± $0.006/kWh is to be compared with our estimate from Chapter 3 of $0.050 ± $0.015/kWh for the avoided costs associated with baseload electricity generation in connection with plutonium disposition in new plants. Combining the ranges based on the square root of the sum of the squares gives an expected net cost of $0.013 ± $0.016/kWh, or, stated another way, our 70-percent judgmental confidence interval for the net economic effect of the use of WNP- 1 or WNP-3 for WPu disposition under the indicated assumptions, on a levelized-annualized basis, extends from a profit of $0.003/kWh to a loss of $0.029/kWh. Over 30 years of operation at 75-percent capacity factor the range extends from a profit of $0.7 billion to a loss of $7.1 billion. (These are sums of constant-dollar annual figures; the corresponding net present values at start of reactor operation, in 1992 dollars, extend from a profit of $0.3 billion to a loss of $2.9 billion.)

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

If it were taken into account that gas-fired combined-cycle generation costs in the Northwest (SAIC 1993a) appear to be at the low end of the range cited in Chapter 3, however, the result would tilt even farther toward the net cost side. If the 70-percent confidence range on the avoided cost were $0.04 ± $0.01/kWh, for example, the 70-percent confidence range on the net cost per kilowatt-hour would be $0.023 ± $0.011/kWh, and the range of net present values as of reactor startup would extend from a loss of $1.2 billion on the low end to a loss of $3.5 billion on the high end. Similarly, an increase of acquisition cost from $1 to $1.2 billion would pull the optimistic end of the 70-percent confidence range down from a modest profit to bare break-even (for the $0.050 ± $0.015/kWh avoided-cost range), while the acquisition of the reactor for nothing would make the 70-percent confidence range for the net cost of the operation approximately symmetric around the break-even point (net present value at reactor startup from minus $1.6 billion to plus $1.6 billion for the $0.050 ± $0.015/kWh avoided-cost range).

If the high residual plutonium content in spent fuel associated with a 6.8-percent initial plutonium loading were considered problematic, so that both reactors needed to be used in order to complete the disposition campaign in 30 years or less, the up-front costs of the option would roughly double, but the per-kilowatt-hour costs would fall because of economies in scale in MOX fabrication. At 4-percent initial plutonium loading and 42-MWd/kgHM average burnup, it would take about 50 reactor-years to process 50 tons of plutonium, hence 25 years each for the two WNP reactors. If the reactors were financed over the 25-year period, the range of expected costs of the operation would be $0.0615 ± $0.0039/kWh and the differential over the assumed avoided cost of baseload electricity generation would be $0.0115 ± $0.0155/kWh, or $189 ± $255 million/yr for the two-reactor system. The discounted present value at start of reactor operation of a 25-year stream of costs or revenues of this magnitude would range from a $0.8-billion profit to a $5.2-billion loss. If the two reactors were assumed to operate for the remaining 5 years of a nominal 30-year lifetime using LEU fuel, the levelized net income from the operation of the reactors in this period would be $0.030 ± $0.015/kWh; the effect of the resulting 5-year profit stream at the end of the 30-year period on the discounted present value of the enterprise at the start of reactor operation would improve the profit by about $550 million (to some $1.3 billion) if the electricity value is at the high end of the range and would reduce the loss by about $185 million (to about $5.0 billion) if the electricity value is at the low end of the range.

We note, finally, that the somewhat complex institutional and financial arrangements put forward in the Isaiah proposal appear at first glance to yield a more favorable financial picture than portrayed above for one crucial reason: it is assumed in that proposal that the federal government pays all of the fuel-cycle costs and all of the nonfuel operation and maintenance costs for the reactors, exclusive of the O&M costs of the turbine generator. Taking the turbine genera-

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

tor O&M costs to be about 15 percent of the fuel plus nonfuel O&M costs for the plant (SAIC 1993b), this means—in the case of our example—that the government would be paying $0.020 ± $0.002/kWh of the levelized-annualized costs of the operation while the various investors would be paying $0.043 ± $0.003/kWh. The discounted present value in 1992 dollars, at the start of reactor operation, of the government's 30-year cost stream in this one-reactor case would be, under the central estimate, about $2.0 billion—which is just the amount that the Isaiah Project proposes to provide the government.

Building New Reactors for Plutonium Disposition

In Phase I of the U.S. Department of Energy's Plutonium Disposition Study, carried out in 1992 and 1993 (USDOE 1993b), the only options considered were those involving construction of new reactors of evolutionary and advanced types. These would be dedicated to the task of WPu disposition, although they would also generate electricity—the sale of which would defray the costs of the activity. Five reactor types that could be used for this purpose were examined in detail in the Phase I PDS, based on studies commissioned from the companies that would provide the reactors if one of them were selected for the plutonium-disposition mission: the ABB-Combustion Engineering System-80+, an evolutionary pressurized-water reactor with net electrical output of 1,256 MWe per reactor (ABB-CE 1993); the General Electric Advanced Boiling-Water Reactor (ABWR), an evolutionary BWR with net electrical output of 1,300 MWe per reactor (GE 1993); the Westinghouse PDR-600, an advanced PWR with net electrical output of 610 MWe per reactor (Westinghouse 1993); the General Atomics Modular High-Temperature Gas-Cooled Reactor (MHTGR), with net electrical output of 169 MWe per reactor (GA 1993); and the General Electric/Argonne National Laboratory Advanced Liquid-Metal Reactor (ALMR), with net electrical output of 303 MWe per reactor (GE-ANL 1993).

These contractor studies included detailed economic assessments, which were supposed to follow a set of ground rules provided in advance by DOE. The Technical Review Committee (TRC) set up by the DOE to evaluate and compare the contractor studies found that the ground rules were not always consistently followed. Problems mentioned by the TRC included (USDOE 1993a, pp. SC6-1/30):

  • understatement (by all of the contractors) of site-support/infrastructure costs;

  • understatement (by all of the contractors) of preoperational costs, which include research and development and conceptual design, safety and environmental impact analyses, operational readiness review, plant startup and testing, and more;

  • "overly aggressive" schedule assumptions (by all of the contractors);

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×
  • "overly aggressive" assumptions about cost reductions from learning in connection with construction of multiple units (by some contractors);

  • widespread understatement of cost and schedule risks associated with licensing of MOX fuel fabrication facilities and of reactors for MOX use;

  • nonuniform treatment of indirect costs;

  • contingency factors lower in some cases than specified in the ground rules;

  • incomplete and nonuniform estimates of the costs of treatment of the waste streams from fuel fabrication, and of costs of interim spent fuel storage;

  • nonuniform assumptions about whether WPu feed would be provided as metal or as oxide;

  • nonuniform treatment of seismic considerations; and

  • nonuniform treatment of dismantling and disposal (D&D) costs.

Almost without exception, the effects of these deficiencies were to understate the costs of the options in question.

The TRC produced a set of modified cost estimates and comparisons in which some—but far from all—of the above-mentioned deficiencies were corrected (mainly, the variations from guidelines in contingency factors and indirect costs, and some but not all of the overoptimism about schedule and learning). As noted in Chapter 3, however, the TRC's modified economic assessment still suffered from use of an inappropriately low interest rate, failure to account for interest during construction, and frequent lack of clarity about which deficiencies in the contractor studies had been remedied and how.

These shortcomings of the TRC report were criticized in the Peer Review Report (USDOE 1993b) commissioned by DOE as part of the PDS. The Peer Review Report also criticized the TRC report for inadequate analysis of cost uncertainties, insufficient attention to the implications of schedule differentials between more advanced and less advanced reactor types, failure to make comparisons with the economic costs of using existing reactors for the plutonium disposition mission, and failure to compare the costs of electricity generation using plutonium versus uranium fuels. While the Peer Review Report was criticizing a version of the TRC report different from the one publicly released, the panel believes that many of these criticisms apply to the final report as well. Phase II of the PDS—in which, it is to be hoped, some of these shortcomings will be remedied—is reported to be essentially complete, but was not available at the time the panel finished its work.

In the meantime, we have prepared our own set of economic comparisons for the new-reactor options, basing these in part on the fuel-cycle analyses and direct and indirect cost estimates of the Phase I PDS contractor studies as modified by the TRC report. In so doing, we have attempted to improve on the PDS

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

contractor and TRC reports in a number of respects, including development and consistent treatment of uncertainty ranges, use of a more appropriate real cost of money, consistent treatment of preoperational costs, inclusion of interest on construction and on other preoperational costs, and, in general, adherence to the generally accepted practices described in Chapter 3 and DOE guidelines (USDOE 1988b, Delene and Hudson 1993) for economic evaluation of projects in terms of levelized-annualized costs and net discounted present values.25

Our cost calculations for the five types of reactors considered in the contractor studies and the TRC report are based on reactor specifications and operating modes summarized in Table 6-14. The contractor studies and TRC report examined a variety of operating modes for each reactor-corresponding to "spiking," "spent fuel," and "destruction" variants, all subject to the very severe constraint that 100 tons of WPu be processed within 25 years of a program start date taken to be January 1, 1994. We do not consider this constraint a useful basis for a comparative evaluation; it is virtually a prescription for unrealistic deployment schedules for advanced-reactor options, which under realistic schedules would not meet the constraint. Instead, we evaluated, for each of the five reactor types, the economics of a reactor array with sufficient capacity to process 50 tons of WPu within approximately one nominal reactor lifetime of 30-40 years from the start of reactor operation, using the contractor-analyzed fuel-cycle variant that we considered most representative of that reactor type's near-term plutonium disposition capabilities. (We chose the highest burnup variants that used the reactor type's “standard" fuel, which were the "spent fuel" variants for the PDR-600, System-80+, and ALMR, and the "destruction" variants for the ABWR and MHTGR.)

Our cost calculations are based, moreover, on reactor operational lifetimes coinciding exactly with the period of operation required to irradiate 50 tons of WPu to the average burnups indicated. (These time figures were derived using the simplifying assumption that fuel shuffling is used in the beginning and ending phases of reactor operation in order to achieve the indicated average burnup for all of the plutonium-based fuel, without introducing any uranium-based fuel.) As can be seen in Table 6-14, the operating times actually required— under the stated assumptions about plutonium percentages in spent fuel, average burnups, numbers of reactors of a given type, and lifetime-average capacity factors—range from 28 to 36 years.

In contrast, then, to the approach of DOE's TRC, which assumed that the plants would complete an assumed 40-year operating lifetime (running on uranium fuel for the part of that period after WPu disposition has been completed,

25  

We have omitted, however, a substantial amount of analytical complexity in the form of aspects of electric-utility accounting practices that depend on debt-to-equity ratios, accelerated depreciation schemes, tax policy, and allocation of electricity value between capacity credit and energy credit, since these context-specific complications add much speculation and calculation, but little value in a general comparison of the sort undertaken here.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

and earning additional revenues from electricity generation during this period of uranium-fueled operation), we have assumed in the calculations for the table that the plants do not operate after the plutonium disposition mission has been completed. Accordingly, our fixed charge rates for determining levelized-annualized capital charges are based on just the operating lifetimes needed to complete that mission.

There are several reasons for choosing this approach for our baseline economic comparison of the five options: it removes the unfortunate characteristic of the TRC approach that economic benefits are attributed to plutonium disposition in proportion to the amount of time that reactors operate with nonplutonium fuel; it reduces the sensitivity of the results to highly uncertain assumptions about distant-future electricity prices; it avoids uncertainty about whether these reactor types would in fact be able to achieve an operating life as long as 40 years; and it reflects the possibility that society might choose, for economic or other reasons, not to use the plutonium disposition reactors for electricity generation once the plutonium mission has been completed. At the end of this section, however, we do show how the results would change under the TRC approach of assuming continuing uranium-based electricity generation to a total operating life of 40 years.

The cornerstones of a calculation of the monetary cost of any reactor option are the estimates of direct plus indirect construction costs, of fuel-cycle costs, and of nonfuel O&M costs. Rather than accepting the point-value estimates for these quantities provided in the PDS—and accepting along with these estimates the inevitable doubts about whether the different contractors derived these estimates in comparable ways, with comparable degrees of optimism or conservatism—we have used the PDS estimates together with cost estimates for these fuel cycles and reactor types from other sources to develop ranges of values for use in our own economic calculations. Central estimates and ranges for MOX fuel-cycle costs were obtained in this way in "Weapons Plutonium Versus Uranium as Power Reactor Fuel" above. Here we do so for the construction and nonfuel O&M costs and for the fuel-cycle costs in the non-MOX cases.

Because our choices about the time frame for plutonium disposition, and hence the scale of the facilities needed to accomplish it, are different than those employed in DOE's PDS, we have had to adjust the cost estimates provided by the PDS contractors and the TRC, in some instances, to apply them to different numbers of reactors. In contrast to the extremely optimistic "learning" assumptions applied by some of the contractors to multiple-reactor systems, we assume (consistent with the TRC) that the overnight construction costs and nonfuel O&M costs of such systems increase with the 0.9 power of the number of reactors (costs per reactor decline with the -0.1 power of the number of reactors). For the non-MOX fuel fabrication facilities (for the MHTGR and ALMR) analyzed here, we assume as before that total costs in a facility of a given type scale

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-14 New-Reactor Specifications and Reference Operating Modes

Characteristics

Evolutionary BWR (GE ABWR)a

Evolutionary PWR (ABB-CE System-80+)b

Advanced PWR (Westinghouse PDR-600)c

Advanced Gas Reactor (GA MHTGR)d

Advanced Metal Reactor (BE/ANL ALMR)e

System Performance

Thermal power, MWt/reactor

3,926

3,817

1,940

450

840

Net electric power, MWe/reactor

1,300

1,256

610

169

303

Conversion efficiency

0.331

0.329

0.314

0.376

0.361

Assumed capacity factor

0.75

0.75

0.75

0.79

0.75

Reactor output, 109 kWh/yr

8.55

8.26

4.01

1.17

1.99

Operating Mode

Pu as percent of heavy metal

3.0

6.8

5.5

100.0

10.5

Average burnup. MWd/kgHM

37.1

42.2

40.0

580.0

69.1

Mean fuel life, yr

5.3

4.0

5.0

2.0

6.0

Fuel loaded, MTHM/reactor-yr

29.0

24.8

13.3

0.2

3.3

Pu loaded, kgPu/reactor-yr

870

1,685

731

224

350

Array Characteristics

Reactors × yrs for 50 MT Pu

2 × 29

1 × 30

2 × 34

8 × 28

4 × 36

Array capacity, MWe

2,600

1,256

1,220

1,352

1,212

Array output, 109 kWh/yr

17.09

8.26

8.02

9.36

7.97

Fuel fabrication requirement MTHM/yr

58.0

24.8

26.6

1.8

13.3

Pu loaded, kgPu/yr

1,739

1,685

1,461

1,791

1,399

a GE ABWR: General Electric Advanced Boiling-Water Reactor. In destruction option, 26.6 percent of core amounting to 41. MTHM is discharged every 17 months at an average irradiation of 37,090 MWd/MTHM. Fuel life is (17/12) / 0.266 = 5.3 years and annual plutonium loading is 0.030 kgPu/kgHM × (12/17) × 41,100 kgHM/reactor-yr = 870 kgPu/reactor-yr.

b ABB-CE System-80+: ABB-Combustion Engineering System-80+. In spent fuel option, full core amounting to 99.2 MTHM is discharged every 48 months at an average irradiation of 42,200 MWd/MTHM (achieved by fuel shuffling annually). Fuel lifetime is 4.0 years and annual plutonium loading is 0.068 kgPu/kgHM × 99,200/4 kgHM/reactor-yr = 1,686 kgPu/reactor-yr.

c Westinghouse PDR-600: In spent fuel option, one-third of core amounting to 22.15 MTHM is discharged every 20 months at an average irradiation of 40,000 MWd/MTHM. Fuel life is thus (20/12)/0.333 = 5.0 years, annual fuel fabrication requirement is 22.15 MTHM/cycle / (1.667 yr/cycle) = 13.29 MTHM/reactor-yr, and annual plutonium loading at 0.055 kgPu/kgHM × 13,290 kgHM = 731 kgPu/reactor-yr.

d GA MHTGR: General Atomics Modular High-Temperature Gas-Cooled Reactor. In destruction option, 50 percent of core amounting to 224 kgHM irradiated to an average of 580,000 MWd/kgHM is discharged annually. Fuel lifetime is 2 years and plutonium loading is 224 kgHM/reactor-yr since heavy metal in this fuel is 100 percent plutonium. The indicated 79-percent capacity factor for this reactor is the value assumed by GA in its analyses; the panel chose not to redo the analysis for the 75-percent capacity factor assumed for the other reactor types. After most of our economic analyses were completed, GA proposed a modified high-temperature gas reactor known as the modular helium reactor (MHR) for the plutonium disposition mission. This modification included the use of a direct (Brayton) cycle in which the high-temperature helium is sent directly to the turbine, eliminating the need for a steam generator. This cycle is estimated by the vendor to lead to a 25-percent increase in thermal efficiency, as well as reducing the capital cost of the reactor. The vendor also proposes to increase the size of the reactor to 600 MWt, further reducing the cost per megawatt. We have not used, for our analysis, the revised cost estimates provided by GA for this modification. The direct-cycle HTGR may prove to meet the vendor's expectations, but we think the information now available is too preliminary to support our using the revised

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

figures. Licensing review, in particular, will be more difficult for the new design, as the vendor's own analysis concedes (GA 1994, p. 13-1).

e GE/ANL ALMR: General Electric/Argonne National Laboratory Advanced Liquid-Metal Reactor. In spent fuel option, one-third of core fuel assemblies amounting to 3,480 kgHM are discharged every two years at an average irradiation of 106,300 MWd/MTHM and one-fourth of blanket fuel assemblies amounting to 3,185 kgHM are discharged every two years at an average irradiation of 28,300 MWd/MTHM. Fuel life for core fuel is 2/0.333 = 6 years. Plutonium is loaded only into core fuel at a rate of 0.202 kgPu/(kgHM in core fuel) × 1,740 kgHM core fuel/reactor-yr = 351 kgPu/reactor-yr. Plutonium loading averaged over core and blanket fuel is 351 kgPu/yr / (3,480 kgHM + 3,185 kgHM) / (2 yr) = 0.105 kgPu/kgHM. Average irradiation at discharge for all fuel is (840 MW × 365.25 days/yr × 0.75) / (3,480 kgHM + 3,185 kgHM) / (2 yr) = 69.05 MWd/kgHM.

with the 0.6 power of output, so that unit costs decline with the -0.4 power of output.

The resulting central estimates and ranges for direct plus indirect costs of power-plant construction, as well as nonfuel O&M costs, are presented and compared to the PDS point estimates in Table 6-15. In using estimates from the literature to develop our construction-cost ranges, we have applied where appropriate an increment of 20 percent on direct plus indirect costs for first-of-a-kind costs, since the new reactors under consideration would indeed be first-of-a-kind if built for the immediate plutonium disposition need; the 20-percent figure is based on arguments and examples given in DOE costing-guideline documents (USDOE 1988b, Delene and Hudson 1993). The ranges in the table have been chosen to extend a uniform ±20 percent on construction costs for the evolutionary reactors, ±25 percent on construction costs for the advanced PWR, ±30 percent on construction costs for the MHTGR and ALMR, and ±15 percent on operating costs for all reactor types; these are our judgmental 70-percent confidence intervals. Further details on the derivations of the estimates are given in the notes to the table.

The estimates of direct plus indirect power-plant construction costs and nonfuel O&M costs in Table 6-15, plus the plutonium-disposition-array characteristics from Table 6-14 and the estimates of MOX-fuel-associated costs developed in the preceding section and summarized in Table 6-10, provided the starting point for comparative economic calculations we carried out in spreadsheets developed for the purpose and which are summarized in Table 6-16. The subparagraphs that follow here provide further information on assumptions and parameter choices used in those calculations. (A more detailed discussion of the concepts and methods relevant to economic comparisons of this sort was provided in "Issues and Criteria in Economic Evaluation of Alternatives" in Chapter 3. Further details specific to the individual reactor types are provided in notes to the table.)

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-15 Panel Estimates of Construction and O&M Costs for New Reactors (in 1992 U.S. dollars)

 

Direct + Indirect Construction Costs, $/kWe

Nonfuel O&M Costs, Million $/GWe Capacity per Year

New Reactors

TRC Estimate

Panel Estimate with Range

TRC Estimate

Panel Estimate with Range

Evolutionary BWR,a 2 × 1,300 MWe

1,805

1,800±360

43

75±11

Evolutionary PWR,b 1 × 1256 MWe

2,246

1,900±380

74

80±12

Advanced PWR,c 2 × 610 MWe

1,824

2,100±525

88

90±14

Advanced Gas Reactor,d 8 × 169 MWe

2,063

2,400±720

78

85±13

Advanced Metal Reactor,e 4 × 303 MWe

2,311

2,500±750

92

90±14

NOTES: TRC estimates have been adjusted for scale as described in the footnotes. Two- to four-digit precision is illusory but has been preserved to facilitate checking results.

a Evolutionary boiling-water reactor: PDS TRC estimates direct plus indirect capital costs at $2,515.0 million for one 1300-MWe power plant (USDOE 1993a, p. A-22). Assuming costs scale with the 0.9 power of the number of units, this corresponds to $2,515 × 20.9 = $4,693 million for two units, or $4,693 million / 2.6 GWe = $1,805/kWe. In the absence of other estimates for an ABWR per se, we refer to an estimate of the Fission Working Group Review Committee of the PDS (Omberg and Walter 1993, p. 23) of $1,600/kWe direct plus indirect capital costs, excluding first-of-a-kind costs, for a generic evolutionary LWR, and to a Nuclear Energy Cost Data Base (NECDB) (USDOE 1988b, p. 26) figure of $1,147/kWe in 1987 dollars, hence $1,394/kWe in 1992 dollars, for an 1,100-MWe evolutionary PWR. Based on examples given in DOE costing-guideline documents (USDOE 1988b, Delene and Hudson 1993), we take first-of-a-kind costs, as would apply to the WPu disposition case we are considering, to be an increment of 20 percent on the indicated direct plus indirect costs; applying this factor and the two-unit, unit-cost learning factor of 1/20.1 makes the two indicated estimates $1,560/kWe and $1,790/kWe for our 2 × 1,300 MWe case. We take as our central estimate $1,800/kWe, with a judgmental 70-percent confidence interval of ±20 percent, hence ±$360. The TRC estimates nonfuel O&M costs less D&D annuity at $60.4 million/reactor-yr for one reactor, which becomes $60.4 million × 20.9 = $112.7 million/yr for a two-reactor plant, or $43 million/GWe/yr. The NECDB (USDOE 1988b, p. 32) gives nonfuel O&M costs for an 1,100-MWe conventional LWR as $62 million/yr fixed and $0.0005/kWh variable in 1987 dollars, and says this figure is assumed applicable to evolutionary and advanced plants as well. Converting to 1992 dollars and assuming the fixed part scales with capacity, the relation for a 1,300-MWe plant would be $88 million/yr + $0.0006/kWh, or at 75-percent capacity factor $93 million/yr; with a 25-percent increment as appropriate on nonfuel O&M costs for first-of-a-kind (USDOE 1988b, Delene and Hudson 1993), this would be $116 million/yr, or for two plants $116 million/yr × 20.9 = $216 million/yr, hence $83 million/GWe/yr. Even allowing for reduced labor requirements in an evolutionary plant, the GE/TRC estimate seems very low. Our central estimate is $75 million/GWe/yr, with a judgmental 70-percent confidence interval of ±15 percent or ±$11 million/yr.

b Evolutionary pressurized-water reactor: The PDS TRC estimates direct plus indirect capital costs at $2,820.9 million for one 1,256-MWe power plant (USDOE 1993a, p. A-10), or $2.246/kWe. The contractor estimate of indirect + indirect costs submitted to the PDS for this case (ABB-CE 1993, p. VI-14) was $2,348 however, or $1,869/kWe, and the basis on which the TRC increased the estimate is not clearly explained in the TRC report. The estimates in the NECDB and the report of the Fission Working Group Review Committee, mentioned above, would become

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

$1,920/kWe and $1,673/kWe, respectively, for this case after the addition of 20 percent first-of-a-kind costs. One expects a PWR to be a bit costlier than a BWR, all else being equal. We take as our central estimate $1,900/kWe, with a judgmental 70-percent confidence interval of ±20 percent, hence ±$380. The PDS contractor estimate for nonfuel O&M costs in the spent fuel mode are $92.8 million/reactor-yr without the D&D annuity (ABB-CE 1993, p. VI-24), hence $74 million/GWe/yr. The NECDB estimate would be the same as for the evolutionary BWR, hence, with 25 percent first-of-a-kind increment, $116 million/reactor-yr or $92 million/GWe/yr. Our central estimate is $80 million, with judgmental 70-percent confidence interval of ± 15 percent or ±$12 million/yr.

c Advanced pressurized-water reactor: TRC estimates direct plus indirect costs at $1192.6 million for one 610-MWe power plant (USDOE 1993a, p. A-1). Assuming costs scale with the 0.9 power of the number of units, this corresponds to $1,192.6 million × 20.9 for two units, or $2,225.5 million / 1.22 GWe = $1,824/kWe. The NECDB (USDOE 1988b, p. 26) estimates direct plus indirect costs for a 550-MWe APWR at $743 million, hence $1,351/kWe in 1987 dollars or $1,630/kWe in 1992 dollars; with first-of-a-kind costs of 20 percent and unit costs lower by 20.1 for two units, this estimate would be $1,825/kWe. Despite the close coincidence of this number and that of the PDS, we think it rather unlikely that cost-saving advances between evolutionary and advanced PWRs will offset the diseconomies of scale suffered by 600-MWe units in comparison to 1,200-MWe ones. Thus we take as our central estimate $2,100/kWe, with a judgmental 70-percent confidence interval of ±25 percent or ±$525. The contractor estimates nonfuel O&M costs less D&D annuity as $57.3 million/yr for one reactor (Westinghouse 1993, p. 3-19), hence $57.3 million × 20.9 = $106.9 million/yr for a two-unit plant, or $88 million/GWe/yr. The NECDB estimates nonfuel O&M costs for a 2 × 550-MWe advanced LWR at $71 million/yr fixed and $0.0005/kWh variable in 1987 dollars; adjusting for 2 × 610 MWe and 1992 dollars makes this $95 million + 0.0006/kWh, which at 75-percent capacity factor is $100 million/yr or $82 million/GWe/year: with a 25-percent first-of-a-kind increment, this becomes $102.5 million/GWe/yr. Our central estimate is $90/GWe/yr, with judgmental 70-percent confidence interval ± 15 percent or ±$14.

d Advanced gas reactor: The PDS TRC estimates direct plus indirect costs at $2,789.4 million for an eight-module, 1,352-MWe power plant (USDOE 1993a, p. A-37), or $2,063/kWe. The Fission Working Group Review Committee estimates direct plus indirect costs for the MHTGR at $2,200/kWe, which a first-of-a-kind increment of 20 percent would increase to $2,640/kWe. There is no MHTGR estimate in the 1988 NECDB, but a late 1980s DOE-sponsored comparison of advanced fission and fusion reactors (Holdren et al. 1989), which used NECDB methods to generate its cost estimates, produced a direct plus indirect cost estimate for an eight module MHTGR that translates to $1,786/kWe in 1992 dollars, which with a 20-percent first-of-a-kind increment would be $2,143/kWe. We take as our central estimate $2,400/kWe, with a judgmental 70-percent confidence interval of ±30 percent or ±$720. The contractor estimate for nonfuel O&M costs for the eight-module case (GA 1993, Table A-7) is $105.9 million/yr or $78.3 million/GWe/yr. The NECDB contains no estimate for MHTGR operating costs. The fusion/fission study estimated nonfuel O&M costs for an eight-module MHTGR at $0.0083/kWh in 1986 dollars, hence $0.0103/kWh in 1992 dollars, which translates at the assumed capacity factor of 0.79 to $96.4 million/yr or $71.3 million/GWe/yr; with 25-percent first-of-a-kind increment this is $89.1 million/GWe/yr. Our central estimate is $85 million/GWe/yr, with a judgmental 70-percent confidence interval of 15 percent or ±$13 million.

e Advanced metal reactor: PDS estimates direct plus indirect costs at $1,501 million for a two-reactor module, 606-MWe power plant (USDOE 1993a, p. A-28). Assuming costs scale with the 0.9 power of the number of units, this corresponds to $1,501 million × 20.9 = $2,801 million for two such plants, or $2,801 million / 1.212 GWe = $2,311/kWe. The Fission Working Group Review Committee estimates direct plus indirect costs for the ALMR at $2.100/kWe, which a first-of-a-kind increment of 20 percent would increase to $2,520. The NECDB direct plus indirect cost estimate for an 1,100-MWe LMR is $1,988/kWe in 1987 dollars, or $2,398/kWe in 1992 dollars, which with a 20-percent first-of-a-kind increment would be $2,878. (This LMR is not of the modular design considered here.) The fission/fusion study estimated modular and conventional LMRs of 1,200- to 1300- MWe total capacity to have direct plus indirect costs nearly identical at $1,700/kWe (after translation to 1992 dollars), or $2,040/kWe after addition of a 20-percent first-of-a-kind increment. We take as our central estimate $2,500/kWe, with a judgmental 70-percent confidence interval of ±30 percent or ±$750. The TRC estimate of nonfuel O&M costs less D&D annuity is $60.0 million/yr for a two-reactor power block, hence $60.0 million/yr × 20.9 = $112

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

million/yr for a four-reactor plant, or $92.4 million/GWe/yr. The NECDB estimate of LMR operating costs is $67 million/yr fixed plus $0.0006/kWh variable for an 1,100-MWe (nonmodular) unit, in 1987 dollars, hence $86 million/yr in 1992 dollars at 75-percent capacity factor, or $71 million/GWe/yr; with 25-percent first-of-a-kind increment this is $89 million/GWe/year. The fission/fusion study's estimated nonfuel O&M costs for a modular LMR similar in size to our case were $0.0079/kWh in 1986 dollars, hence $0.0098/kWh in 1992 dollars, translating to $78 million/yr or $64 million/GWe/yr; with 25-percent first-of-a-kind increment this is $80 million/GWe/yr. Our central estimate is $90 million/GWe/yr, with a judgmental 70-percent confidence interval of ±15 percent of $14.

Constant Dollars

All figures are reported in January 1, 1992 constant dollars.

Contingency

The TRC for the DOE's Plutonium Disposition Study applied a contingency of 0.20 of direct plus indirect costs of a power plant's nuclear equipment and 0.15 of direct plus indirect costs of the ECA (Energy Conversion Area) for the evolutionary systems; 0.25 of nuclear reactor direct plus indirect costs and 0.20 of direct plus indirect costs of the ECA for the advanced systems; and 0.25 of direct plus indirect costs for fuel fabrication plants for both evolutionary and advanced systems. Since many sources of power-plant cost estimates do not disaggregate nuclear equipment from the ECA, we assume a typical 70-30 percent division so that the combined contingency, using the same values as the TRC, is 0.185 for evolutionary systems and 0.235 for advanced systems.

Preoperational Costs

To replace the drastically divergent treatment of preoperational costs of different options in the contractor studies and the TRC report, we have used a simple rule of thumb that approximates these costs as the following fractions of overnight construction costs: 10 percent for evolutionary reactor systems, 20 percent for the fuel fabrication plants for LWRs (whether current, evolutionary, or advanced), and 25 percent for advanced reactor systems26 and for the fuel fabrication plants for the MHTGR and ALMR.

26  

In the interest of simplicity, no distinction has been made in reactor-system contingency or preoperational costs, expressed as a percentage of direct plus indirect construction costs, between the advanced LWR and the other advanced reactor types for which detailed estimates were made, the MHTGR and the ALMR. It can be argued, however, that the advanced LWR is based on more proven technology and that this would be likely to be reflected in smaller contingency and preoperational costs. Incorporating such a distinction would not significantly affect any of the conclusions drawn.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×
Interest During Construction

We have assumed an S-curve distribution of construction and preoperational costs over a period of nine years preceding full operation in the case of reactors and advanced fuel fabrication plants. At the OMB-recommended real annual interest rate of r = 0.07, this procedure gives IDC equal to 41 percent of overnight costs (see section in Chapter 3 "Issues and Criteria in Economic Evaluation of Alternatives").

Fixed Charge Rates

These are computed according to the prescriptions described in Chapter 3, with r = 0.07 and depreciation periods equal to the operating lifetimes shown in Table 6-14 for irradiation of 50,000 kg of plutonium, both with and without an allowance of 0.02/yr of total beginning-of-life capital costs for property taxes and insurance.

Dismantling and Disposal Costs

For D&D costs of reactors we use contractors' estimates where available and otherwise use the DOE guideline (see "Principles and Pitfalls in Cost Comparisons" in Chapter 3) whereby end-of-life D&D cost = B + 0.02(P-1,200) million 1992 dollars where P is the unit thermal power and B is 145 for PWRs, 185 for BWRs, and 165 for other reactor types. For fuel fabrication plants we use the DOE guideline suggested in the PDS (USDOE 1993a, p. SC6-7) whereby the end-of-life D&D cost is taken as 10 percent of the beginning of life capital cost (direct plus indirect plus contingency), not including IDC. In each case the annual payment needed during operation in order to produce the indicated sum at the end of operating life is computed (assuming the payments are invested at 3 percent per year real) and added to the operating costs.

Fuel Costs

Estimated costs of MOX fuel in $/kgHM for the evolutionary and advanced LWRs are taken from Table 6-10 for the three cases presented there— fabrication at FMEF, paying no property tax or insurance, and fabrication at new plants both with and without property taxes and insurance—at the scales most closely corresponding to the plutonium disposition arrays described in Table 6-14, and corrected for plutonium metal-to-oxide conversion and incremental transport and storage costs where the plutonium loading differs from the reference loading of 4.8 percent used in deriving the Table 6-10 figures. For the MHTGR and ALMR, fuel costs per kilogram of heavy metal are derived starting

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-16 Costs of Plutonium Disposition with Evolutionary and Advanced Reactors (in 1992 U.S. dollars)

 

Evolutionary BWR (GE ABWR)

Evolutionary PWR (ABB-CE System 80+)

Advanced PWR (Westinghouse PDR 600)

Advanced Gas Reactor (GA MHTGR)

Advanced Metal Reactor (GE/ANL ALMR)

Array Characteristics

Reactors × net MWe

2 × 1,300

1 × 1,256

2 × 610

8 × 169

4 × 303

Array annual TWh

17.09

8.26

8.02

9.36

7.97

MTHM/yr

58.0

24.8

26.6

1.8

13.3

kgPu/yr

1.739

1,685

1,461

1,791

1,399

yr for 50 MTPu

29

30

34

28

36

Power-Plant Construction Costs

Array dir+indir, $/kWe

1,800±360

1,900±380

2,100±525

2,400±720

2,500±750

Array overnight, $/kWe

2,133

2,252

2,594

2,964

3,088

Array w IDC, $/kWe

3,008

3,175

3,657

4,179

4,353

Array w IDC, M$

7,820

3,987

4,461

5,650

5,276

Array pre-op, M$

782

399

1,115

1,413

1,319

Array total, MS

8,602±1720

4,386±877

5,577±1,394

7,063±2,119

6,595±1,979

Power-Plant Array Capital Charges

FCR, no tax, yr-1

00814

0.0806

0.0778

0.0824

0.0767

FCR, w tax, yr-1

0.1014

0.1006

0.0978

0.1024

0.0967

LACC no tax, M$/yr

701

353

434

582

506

LACC w tax,

873

441

545

723

638

c/kWh no tax

4.10±0.82

4.28±0.86

5.41±1.35

6.22±1.86

6.35±1.90

c/kWh w tax

5.10±1.02

5.34±1.07

6.80±1.70

7.72±2.32

8.01±2.40

Power-Plant Array Nonfuel O&M Costs

O&M w/o D&D, M$/GWe-yr

75±11

80±12

90±14

85±13

90±14

O&M w/o D&D, MS/yr

208

94

110

115

109

Array D&D, MS

447

197

298

330

326

D&D annuity, M$/yr

9.9

4.1

5.2

7.7

5.1

O&M w D&D, M$/yr

218±31

98±14

115±17

123±17

114±16

O&M w D&D, c/kWh

1.27±0.18

1.19±0.17

1.43±0.21

1.31±0.18

1.43±0.21

MOX Fuel Costs,a $/kgHM

FMEF

1,900±300

2,500±400

2,500±400

NA

NA

New plant, no tax

2,600±400

3,400 500

3,400±500

38,000±7,600

3,800±760

New plant, w tax

2,900±400

3,800±500

3,800±500

44,000±8,800

4,100±820

Pu conversion correction

-202 ±66

284±90

99±31

 

0

Adj FMEFb

1,799±302

2,642±403

2,550±400

NA

NA

Adj new plant no tax

2,398±405

3,684±508

3,499±501

38,000±7,600

3,800±760

Adj new plant w tax

2,698±405

4,084±508

3,899±501

44,000±8,800

4,100±820

Fuel Cost Contribution to Electricity Cost, Including Repository Fee

Carrying-charge factor

1.271

1.219

1.259

1.143

1.219

FMEF, c/kWh

0.88±0.12

1.07±0.15

1.16±0.17

NA

NA

New plant no tax, c/kWh

1.13±0.17

1.45±0.19

1.56±0.21

093 ±0.17

0.93±0.17

New plant w tax, c/kWh

1.26±0.17

1.59±0.19

1.73±0.21

1.06±0.19

0 990.18

Total Levelized Generation Cost. c/kWh

Reactor no tax, FMEFc

6.25±0.85

6.54±0.89

8.01±1.38

8.46±1.88

871±1.92

Reactor, new plant both tax

7.64 1.05

8.13±1.10

9.96±1.72

10.10± 2.33

10.43±2.42

Levelized Costs Net of Electricity Revenues @ 5.0± 1.5 c/kWh, M$/yr

Reactor no tax, FMEFc

213±295

127±144

241±163

324±225

296±194

Reactor, new plant both tax

452±313

258±154

398±183

477±260

433±227

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

 

Evolutionary BWR (GE ABWR)

Evolutionary PWR (ABB-CE System 80+)

Advanced PWR (Westinghouse PDR 600)

Advanced Gas Reactor (GA MHTGR)

Advanced Metal Reactor (GE/ANL ALMR)

Net Cost of Campaign Discounted to Start of Reactor Operation, billion $

Reactor no tax, FMEFc

2.6±3.6

1.6±1.8

3.1±2.1

3.9±2.7

3.9±2.5

Reactor, new plant both tax

5.5±3.8

3.2±1.9

5.1±2.4

5.8±3.2

5.6±3.0

Cost Excess Over Same Reactor Using LEU Fuel, c/kWh

LEU fuel-cycle cost

0.51±0.09

0.61±0.07

0.68±0.08

NCd

NCd

FMEF

0.37±0.16

0.45±0.16

0.48±0.19

NCd

NCd

New plant w tax

0.76±0.19

0.98±0.20

1.04±0.22

NCd

NCd

Cost Excess Over Same Reactor Using LEU Fuel, $M/yr

FMEF

63±27

38±14

39±15

NCd

NCd

New plant w tax

130±33

81±17

84±18

NCd

NCd

Alternative Net Cost for Campaign, Discounted to Start of Reactor Operation, M$

FMEF

777±328

466±169

494±192

NCd

NCd

New plant w tax

1,590±409

1,006±205

1,075±232

NCd

NCd

ABBREVIATIONS:

c/kWh: cents per kilowatt-hour.

D&D: dismantling and disposal.

FCR: fixed charge rate.

LACC: levelized annual capital charge.

M$: million dollars.

O&M: operation and maintenance.

TWh: terawatt-hour.

NOTES: Based on estimates from Tables 6-10 and 6-15, with reactor characteristics and operating modes as indicated in Table 6-14. See also additional assumptions and disclaimers in the text and notes. Two- to four-digit precision is illusory but has been preserved to facilitate checking results. Uncertainty ranges are the Reactor Panel's judgmental 70-percent confidence intervals; these are not shown for many intermediate quantities to avoid clutter.

a Includes capital charges, O&M, D&D, plutonium metal-to-oxide conversion, and incremental plutonium storage and transport).

b Only half of the correction is applied to the FMEF (for operating costs of the additional plutonium conversion but not the capital costs), since the FMEF facility's capital cost includes allowance for a metal-to-oxide conversion operation.

c FMEF not assumed applicable to MHTGR and ALMR. Entries in this row for those reactors are based on a new fuel fabrication plant that does not pay property tax or insurance.

d Not calculated because of lack of information on fuel costs with uranium fuel.

from construction-cost and operating-cost estimates provided by the PDS contractors, and employing contingency, preoperational costs, IDC, FCRs, and D&D costs as indicated above. The $/kgHM fuel costs are then capitalized over an accounting life of n (= core residence time plus one) years, in order to compute cost per kilowatt-hour:

$/kWh =

$/kgHM × n × r(1+r)n / {[(1+r)n - 1] × 1000 × MWd/kgHM × 24 × h},

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

where η is the thermal-to-electric conversion efficiency. With r = 0.07 and n = 4 years, for example, this becomes,

$/kWh = 1.18 × $/kgHM / (24,000 × MWd/kgHM × η) .

The statutory repository fee of $0.001/kWh is then added to this cost per kilowatt-hour.

Total and Net Costs

Total costs per kilowatt-hour, levelized in constant 1992 dollars, are then obtained by summing the power plant capital charges and nonfuel O&M costs (including the annuity to cover power-plant D&D costs) and the fuel costs. The uncertainty bounds are calculated as the square root of the sum of the squares of the bounds on the components of the sum, preserving the notion of a consistent judgmental confidence interval. These are shown just for the high and low bounding cases—on the low side, LWRs paying no property tax or insurance and supplied with MOX from the FMEF, which also pays no property tax or insurance, and, on the high side, reactors and new fuel fabrication plants all paying property tax and insurance.

Net costs on a levelized-annualized basis are obtained by subtracting from the total costs the busbar value of the electricity generated in the course of plutonium disposition, based on a figure of $0.05 ± $0.015/kWh (see “Issues and Criteria in Economic Evaluation of Alternatives" in Chapter 3 and "Completing Existing LWRs" above). The net cost of the campaign figured as a net present value as of the start of reactor operation is then computed by discounting the annualized cost stream at a real interest rate of r = 0.07/yr.

Alternate net costs on a levelized-annualized basis and as a net present value at the start of reactor operation reflect the difference between the calculated cost of plutonium-based operation and operation for the same period using LEU fuels (for the MOX-based, evolutionary systems), based on LEU fuel-cycle costs developed in the section "Weapons Plutonium Versus Uranium as Power Reactor Fuel" above.

Total capital investments as of the start of reactor operation are just the sums of the investments in the power plants and the fuel fabrication plants.

Table 6-16 shows that the central estimates for the net costs of the campaign, measured as 1992-dollar net present values as of the start of reactor operation (which of course would be at a more distant time in the future for the advanced reactor types than for the evolutionary ones), range from $1.6 billion in the most favorable case to about $5.5 billion in each of three least favorable cases. The 70-percent confidence interval extends into the "profit" range only in two cases, both being associated with the use of FMEF to fabricate MOX fuel for LWRs that do not pay property taxes or insurance. A very substantial advan-

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

tage is associated with the use of FMEF to do the MOX fuel fabrication for the LWRs; this benefit arises by virtue of the substantial construction costs already invested in this facility, which are treated as sunk. For MHTGRs and ALMRs that do not pay property taxes and insurance, fed by fuel fabrication plants that also do not pay these items, the 70-percent confidence ranges of the net discounted campaign costs as of the start of reactor operation are $1.2-$6.6 billion and $1.4-$6.3 billion, respectively.

For reactors that do pay property taxes and insurance, fed by new fuel fabrication plants that also pay these items, the 70-percent confidence ranges of the net discounted campaign costs for the five cases are, respectively, $1.7-$9.3 billion for the ABWR, $1.3-$5.1 billion for the evolutionary PWR, $2.7-$7.5 billion for the advanced PWR, $2.6-$9.0 billion for the MHTGR, and $2.6-$8.6 billion for the ALMR. It is difficult to ascribe much significance to the differences among these cases. Within the considerable limitations of available basic cost estimates and the procedures we have used to try to treat uncertainties in a consistent manner, however, a rather robust overall conclusion is that the net campaign costs in 1992 dollars, discounted to the start of reactor operations, for all new facilities paying property taxes and insurance, seem likely to be in the range of $2-$9 billion.

The three clusters of entries at the end of Table 6-16 show, in addition, the alternative measure of net costs obtained by comparing the gross costs of the plutonium disposition campaigns with the costs of operating the same reactors for the same periods using uranium-based fuels. (These calculations were done only for the LWRs; as discussed in "Completing Existing LWRs" above, the MHTGR and ALMR would be expected to show a net economic benefit from using free WPu in place of uranium, but we did not think that the estimates of uranium-based fuel-cycle costs available to us for these reactors were good enough to warrant a numerical calculation.) Discounted to the start of reactor operation, the central estimates of these LWR net costs relative to LEU operation are seen in the table to range from $500-$800 million in the case of reactors that do not pay property taxes and insurance, fed by MOX from FMEF, up to $1,000-$1,500 million for reactors that do pay these items, fed by new fuel fabrication plants that also pay them.

It must be emphasized that, although we believe the procedures by which these figures were calculated represent some improvement in consistency and transparency over those used in Phase I of DOE's PDS, it would nonetheless be a mistake to attribute great accuracy to any of the results shown. Some potentially important shortcomings of the contractor studies—such as inconsistent and sometimes inadequate treatment of site costs, interim spent fuel storage, low-level waste management, and seismic requirements—were not corrected here; costs and delays in safety and environmental analysis, licensing, and the like could be greater than implied in the prescriptions we have used for preoperational costs and IDC; and it is hard to be sure whether our procedures for

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

expanding the uncertainty ranges on contractor estimates of direct plus indirect costs of power-plant and fuel-fabrication-plant construction, and of O&M costs for these facilities, have compensated adequately for noncomparable degrees of comprehensiveness and conservatism in the original estimates.

It should also be emphasized that the per-kilowatt-hour figures in Table 6-16 are not to be confused with estimates of the costs of electricity to be expected from reactors of these types if they were placed in commercial operation on a substantial scale. After all, the estimates given here are based on prorating, over only a few units, substantial preoperational costs associated both with new reactor types and with implementation of plutonium fuel cycles; per-unit and per-kilowatt-hour charges associated with these costs would be smaller if the new reactor types and fuel cycles saw large-scale commercial service. In the other direction, private companies contemplating the use of new nuclear power plants for commercial electricity generation might base their financial calculations on a considerably higher real cost of money than the 0.07/yr figure used here.

We now address briefly the sensitivity of our economic estimates to our assumption that the plants operate only for the duration of the plutonium disposition campaign, rather than continuing to generate electricity using uranium fuel for an assumed 40-year plant lifetime. This question can be addressed (see "Issues and Criteria in Economic Evaluation of Alternatives" in Chapter 3) by considering that the power plant has a residual value at the end of the plutonium disposition campaign equal to the discounted present value at that time of the stream of electricity revenues from its future operation less the discounted present value of the stream of its future operating costs. (There is also a small correction for the postponement of the D&D costs of the power plant, but we neglect it here as small compared to the other uncertainties inherent in the calculation.)

We apply this prescription, as an example, to the case of the evolutionary PWR, using just central estimates. We take its nonfuel O&M costs to be unchanged from those shown in Table 6-16, at about as $0.012/kWh, its fuel-cycle costs on LEU to be $0.006/kWh (including repository charge), and the revenues to be $0.050/kWh. The annual output is 8.26 × 109 kWh, so the net revenue stream is 8.26 × 109 kWh/yr × ($0.050 - 0.018)/kWh = $264 million/yr during the 10 years of plant life beyond the 30 years of operation in the plutonium campaign, which at r = 0.07/yr has a discounted present value, at the end of the plutonium campaign, of $1,854 million. Discounting this sum to the start of reactor operation at the beginning of the plutonium campaign (30 years earlier), for comparison with the net cost of campaign at reactor startup shown in Table 6-16, gives $243 million. This approach, then, would reduce the central estimates of the net cost of campaign for the evolutionary LWR from $1.6 to $1.4 billion (reactor paying no property tax and insurance, fed by FMEF) or from $3.2 to $3.0 billion (reactor and new fabrication plant both pay property taxes

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

and insurance). Thus, while we believe for the reasons mentioned earlier that the most informative economic comparisons are those based only on the period of plutonium disposition, it is apparent that our conclusions would not be changed much if we instead adopted the TRC approach.

Economics of Vitrification

Much less information and analysis are available on the costs of vitrifying surplus WPu along with defense high-level wastes, compared to the amounts of information and analysis we were able to draw on in making our cost estimates for the reactor options. The one economic analysis we have seen (McKibben et al. 1993) is for the incorporation of 50 tons of U.S. surplus WPu into a fraction of the glass logs already scheduled to be produced at the Savannah River site as a means of immobilizing defense high-level radioactive wastes (see Chapter 5 and Table 6-3). The panel believes that variations of the vitrification option that made less use of already planned facilities and operations would cost more, assuming that the full costs of constructing new facilities, and of producing and disposing of thousands of glass logs that would not otherwise be produced, were charged to the WPu disposition mission.

The McKibben et al. analysis of the cost of adding 50 tons of WPu to the Savannah River log-making operation estimated that the preliminary steps, including conversion of plutonium metal to oxide, would cost about $400 million and the other increases in the costs of the log-making campaign would amount to about $600 million. We assume that the preliminary costs are spread over a nine-year period prior to melter operation, which for the usual S-curve pattern and real interest rate of 0.07/yr gives (see Table 3-7) a discounted present value at start of melter operation of $400 million × 1.41 = $564 million. If the remaining $600 million in the McKibben et al. estimate represents additional costs incurred at the rate of $75 million per year during the eight years of log-making that follow the initiation of plutonium additions to the melter, the contribution of this cost stream to the net discounted present value at start of melter operation would be about $448 million. We consider that the resulting $1 billion total for the net discounted present value of the preoperational and operational cost streams as of the start of melter operation (which in our reference scenario would be in 2005) should be considered uncertain by ±50 percent, given the lack of actual operational experience with an activity of this kind. Thus our estimate for this reference vitrification case is $0.5-$1.5 billion.

Economics of Russian Disposition Options

No comparably detailed assessment of the economics of Russian WPu disposition options is possible with the information available to the panel. Costs of large capital projects in Russia, while substantially lower than in the United States, are highly uncertain and, with current substantial inflation, changing

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

rapidly and unpredictably. Labor costs in Russia are lower than those in the United States by large factors at present (the average salary corresponding, as of mid-1994, to approximately $100 per month), but labor rates too are changing quickly and unpredictably. Costs in different sectors can be dramatically different. Electricity prices are still at extremely low subsidized levels, and there is a massive nonpayments problem. This has led to a severe economic crisis throughout the nuclear industry, with plants having too little income to pay workers, purchase fuel, and the like. Nuclear workers have staged well-publicized strikes, and have asserted that labor conditions are now further imperiling nuclear safety. A nuclear regulatory agency and an antinuclear environmental movement have only come into being in recent years, and their full impact on how nuclear operations will be conducted in Russia is not yet established. Given these conditions, calculating likely costs of nuclear projects that would take place years in the future is not practical.

Nevertheless, a few remarks are in order. First, Russia has abundant supplies of cheap uranium and uranium enrichment services; indeed, in today's fuel market, the states of the former Soviet Union are the cheapest available source of these commodities. (With plants already built, the primary marginal cost of enrichment is electricity, which, as just noted, is currently quite cheap in Russia.) Second, like the United States, Russia does not have an operating commercial-scale MOX fabrication facility, and therefore very large capital investments would be required before plutonium could be used as a fuel on a substantial scale. It seems clear, therefore, that the conclusion that a kilogram of LEU fuel is cheaper to provide than an equivalent kilogram of MOX fuel—even if the plutonium in the MOX is itself available "free" as a result of weapons dismantlement—is also true in the Russian case.

Because of the low material and labor costs in Russia, however, it may be that a kilogram of MOX fuel could be produced in Russia for less than the cost of an equivalent kilogram of LEU fuel in the West. This could create opportunities for largely private financing of plutonium disposition in Russia— minimizing or eliminating the need for explicit government subsidies—that do not exist in the case of U.S. WPu disposition. For example, if one assumes (conservatively, the panel believes) that the capital cost of building a MOX plant in Russia would be roughly one-half what it would be in the West, and that all of the cost of MOX is accounted for by capital cost rather than labor cost (where the cost would be reduced by an even larger factor), then the cost of producing MOX in Russia would be one-half the cost we have calculated for the United States using a new facility and paying no property taxes, or some $800/kg. (If the partly completed facility could be completed for a lower cost and meet acceptable safeguards and safety standards, the cost might be lower still.) This compares to the $1,400/kg cost of equivalent LEU projected earlier in this chapter for 2015, the estimated midpoint of plutonium disposition operations.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

Thus, if the large current price differentials between Russian and Western prices of labor and materials were maintained over a long period, one could imagine two alternatives by which MOX production in Russia might be privately financed:

  1. MOX produced in Russia could be sold on the commercial market in Western Europe in competition with LEU. If Western countries or utilities agreed to firm contracts to purchase such MOX, these contracts could serve as the basis for borrowing, on international capital markets, the funds necessary to build and operate the plant.

  2. If there were a desire to keep MOX operations limited to a small number of sites in the country of origin, Western countries might agree to permit additional LEU sales in their markets (to which Russia's access is limited by trade agreements) for each kilogram of MOX Russia produced and used in its own reactors. In other words, rather than producing MOX for sale abroad and LEU for its own reactors, Russia would produce an equal amount of MOX for use in its own reactors and ship that amount of LEU abroad instead. The price available in the West (and therefore the possibility of borrowing on international capital markets) would be similar to that in the previous case.

Needless to say, major nuclear capital projects of this kind in Russia today would involve substantial uncertainties and risks. The major banks would probably have higher confidence in the success of the project (and therefore charge a lower rate on capital) if a Western firm with experience in MOX production (such as Siemens, COGEMA, British Nuclear Fuels, Limited, or Belgonucleaire) were involved. Some of these firms have already been discussing possible co-operation in MOX fabrication with Russia. (Both government and corporate policy-makers, however, will have to take into account the possibility of stiff commercial competition from a Russian MOX producer, once the plant was built and operational.)

The vitrification option would presumably also be cheaper to implement in Russia than in the West, due to lower capital and labor costs, and possibly a less stringent regulatory environment. In the United States, as described above, the likely net subsidies required for the LWR MOX and vitrification options are likely to be in the same general range. In Russia, however, the possibility described above for largely or completely private financing of the operation presents itself only in the case of the MOX option. Since vitrification produces no saleable product, its full cost would have to be provided by the government, as in the West.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

ENVIRONMENT, SAFETY, AND HEALTH

We argued in Chapter 3—dealing with criteria relating to environment, safety, and health (ES&H)—that suitable options for the disposition of WPu in the United States should:

  • comply with existing U.S. regulations governing radioactivity and radiation from civilian nuclear-energy activities;

  • comply with existing international agreements and standards on the disposition of radioactive materials in the environment; and

  • not add significantly to the ES&H burdens that would result, in the absence of programs for WPu disposition, from appropriate management of civilian nuclear-energy generation and of the environmental legacy of past nuclear weapons production.

We argued, further, that disposition activities to take place in other countries should meet the same criteria, with the replacement of U.S. regulations by equivalent regulations of the countries in question. In the present section, we summarize our understanding of the specific issues that will have to be addressed in order to establish that the leading-candidate approaches described in Chapters 4 and 5—namely, the MOX fuel option in power reactors of existing types, and the vitrification-with-high-level-waste option—can indeed meet these criteria.

The main ES&H issues posed by the MOX fuel and vitrification options were identified in "The Main ES&H Issues in Weapons Plutonium Disposition" in Chapter 3. In addressing those issues here, we begin with a synopsis of the characteristics of plutonium that bear on ES&H hazards and then discuss, in turn, the hazards of interim storage, transport, and processing of plutonium; the influence of the use of plutonium-based fuel on reactor safety; and the ways in which the MOX fuel and vitrification options may influence the ES&H characteristics of radioactive wastes. Our treatment of these ES&H issues relates primarily to disposition of WPu in the United States. The arguments would not be substantially different for disposition in Russia or other countries, although the background of existing ES&H risks from civilian and military nuclear-energy activities and the regulatory structures responsible for minimizing these risks would differ.

Relevant Characteristics of Plutonium

Table 6-17 compares some of the hazard-relevant characteristics of principal isotopes of plutonium with the same characteristics of the uranium isotopes that make up natural uranium. Notice that while the half-lives of most of the plutonium isotopes are long enough to make these a long-lived hazard when measured against the scale of a human lifetime, the plutonium half-lives are nonetheless very much shorter than those of the uranium isotopes. As a result,

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-17 Hazard-Relevant Properties of Principal Plutonium and Uranium Isotopes

Isotope

Half-Life (yr)

Curies per Gram

Main Emissionsa (MeV)

Worker Ingestion ALI (µCi)

Worker Inhalation ALI (µCi)

Public DCLW (Ci/m3)

Public DCLA (Ci/m3)

Dilution Volume in Air (m3/g)

Pu-238

8.8x101

1.7x101

5.5 a

9x10-1

7x10-3

2x10-8

2x10-14

8.5x1014

Pu-239

2.4x104

6.2x10-2

5.2a

8x10-1

6x10-3

2x10-8

2x10-14

3.1x1012

Pu-240

6.6x103

2.3x10-1

5.2 a

8x10-1

6x10-3

2x10-8

2x10-14

1.2x1013

Pu-241

1.4x101

1.0x102

0.02b

4x101

3x10-1

1x10-6

8x10-13

1.3x1014

Pu-242

3.7x105

4.0x10-3

4.9a

8x10-1

7x10-3

2x10-8

2x10-14

2.0x1011

Am-241

4.3x102

3.5x100

5.5a, 0.06g

8x10-1

6x10-3

2x10-8

2x10-14

1.8x1014

U-234

2.5x105

6.2x10-3

4.8a

1x101

4x10-2

3x10-7

5x10-14

1.2x1011

U-235

7.1x108

2.1x10-6

4.4a

1x101

4x10-2

3x10-7

6x10-14

3.5x107

U-238

4.5x109

3.3x10-7

4.2a

1x101

4x10-2

3x10-7

6x10-14

5.5x106

ABBREVIATIONS:

ALI: annual limit of intake.

Ci/m3: curie per cubic meter.

DCLA: derived concentration limits for air.

DCLW: derived concentration limits for water.

MeV: million electron volts.

µCi: microcurie

NOTES: Half-lives and emissions are from IAEA (1986a). Allowable intakes and concentrations are from OFR (1992, Appendix B to Sections 20.1001-20.2401) and apply after January 1, 1994. Americium-241 is included here because of its rapid buildup in plutonium from the decay of 14.4-year halflife Pu-241.

a a = alpha particle, b = beta particle, g = gamma ray.

the intensity of the radioactivity of plutonium—measured as "specific activity" in curies per gram—is much larger than that of uranium. For example, the specific activity of the most common plutonium isotope, Pu-239, is about 200,000 times greater than that of the most common uranium isotope, U-238. Since the alpha particles emitted in the radioactive decay of Pu-239 are about 25 percent more energetic than those emitted in the decay of U-238, the former is altogether about 250,000 times more damaging per gram, radiologically, than the latter.

The preceding comparison does not take into account differences in the pathways and residence times of plutonium and uranium in the body or in the environment. The differences with respect to pathways and residence times in the body have been accounted for, however, in the Annual Limit of Intake (ALI) by ingestion and inhalation for workers in nuclear industries and the Derived Concentration Limits for air and water (DCLA and DCLW) for public exposure, as promulgated by the Nuclear Regulatory Commission (last 4 columns of Table

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-18 Hazard Indices for Various Mixtures of Heavy-Metal Isotopes

 

Weight Percent of Isotopes in Mix

Dilution Volume (m3/g)

Surface Gamma Dose

 

Pu-238

Pu-239

Pu-240

Pu-241

Pu-242

Am-241

In Air

In Water

(rem/hr)

WPu

0.01

93.8

5.8

0.13

0.02

0.22

4.2x1012

4.2x106

0.9

RPu

1.3

60.3

24.3

5.6

5.0

3.5

2.9x1013

2.8x107

14.6

 

U-234

U-235

U-238

Natural U

6x10-3

0.72

99.2

 

 

 

1.3x107

2.3x100

1.2x10-5

Reactor U

0.03

3.5

96.5

 

 

 

3.8x107

6.5x100

5.7x10-5

Weapons U

0.12

94.0

5.9

 

 

 

1.8x108

3.2x101

1.5x10-3

NOTES: Dilution volumes based on specific activities and DCLs from Table 6-17. Surface gamma doses calculated for uranium and plutonium metal at highest naturally occurring densities.

6-17). The DCLs and specific activities can be combined to calculate, isotope by isotope and for any mixture of these, the volumes of air or water that would be required to dilute a gram of any radioactive material to the concentration limits specified by these standards. 27 This "dilution volume" index is widely used in preliminary environmental assessment of nuclear technologies, to convey a rough idea of radiologic hazard potential.

Table 6-18 shows the dilution volumes for isotopic mixtures of plutonium corresponding to weapons-grade and reactor-grade material, with comparisons to natural uranium, LEU (at the 3.5 percent enrichment corresponding to typical fuel for a PWR), and HEU (at the 94 percent enrichment corresponding to nuclear weapons application). One sees there that, per gram, weapons-grade plutonium represents a potential inhalation hazard about 23,000 times greater and a potential ingestion hazard about 130,000 times greater than the corresponding potential hazards of weapons-grade uranium; the potential hazards of natural uranium in these respects are about 14 times smaller than those of weapons-grade uranium, and those of reactor-grade plutonium are about 7 times greater than those of weapons-grade plutonium.

27  

The DCLs for public exposure correspond to the concentrations that would produce a steady-state whole-body dose commitment of 0.5 millisieverts (mSv) (50 millirem; mrem) per year to an individual who continuously breathed air or continuously drank water contaminated to the indicated levels. If i different isotopes are present simultaneously in air at concentrations Ci, the requirement is that Σi(Ci/DCLAi) < 1, and, similarly, if j different isotopes are present simultaneously in water at concentrations Cj, the requirement is that Σj (Cj/DCLW j) < 1. Note that dose rates of 0.5 mSv each from air and water, to which the DCLA and DCLW correspond, relate to the total dose rate of 1 mSv (100 mrem) per year permitted to an individual member of the public, under Environmental Protection Agency regulations, from all nuclear facilities combined; the dose permitted from any single nuclear facility, however, is 10 times smaller (see "Some Relevant Standards Limiting Doses and Emissions" in Chapter 3).

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

The DCLs and dilution volumes for the plutonium and uranium isotopes are dominated by the dose potentials presented by these isotopes as internal (inside the body) emitters of alpha particles. As can be seen by comparison of the occupational annual intake limits for plutonium via ingestion and inhalation (Table 6-17, columns 5 and 6), a given number of curies is much more dangerous if inhaled than if ingested. This is because most ingested plutonium passes through the body without being absorbed, whereas plutonium inhaled in the form of particles of small diameter tends to become lodged in the lungs; there the plutonium can do great damage directly, as well as being gradually absorbed into the bloodstream and thus gaining access to bone, kidneys, and other vulnerable parts of the body. Recent reviews of the available information on the carcinogenicity of inhaled plutonium oxide suggest that the damage-to-exposure ratio is in the range of 3 to 12 excess cancer deaths per milligram of weapons-grade plutonium inhaled in oxide form by an exposed population (Fetter and von Hippel 1990, National Research Council 1988).28

Notwithstanding the high carcinogenic potential of alpha particles when emitted inside the body, they cause no harm when emitted outside it because they cannot penetrate the dead layer of the skin. Gamma rays can penetrate the body from outside it, however, and while plutonium and uranium isotopes and most of their daughter products emit very little in the way of gamma rays, the americium-241 (Am-241) daughter of Pu-241 emits a 60-kiloelectron volt (keV) gamma ray that can be a significant source of external radiation dose. The amount of Am-241 in a given quantity of plutonium depends both on the initial Pu-241 concentration (which is relatively high in reactor-grade plutonium and low in weapons-grade plutonium) and on the time elapsed since the plutonium was separated (since the buildup of Am-241 is governed by the 14-year decay half-life of the Pu-241). The gamma dose rates at the surface of metallic spheres of plutonium and uranium of different isotopic compositions are shown in the last column of Table 6-18. It can be seen there that even the modest percentage of Am-241 in weapons-grade plutonium (about 0.2 percent for the nominal composition considered here) is enough to yield a gamma dose rate about 600 times higher than that from weapons-grade uranium, and the gamma dose rate from reactor-grade plutonium containing a nominal 3.5 percent Am-241 is another 15 times higher still.

The ratios of dilution volumes and of surface gamma doses are instructive in clarifying why the handling and processing of plutonium require far more stringent precautions for the protection of workers and the public against radia-

28  

Under the usual linear hypothesis (see Chapter 3, Appendix D), this would mean that 3-12 excess cancer deaths would be expected in a population of 1,000 people inhaling 0.001 milligrams each of weapons-grade plutonium in oxide form, that the same number of excess cancer deaths would be expected in a population of 100 people inhaling 0.01 milligrams each, and, at the extreme, that one individual inhaling 80-300 micrograms could expect to suffer a fatal cancer as a result.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

tion than do the handling and processing of uranium. Some additional factors, relating to the pathways by which plutonium can become available for intake by humans, are also relevant. One such factor is that plutonium metal shavings or filings can be pyrophoric (that is, they can ignite spontaneously in air); this phenomenon has been responsible for a number of serious fires in the U.S. nuclear-weapons-production complex (IPPNW/IEER 1992, National Research Council 1989). Plutonium fires, of course, produce plutonium-oxide smokes representing an inhalation hazard for workers and, if the fire is uncontained, for the public.

Another set of pathways by which plutonium could reach humans is propagation through food chains following dispersal in the environment by, for example, a fire or other accident during transport. Plutonium appears to bind strongly to soil particles and is not readily taken up by terrestrial plants, with concentration factors (parts per million [ppm] plutonium in dry plant material divided by ppm plutonium in dry soil) reported to be typically less than 0.01 (Eisenbud 1973). This means that resuspension of respirable plutonium particles from contaminated soil is likely to be a more important pathway to human exposure than ingestion of plutonium taken up from soil by food crops. In marine ecosystems, by contrast, plutonium is strongly concentrated by plankton, seaweed, and shellfish: marine concentration factors for plutonium used by the International Atomic Energy Agency in connection with the regulation of ocean dumping of radioactivity include factors of 100,000 (ppm plutonium in wet biomass divided by ppm plutonium in seawater) in plankton, 2,000 in seaweed, 3,000 in molluscs, and 40 in fish; the corresponding concentration factors for uranium are 20, 100, 30, and 10, respectively (IAEA 1986b). These figures suggest that marine food chains could be important pathways to human exposure if plutonium found its way in significant quantities into the oceans.

In addition to the characteristics of plutonium described in the preceding paragraphs, all of which bear on the radiological hazards presented by the handling and possible dispersal of this element, the possibility of an accidental chain reaction involving plutonium must also be mentioned. The possibility of plutonium's “going critical" as a result of mishandling or other mishap in processing, transport, or storage represents a danger of direct irradiation of humans in the vicinity by the intense flux of neutrons and gamma rays produced by criticality (see section on "Physics and Technology of Nuclear Fission" in Chapter 2) as well as a secondary radiologic hazard from the fission products produced by the chain reaction; and it could represent, in some circumstances, a source of sufficient energy release to disperse the plutonium itself and any accompanying radioactive material more rapidly or widely than would be likely in the absence of a chain reaction.

It must be emphasized that the kinds of criticality accidents to which we are referring here—that is, excluding the accidental detonation of a nuclear weapon (which is outside the scope of our responsibilities in this report) and criticality

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

accidents in nuclear reactors (which are discussed separately below)—are not expected to entail actual nuclear explosions or even the smaller but still very impressive energy releases associated with severe nuclear reactor accidents.29 The few criticality accidents known to have occurred over the years in the U.S. nuclear-weapons complex, for example, have involved nuclear-energy releases ranging from a few megajoules to a few hundred megajoules, in contrast to about 85 million megajoules in a 20-kiloton nuclear weapon and perhaps I million megajoules in the criticality accident that set off the Chernobyl reactor disaster (National Research Council 1989, IPPNW/IEER 1992). The nuclear energy releases in the nonreactor criticality accidents were enough, however, to produce potentially lethal radiation doses at a distance of several meters; and it is certainly possible to imagine circumstances—e.g., with the reactants confined by soil or rock in a shallow burial site or deep geologic repository—where the energy release could continue for long enough to become a significant driver of dispersal of radioactivity.

Hazards in Interim Storage of Plutonium

All options for the disposition of WPu will entail some period of interim storage of this material subsequent to the dismantling of the warheads. The 1994 CISAC study, for which our own work was conducted, concluded that the preferred approach for this interim storage is to store the plutonium in the form of the nuclear-weapon cores or "pits" in which it is removed from the weapons during dismantling, perhaps with the addition of some deformation or other modifications to reduce the immediate reusability of the pits in new nuclear weapons. Destroying the pits and converting their plutonium metal to alloys with other metals, or to plutonium oxide or plutonium nitrate or other compounds, was not recommended for this interim-storage purpose because (1) such processing produces security liabilities that at least partly offset the gains;30 (2) the processing would entail costs, which it is more efficient to defer until it has become clear what form of processing is needed for whatever subsequent disposition option is chosen; and (3) the processing would entail additional ES&H hazards, beyond those associated with storing the pits, with no clear reason to want to face them in the absence of substantial security gains from the processing and in view of the possibility that a different form of processing might turn out later to be what is required.

29  

Achieving a true nuclear explosion requires bringing the fissile material from a subcritical to a highly supercritical configuration in a very small fraction of a second. The considerable difficulty of accomplishing this intentionally in a nuclear weapon, in which a substantial quantity of chemical high explosive and carefully configured fissile components are employed for the purpose, suggests that its accidental occurrence in the absence of these ingredients is extremely improbable.

30  

Processing itself provides opportunities for diversion or theft, and the plutonium pits are easier to count and to guard than, for example, plutonium-oxide powders.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

The problem of pit storage, being common to all disposition options, is not strictly speaking a part of the Reactor Panel's charge, but we nonetheless will offer some observations here on the ES&H aspects of this operation, in the interest of completeness. The main ES&H issues associated with pit storage would appear to be:

  1. avoidance of criticality accidents resulting from excessive proximity and inadequate shielding of two or more pits in combination, as could occur in the course of bringing the pits into the facility, or as a result of insufficient care in designing the array in which they are to be stored, or as a result of subsequent unintended rearrangement of this array, for example, by flood, earthquake, or aircraft impact;

  2. avoidance of accidental plutonium dispersal, particularly by fires in the storage facility (which could mobilize the plutonium metal in the pits as plutonium-oxide smoke); and

  3. avoidance of excessive doses to the workers handling the pits at the time of their emplacement in storage or in subsequent monitoring of them (such avoidance being mainly a matter of a combination of shielding and restricted exposure time, with particular reference to the gamma emissions from the Am-241 daughter of Pu-241).

Careful attention will have to be paid to all of these issues, and assurances about how the first two in particular will be managed will have to be provided to the communities in the vicinity of the storage facilities as well as to the responsible regulatory authorities.

At the same time, we see no reason that these risks cannot be restricted—given appropriate diligence and the provision of adequate funds—to very low levels, well within our criteria that existing U.S. regulations about public and worker exposures to radioactivity would be met and that the total ES&H burdens of this activity would be small compared to those associated with appropriate management of other civilian and military nuclear-energy activities. The safe storage of pits is not, after all, a more demanding task than the safe storage of comparable (and even larger) numbers of intact nuclear weapons, and the latter task has been accomplished in the United States for decades with few ES&H problems.31

Temporary storage of plutonium in forms other than pits will of course be involved at the later stages of any disposition option, and ultimately all the plutonium that has not been fissioned will have to be stored essentially permanently

31  

The ES&H problems that have occurred in some parts of the U.S. nuclear-weapon complex over the years (about which more below) have been associated mainly with steps other than weapon storage; see National Research Council (1989), OTA (1991, 1993), and IPPNW/IEER (1992). While less is known about the corresponding history in the nuclear-weapon complex of the former Soviet Union, the main ES&H problems there that have come to light so far also involved operations other than weapon storage (IPPNW/IEER 1992).

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

in one or another form of radioactive waste. The ES&H issues connected with radioactive wastes are taken up below in "Radioactive Waste Issues." The prior forms needing temporary storage are likely to include plutonium-dioxide powder and may also include liquid plutonium nitrate solutions and fabricated MOX nuclear fuel. Of these, the plutonium dioxide and the plutonium nitrate will require the greatest care to avoid accidental criticality and release modes that could produce significant worker exposures. (The fabricated MOX fuel is less problematic because (1) the plutonium is diluted by about 20-to-1 with nonfissile and much less radiotoxic U-238, (2) the plutonium is confined by the ceramic-pellet fuel matrix and by metal cladding, and (3) the cladding attenuates somewhat the gamma rays from the contained Am-241.)

Handling of plutonium dioxide and plutonium nitrate in the nuclear-weapon complex has not always been exemplary (National Research Council 1989, IPPNW/IEER 1992, OTA 1993)—all of the criticality accidents that have been publicly described involved plutonium nitrate, for example—but the technologies and procedures for avoiding such problems are well established (see, e.g., PNL 1988, OTA 1993). Levels of awareness and sophistication about ES&H issues are greater in the 1990s, in the relevant government organizations and indeed throughout society, than was the case in earlier decades when ES&H problems in some parts of the nuclear-weapon complex materialized; and specific recent initiatives give further reason to expect that the diligence required for safe plutonium handling in dealing with surplus nuclear weapons will in fact be applied.32 The ES&H risks associated with storage of plutonium from dismantled surplus nuclear weapons must be compared, in any case, with the risks of storage of a wider array of plutonium forms and plutonium-contaminated materials that have been produced over the years in both the civilian and military nuclear complexes; the forms and locations of much of this material make it more problematic from the ES&H standpoint than stored plutonium from surplus weapons is likely to be.

32  

For example, DOE has recently carried out an extensive Plutonium Vulnerability Study identifying the most important ES&H problems associated with the various forms of plutonium at the sites in the DOE complex, and has developed an extensive action plan to resolve the problems found. In April 1994, before the vulnerability study was complete, the Defense Nuclear Facilities Safety Board (DNFSB) recommended a wide range of improvements in handling of plutonium in the DOE complex. Intensive work is now underway at several DOE sites to correct the deficiencies identified by the DNFSB and the vulnerability study. In addition, although DOE's facilities have been exempted until now from regulatory oversight by the Occupational Safety and Health Administration (OSHA), DOE's May 1993 Health and Safety Initiative commits the DOE to a transition to regulation by OSHA over the next 3-5 years. The background leading up to these developments is described in considerably greater detail in the recent study of management of nuclear-weapons materials by the Office of Technology Assessment of the U.S. Congress (OTA 1993).

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

Hazards in Plutonium Transport

The transport links likely to be associated with plutonium disposition were detailed above in "Economic Comparisons," in connection with discussion of the associated security risks. Truck and rail are the modes most likely to be chosen for these transport activities, and in either case special precautions (protective containers, carefully selected routes) will be needed to minimize the probability of a release and the extent of public exposure if a release occurs.

In the case of the transport of intact or deformed pits from the warhead dismantlement plant to the (future) central storage facility and the subsequent transport of the pits from there to a MOX fuel fabrication or vitrification plant (assuming that conversion to oxide takes place at these plants and not adjacent to the storage facility), the obvious comparison is with transport of intact nuclear weapons. Such transport has been managed in the United States for tens of thousands of nuclear weapons over the years with almost no serious ES&H incidents; the only exceptions have involved aircraft crashes. (Less is known about the record in the former Soviet Union.)

Air transport is not likely to be contemplated for the transfer of pits from interim storage to a MOX fabrication or vitrification plant on the same continent (mainly because of the difficulty of designing reasonably lightweight containers that could prevent the dispersal of the plutonium in the event of a crash); and transport of pits between continents is, for political and security reasons, unlikely to be undertaken at all. Truck and rail transport are therefore the modes most likely to be used, and we see no reason why such transport cannot be managed for pits with the same high degree of safety and reliability that has characterized truck and train transport of intact warheads in the United States. If, nonetheless, a transport accident severe enough to breach the shipping containers did occur, the possibility of a fire's converting the plutonium metal in the pits to plutonium-oxide smoke would pose a public health hazard.33

In the case of the vitrification option, co-location of the processes for converting pits to plutonium oxide with the vitrification plant would avoid further transportation steps except for the eventual transport of the plutonium-bearing glass logs to a permanent geologic repository. Inasmuch as the nominal 50 tons of surplus U.S. WPu could be embedded in fewer glass logs than will need to be produced in any case in order to stabilize the accumulated high-level radioactive wastes from this country's defense programs, choosing this means of WPu disposition would not entail any additional transportation of logs or, hence, any additional probability of transportation accidents.

Nor would the addition of WPu to the glass logs affect significantly the routine doses of radiation to the workers involved with transporting the logs. The routine doses to transport workers would come almost entirely from gamma

33  

Some pits were designed to be fire resistant and would, therefore, be less susceptible to this accident scenario.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

radiation; in logs containing about 20 percent by weight of 20- to 40-year-old fission products and 1-2 percent by weight of WPu, the gamma dose from fission product cesium-137 will exceed that from the Am-241 in the plutonium by several orders of magnitude (see Table 6-5 and accompanying notes).

As for doses to the public in the event of an accident, even a crash severe enough to break the transportation cask, followed by a long and intense fire, would be unlikely to mobilize much of the highly refractory glass log. To the extent that such an accident did mobilize radioactive material from the log, however, the contribution of the plutonium to the dose potential to the public could add very substantially to the contribution from the fission products.34 This means that, in order to meet the criterion that the disposition of the WPu should not add substantially to the ES&H risks from the nuclear-energy activities that would be going on in any case, one must ensure that the total expected dose to the public from accidents in the transport of the glass logs will be small compared to the doses to the public from other aspects of civilian and military nuclear-energy activities, if those other aspects are properly managed. Given the relatively small number of glass logs needed to accommodate all of the surplus WPu, the highly refractory character of the glass, and the capacity to make shipping casks that greatly reduce the probability of significant releases even in severe accidents, it seems very likely that this will be the case. Further study of the mobilizability of plutonium from borosilicate glass logs in severe transport accidents is warranted, however.

In the case of the spent fuel option, it is reasonable to assume that conversion of pits to plutonium oxide would be co-located with other fuel fabrication steps, so that the remaining transport links would include only the transport of the fabricated fuel to the reactor site(s)—if the reactors are not co-located with the fabrication plant—plus transport of the spent reactor fuel to any intermediate

34  

According to Table 6-5, note 9, the logs to be produced at Savannah River will contain about 13 curies (Ci) each of cesium-137 (Cs-137) and strontium-90 (Sr-90) per kilogram of glass— these two isotopes dominate the hazard of a fission-product mixture of the age of this material— and at 2-percent WPu by weight the logs would contain 20 gWPu/kg. Suppose the respective fractional release rates of plutonium and fission products were similar to those expected from molten spent fuel in a severe reactor accident. The "worst-case" accident releases from oxide reactor fuel are generally estimated to be 10-40 percent for cesium, 5-10 percent for strontium, and 0.3-3.0 percent for plutonium (USNRC 1975, Hohenemser 1988). Taking 25 percent for cesium, 5 percent for strontium, and 1 percent for plutonium would give releases from the glass of about 3 Ci Cs-137, 0.7 Ci Sr-90, and 0.2 gWPu/kg of glass. (It is not necessary to imagine that absolute releases of this magnitude are possible; we are interested just in the relative magnitudes of the fission product and plutonium hazards, assuming they are released in these proportions.) Taking the dilution volume for WPu from Table 6-18 as 4 × 1012 m3/g and those for Cs-137 and Sr-90, based on the January 1, 1994 DCLAs from the U.S. Code of Federal Regulations (OFR 1992), as 5 × 109 m3/Ci and 2 × 1011 m3/Ci, respectively, gives "hazard measures" of 1.4 × 1011 m3 for the released cesium and strontium (combined) versus 8 × 1011 m3 for the released plutonium. If the material were totally vaporized, releasing plutonium and fission products in direct proportion to their weight fractions in the glass, plutonium's dominance of the hazard (by this crude measure) would be even greater.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

storage site and, later, its transport to a final repository. Fresh-fuel transport is a link with no counterpart in the vitrification option, but it probably poses only moderate ES&H risks. Because of the absence of fission products, the relatively low gamma dose rate from WPu, and the integrity of the fuel pellets and cladding under normal conditions, the routine radiation exposures to the workers in transport of fresh MOX fuel should be extremely low.

As for the risk of public exposures in severe transport accidents involving fresh fuel, the use of suitably sturdy and fire-resistant shipping containers should be able to reduce to a very low level the probability of accidents severe enough to mobilize significant quantities of plutonium. That plutonium-oxide pellets have a very high melting point and do not burn is helpful in this respect. (The combustibility of the graphite matrix in MHTGR fuel in a severe accident that includes a fire needs clarification. In the case of those LMRs that use metallic fuel, the fuel would certainly be combustible in a sufficiently severe fire; but since fabrication of such fuel under the IFR scheme is likely to be co-located with the reactor, there would be no highway or railroad accidents in fresh-fuel transport to serve as causes for such fires.) If significant quantities of fresh fuel could be mobilized so as to become airborne as particulate matter, the resulting public health risks would be substantial.35 This puts a heavy premium on prevention.

Transport of spent MOX fuel under the spent fuel option would pose transportation risks qualitatively similar to those from transport of glass logs under the vitrification option. The quantitative risks of transport depend on the number of shipments and the distance of transport, on the probability of shipping-container failure in the event of an accident, and on the radiologic hazard potential of the material in each shipment and its mobility under accident conditions.

The spent fuel assemblies will have larger inventories of fission products and larger plutonium hazard potential per kilogram than the glass logs, and some of the fission products in the spent fuel may be in forms more volatile under accident conditions than any of the fission products in the glass logs.36 On

35  

Recall from Table 6-1 that the amounts of plutonium in fresh MOX or MHTGR fuel as a fraction of the fuel matrix fall in the range of 1 to 5 percent by weight, with plutonium fractions in LMR fuels being several times higher. A gram of MOX fuel containing 4 percent by weight WPu and mobilized as fine airborne particulate matter would require a dilution volume of 1.6 × 1011m3 of air to reach the DCLA corresponding to a whole-body-equivalent annual dose of 0.5 millisieverts (50 millirem) from continuous inhalation. This implies that the whole-body-equivalent dose-commitment from one-time inhalation of 25 micrograms of this material would exceed the 0.25 sievert (25 rem) once-in-lifetime emergency dose limit specified in Nuclear Regulatory Commission regulations. (To derive this result, note that the steady-state annual dose from continuous intake of a curie of a radioactive mixture per year is approximately equal to the lifetime dose commitment from a one-time intake of a curie of the same mixture; see, e.g., APS 1978.)

36  

Thermal fission of U-235 makes 3.3 Ci Cs-137 and 2.9 Ci Sr-90 per MWd of thermal energy release, hence per gram of fission products, while thermal fission of Pu-239 makes 3.3 Ci Cs-137 and 1.1 Ci Sr-90. Thus fresh spent fuel from the thermal fission of MOX fuel irradiated to 40

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

the other hand, the brittleness of the glass logs could contribute to the potential for dispersion of the material in severe accidents, and, at our assumed loadings of plutonium in glass logs and in spent MOX fuel, the tonnage of material to be shipped in the course of disposition of 50 tons of WPu will be about 2.5 times smaller for the MOX option than for the glass option.37 The average shipment distance would be on the order of three times greater for the glass logs than for the spent fuel, assuming that the MOX-burning reactors would be in the western United States, the glass-log manufacture would be at the Savannah River site, and the repository for either waste form would be at Yucca Mountain; other assumptions are possible, of course, and would yield different conclusions about comparative distances of transport.

The actual number of shipments that will be required in each case will depend not only on the mass and bulk of the spent fuel assemblies and of the glass logs in their canisters, but also on the mass and bulk of the respective shipping containers. Such containers must provide shielding to protect drivers, other workers, and the public from the gamma radiation originating in the material inside, as well as providing containment in the event of transport accidents. Shipping casks for LWR spent fuel have undergone extensive development and testing over a period of many years in several countries, and the resulting designs have demonstrated their integrity under the severe stresses imposed by

   

MWd/kgHM will contain 132,000 Ci Cs-137 and 44,000 Ci Sr-90 per MTHM, or 115,000 and 40,000 Ci of Cs-137 and Sr-90, respectively, per ton of MOX/fission product matrix. This spent fuel will also contain 20-50 kg Pu/MTHM (depending on the plutonium loading in the fresh fuel)—hence 18-44 kg Pu/ton of MOX matrix—which in most cases will be similar in isotopic composition to reactor-grade plutonium produced in once-through LEU fuel. If the spent fuel averages 15 years old at the time of transport, the fission product quantities will be about 80,000 Ci Cs-137 and 30,000 Ci Sr-90 per ton of MOX/fission product matrix. These figures compare to about 13,000 Ci each of Cs-137 and Sr-90 per ton of the borosilicate glass to be produced at Savannah River, and 13 kg of weapons-grade plutonium per ton in this glass (at the 1.3 percent by weight plutonium loading assumed in the reference case of Table 6-3). Then the Cs-137 hazard per kilogram of spent fuel would be about six times greater than that per kilogram of glass, the ratio for Sr-90 would be about a factor of 2, and the ratio for plutonium would be a factor of 7-20 (a factor of 1.4-3.4 from the greater mass of plutonium in the spent fuel and another factor of 5-6 from the greater radiotoxicity per gram of the post-irradiation plutonium compared to the still-weapons-grade plutonium in the glass). A modest additional factor of excess radiotoxicity in the spent fuel compared to that in the glass would result from the lower age of the fission products in the former, meaning that the accompanying fission products other than Cs-137 and Sr-90 would have had less time to decay or leak away. The relative mobility of fission products in spent fuel and glass under accident conditions needs more study.

37  

Disposition of the nominal 50 tons of WPu would utilize 2,200 glass logs containing 2,200 × 625 liters × 2.7 kg/liter = 3.7 × 10 6 kg glass; the mass of the logs including their stainless steel canisters would be 4.7 × 106 kg. For the reference MOX fuel case (CLWR, 100-percent MOX core, 40 gWPu/kgHM in fresh fuel, 42-MWd/kgHM irradiation), the disposition of 50 tons of WPu would utilize 2,700 fuel assemblies containing 1.25 × 106 kgHM, with a total mass, including cladding, spacers, and so on, of 1.8 × 106 kg, or about 2.5 times less than that of glass logs with their canisters.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

collisions, falls, fires, and the like.38 As far as we know, comparable testing of shipping containers for glass logs has not yet been carried out. Thus, we cannot yet determine whether glass-log containers designed to the same shielding and accident-integrity standards as are met by spent fuel casks will be more or less massive, in relation to the mass of the contents, than the spent fuel casks. If the ratio of mass of container to mass of contents were the same in both cases, then the number of shipments of glass logs would be larger than the number of shipments of spent fuel by approximately the factor of 2.5 ratio of the masses of the plutonium-bearing material to be shipped. We assume, then, that a somewhat smaller number and shorter average distance of shipments of spent fuel as compared to glass logs will tend to offset a somewhat larger spent fuel risk per shipment (based on higher radiologic hazard potential per kilogram, comparable container integrity, and, perhaps, higher mobility of radioactivity in the event of container failure). Thus we judge the overall transport risk to be roughly comparable.

The spent fuel from the use of WPu in MOX fuel may be somewhat more hazardous, in terms of potential public exposure in transport accidents, than spent fuel from the once-through use of LEU fuel. This is because there is two to four times more plutonium per ton in the WPu spent fuel than in comparably irradiated LEU spent fuel, while the isotopic compositions, and hence the hazard per gram, are practically the same. (The extra plutonium hazard in the WPu spent fuel is partly offset by the threefold smaller production of Sr-90 from thermal fission of Pu-239, compared to that from thermal fission of U-235. The Cs-137 contents are about the same in the two fuel types, for equal burnup.) These figures suggest that the overall extra public risk from transport of WPu-MOX spent fuel compared to that from transport of LEU spent fuel of equal burnup is not more than a factor of two or three.39 Two important conclusions follow from this result:

  1. Since the volume of spent fuel transport from ordinary civilian nuclear-power generation over the next few decades will be greater than the

38  

In the United States, for example, shipping casks for spent fuel must be able to survive, in sequence, a drop through a distance of 9 meters onto a hard surface, a drop of I meter onto a vertical, 15-cm-diameter steel bar, a 30-minute fire at 800° C, and 8 hours of submersion under a meter of water (OFR 1992, Sec. 71.73).

39  

Consider 15-year-old spent fuels that were irradiated to 42 MWd/kgHM and that contain 1.0 percent plutonium in the case of the LEU fuel and 2.6 percent plutonium in the case of the WPu-MOX fuel (see Table 6-1). The potential hazards from volatilization of portions of these fuels in accidents will be dominated by Cs-137, Sr-90, and the various plutonium and americium isotopes. Based on the production rates for Cs-137 and Sr-90 given in note 36 and the release fractions for cesium, strontium, and plutonium given in note 34, the dilution volume for the radioisotopes releasable from a kilogram of fuel is 1.9 times higher for the WPu-MOX fuel than for the LEU fuel. If one assumes, instead, that mobilization fractions are identical for all elements, the dilution volume per gram of released material is 2.4 times higher for the WPu-MOX fuel than for the LEU fuel.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

volume of spent fuel transport from WPu disposition by much more than a factor of two or three,40 we can conclude that the extra risk associated with transport of spent fuel from WPu disposition will be small compared to the risks of the same kind that society will be incurring anyway from civilian nuclear-energy activities. (That is, it is not necessary to invoke the fact that spent fuel transport risks are much smaller than some other fuel-cycle risks41 in order to reach the conclusion that the third of our ES&H criteria is met.)

  1. Since we have concluded that the transport risks of glass logs containing 1.3 percent WPu would be roughly comparable to the risks of the spent fuel transport connected with disposition of the same total quantity of WPu, it follows from (a) that the glass-log transport risks would also be small compared to risks of the same general kind that society will be incurring anyway from civilian nuclear-energy activities. So the third of our ES&H criteria is met for the transport risks from this option too.

Of course, much more thorough risk assessments for the transport of various plutonium forms will need to be done before such activities are actually undertaken on a substantial scale. We think the rough quantitative comparisons presented here for the MOX spent fuel and vitrification options are sufficiently robust, however, that our conclusion about the incremental transport risks for these materials being small compared to other risks of the same kinds is not likely to be overturned by further work.

Hazards in Plutonium Processing

Metal-to-Oxide Conversion

Conversion of the plutonium metal pits to plutonium oxide would be required as an initial processing step for the vitrification option as well as for the spent fuel option with most reactor types—LWRs, CANDUs, and MHTGRs using the current leading-candidate fuel formulation, and some LMRs. In the case of other HTGR designs using carbide fuels, and in the case of those LMR types that use metallic fuels in which plutonium is alloyed with other metals, the initial processing step would differ in details but would involve similar ES&H hazards. These arise mainly from the radiological toxicity of plutonium and

40  

Current U.S. nuclear-generating capacity of 99,500 MWe discharges more than 2,500 MTHM in spent fuel per year, hence more than 25,000 MTHM per decade and more than 75,000 MTHM in the nominal 30-year operating lifetimes of this generation of reactors. A campaign to process 50 tons of WPu at a MOX-fuel loading of 5 percent WPu in heavy metal would produce 1,000 MTHM in spent fuel.

41  

See, for example: Smith 1978, Fischer et al. 1987, and Lahs 1987.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

from its potential to achieve criticality, as discussed in general terms above in “Relevant Characteristics of Plutonium."

The most widely used process for plutonium oxide production entails dissolution of the plutonium metal in acid to form plutonium nitrate, followed by steps to precipitate plutonium dioxide from the solution. The main nonroutine ES&H hazards associated with these operations—that is, events of low frequency that, by virtue of their severe consequences, one strives to make as improbable as possible—are mainly criticality accidents and fires. The plutonium nitrate solution is particularly problematic in terms of criticality, because it is easy for a liquid to flow accidentally into a more critical geometry and because the reactivity of any given plutonium concentration in the solution is increased by the presence of water, which serves as a moderator. Criticality accidents in plutonium processing pose dangers almost exclusively to workers, since the energy releases involved are not large enough to breach the buildings in which the operations are taking place. In any case, these accidents can be avoided by scrupulous adherence to appropriate procedures and proper design of the containers and transfer systems for plutonium solutions.

Fires, by contrast, have the potential to mobilize plutonium in ways that can lead to public as well as worker exposures, and fires are probably harder to prevent altogether than are criticality accidents. Nonetheless, proper design and operation of the facilities, combined with adequate onsite fire-fighting capabilities, surely can hold the occurrence of fires—and the mobilization of plutonium from those that do occur—to very low levels. This is a matter that certainly will require the most careful attention from the designers, operators, and regulators of any plutonium disposition option.

The main routine hazards of the plutonium-processing operations in oxide production are to workers: the mobilization of respirable plutonium aerosols and exposure to gamma irradiation from the Am-241 contained in the plutonium. While carelessness can certainly lead to excessive doses from these sources, available technologies and processes (involving the use of glove-boxes, shielded processing cells, and the like) appear to be adequate—if conscientiously applied—to keep the doses well within existing regulatory guidelines (see, e.g., OTA 1993). Of course, the regulatory authorities with oversight of the WPu disposition campaign will have the task of ensuring that the needed conscientiousness materializes, and that it does not weaken over time under pressures to cut costs and meet schedules.

There is no obvious reason that the ES&H hazards from plutonium metal-to-oxide conversion in connection with the spent fuel and vitrification disposition options should be any larger than similar plutonium-handling hazards encountered in the course of nuclear-weapon production—indeed, with the improved technology and regulatory oversight that can now be expected in comparison to the practices that prevailed when most weapons production was taking place, the hazards should be smaller. Nonetheless, it may well be that these

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

exposures will prove to be the hazard of WPu disposition most difficult to make small compared to analogous hazards in the other military and civilian nuclear-energy operations that will be occurring in the future—that is, these exposures may be the most problematic ones in relation to our third criterion—so they should (and we assume will) receive particular scrutiny.

Further Plutonium Processing for Spent Fuel Options

In the case of spent fuel options for disposition of WPu, processing steps beyond metal-to-oxide conversion would include: (1) mixing the plutonium in the appropriate proportions with the other constituents of the fuel (e.g., uranium dioxide in the case of MOX fuels, alloying metals for metallic fuels); (2) the actual production of the basic fuel entities (pellets, particles, etc.); and (3) the fabrication of the completed fuel elements. In all of these steps, the fact that plutonium and not just uranium is being used is decisive in governing the ES&H hazards, both because of the much higher radiological toxicity of plutonium compared to uranium and because the plutonium poses greater criticality problems.

The resulting need for special equipment and precautions in the fabrication of MOX fuel, as compared to what is involved in uranium fuel fabrication, is obvious, and this accounts for much of the difference in cost between MOX fuel fabrication and LEU fuel fabrication, as summarized above in "Weapons Plutonium Versus Uranium as Power Reactor Fuel." At the same time, the hazards at the fuel fabrication stage should be smaller in several respects than those encountered in plutonium metal-to-oxide conversion: once in oxide form, the plutonium is not flammable like the metal nor as prone to criticality accident as a plutonium nitrate solution; and, once diluted with uranium oxide, its radiologic and criticality hazards are further reduced.42 If, as we have concluded above, the ES&H hazards of plutonium metal-to-oxide conversion can be expected to satisfy the ES&H criteria set forth in this report, this should also be true of the additional plutonium-processing steps in MOX fuel fabrication.

Further Plutonium Processing for the Vitrification Option

In the case of the vitrification approach to plutonium disposition, the only plutonium-processing step beyond oxide production would consist of incorporation of the oxide into the mix of borosilicate glass and fission products either in the melter, or before the glass frit and fission products are introduced into the melter. In the subsequent pouring, cooling, and handling of the glass logs,

42  

We note also that the radiologic toxicity of MOX made with WPu will be considerably less than that of the MOX now being made with recycled power-reactor plutonium in France and Belgium.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

the presence of 1-2 percent by weight plutonium in the glass would add little to the ES&H risks of these operations, which would be dominated by the fission products in the high-level wastes (HLW) constituting about 20 percent of the mixture. (The dominance of the fission-product hazards in these steps arises from the combination of their larger quantities plus the presence among them of intense gamma emitters, above all Cs-137.) Also, plutonium criticality in intact glass logs will not be a problem, because of the combination of low plutonium concentration and the substantial content of neutron-absorbing boron in the glass. Possible criticality problems in the melter are discussed below, in this subsection, and criticality hazards after thousands of years in a geologic repository, when the boron may have leached away, are discussed below in "Radioactive Waste Issues."

The hazards presented by the presence of, and operations with, plutonium in the vitrification plant prior to mixing with the melt would be mainly those of plutonium-oxide production, as discussed above. In the vitrification case, however, the context to which these plutonium hazards would be added would be a riskier one than in the spent fuel case—a vitrification plant full of HLW as opposed to a fuel fabrication plant containing only LEU—so the relative change in ES&H risks caused by the addition of the plutonium would be smaller.43

An ES&H risk unique to the vitrification case, however, is posed by the possibility of criticality accidents in the melter. These might occur if a substantial part of the added plutonium dioxide somehow coalesced in one part of the melt rather than being well mixed throughout it. A number of approaches to minimizing this risk can be envisioned, such as by using melters small enough that, at the highest plutonium loading contemplated, they would never contain as much as one critical mass of plutonium; it would also be necessary to ensure against malfunctions in the plutonium feed system that might add a higher proportion of plutonium to the melt than is intended. The obvious alternative approach of avoiding criticality by means of mechanical stirring to ensure adequate mixing is not permitted under current U.S. safety regulations.

Determining how best to ensure that melter criticality problems do not contribute significantly to the ES&H hazards of plutonium disposition by the vitrification route is a technical issue that needs to be resolved before this disposition option is embraced. We do not think it will be so difficult to resolve, however, as to pose a significant obstacle to proceeding with this option if that

43  

For example, a major fire in a vitrification plant being used for plutonium disposition might mobilize a similar quantity of plutonium to that mobilized by a major fire in a MOX fuel fabrication plant, but in the former case the extra hazard posed by the plutonium probably would be smaller than in the latter case because the fire in the vitrification plant would also mobilize significant quantities of fission products. Similarly, the risk of worker radiation exposure from accidental criticality in the course of plutonium-oxide production would be added, in a vitrification plant, to an already significant risk of worker radiation exposure from accidents in the handling of fission products, to which there would be no counterpart in the MOX fuel fabrication plant.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

seems otherwise desirable. As for the other ES&H impacts of the addition of plutonium to the vitrification operation (aside from those of metal-to-oxide conversion, which is common to vitrification and most spent fuel options and was discussed separately above), these effects seem unlikely, for the reasons given earlier, to represent large additions to the ES&H impacts that will be encountered in any case in the use of vitrification to stabilize military HLW.

Elimination Options

When and if it is desired to destroy, by fission, a larger fraction of the WPu than is consumed in the once-through reactor operations associated with the spent fuel option, there will have to be added to the plutonium-processing operations described above the additional operations associated with reprocessing spent fuel. (As noted earlier, fissioning more than about 80 percent of a given initial quantity of plutonium can only be achieved by repeated recycle of the plutonium through reactors, and this in turn requires reprocessing.)

Nuclear fuel reprocessing, if carried out using technologies of the type that have been developed on a commercial scale up until now, 44 would entail routine exposures of workers and the public to radiation that, while presumably within regulatory standards, would not necessarily be as small—per unit of electricity generated-as the analogous exposures from other operations in the commercial nuclear fuel cycle (USNRC 1976, APS 1978, NAS 1979). The plutonium that emerges from these reprocessing operations, moreover, will be more similar to reactor-grade plutonium than to weapons-grade plutonium in isotopic composition, hence will amplify the hazards ascribed above to MOX fuel fabrication with the weapons-grade material. On the other hand, plutonium recycle would reduce the amount of uranium mining and milling required to generate a given quantity of electricity, and hence would reduce the ES&H impacts of those operations. (Whether reprocessing and plutonium recycle would reduce waste management burdens is unclear; see "Radioactive Waste Issues" below.)

Even if reprocessing and recycle of plutonium turn out, on balance, to increase the total worker and public exposures per unit of nuclear electricity generation, that is not to say that the additional exposures from reprocessing and from fuel fabrication with recycled plutonium would exceed regulatory standards (the plants will have to be designed to avoid that) or that the public will deem the exposures unacceptable (it is not easy to judge, in advance, what exposures the public will deem acceptable in exchange for what benefits); it is only to say that their being potentially non-negligible compared to the exposures

44  

These are based on the PUREX process, which entails mechanical chopping up of the fuel elements (after a cooling-off period of months to years following discharge from the reactor), followed by dissolution in acid and a series of chemical operations to separate the fission products from the actinides and, within the actinides, the plutonium from the rest (see, e.g., APS 1978).

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

from other nuclear-energy operations means that careful attention should and will be paid to them and the means for their minimization. This, in turn, could be a source of delay before such activities are licensed in the United States—they already are licensed in some other countries—as well as a source of additional costs.

As noted in Chapter 4, some of the advanced-reactor options for plutonium destruction would employ improved reprocessing technologies that are less developed than the PUREX approach now in use where reprocessing is practiced. Among the possibilities are the pyrometallurgical processing techniques envisioned for use with the integral fast reactor (IFR) and the online molten-salt reprocessing technology that might be developed for an accelerator-based conversion (ABC) option employing a molten-salt blanket. These approaches appear to offer some advantages in the security realm compared to conventional reprocessing technology. Perhaps they will offer ES&H advantages, as well, but this will be difficult to confirm until these approaches have been further developed and tested.45

It is difficult, in any case, to compare the ES&H hazards of plutonium reprocessing and recycling for WPu disposition with the ES&H hazards that would be associated with civilian and military nuclear-energy activities in the absence of WPu disposition, because of the way in which disposition of WPu and the management of civilian plutonium are linked. If plutonium reprocessing and recycle were practiced only for excess WPu, then even if the ES&H hazards of reprocessing and recycle were a significant addition to other nuclear fuel-cycle risks on a per-unit-energy basis, the small quantity of WPu compared to the scale of civilian nuclear-energy generation would ensure that the plutonium's contribution to total exposures would be modest. Similarly, if society were reprocessing and recycling civilian plutonium on a large scale anyway, then the addition of the WPu would not produce a significant increment to the ES&H hazards of those civilian operations.

But neither of these scenarios seems very relevant: as discussed elsewhere in this report and in the report of the full committee (NAS 1994), there is little security benefit in actually fissioning a large fraction of the atoms of excess WPu (as opposed to merely embedding the WPu in forms that meet the spent fuel standard) if the much larger inventory of civilian plutonium is not fissioned; on the other hand, there is currently no economic incentive in the United States to reprocess and recycle civilian plutonium for energy generation, and there are reasons of proliferation policy not to do so. There will probably not be a cost incentive until uranium is several times more costly than it is today. Reprocessing and recycling plutonium, then, is unlikely to be undertaken in the United

45  

The forthcoming report of the National Research Council's Panel on Separations Technology and Transmutation Systems (STATS) treats the possibilities in considerably more detail than is possible here (National Research Council forthcoming).

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

States in the next few decades unless it is for purposes of making the plutonium less accessible for weapons use than it would be in spent fuel, and unless it is done for civilian as well as for WPu. That this might entail ES&H impacts that are not insignificant compared to those of operating civilian nuclear-energy generation without reprocessing and recycle is a potential liability which, like the additional costs and time requirements of this approach, weighs against the security gain from actually destroying most of the plutonium.

Reactor Safety Issues

The potential influences on safety of the use, in LWRs, of MOX fuel containing reactor plutonium were extensively studied in the United States in the 1970s, when large-scale use of this technology was being contemplated for commercial electric-power production (USNRC 1976). These influences have also been studied in Europe (where considerable operating experience with one-third MOX cores in LWRs has been accumulated), in Japan, and in Russia;46 and safety issues for liquid-metal reactors (LMRs) fueled with reactor plutonium have likewise been investigated quite extensively in the United States, Europe, Japan, and Russia, not only in theoretical studies but also in large-scale experiments and in prototype-reactor operations (see, e.g., Planchon et al. 1987, D. Lucoff 1989, Atomic Energy Society of Japan 1992).

In connection with recent interest in the use of reactors for disposition of WPu, additional studies of the associated reactor safety issues for reactors and fuels of various types have been undertaken by reactor manufacturers (e.g., in the vendor reports for the Plutonium Disposition Study of the U.S. Department of Energy, summarized in USDOE 1992), in part for the purpose of clarifying ways in which the different isotopic composition of weapons-grade as opposed to reactor-grade plutonium could affect safety characteristics. Before reactors are licensed to operate with weapons-grade plutonium in any country, moreover, it can be assumed that there will be a further reexamination of the safety implications by the relevant national reactor safety authorities.

We cannot comprehensively summarize, in the time and space available here, the findings of the studies that have been conducted up until now of the safety of the use of plutonium fuels, much less anticipate in detail what may be found in the further layers of studies that will be done before reactors are actually used for WPu disposition. We merely sketch out, in what follows, enough of an overview of the issue to support our belief that it is likely, based on current information, that power reactors of a variety of designs will prove to be operable with WPu fuels without adding significantly to the safety risks that would be associated with operating the same reactors with uranium fuels. (Our primary focus is on LWRs, however, about which the most is known, and which are the

46  

See, e.g., Kudriavtsev (1993), Schlosser et al. (1993), Shiratori et al. (1993), Levina et al. (1994), Mikhailov et al. (1994), Murogov et al. (1994), and Novikov et al. (1994).

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

most likely to see use for disposition of WPu in the relatively near future.) We note that the increase in safety risks from operating with WPu fuel would have to be very substantial in order for such operations, in the relatively few reactors that would be needed for the disposition of 50 or 100 tons of surplus WPu, to represent a significant increase in the total risk from civilian nuclear-energy generation—that is, in order to violate our third criterion for the ES&H characteristics of WPu disposition.

The risk from reactor accidents is a matter of both probabilities and consequences—the probabilities of various combinations of mechanical and human failures and the consequences to be expected when and if these failure modes actually occur. The influences on reactor safety from the substitution of WPu for U-235 in fresh fuel—whether as MOX in the case of LWRs, CANDUs, and some LMRs, or as metal in other LMRs, or as pure plutonium oxide in the MHTGR—could take the form of changes in probabilities of particular accident sequences or changes in the consequences of accidents of particular types. Most studies of the reactor safety implications of the use of plutonium fuels have focused predominantly on the potential effects on accident probability, on the supposition that the effects on accident consequences are modest at most. In what follows, we consider probabilities first and then turn briefly to consequences.

Effects on Accident Probabilities

The two main classes of nuclear-reactor accidents with the potential to release significant quantities of radioactivity to the environment are reactivity excursions and loss-of-coolant/loss-of-coolant-flow accidents. The Chernobyl accident in 1986 was of the first type; accidents in this category arise when the nuclear chain reaction in part or all of the reactor core accelerates out of control to reach rates of energy release exceeding the core's capacity to absorb and remove heat. The Three Mile Island accident in 1979 (which, unlike Chernobyl, did not produce a large release of radioactivity to the environment, but which did destroy the reactor core) was of the second type; accidents of this type arise not from an unexpectedly high rate of energy release in the core but from an unexpected reduction in the capacity of cooling systems to remove the decay heat generated in the core (as, for example, when a faulty valve or a break in a coolant pipe allows a substantial fraction of the primary heat-transfer medium to escape and the emergency coolant injection systems fail to replenish the escaping coolant).47

47  

In the most probable loss-of-cooling-accident scenarios, in fact, the main source of energy leading to overheating of the core after cooling fails is not the energy from the nuclear chain reaction, which usually would be promptly quenched. The problematic energy comes rather from "afterheat"—the result of the radioactive decay of fission products, which diminishes over time according to their half-lives but cannot be shut off more rapidly.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

As noted in Chapter 2, the characteristics of plutonium in a chain reaction are subtly different from those of uranium in ways that influence the task of protecting against accidents of the reactivity-excursion type: plutonium has a smaller delayed-neutron fraction and a higher thermal-neutron absorption fraction than uranium, both of which factors increase the need for control-absorbers in a plutonium-fueled reactor core compared to what is needed when uranium-based fuel is used; and Pu-239 has a resonance in its fission cross-section that produces a tendency toward a positive temperature coefficient of reactivity that is not present in uranium fuel.48 It was also noted in Chapter 2 (see also Wiese 1993) that MOX fuel exhibits levels of radioactive afterheat modestly higher than those exhibited in LEU fuel irradiated to the same burnup, which affects the task of protecting against core damage in the event of loss-of-coolant accidents. Quantitative analysis of these phenomena, however, combined with operating experience using MOX fuels in a number of countries, has shown that most LWR designs can accommodate MOX fuel made from reactor plutonium in at least one-third of their cores, without modification to the reactor, while remaining well within the capabilities of their control systems to safely limit reactivity excursions and the capabilities of their cooling systems to keep the fuel within safe thermal limits.

As discussed in Chapter 4, the prospect of the availability of significant quantities of WPu for use as nuclear fuel following the end of the Cold War led, at the beginning of the 1990s, to a number of studies of the adaptability of LWRs to the use of MOX fuel made from this WPu rather than from the reactor-grade plutonium (RPu) that therefore had been studied and, to a modest extent, utilized. These initial studies (see, e.g., Omberg and Walter 1993, USDOE 1993a) assumed that most LWR types already in operation would be able to use safely about the same fraction of WPu-MOX fuel in their cores as RPu-MOX (that is, about one third), possibly with some addition of burnable neutron absorbers to compensate for the increased proportion of Pu-239 (with its 0.3 electron volt resonance in fission cross-section) in WPu. This addition could be accomplished in the manufacture of the fuel and would not necessitate modifications to the reactor itself.

48  

A resonance is a sharp peak in the value of a cross-section—that is, in the probability of a particular reaction—in a narrow range of values of the relative velocity between the target nucleus and an incident neutron. The key resonance in the fission cross-section of Pu-239 occurs at a neutron energy of about 0.3 electron volts, which is well above the average energy of the neutrons in the core of a normally operating LWR reactor. Heating the core results in an increase in the number of neutrons energetic enough to experience the resonance, and it also widens that energy range through the phenomenon called Doppler broadening. Both effects tend to increase the reaction rate as the temperature of the core increases, which if not compensated by other effects acting in the opposite direction would produce what is called a positive temperature coefficient of reactivity (see also Chapter 2).

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

It was further assumed that most existing LWRs, if they were able to utilize MOX fuel in 100 percent of their cores safely, would need modifications to increase the capabilities of their control systems. (As noted earlier, one existing U.S. LWR type—the ABB-Combustion Engineering System-80, of which three are in operation in Arizona and another has been mothballed in a partial state of completion in Washington state—was designed from the outset to be able to use a 100-percent MOX core. The vendor has indicated that this capability would apply to WPu as well as to RPu; see ABB-CE 1993.) Any new LWR constructed for the purpose of WPu disposition could be designed to take a 100-percent MOX core, as were the "evolutionary" and "advanced" LWR designs presented by vendors in the first phase of the U.S. Department of Energy's Plutonium Disposition Study (USDOE 1993a).

Analyses performed by vendors in the second phase of that study, which became available late in the panel's deliberations, suggest—contrary to previous assumptions—that several existing LWR types besides the System-80 could in fact use 100-percent WPu-MOX cores without undergoing significant modifications and without compromising safety. Further analysis and review will be required before this conclusion can be considered firm. In any case, the still unfolding understanding of the circumstances under which existing LWRs would be able to use 100-percent WPu-MOX fuel only underlines the key point that any approach to the use of MOX fuel in U.S. power reactors must and will receive a thorough, formal safety review before it is licensed.

While we are not in a position to predict what if any modifications to existing reactor types—or to the designs of newer types that have not yet operated—will be required as a result of such licensing reviews, we expect that the final outcome will be certification that whatever LWR type is chosen will be able, with modifications if appropriate, to operate within prevailing reactivity and thermal margins using sufficient plutonium loadings to accomplish the disposition mission in a small number of reactors. We believe, further, that under these circumstances no important overall adverse impact of MOX use on the accident probabilities of the LWRs involved will occur; if there are adequate reactivity and thermal margins in the fuel, as licensing review should ensure, the main remaining determinants of accident probabilities will involve factors not related to fuel composition and hence unaffected by the use of MOX rather than LEU fuel.

Addressing the implications for accident probabilities of using evolutionary or advanced reactor types other than LWRs for WPu disposition is more difficult, because, necessarily, it is difficult to compare the safety of reactors that have not yet been built with the safety of those that have. (It is usually easy enough to identify some features of alternative reactor types that complicate the task of ensuring safety, compared to the situation with LWRs, while identifying other features that ease the task of ensuring safety; what is hard is predicting what the net difference in safety will be, all things considered.) There is, none-

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

theless, reason to think that advanced reactors for energy supply will not be deployed unless society concludes—through analysis, testing, debate, and the licensing process—that they are at least as safe as current reactors. And since, as we have argued elsewhere in this report, there is no compelling reason to deploy advanced reactors solely for the purpose of WPu disposition, there are only likely to be available for that purpose if they have also passed the safety "test" for use in large-scale electricity generation. If that is so, their use for WPu disposition seems unlikely to so alter their safety characteristics as to perturb significantly the safety of the whole nuclear-energy enterprise, all the more so because advanced-reactor types henceforth will probably all be designed from the outset to minimize any safety problems of plutonium use.

Effects on Accident Consequences

In the event of a large release of radioactivity as the result of a severe nuclear-reactor accident, the consequences generally will include relatively large doses of radiation (say, above 0.1 Sv or 10 rem) delivered in the hours and days immediately after the accident to relatively small numbers of people who are in the path of the radioactive plume near the reactor, smaller doses (0.01-0.1 Sv or 1-10 rem) delivered over a longer period of time to larger numbers of people in the plume path out to distances of some hundreds of kilometers from the reactor, and very small doses (less than 0.01 Sv or 1 rem) delivered to much larger numbers of people at even greater distances and lower dose rates (see, e.g., APS 1975, USNRC 1975, 1987; Hohenemser 1988.) The dominant mechanism for the close-in exposures is inhalation of the mixture of suspended radionuclides in the plume, with a substantial additional contribution by external irradiation from radionuclides deposited on the ground. With increasing distance from the reactor, inhalation from the plume becomes relatively less important and external irradiation from ground-deposited material—along with inhalation of resuspended material and ingestion of radionuclides in food and water—become more important.

It is to be expected that about 95 percent of the total population exposure in the 50 years following a severe accident (measured in person-sieverts or person-rem, the product of the total number of persons exposed times the average exposure received) would result from one or another form of "ground dose" external irradiation and inhalation of resuspended material—and from ingestion with food and water. The doses from inhalation of radionuclides from the passing plume and from external irradiation as the plume passed would contribute less than 5 percent to the total population dose. Most of the population dose would be distributed over a large number of people, at substantial distances from the reactor, each experiencing a low individual dose. The details of this pattern of doses would depend on the quantities of various radionuclides released by the accident, the temperature of the plume in which they were re-

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

Table 6-19 Contributions to Doses from Severe Reactor Accidents

 

Calculated LWR Release Fractions

Estimated Chernobyl Release Fractionsc

Nominal Severe-Accidentd 50-year Population Dose (104 pers-Sv) Delivered Via:

Elements

WASH 1,400a

NUREG 1,150b

 

Plume Inhalation

Ground Dose

Ingestion

Kr,Xe

0.8-1.0

0.9-1.0

1.0

negl

0

0

I

0.4-0.9

0.2-0.9

0.2-0.6

0.5

1

1-5

Cs

0.4-0.5

0.05-0.8

0.1-0.6

0.7

50

20-50

Te

0.3-0.7

0.01-0.7

0.1-0.6

0.3

small

1-2

Sr

0.05-0.1

0.005-0.3

0.04-0.06

0.6

small

2-20

Ba

0.05-0.1

0.005-0.3

0.04-0.06

0.3

1

1-4

Ru,Mo

0.02-0.5

0.001-0.1

0.02-0.03

small

1

small

Pu,Am,Cm

3×10-3/5×10-3

1×10-5/0.05

0.03

0.3

1

1-5

ELEMENTS:

Kr, Xe: krypton, xenon.

I: iodine.

Cs: cesium.

Te: tellurium.

Sr: strontium.

Ba: barium.

Ru, Mo: ruthenium, molybdenum.

Pu, Am, Cm: plutonium, americium, curium.

a USNRC (1975), with ranges covering release categories PWR 1-2, BWR 1-2.

b USNRC (1987), with ranges for most severe releases considered.

c See Hohenemser (1988). The Chernobyl reactor was a graphite-moderated, light-watercooled reactor, not an LWR, and lacked a real containment building. Releases of the indicated magnitude are considered much less probable for LWRs.

d Assumed release fractions are Kr,Xe = 0.9, I = 0.7, Cs = 0.5, Te = 0.3, Ba,Sr = 0.06, Ru,Mo = 0.02, Pu,Am,Cm = 0.004. Total population exposure is about 100 × 104 pers-Sv or 100 × 106 pers-rem, divided about equally between ground dose and ingestion dose. Dose contributions estimated based on APS (1975), USNRC (1987), and Hohenemser (1988).

leased, the weather conditions at the time of the accident and afterwards, the distribution of population in relation to the location of the reactor and the wind direction following the accident, the habits of the population (fractions of time spent indoors and outdoors, diet), and so on.

We assume, consistent with the discussion in the preceding section, that once an accident of a given type occurs the conditions governing what proportions of the radionuclides in the reactor will be released will not depend on the percentage or type of plutonium in its fuel. (That is to say, for example, that the modest differences in afterheat among fuels with different quantities and isotopic qualities of plutonium are not big enough, in relation to other factors affecting the post-accident condition of the fuel, to affect significantly the release fractions—i.e., the percentages of the various classes of elements present that

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

actually escape the reactor into the environment.) In that case, the only factor governing accident consequences that does depend on how much and what kind of plutonium was used in the reactor fuel is the inventory of radionuclides in the reactor at the time of the accident.

We are interested, particularly, in the quantities of those radionuclides that would tend to make important contributions to the total population exposure resulting from severe accidents. Previous studies have estimated release fractions, for different classes of elements in the fuel, that might be possible in severe accidents at LWRs (see, e.g., USNRC 1975, 1987).49 Combining such release-fraction estimates with population-dose models indicates that the largest contribution to the population dose comes from cesium isotopes, followed in importance by strontium, iodine, plutonium and isotopes derived from it, barium, ruthenium, and tellurium, as shown in Table 6-19. Fission of plutonium produces about the same quantity of cesium as does fission of uranium, so the use of plutonium-based fuel will not change very much the releasable inventories of this dominant contributor to accident consequences. Plutonium fission produces somewhat less strontium than does uranium fission, which effect would contribute to a modest reduction in the consequences of a release (for fixed release fractions) from plutonium-based fuel, but the increased quantities of plutonium itself in such fuel—typically 1.5-5 times more than in uranium-based LWR fuel just before discharge—could offset this advantage partly or entirely.

Based on the foregoing considerations and the figures in Table 6-19, it seems unlikely that the switch from uranium-based to plutonium-based fuel could worsen the consequences of a postulated (and very improbable) severe accident in a LWR by more than 10 or 20 percent.50 The influence on the consequences of less severe accidents, which probably dominate the expectation value of population exposure per reactor-year of operation (USNRC 1975, 1987), would be even smaller, because less severe accidents are unlikely to mobilize any significant quantity of plutonium at all.

49  

Notwithstanding significant differences in reactor characteristics, these calculated severe-accident release fractions for LWRs are in rather good agreement with estimates of what was actually released in the April 1986 accident at the graphite-moderated, water-cooled Chernobyl reactor near Kiev (see Table 6-19).

50  

Using the geometric means of the ranges of values in Table 6-19 for the contributions of different elements to the 50-year population dose from a hypothetical severe accident at an LEU-fueled LWR, one concludes that strontium contributes about 6 × 104 person-sieverts and plutonium and the isotopes derived from it contribute about 4 × 104 person-sieverts to a total, for the accident, of about 105 × 104 person-sievert. Based only on changes in the fuel composition, with release fractions for each element held constant, the substitution of MOX for LEU fuel would shrink the strontium dose about threefold while increasing that from plutonium, in the highest plutonium case, about fivefold. This would produce a net 10-percent increase in the total population dose. Saying. as we do, that an increase of “more than 10 or 20 percent" is "unlikely" allows some leeway for the considerable uncertainty in the severe-accident release fractions for both plutonium and strontium.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

We expect that preliminary conclusions of this sort would be revisited in the course of more detailed safety calculations performed for licensing purposes, with up-to-date calculations of radionuclide inventories for different fuel compositions and irradiation histories, and reexamination of population-exposure models for sensitivity to changes in those inventories resulting from the use of WPu in the fuel. We endorse such further investigation, but we would consider it very surprising if the results altered our conclusion that use of plutonium fuels will not change by very much the consequences of such LWR accidents as may occur. Although accident modes and probabilities may differ among reactor types, moreover, it is hard to see how use of WPu in the fuel of, for example, liquid-metal reactors or high-temperature gas reactors could have a large influence on the consequences of accidents with these reactors. And unless there is a very large such influence, the use of WPu in a modest fraction of the world's operating power reactors, of whatever types, could not affect significantly the accident risks of the nuclear-energy system as a whole.

Radioactive Waste Issues

All options for the disposition of plutonium from surplus nuclear weapons will generate radioactive wastes of one or more types, and some options will affect the quantities or other characteristics of radioactive wastes that have been or will be produced by other military and civilian nuclear-energy activities. As a basis for surveying the radioactive waste implications of alternative approaches to plutonium disposition, the categories according to which radioactive wastes are described and regulated in the United States—and the approaches for managing the wastes in these categories—can be summarized as follows (APS 1978; OTA 1985, 1989; Holdren 1992; OFR 1992, Pts. 61.2 and 72.3):

  • Spent fuel consists of fuel rods that have been removed, after nuclear-energy generation, from commercial, defense, or research reactors. It is initially highly radioactive, generates considerable heat, and requires heavy shielding. The volume of spent fuel is 20-30 m3 per year from a large (1,200-MWe) LWR. Current practice is interim storage of the spent fuel in suitably designed pools of water (providing cooling and shielding) at reactor sites, followed in some cases by a second interim storage phase in dry casks. In "once-through" fuel cycles, wherein reprocessing of spent fuel is foregone, the expectation is that such interim storage of the spent fuel would ultimately be followed by emplacement in a geologic repository.

  • High-level wastes (HLW) are defined, in current U.S. regulatory practice, as the concentrated radioactive waste forms generated when spent

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

fuel is reprocessed to recover uranium and plutonium.51 HLW are initially highly radioactive, generate considerable heat, and require heavy shielding. A distinction is made between "defense" and "commercial" HLW, according to whether the material originates in military activities (weapons production and naval reactor operations) or commercial electricity-production activities. The physical form of such wastes depends on the reprocessing technology used and on the choices made about how to prepare the waste for ultimate disposition. Most U.S. defense HLW were initially in the form of slurries, as were the much smaller quantities of commercial HLW produced in this country before commercial reprocessing was suspended. The slurries have since separated into solid and liquid phases. The HLW are expected to be vitrified into highly leach-resistant glass logs—as is already being done in a few other countries—before emplacement into a geologic repository. The volume of HLW resulting from reprocessing the fuel from a 1,200-MWe LWR and vitrifying it would be 5-8 m3 per year, including the glass matrix and metal canisters.

  • Uranium mill tailings are the residues from the process by which uranium is extracted from mined uranium ore. They contain very low concentrations of radioactive material, but the half-life of the nuclide that governs the longevity of their hazard (thorium-230, parent of radon-226) is about 80,000 years and the volume of these wastes is very large—100,000-200,000 m3 per 1,200-MWe LWR per year for uranium ore of typical characteristics, assuming a once-through fuel cycle, and about 30 percent less for a reprocessing and recycle fuel cycle. (This radioactivity was in the ground all along, but was brought to the surface by mining and rendered more mobilizable by milling.) Current U.S. regulations require covering of newly produced mill tailings with soil and rock to reduce radon release, control erosion, and minimize water infiltration.

  • Low-level wastes (LLW) are defined in the United States to mean all radioactive waste that is not spent fuel, HLW, or uranium mill tailings. LLW originate in nuclear-weapons-related activities, in commercial nuclear fuel-cycle operations (e.g., uranium enrichment, nuclear fuel fabrication, reactor operation and decommissioning, spent fuel handling, and fuel reprocessing), and in medical, industrial, and research applications of radionuclides. The quantities of LLW being generated in the United States in the late 1980s from the operation of LWRs and

51  

The fission products that dominate the radioactivity of HLW are the same ones that were in the spent fuel; only the physical and chemical matrix in which they are embedded has changed. Some authorities define HLW more broadly, as an overarching category that includes both spent fuel and the concentrated waste forms arising from reprocessing.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

associated fuel-cycle activities amounted to about 250 m3 per 1,000-MWe PWR per year and 750 m3 per 1,000-MWe BWR per year, not including decommissioning wastes (which would add about 500 m3 of LLW per year, prorated over a nominal 30-year reactor lifetime, for either reactor type) and not including LLW from reprocessing (which, if carried out, would add another 150-300 m3 of LLW per reactor-year).52 Most types of LLW in the United States have been disposed of by shallow burial at a modest number of sites licensed for this purpose. LLW containing concentrations of transuranic and certain other long-lived radionuclides above specified limits are not acceptable for shallow burial under current U.S. regulations, and such wastes are being held in interim storage pending the availability of a geologic repository for them.

  • Transuranic (TRU) wastes are wastes, other than spent fuel or HLW, that contain more than 100 nCi/g of radionuclides with atomic number greater than 92. TRU wastes may be generated in the production and handling of plutonium for nuclear weapons, in the manufacture of sealed radioactive sources, and in the refurbishing or decommissioning of nuclear power plants. TRU wastes are considered a subcategory of LLW. They amount to 100-150 m3/yr for a 1,200-MWe LWR with reprocessing and recycle (essentially entirely from the reprocessing and MOX fuel fabrication operations), none from an LWR using LEU fuel once-through. Although some TRU wastes were formerly disposed of in the United States by shallow burial, this practice is no longer permitted; these wastes are now being held for eventual emplacement in deep geologic storage.

A site at Yucca Mountain, Nevada, is under investigation by the U.S. Department of Energy for use as a mined geologic repository for civilian spent fuel, vitrified defense HLW, and probably civilian TRU waste; actual emplacement of waste at this site, which is in a region of unsaturated volcanic tuff, is not likely to begin before 2015, if then. A mined repository in salt beds near Carlsbad, New Mexico—the Waste Isolation Pilot Plant (WIPP)—probably will be in operation considerably sooner and is to be used mainly for disposal of defense TRU wastes. A few shallow burial sites are in operation at locations around the United States for those parts of commercial and defense LLW that qualify for such treatment; the capacities of the existing sites and the locations of new ones are under continuing and often contentious discussion.

In the subsections that follow, we summarize the ways in which the two primary classes of options discussed in this chapter—(l) once-through irradiation in MOX fuel in reactors of current commercial types and (2) vitrification

52  

See USDOE (1988a). Current LLW volumes per reactor-year may be smaller because of the incentives for compaction produced by a shortage of licensed LLW disposal sites.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-20 Effects of Plutonium Disposition on Radioactive Wastes

 

Disposition Option

Waste Type

MOX to Spent Fuela

Vitrification with HLWb

Spent Fuel

Quantity unchanged, plutonium content increases significantly, fission-product composition changes slightly

No spent fuel involved

HLW

No reprocessed HLW involved

Number and size of HLW-bearing glass logs unchanged, plutonium content increases greatly

TRU

Additional TRU produced in conversion of plutonium metal to oxide and fabrication of MOX fuel

Additional TRU produced in conversion of plutonium metal to oxide and in vitrification process

non-TRU LLW

Some additional plutonium-bearing LLW below 100 nCi/g of TRU isotopes may be produced in plutonium conversion and MOX fabrication

Some additional plutonium-bearing LLW below 100 nCi/g of TRU isotopes may be produced in plutonium conversion

a We assume the MOX is used in reactors that would be operating in any case, so total amount of nuclear electricity generation is unchanged by this choice of disposition option.

b We assume that a sufficient plutonium loading is possible from the criticality standpoint to permit accommodating all of the surplus WPu without increasing the number of glass logs already scheduled to be produced for the stabilization of existing defense HLW.

together with military HLW—would affect the quantities and other relevant characteristics of the radioactive wastes in the foregoing categories that would be associated with civilian and military nuclear-energy operations in this country in the absence of WPu disposition. (The main effects are summarized in general terms in Table 6-20.) Implications for radioactive wastes of alternative options are considered more briefly.

Spent Fuel and High-Level Wastes: Dose Potentials

Disposition of WPu by fabricating it into MOX fuel and irradiating that fuel once-through in power reactors of existing commercial types (e.g., LWRs or CANDUs) would produce spent fuel generally similar to the spent fuel that these same reactors would be producing if operated with their usual uranium-based fuels. The characteristics of spent fuel that govern the magnitude of the waste management task and the risks associated with it are mainly its volume, its fission-product content, its content of actinides (including especially those that are fissile), and the chemical and structural characteristics that influence the rate at which it would release these radioactive inventories under conditions encountered in transport and storage.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

The volume of spent fuel associated with a given amount of nuclear-energy generation depends only on the burnup, which, in LWRs, is not expected to change as a result of the incorporation of WPu in the fuel. In CANDUs, which ordinarily use unenriched uranium fuel, the burnup would probably be increased—and the spent fuel volume for a given amount of energy generation correspondingly decreased—if WPu were used in the fuel, since exploiting the higher burnups made possible by the addition of the plutonium would partly offset the higher fuel fabrication costs. Thus, disposition of WPu by means of the MOX option in reactors of existing commercial types would not increase spent fuel volume if LWRs were used, and would decrease this volume if CANDUs were used, assuming only that the disposition occurs in reactors that would have operated in any case. If new reactors were commissioned for the purpose of WPu disposition, of course, there would be a net increase in spent fuel volume; but this increase would be just proportional to the fraction of total nuclear electricity generation accomplished by the new reactors, and, as we have seen, the extra electricity generation needed to process 50 or 100 tons of WPu would be only a modest addition to the nuclear energy generation expected to occur in any case over the same period.

The quantity of fission products in spent fuel depends mainly on the total amount of nuclear energy generated from that fuel—that is, on the burnup— and, secondarily, on the neutron-energy spectrum and on how much of the energy came from fissioning U-235, how much from fissioning Pu-239, and so on. The differences in fission-product production between uranium-based fuels and MOX fuels are quite modest (all the more so because, after all, a significant part of the fission even in uranium-based fuels occurs in plutonium that has been produced in these fuels by neutron absorption in U-238), and, in particular, the inventories of the fission-product isotopes that contribute the most to the radiological risks from spent fuel are practically the same in MOX as in uranium-based spent fuels. Thus, as with the volume of the spent fuel, the quantities of important fission products it contains would not be much affected by the use of the MOX fuel option for WPu disposition. (The most important such change might well be the reduction in the quantity of strontium-90 in the spent fuel, compared to that from using LEU in the same reactors, as discussed above in connection with hazard potential from reactor accidents.)

The quantity of actinides in spent fuel would be substantially different under the MOX option for WPu disposition than for nuclear-energy generation with ordinary uranium-based fuels, however. Typical spent LWR fuel contains about 1 percent plutonium and typical spent CANDU fuel about 0.4 percent; the spent fuel under the MOX option for WPu disposition would contain, for reasonable initial plutonium loadings, from 2.5-5 percent plutonium if LWRs were used and 0.8-1.4 percent if CANDUs were used (see Table 6-1). The amount of americium in discharged MOX fuel would also be greater than in LEU fuel at the same burnup. On the other hand, spent LWR fuel from the MOX option

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

would contain considerably less of the important actinide, neptunium-237 (Np-237), than ordinary spent LWR fuel, because Np-237 is produced by successive nonfission neutron captures by U-235, of which there is several times more in the uranium-based fuel than in the MOX. (If CANDUs were used, the magnitude in the change of Np-237 content in spent fuel from MOX compared to spent fuel from the usual natural uranium CANDU fuel would depend on whether the MOX was made from depleted or natural uranium.)

The spent fuel characteristics governing releasability of the contained fission products and actinides must be considered in the context of the specific release modes that could be responsible for significant doses to humans. Two such modes are generally assumed to constitute the main hazards from geologic repositories:

  • The release mode generally considered to be the most troublesome one is dissolution of the radioactive materials into groundwater that has entered the repository, followed by hydrogeologic transport of the radionuclides into an aquifer used by humans for drinking or irrigation.

  • The second release mode usually taken into account is inadvertent intrusion into the repository in connection with mining operations by a future society unfamiliar with the nature or location of the radioactive waste. In such a scenario, waste-bearing material might be handled by miners and ore processors, as well as being piled up on Earth's surface where it could be further dispersed by wind, rain, and surface water.

The details of the chemical properties of the waste are more important in the first case than in the second, where the radiologic potency per unit volume of waste is the key factor.

In the groundwater-intrusion/hydrogeologic-transport release mode, the potential for doses to the public is dominated by relatively soluble long-lived fission products—technetium-99 (half-life 211,000 years), iodine-129 (half-life 15.7 million years), and cesium-135 (half-life 2.3 million years)—the inventories of which are about the same in MOX as in LEU fuel. The rate of dissolution of these radionuclides into groundwater is determined in part by the rate of oxidative solid-state alteration of the fuel matrix (Sadeghi et al. 1991). Although the fuel matrix of MOX fuel is mainly uranium dioxide, the same as for uranium fuel, experimental investigation would be necessary to determine whether the dissolution rates of these fission products is affected by the larger quantities of plutonium and by the presence of distinct grains of plutonium oxide in the MOX fuel. Thermodynamic considerations suggest that the dissolution rates of the soluble fission products could be less than for LEU fuel.

The higher actinide content of the MOX fuel as compared with LEU fuel does not contribute directly to the risk from the groundwater-intrusion/hydrogeologic-transport release mode, in the form of extra actinide contributions to the potential doses, because the rate of removal of actinides from the emplaced

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

wastes is limited by the solubility of the actinides in the local groundwater conditions rather than by the concentration of actinides in the wastes or the rate of alteration of the waste matrix. There could, however, be an indirect effect in which a higher actinide concentration enhances the rate of alteration of the fuel matrix and hence the rate of dissolution of the fission products by virtue of the extra heat generation contributed by the actinides. This effect could only be significant in the long term—that is, at times beyond 50 years from the time the fuel is discharged from the reactor—because at shorter times the heat generation is dominated by the fission products. It is not even completely clear that a higher heating rate is always disadvantageous: it would tend to accelerate the processes tending to break down the fuel matrix over time, but in unsaturated conditions the extra heat could actually confer some advantage in helping to maintain dry conditions in the repository. In some repository concepts, it appears that the increased alpha activity resulting from the higher concentration of plutonium in MOX fuel could modestly increase the dissolution rate of solubility-limited nuclides.

In the case of the mining intrusion scenario for waste mobilization, the transport of plutonium in surface water exposed to the exhumed waste would still probably be solubility-limited, hence not worse for MOX than for LEU spent fuel, but the hazard from weathered airborne plutonium could increase by a factor of three to five for MOX fuel compared to LEU, depending on the plutonium loading of the former. For a repository to be considered satisfactory even for LEU fuel, however, the probability of intrusion by future miners needs to be very low, and given a suitably low intrusion probability the overall hazard from the repository is likely to be dominated by the fission products in the groundwater/hydrogeologic transport scenario-which, as noted above, will not depend strongly on whether the spent fuel is MOX or LEU.

In the case of the vitrification option, the disposition of 50 tons of WPu could be accomplished through the addition of 1.35 weight percent plutonium to 50/0.0135 = 3,700 tons of borosilicate glass, which would correspond to 2,200 of the 1,680-kg glass logs scheduled to be produced in the large melter at Savannah River. (This combination of plutonium concentration and number of logs is only illustrative; larger numbers of logs at lower plutonium concentration, or smaller numbers at higher concentration, are also possible, subject to the criticality constraint discussed below. The total number of 1,680-kg logs needed to accommodate all of the U.S. defense HLW has been estimated at 6,0007,000.) These glass logs would contain about 20 weight percent fission products; their plutonium content if no WPu were added would be about 0.05 percent.

As in the case of spent fuel, the potential rate of dispersion of the plutonium contained in the glass logs, upon intrusion of groundwater into the repository, would probably be limited by the plutonium's solubility rather than by the plutonium-retention properties of the glass, and the risk to the public in the kinds of

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

scenarios likely to dominate the overall hazard would be controlled by the rates of release and dispersion of the fission products. There is some evidence that the relevant release rates from the glass would actually be diminished by the addition of plutonium to the glass matrix (Berkhout et al. 1993); if this is so, the use of this approach for the disposition of WPu would diminish the otherwise extant risks from defense HLW.

Spent Fuel and High-Level Wastes: Criticality Issues

As already noted, spent MOX fuel or plutonium-laden glass resulting from plutonium disposition would have higher plutonium concentrations than would the corresponding waste forms produced in the absence of plutonium disposition. Unless the effect of this additional plutonium is balanced by additional neutron poisons, the plutonium disposition waste forms will have higher nuclear reactivity and hence higher potential for achieving criticality accidentally.

Prevention of criticality in nuclear waste forms relies on a combination of physical configuration (geometry, dilution, spacing), neutron-absorbing materials in the waste and intervening material, and absence of an efficient moderator. (All three factors do not necessarily have to be present at once: spacing and neutron-absorbing materials suffice to prevent criticality in the pools typically employed for preliminary storage of spent fuel at reactor sites, despite immersion of the spent fuel in water—a good moderator—for purposes of cooling and shielding.) Over long spans of time in a waste repository, many factors affecting the possibility of criticality could change: radioactive decay will change the mix of fissile and neutron-absorbing materials; rock movements and migration of the wastes within the repository could change the physical configuration; differential migration could separate the fissile and neutron-absorbing materials; and intrusion of groundwater into the repository could enhance neutron moderation.

Nuclear Regulatory Commission (NRC) regulations require that waste emplacements be designed so that the effective multiplication factor in the repository (keff) be less than 0.95—meaning that a nuclear chain reaction could not be sustained—in both normal and accident conditions, and that keff would not reach 1, leading to a chain reaction, "unless at least two unlikely, independent, and concurrent or sequential changes have occurred" (OFR 1992, Sec. 60.131). It has not yet been determined, however, over what period criticality control must be guaranteed,53 or what factors prospective licensees will be permitted to rely on in guaranteeing the absence of a criticality concern. Thus, crucial parts of the

53  

The period of potential concern extends to many millions of years. Although the main plutonium isotope, Pu-239, has a half-life of only about 24,000 years—hence would have vanished, for all practical purposes, after a few hundred thousand years in the repository—the product of its radioactive decay is U-235, which also poses a potential criticality problem and has a half-life of over 700 million years.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

regulatory basis for examining the licensing of MOX spent fuel and plutonium-bearing HLW glass remain incomplete.

Criticality in the repository would not necessarily constitute a hazard to long-term public health and safety. The result, if the WPu wastes did go critical, would be to create an underground reactor, similar to the Oklo natural reactor that operated in Africa over a billion years ago (Cowan 1976). The Oklo natural reactor did not "blow up," and neither have any plausible circumstances been identified in which a waste repository containing spent fuel or plutonium-bearing glass could do so. In a moist environment, which is the worst case, 54 it is expected that waste materials would exhibit a strong negative temperature coefficient of nuclear reactivity, meaning that the nuclear reaction would slow as it heated the surrounding material. The rate of heat generation of the critical system would then be limited to the cooling capacity of the surrounding rock. The heat generation, limited in this way, would likely be less than the fission-product decay heat at the time of emplacement of the wastes now scheduled for Yucca Mountain, and the quantity of fission products that might be generated in this way would be substantially smaller than the amount already scheduled for emplacement in Yucca Mountain.

Nonetheless, the new fission products would be generated at a time thousands of years in the future, when nearly all of the originally emplaced fission products would have decayed away, and when the various engineered barriers to release of these fission products might have failed. Hence it is prudent to avoid long-term criticality, and NRC licensing is likely to require a respectable argument that such criticality will be avoided for very long times. It appears that approaches that can provide the needed assurance are attainable, but further research is required to confirm this. In what follows, we examine the relevant factors more closely for the specific case of the proposed Yucca Mountain repository, and we identify the specific issues on which further work is required.

Waste Emplacements in Yucca Mountain. There are two distinct concepts under consideration for loading waste packages in the proposed Yucca Moun-

54  

A moist environment is the worst case because an effective moderator is required to achieve criticality at the low fissile-material concentrations characteristic of spent fuel (whether from LEU or MOX) and plutonium-bearing glass logs—see Chapter 2—and water is the best moderator that could plausibly occur in relevant quantities in a repository. (Very pure graphite is a better moderator, but there is no mechanism for such a substance to materialize unexpectedly in a waste repository.) As our report was in the final stages of preparation, a scenario was proposed by C.D. Bowman and F. Venneri of the Los Alamos National Laboratory (LANL) in which Pu-239 from spent PWR fuel, or from glass logs containing vitrified WPu, was hypothesized to separate from the accompanying materials and spread into surrounding rock in a way that would lead to supercriticality and explosive energy release of hundreds of tons of high-explosive equivalent (Bowman and Venneri 1995). A LANL technical review of the Bowman and Venneri analysis, released March 7, 1995, challenged each of the major hypotheses leading to this result and concluded that the probability of such explosions occurring was "essentially zero" (Canavan et al. 1995).

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

tain repository. One is borehole emplacement, wherein holes large enough to accept a waste container are drilled in the floor of mined cavities, or “drifts," in the repository. To provide for adequate conduction of decay heat into the surrounding rock, the boreholes would be separated by about 10-20 meters. With this spacing, little neutron interaction between emplaced containers is expected, so that the criticality of a single container can be examined without considering the entire system. Recently, attention is turning to drift emplacement, wherein clusters of waste containers would be loaded horizontally on the floor of drifts and later surrounded by clay or crushed rock backfill. In this case, neutron interaction between adjacent containers is likely, and, in considering criticality, all the containers in one drift would have to be considered as an integrated system.

Glass Criticality and Preferential Leaching. In current planning, criticality of HLW glass is not a major concern, as this glass would have only tiny quantities of fissile plutonium. In a glass containing one or a few percent plutonium by weight, such as might be used for disposition of WPu, neutron multiplication would be held down by the presence of large quantities of neutron-absorbing boron in the borosilicate glass. Over the very long term, however, one must consider the possibility that groundwater could intrude into the repository, that the waste container might ultimately fail, and that in the presence of water, the boron and lithium in the glass might leach away more rapidly than the plutonium. In current performance analyses of conceptual geologic repositories, the processes of dissolution55 and mass transfer into the surrounding rock and groundwater are analyzed for times extending to hundreds of thousands of years and more. This type of analysis must be applied to these plutonium-bearing waste forms, to examine whether the boron and lithium neutron absorbers would leach away before the plutonium would, causing the neutron multiplication to increase with time.

First, it is important to consider the likely presence of water, which acts both as a leaching agent and as a neutron moderator, increasing criticality. The glass itself will have some porosity initially (and some cracking), allowing some water to intrude within the glass itself. Over time, water would react with silica and other constituents in the glass, resulting in lower-density hydrated reaction products, and potentially increasing the concentration of hydrogen atoms compared to plutonium in the glass (which is very important to the criticality of the system). The area surrounding the waste container may also fill with water, and the surrounding rock has a porosity in the range of about 15-30 percent, which may be partly or completely filled with water. Thus it is possible for a wide

55  

Here we use "dissolution" of a component to mean the net dissolution of that component, with the dissolved species either increasing the concentration of that species in the water surrounding the waste or being transported away from the dissolution front by diffusion and convection. Other writers in this field sometimes take "dissolution" to mean the reaction of the waste solid with water, even if the reaction product is a new solid phase.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

range of densities and amounts of water to exist in the glass and its surroundings over thousands of years. Calculations demonstrate that neutron multiplication will be very sensitive to the amount of water present. Future calculations on these subjects should explore the various possibilities to find the worst case, which can then form a basis for design.

How fast might the boron, lithium, and plutonium in the glass dissolve? The criterion for preferential dissolution can be specified in terms of the fractional dissolution rates, that is, the ratio of the mass dissolution rate of a species to the inventory of that species in the waste. If the atom ratio of two dissolved species is the same as the atom ratio in the undissolved solid at the time of dissolution, we have "congruent dissolution." Here the fractional dissolution rates of the two species are equal, and no preferential dissolution that would increase criticality would occur. If the fractional dissolution rate of species A is greater than that of species B, the ratio of the amount of A to the amount of B in the undissolved solid decreases with time. If A is the absorber and B is the plutonium, decreasing the ratio of A to B could lead to criticality.

Laboratory experiments show that lithium in glass dissolves at a greater fractional rate than does the silica. As the silica in glass reacts with water to form a hydrated siliceous compound, the boron may be incorporated in the solid reaction product, or it may dissolve congruently with the hydration reactions of the glass. However, according to Wicks (1992) boron will be leached from the plutonium glass rubble relatively quickly.

When glass containing plutonium reacts with water to form a hydrated siliceous solid, the plutonium also reacts and forms a hydrated oxide precipitate. These solid phases will be distributed throughout the glass-water reaction products that will eventually fill the volume once occupied by the glass; because of their lower density, they will expand. Plutonium precipitates are expected to have a far lower solubility than the siliceous reaction product. Even though the inventory of plutonium is far lower than that of silica, the plutonium solubility is low enough that its net fractional dissolution rate is expected to be lower than that of silica. Predictions of the net fractional dissolution rate that appear in the literature (National Research Council 1983) are 10-6 per year for silica and 4 × 10-7 per year for plutonium, for a waste glass containing 0.007 percent plutonium. From these data we would expect a 140-fold decrease in the net fractional dissolution rate of plutonium for a glass containing one percent plutonium. Therefore, there would be greater chance for preferential dissolution of boron in glass containing higher concentrations of plutonium.

It is possible that most of the boron will have leached out before the Pu-239, with a half life of 24,000 years, will have decayed. If not, the issue of criticality may become that of criticality of U-235, the decay daughter of Pu-239. Even though the U-235 thermal-neutron cross section is less than that of plutonium, criticality can eventually occur if there is sufficient initial concentration of plutonium. After Pu-239 has decayed, and if sufficient boron is still present, the

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

possibility of later criticality will be affected by the relative fractional rates of net dissolution of boron and uranium. In the Yucca Mountain environment the solubility of uranium may be greater than that of plutonium, because of the oxidizing environment. Therefore, there may be less chance that preferential dissolution of boron will enhance criticality after the Pu-239 has decayed.

The lithium that is present in the glass may or may not form precipitates during the hydration reactions of glass and water. Most lithium compounds are soluble in water, so preferential dissolution of the lithium is also possible.

If the possibility that criticality could arise tens of thousands of years after the waste was emplaced is considered a significant problem, a number of potential solutions seem possible. One possibility would be to limit the plutonium concentration in the glass to a level low enough that the package would remain noncritical even if all of the boron and lithium in the glass leached away. Within the constraints of our study, we have only been able to begin exploring what the maximum concentration of plutonium that would be reliably noncritical in the worst case would be.56 In a dry system, glass logs with no boron or lithium would remain non-critical even with concentrations of 3 percent or more plutonium by weight (corresponding to 60 kg or more of plutonium in each log).57 As noted above, however, adding water to the system can greatly increase the neutron multiplication. At the other extreme, if all the glass leached away, leaving only plutonium (or its uranium decay daughter) dissolved in water in the volume that once contained the glass container, 10 kg of plutonium would be critical.58 That system, however, may be overmoderated; there may be an intermediate water percentage in the glass that would result in even lower maximum plutonium concentrations. If we assume, for example, that for this reason only 5 kg of plutonium could be incorporated in each log, disposing of 50 tons of plutonium would require 10,000 logs, compared to a total of just over 6,000 currently scheduled for production at the Savannah River Site (McKibben et al. 1993).

Another possible approach is to add a neutron-absorbing material which, unlike boron, would not dissolve away more rapidly than the plutonium. From the theories of mass transfer described above, it is possible to design such a

56  

We would like to thank several parties for their help in providing calculations and advice on this subject, including the group at Westinghouse Hanford Company led by Ron Omberg; William Culbreth of the University of Nevada, Las Vegas; Jor-Shan Choi of the Lawrence Livermore National Laboratory; and Pete McGrail of Pacific Northwest Laboratories. Responsibility for any errors or omissions in this discussion, however, is our own.

57  

This was confirmed in calculations by William Culbreth using the Keno code (version 4) and in calculations by the Hanford group using the MCNP code (Culbreth 1993, Omberg 1993). The results of the two codes were consistent. Because of the lower fission cross-section of U-235, the system would be even less critical if the plutonium had decayed to U-235 by the time the system had reached such a point.

58  

The minimum amount of plutonium that can reach criticality in an idealized system is about half a kilogram. The amount of plutonium's decay daughter, U-235, needed to reach criticality is substantially larger.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

system: what is needed is to ensure that the ratio of solubility to inventory of the absorber is lower than that of the plutonium in the glass. There are several rare-earth oxides that are possible candidates. Gadolinium, for example, appears particularly promising, as it has a neutron-absorption cross-section much larger than that of boron, and a solubility believed to be comparable to that of uranium in the chemical environment expected in Yucca Mountain. Within the constraints of our study, we have not been able to explore the full range of such possibilities, or develop reliable information on the solubility of these potential additives, but we believe that this approach could be successful in alleviating long-term criticality concerns relating to plutonium in borosilicate glass. In combination with approaches limiting the amount of plutonium in each log (either by limiting the concentration or reducing the size of the logs), we feel confident that an acceptable solution to this potential problem could be developed within a few years at quite modest expense.

Criticality of plutonium dissolved in the water surrounding the waste packages would not appear to be an issue for the proposed Yucca Mountain repository. A typical value of the plutonium solubility used in analyses for unsaturated tuff is about 10-3 grams per cubic meter of water. This is far less than the single-parameter concentration limit for criticality of plutonium in water. Sorption of plutonium on the porous rock could increase the local concentration of plutonium in the rock-water mixture. However, even adopting a conservatively high sorption retardation coefficient59 of 1,000 for plutonium, the effective concentration of plutonium would be well below the single-parameter concentration limit. Neutron absorption in the solid phase would further reduce multiplication. Another possibility is that plutonium might be precipitated by a local redox front. While this needs to be examined, we do not expect that this will be a serious problem—and if it is, the problem would also arise with plutonium that had leached away from ordinary spent fuel over very long times.

Spent Fuel Criticality. Potential criticality of normal LEU LWR spent fuel is being examined as part of the development of the Yucca Mountain repository, but planning is still in the early stages. As currently envisioned, as many as 20 spent fuel assemblies might be placed in a single large waste container. There is some concern that if water found its way into the repository, and ultimately into the container, and if the structure of the assemblies failed—so that the plutonium-bearing materials were in a water-moderated system concentrated at the bottom of the container—a criticality problem might arise. The Yucca Mountain project is examining a variety of approaches to address this concern, including means for filling the volume surrounding the assemblies (so that less water could enter the container, and failure of the assemblies would not result in con-

59  

The retardation coefficient is the ratio of the amount of plutonium in a unit volume of saturated porous rock to the amount of plutonium in the pore water in the rock.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

centration of the material) and the addition of neutron-absorbing materials such as boron or gadolinium.60

Our study is not the place, however, to examine issues concerning the spent fuels already scheduled for emplacement in the Yucca Mountain repository. On the assumption that some combination of the methods just described will be successful in demonstrating the absence of a criticality problem for normal LEU LWR spent fuel, we believe that an extension of the same methods should make it possible to resolve similar issues for MOX spent fuel. If necessary, a smaller number of MOX assemblies could be placed in each waste container, or the MOX assemblies could even be emplaced individually. We note that the quantity of plutonium in this fuel (as in the case of the glass), would be a small fraction of the total of more than 600 tons of plutonium in spent fuel already scheduled for emplacement in the Yucca Mountain repository.

Low-Level Wastes, TRU Wastes, and Tailings

In the case of the once-through MOX spent fuel option, the primary influence of plutonium disposition on the character of the LLW that would otherwise be produced by the corresponding amount of electricity generation would be the production of TRU wastes from conversion of plutonium metal to oxide and from MOX fuel fabrication, to which steps there is no counterpart in a once-through LEU fuel cycle. In the 1992-1993 Plutonium Disposition Study of the U.S. Department of Energy, the following estimates of TRU waste production in MOX fabrication were developed by vendors:

  • ABB-Combustion Engineering (ABB-CE 1993): ABB-CE estimated that "total alpha contaminated wastes" would be 100 m3/yr for a 15-year campaign absorbing 50 tons of WPu, hence 30 m3 per ton of WPu. (How much of this total would be TRU waste by the 100 nCi/g definition is not clear from their report.) The ABB-CE study estimated that 0.5 percent of the MOX would be scrap, yielding 250 kg plutonium in MOX scrap for a campaign absorbing 50 tons of WPu. MOX scrap has a density of about 2 g/cm3, which if plutonium were 4 percent by weight of the MOX would imply 0.08 gPu/cm3, or 0.08 tons plutonium per m3, or 0.25 tons Pu / 0.08 tons Pu/m3 = 3.1 m3 of this particular form of TRU waste for the 50-ton campaign.

  • General Electric (GE 1993): GE's fuel fabrication complex was said to produce 60 m3/yr of LLW in connection with a MOX output of 58

60  

DOE also plans to use Yucca Mountain for disposal of a variety of spent fuels with greater enrichments than ordinary LWR spent fuel, which have been generated in defense and research programs. These spent fuels, which include highly enriched naval-reactor fuel and certain research-reactor fuels, will pose greater criticality difficulties and are likely to pose greater criticality difficulties and may require additional compensating measures.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

MTHM/yr (sufficient to absorb 50 tons of WPu in 29 years), hence 35 m3 per ton of WPu. Of the 60 m3/yr, "a small fraction" was said to be TRU wastes. The small production of such wastes, compared to past experience with MOX fabrication, was attributed to minimizing the usual main sources of TRU waste production at MOX fuel fabrication plants, namely the chemical analysis lab and the scrap fuel area. Specifically, GE argued that: (1) modern nondestructive assay techniques will reduce use of chemical analyses and thus reduce the amount of TRU-contaminated chemicals; (2) since the prime objective is disposal of plutonium and not maximum fuel burnup or lifetime, nearly all pellets fabricated will be accepted for irradiation, minimizing scrap waste.

We have not found detailed estimates of the production of LLW or its TRU component from the conversion of plutonium metal to oxide. (In arriving at the above-cited estimates, the vendors assumed that the government would provide the plutonium to the fabrication plant in oxide form. See GE 1993, p. 3.6, and ABB-CE 1993, p. III-82.) As an upper limit, the volume of TRU wastes from metal-to-oxide conversion could hardly be greater than the volume associated with dealing with the same amount of plutonium in a fuel reprocessing plant, less the part of the reprocessing-plant TRU wastes that occur in the form of fuel-cladding hulls. Based on past reprocessing experience using the PUREX process, this volume was on the order of 50 m3 per 1,000-MWe reactor-year, hence per 250 kg of plutonium, thus 200 m3 per ton of plutonium (APS 1978, Holdren 1992). Non-TRU LLW from fuel reprocessing has amounted to another 50-100 m3 per 1,000-MWe reactor-year, or 200-400 m3 per ton of plutonium. We believe the actual totals are likely to be considerably smaller than these upper limits.

The amount of defense TRU wastes accumulated in the United States through 1990 was about 290,000 m3 and the total amount of defense LLW about 250,000,000 m3. Even if the above-cited vendor estimates of the production of LLW and TRU wastes from MOX fuel fabrication were to prove to be too optimistic by severalfold, and even if our upper-limit estimates of LLW and TRU wastes for plutonium metal-to-oxide conversion were correct, the contribution of WPu disposition by the MOX/spent-fuel route to the preexisting burden of defense LLW and TRU wastes would be tiny in relative terms. Depending on what fraction of the total LLW from MOX fuel fabrication actually turns out to be TRU wastes, and on the actual TRU-waste production from plutonium metal-to-oxide conversion, the corresponding TRU claim on repository volume could range from about the same to several times larger than that from the spent fuel produced by this disposition option, but in any case it would be a small fraction of the repository volume claimed by spent fuel from civilian nuclear-energy generation in total in the same period.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

We have not seen or developed any detailed estimates of the production of LLW and TRU wastes for the vitrification option. Starting with plutonium oxide, the number and complexity of the processing steps needed to incorporate the PuO2 into glass logs appear to be smaller than in the case of MOX fuel fabrication, which would suggest that the volume of LLW and TRU wastes should also be smaller. This tentative conclusion would need to be verified by closer investigation, but in any case it does not seem possible that the LLW and TRU wastes from this operation could exceed a percent or so of those already produced by the U.S. defense complex. It may well be, for the vitrification option as well as for the MOX/spent fuel option, that the largest quantities of LLW and TRU wastes will come from the plutonium metal-to-oxide conversion step that is common to both options.

With respect to uranium-mill tailings, use of the MOX/spent fuel option would obviate the need for the uranium mining and milling associated with 3060 1,200-MWe reactor-years of electricity generation, hence would reduce by

(30-60 reactor-years) × (100,000-200,000 m3/reactor-year) = 3-12 × 106 m3

the quantity of mill tailings that otherwise would be produced. This reduction would be an environmental benefit, albeit a modest one in relation to the quantities of tailings that already have been produced in connection with civilian and military nuclear-energy operations and that continue to be produced in connection with the civilian ones. If the vitrification option were chosen, there would be no effect on quantities of mill tailings.

Waste Implications of Other Reactor Approaches

If an advanced-reactor type—e.g., a liquid-metal or high-temperature gas reactor—were employed for WPu disposition, the characteristics of the wastes would differ in greater or lesser respects from those of the wastes of LWRs, depending on the reactor type and the fuel-cycle in which it was used. Differences could include the volume, physical form, chemical properties, and to some extent even isotopic composition of both high-level and low-level wastes. Some of these changes could be advantageous in reducing waste management burdens and the associated risks to workers and the public, while others could be disadvantageous. Given that the repository behavior of even the most familiar and much-studied types of spent fuel and HLW—for instance, those from existing commercial reactor types, reprocessing options, and vitrification processes—has not yet been fully characterized, it is only possible at this point to discuss in general terms the implications for waste management of other reactor and fuel-cycle approaches.

In general, it would be the case that approaches that actually destroy a high fraction of the WPu would be associated with some reduction in the criticality concerns connected with waste management, since the amount of plutonium

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

consigned to wastes would be relatively low; but because most of the high-destruction options would entail repeated reprocessing and MOX fuel fabrication steps, there would be a tendency to offset this criticality advantage with the production of greater quantities of TRU waste per ton of initial WPu treated than would be associated with the once-through MOX approach.61 On the other hand, some of the advanced-reactor/fuel-cycle combinations reviewed earlier in this report—notably those with online or integrated pyrometallurgical fuel reprocessing—have been designed to minimize the production of radioactive wastes associated with fuel reprocessing; as noted earlier, however, the technological practicality and economic feasibility of these approaches, not to mention the details of their waste-generating performance, remains to be proven. And, while it may be argued that greatly improved performance in waste generation—if it should turn out to be achievable—would be a great advantage in a reactor system intended for a major role in power production, the radioactive waste burdens associated with more conventional approaches to the disposition of 50 or 100 tons of WPu are not large enough to constitute an important incentive to develop advanced reactors and fuel cycles for the narrower and smaller-scale purpose of plutonium disposition.

Summary of Waste Issues

Spent fuel resulting from the use of MOX in LWRs and borosilicate glass containing WPu as well as defense HLW would be different enough from spent fuel derived from LEU and the borosilicate-glass/HLW combination, respectively, that separate licenses would be required to certify these new forms as acceptable for waste disposal; it would not be possible to rely on the licenses and associated reviews required for commercial LEU spent fuel and currently planned HLW glass. The licensing processes for the new waste forms would entail investigations of the mobilization of radioactivity and of criticality potential under conceivable repository conditions that would be much more extensive and thorough than the preliminary considerations undertaken so far and summarized above. If the outcome of these investigations were unsatisfactory, the option in question would not proceed.

Based on the considerations summarized here, however, we think that an irreparably unsatisfactory outcome is not likely. Most probably, if it proves possible to certify spent fuel from LEU and borosilicate glass with defense HLW for disposal in a geologic repository, it will also prove possible to certify spent fuel from MOX and borosilicate glass containing WPu as well as defense HLW. It could turn out, of course, that some modifications in repository design or waste form are desirable in order to accommodate the extra plutonium, but we

61  

There would also be a trade-off within the category of criticality issues, since there are criticality hazards associated with reprocessing.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

doubt that these would be of a sort to impose substantial additional delays on a repository program that is unlikely to be in operation before 2015 in any case.62

As for LLW, including especially their TRU component, the disposition of WPu will certainly produce some additional TRU (compared to what would be produced in the absence of the disposition program), and under current regulations these will require disposal in a geologic repository. Conversion of plutonium metal to oxide, a common step for both the MOX/spent fuel and the vitrification options, might well be the dominant source of this increase, minimizing the importance of any difference in TRU generation between the two options in their subsequent steps. In any case, neither the absolute magnitude of the TRU wastes that could plausibly be generated by these two disposition options, nor any plausible differences between them, are likely to be large enough to constitute a significant impediment to choosing either one, or a basis for choosing between them. With respect to uranium-mill tailings, the MOX/spent fuel option has the advantage of reducing the tailings burden from what it would be if all nuclear electricity generation were based on LEU fuel, while the vitrification option would have no impact on tailings quantities; but the magnitude of this advantage of the MOX option is too small, given its tiny impact on the total tailings burden that will exist in any case, to have any significant influence on the choice between MOX and vitrification.

THE COMPARISONS IN SUMMARY

We summarize here the principal conclusions from the comparisons we have undertaken in this chapter. For brevity's sake, we will refer to the three main classes of disposition options we have considered as (1) "current-reactor" options (meaning the use of currently operating light-water or heavy-water reactor types, or evolutionary adaptations of them, to incorporate WPu into spent fuel, at typical commercial burnups, on a once-through basis), (2) "vitrification" (meaning incorporation of the WPu into glass logs containing high-level fission-product wastes), and (3) "advanced-reactor" options (meaning use of such reactor types as LMRs, MHTGRs, MSRs, PBRs, and ABCs either to incorporate

62  

It has been asserted in at least one analysis (Shaw 1992) that borosilicate glass is less complex and easier to analyze than spent fuel in terms of repository behavior, which could be taken to imply that the possibility of reaching an earlier conclusion about its suitability constitutes a significant advantage of choosing the glass route for plutonium disposition. We are not convinced that this distinction is either clear or significant—or if it were that it would remain so when the complication of additional plutonium content is imposed. Further, the intensive effort that has been mounted by DOE to design and evaluate the repository package for LEU spent fuel probably will give a licensing edge to the MOX/spent fuel option over the WPu/borosilicate-glass option. In any case, we think that attempts to characterize the repository behavior of both waste types in their plutonium disposition forms should proceed energetically and in parallel, as part of a process of intensive parallel investigations of both aimed at choosing the one that is best overall.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

the WPu into spent fuel on a once-through basis or to destroy it to a greater extent by means of multiple recycles).

Security

Security considerations include both "direct" and "indirect" effects of a disposition option. The former are the ways an option influences the barriers against diversion or theft of the WPu. The latter are the ways it influences nonproliferation and arms reduction incentives and the protection afforded stocks of nuclear-explosive materials other than those dispositioned.

Timing is crucial to both of these dimensions of security. Minimizing the time until the start of operations to transform the surplus WPu into less accessible forms, and minimizing the time until this transformation is completed, are of obvious value in reducing the direct risks of diversion and theft. An expeditious approach brings major indirect security benefits, moreover, by signaling both commitment to irreversible arms reductions and seriousness in addressing proliferation hazards.

In terms of this crucial timing aspect of security, the current-reactor options and the vitrification option are roughly comparable to each other, and both are greatly superior to the advanced-reactor options. Under the most optimistic assumptions that are defensible, loading of WPu into current-reactor types could begin between 2002 and 2004 and be completed between 2015 and 2025; loading of WPu into waste-bearing glass logs could begin around 2005 and be completed as early as 2013. The timing uncertainties in both cases-relating more to resolution of institutional issues in the reactor case and to resolution of technical issues in the vitrification case-are bigger than the differences in the best-case point estimates we have provided; thus it would not be meaningful to say more than that the two sets of options are comparable.

Under the most optimistic assumptions that we consider defensible, any of the advanced-reactor options would be at least a full decade slower to get started; the delay could easily be longer even for the most well developed of these options (MHTGRs, LMRs), and it would probably be two decades or more for the least well developed of them (MSRs, PBRs, ABCs). We believe that the direct and indirect security risks of delays of this magnitude should be considered unacceptable, given that the current-reactor options and vitrification option provide the means to avoid these risks and given that the advanced-reactor options do not appear to offer advantages in other aspects of security, economics, or ES&H (as summarized in what follows) that could offset their timing liability.

The main factors besides timing that affect the comparative security of disposition options are (1) the extent of exposure to theft or diversion in the processing and transport steps that an option entails and (2) the theft and diversion risks posed by the plutonium in its final form and location. Among processing

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

and transport steps, those involving forms of material with low barriers to use in weapons and high portability are most problematic. Vitrification entails fewer such steps than the current-reactor options but, for the circumstances likely to apply in the United States, longer transport links. (As discussed in the preceding sections, there can be substantial differences—in these respects and in others—among variants within an option, depending for example on how many reactors at how many sites are employed, whether fuel fabrication facilities and reactors are co-located, and so on.) Advanced-reactor options do not appear to offer significant advantages, with respect to vulnerability of processing and transport steps, over the better variants among once-through current-reactor options, or over vitrification.63

With respect to the security of the final plutonium forms that disposition options produce, we have concluded that meeting the “spent fuel standard" is both necessary and, for the decades immediately ahead, sufficient: if the WPu in its final form is not substantially more accessible for weapons purposes than the larger quantities of plutonium that will continue to exist in spent fuel from commercial electricity generation in this period, it will not represent a significant additional security hazard; but there is no great security advantage to be had from making the WPu much less accessible than the rest of the plutonium in commercial spent fuel, since the latter would then dominate the overall risk.

The current-reactor options obviously meet the spent fuel standard, and we judge that the vitrification option meets this standard also. The plutonium in the spent fuel assembly would be of lower isotopic quality for weapon purposes than the still weapons-grade plutonium in the glass log, but since nuclear weapons could be made even with the spent fuel plutonium this difference is not decisive. Under middle-of-the-road assumptions,64 the radiological barrier presented by glass logs would be about three times smaller than that presented by a fuel assembly (but still very high), and the mass of a glass log—containing, coincidentally, about the same amount of plutonium as a fuel assembly—would be about three times greater. The difficulty of separating the plutonium from the accompanying materials would be roughly comparable in the two cases.

Use of advanced reactors could produce reductions in the quantity of residual plutonium from the disposition process (even on a once-through basis in the

63  

If recycling of plutonium in order to burnup more of it were deemed important, some of the advanced-reactor options might offer security advantages by being able to recycle plutonium without separating it completely from fission products. In the once-through mode we favor to reduce the clear and present security danger of WPu, however, advanced reactors do not offer reductions in handling and processing of vulnerable plutonium forms.

64  

Here we compare PWR fuel assemblies initially containing 5.5 percent WPu (55 g WPu per kgHM, 461 kgHM per assembly, 658 kg MOX plus hardware per assembly), irradiated to 40 MWd/kgHM, with the large logs currently planned to be produced at the Savannah River site (1.3 percent WPu and 20 percent fission products in 1,700 kg of glass, 2,200 kg of glass plus canister) (see Table 6-5).

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

case of the MHTGR, and with the assistance of plutonium recycle in other cases) and in some cases (particularly the MHTGR) could produce decreases in the plutonium's isotopic quality for weapons purposes to below that characteristic of spent fuel. As just noted, however, these changes would not bring much reduction in overall security risk unless commercial spent fuel stocks were similarly transformed. Although society might eventually decide to do this and might choose advanced reactor types for the purpose, transforming today's very dangerous stocks of surplus WPu to meet the spent fuel standard does not require advanced reactors and should not wait for them.

Economics

Estimates of the economic consequences of alternative reactor-related disposition options depend strongly on the assumptions that are made about a wide array of factors: the real interest or discount rate; the time required for design, construction, and licensing; the treatment of interest during this preoperational period; the market value of any electricity produced; whether the facilities are owned by the government or the private sector; and so on. What should be assumed about these matters is not always obvious, and any analyst's choices can be criticized. If different disposition options are to be compared on an even-handed basis, however, the assumptions made about these factors must at least be consistent across the cases analyzed, and, similarly, the estimates of construction costs and operating costs that are at the core of any economic analysis must have been generated for the different options with a consistent degree of conservatism or optimism. We have sought, in our economic analysis, to impose a degree of consistency in these respects and to explain the associated assumptions in a way that will permit interested readers to determine the origins of any differences between our results and those of other analysts.

Among many ways to present the results of economic analyses, we have chosen as our primary figure of merit the discounted present value, as of the start of WPu disposition operations at a reactor (or melter, in the case of the vitrification option), of the stream of costs produced by the associated WPu disposition activities before and after this time, after subtracting revenues from associated electricity generation where appropriate. We express this net discounted present value at start of reactor (or melter) operation in 1992 dollars. Table 6-21 summarizes the results for all of the options for which such estimates were developed.

The most important conclusions that emerge from the cost calculations summarized in Table 6-21 are as follows:

  • The central estimates of the net costs of the reactor-related options considered, expressed as net present value as of the start of reactor or melter operations, range from about $0.5 to about $6 billion (1992 dollars) for disposition of 50 tons of WPu. When theuncertainty ranges

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-21 Net Economic Impact of Options for WPu Disposition (in millions 1992 dollars)

MOX Use in Currently Operating U.S. LWRs:a

 

No Modifications Required

Modifications Required

Fuel Source:

FMEF

New Plant no Tax

New Plant with Tax

FMEF

New Plant no Tax

New Plant with Tax

 

450±250

900±300

1,100±300

1,500±400

1,900±400

2,100±400

MOX Use in Currently Operating CANDU Reactors in Canada, Fuel from FMEF: 950b

Completion of Mothballed U.S. PWRs for MOX Use (fuel from FMEF):c

 

 

1 reactor, 6.8% Pu, 30 years:

1,300±1,600

 

 

2 reactors, 4.0% Pu, 25 years:

2,200±3,000

Construction of New Evolutionary or Advanced Reactors for MOX Use: d

 

 

 

Reactor Pays no Ins, Prop Taxes, Fuel from FMEFe

Reactor Pays Ins, Prop Taxes, New Fuel Plant

ELWR-1 (GE ABWR)

2,600±3,600

5,500±3,800

ELWR-2 (ABB-CE System-80+ PWR)

1,600±1,800

3,200±1,900

ALWR (Westinghouse PDR-600)

3,100±2,100

5,100±2,400

MHTGR (GA)

3,900±2,700

5,800±3,200

ALMR (GE)

3,900±2,500

5,600±3,000

Vitrification with Defense HLW Savannah River Site:f 1,000±500

NOTES: Figures are discounted present value of net costs of disposition of 50 tons of WPu by the indicated means, accounting for electricity revenues where relevant. Present values are calculated as of start of plutonium disposition operations at the reactor (or melter, in the vitrification case). For details of calculations and justifications of assumptions, see the economic evaluation section in Chapter 3 and "Economic Comparisons" in this chapter. The ± ranges are the panel's judgmental 70-percent confidence intervals.

a Two 1,200-MWe PWRs dispositioning 50 tons of WPu in 21 years (rounded from Table 6-13).

b Two 769-MWe CANDUs dispositioning 50 tons of WPu in 24 years (no confidence interval estimated) (see section "Weapons Plutonium Versus Uranium as Power Reactor Fuel" in this chapter).

c One or two 1,256-MWe PWRs (see "Completing Existing LWRs" in this chapter).

d For the MHTGR and ALMR, the figures in this column correspond not to FMEF but to a new fuel fabrication plant that does not pay property taxes and insurance.

e See "Building New Reactors for Plutonium Disposition" in this chapter.

f See "Economics of Vitrification" in this chapter.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

on these estimates-here expressed as judgmental 70-percent confidence intervals-are included, the ranges extend from a negative $1.2 billion (i.e., a profit in this amount) to a cost of $9.0 billion.

  • The lowest central estimate, at about $0.5 billion, is for the MOX/spent fuel option using currently operating U.S. LWRs that need no modification to use MOX safely, with the fuel fabricated at the FMEF at the Hanford site. Four of the options studied have central-estimate costs around $1 billion: use of MOX fuel from FMEF in currently operating CANDU reactors in Canada; use of MOX from FMEF in a single, currently mothballed, partly completed PWR that would be completed for this purpose; use of MOX from an entirely new fuel fabrication plant in currently operating U.S. LWRs that need no modification to use MOX safely; and vitrification with defense HLW at the Savannah River site.

  • Although the central estimates in all cases considered correspond to net costs, our judgmental 70-percent confidence intervals include a possibility of profits from WPu disposition for the case in which currently mothballed, partly completed PWRs are completed for the purpose of plutonium disposition and these PWRs use MOX fuel from FMEF, and for cases when new, evolutionary LWRs are built for this purpose at government facilities (paying no property taxes or insurance) and use MOX from FMEF. These profit possibilities depend not only on the costs associated with MOX use falling at the low end of our judgmental 70-percent confidence ranges, but also on the additional electricity generated by the plutonium disposition reactors being marketable at a price of 5.5 cents per kilowatt-hour or higher (1992 dollars) at the busbar.

  • Central estimates for the net costs of the use of advanced reactors for WPu disposition, at $3-$6 billion, are considerably higher than the central estimates of using currently operating or mothballed reactors for this purpose. The circumstances needed to realize costs at the low end of the uncertainty ranges for the advanced reactors include a high market value for the electricity they produce—a circumstance that would also lower the net costs of the other options.

It is important to note that all of these cost estimates represent sums of money that are modest in relation to the security benefits of plutonium disposition. For example, the best-estimate, net-cost figure for a typical current-reactor MOX/spent fuel option—say, $1-$2 billion (1992 dollars) net present value at the start of reactor operations in, say, 2002—could be paid for by setting aside less than 0.4-0.8 percent of the fiscal year 1995 U.S. defense budget (even without allowing for interest on this sum between 1995 and 2002). The range of $0.5-$5 billion ( 1992 dollars)—covering the best estimates of net present value, at reactor or melter startup, of most of the options considered—corresponds to

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

$10,000 to $100,000 per kilogram of WPu, or $40,000 to $600,000 for a nominal "bomb's worth" of 4-6 kg. Even the higher figure is probably less than what this weapon material once cost to produce, as well as much less than would be spent in the attempt to recover such material if it went astray and incomparably less than would be spent to try to deter or otherwise prevent its use in the form of a bomb in the hands of a potential adversary.

Environment, Safety, and Health

The ES&H characteristics of the current-reactor options and the vitrification option for disposition of surplus are summarized in Table 6-22, based on the treatment of this topic in "Building New Reactors for Plutonium Disposition" above. Organized by types of activities common to the two classes of option, this table compares the options to each other with respect to the ES&H issues they raise in each activity, and then places these issues in the larger context of the ES&H impacts of analogous nuclear-weapon and nuclear-energy activities in which the United States has been and will continue to be engaged irrespective of a WPu disposition program.

The main conclusions that emerge from this table and from the more detailed analysis in the section "Building New Reactors for Plutonium Disposition" are that:

  • most of the ES&H impacts of WPu disposition using either of these options can be expected to represent modest additions, at most, to the routine exposures to radiation and risks of accident associated with other civilian and military nuclear-energy activities underway in the United States;

  • there is no apparent reason that the activities involved in WPu disposition using either of these approaches should not be able to comply with all applicable U.S. ES&H regulations and standards;

  • while there are differences in detail in the ES&H challenges and risks posed by the two options in some of the activity categories—e.g., a somewhat more complicated set of plutonium-handling operations for the reactor options than for the vitrification option, and a greater relative increase in plutonium content of the final waste form for the vitrification option than for the reactor options—these differences do not consistently favor one class of options or the other, and none is large enough in relative or absolute terms to justify choosing one class of options over the other;

  • the ES&H issues to which the greatest attention ought to be given in the next phase of study of these options are (1) ensuring adequate safety against criticality accidents in the melter for the vitrification option, (2) confirming the conditions under which full-MOX cores can be

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

TABLE 6-22 Summary of ES&H Issues for Current-Reactor and Vitrification Options

Activity

Current-Reactor Options vs. Vitrification Options

WPu Disposition vs. Other Nuclear-Energy and Nuclear-Weapon Activities

Pit storage

No difference

Not more demanding than intact-weapon storage already done on a larger scale

Transport of pits or plutonium metal

No difference

Not more demanding than intact-weapon transport already done on a larger scale

Plutonium metal-to-oxide conversion

No difference

No equivalent in current U.S. nuclear-weapon or nuclear-energy practice, but technology for doing it safely is well established

Storage of plutonium oxide or nitrate

No difference

Less demanding than storage of wide array of plutonium forms and plutonium-contaminated materials in nuclear-weapon and nuclear-energy complexes

MOX fuel fabrication

Vitrification option lacks an equivalent step

No equivalent in current U.S. practice, but technology well established and less demanding in ES&H terms than metal-to-oxide conversion

Transport of fresh MOX fuel

Step does not occur in vitrification option

Plutonium in fresh fuel adds some risk to current nuclear-energy practice, but risk is small compared to transport of pits and weapons

Irradiation in reactor

Effects of plutonium addition more complex than in vitrification counterpart, but well studied

Plutonium addition unlikely to significantly affect safety of the small subset of commercial reactors involved

Addition of plutonium to vitrification melt

See comparison above; main plutonium effect is on criticality potential, which needs more study

Except for criticality issue, plutonium addition adds little to ES&H risks from melter operation that would occur without WPu program

Transport of spent fuel or glass logs

Spent fuel more radioactive and probably more easily dispersed

Extra plutonium from WPu disposition adds only moderately to spent fuel or glass-log transport risks that would occur without WPu program

Geologic repository storage of spent fuel or glass logs

Relative increase in plutonium content from WPu disposition is greater for glass logs than for spent fuel

WPu adds little to population dose except in very unlikely intrusion scenarios; added criticality needs more study

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

used without adverse impacts on safety in reactors of currently operating commercial types, and (3) clarifying the conditions for avoiding long-term criticality in geologic repositories containing either spent fuel or glass logs from WPu disposition operations.

It is worth emphasizing here that the task of ES&H assessment is considerably simplified by the circumstance that both of the leading-candidate classes of options would simply add WPu to a set of nuclear activities that would be going on in any case. In addition, both would leave the residual WPu in a waste form-spent fuel in one case and HLW-bearing borosilicate glass logs in the other-which will exist in large quantities and will need to be safely managed whether used for WPu disposition or not, and the relevant properties of which will not, for the most part, be very much affected by the addition of WPu in the quantities foreseen.65

Some of the advanced-reactor options could plausibly reduce some of the ES&H impacts associated with use of the current-reactor options, such as the magnitude of the uranium-mill-tailings burden and, perhaps, the quantities of low-level TRU wastes produced in MOX fuel fabrication. Such potential ES&H gains from the use of advanced-reactor options are far too modest, however, to justify the security liabilities of postponing WPu disposition until the advanced-reactor options are available.

65  

The main caveat relates to the potential for criticality in the repository in the very long term. The higher plutonium concentrations in spent fuel associated with using current-reactor/spent-fuel options for disposition, and the much higher plutonium concentrations in glass logs used for WPu disposition compared to those not so used, may add to the repository criticality problem. Nonetheless, the plutonium content in ordinary spent fuel is sufficient to necessitate very careful attention to the avoidance of repository criticality even in the absence of WPu disposition, and it is likely that the effort required to provide assurance against repository criticality for ordinary spent fuel will lead the way to measures that will provide this assurance for WPu-MOX spent fuel and for plutonium-bearing glass logs as well.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

APPENDIX:APPROVAL AND LICENSING ISSUES IN WEAPONS PLUTONIUM DISPOSITION

Regulatory and licensing issues will be a critical pacing factor in accomplishing plutonium disposition in the United States (and, to differing extents, in other countries that might be involved). An administration decision to burn or bury the WPu will involve material that was very costly to produce, and is viewed by some as a potential asset because of its energy value. Such a decision will also raise important issues for nuclear waste disposal. DOE has already determined that a Programmatic Environmental Impact Statement (PEIS) is required. Congressional authorization and appropriation will also be necessary. The process of decision and approval is likely to be extended and controversial.

As noted in the text, the regulatory environment that now exists in the United States affords multiple opportunities for opponents of large nuclear projects to intercede with regulatory agencies, the courts, and Congress to slow or stop their implementation. While some progress has been made in streamlining the licensing process for such projects in recent years, the overall process of pursuing a large project to completion-including not only the licensing process but the fundamentally political process of gaining funding and associated approvals for such efforts-remains a difficult and uncertain one.

Any new activity at a nuclear-reactor site or major DOE nuclear site generates local public interest and, usually, opposition. Plans to process tens of tons of plutonium at a particular site, or to introduce MOX fuel into certain reactors, can be expected to produce such interest. Added to the reactor-related opposition is the strong anti-DOE feeling in many communities. Although the current DOE has made efforts to change these attitudes, any DOE initiative is likely to be scrutinized closely and to face public opposition. Any flaws in a required PEIS or EIS probably would serve as the basis for court challenges. The length of time required, the probability of success, and the cost involved in gaining licensing and regulatory approval for relevant plutonium disposition options are difficult to predict.

Reaching Agreement on an Option

The first and most fundamental steps in the process will be to bring the various differing interests within the administration and Congress to agreement on a specific choice of disposition options, and to gain sufficient funding from Congress. This will take time (the current hoped-for schedule is to reach a decision in the spring of 1996) and is likely to be contentious. Committed advocates

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

of particular options are likely to take whatever opportunities present themselves to attempt to reverse decisions that go against them.66

The National Environmental Policy Act and the EIS Process

The first major regulatory step will be meeting the requirements of the National Environmental Policy Act (NEPA), which requires an Environmental Impact Statement (EIS) for all major federal actions likely to have a significant environmental impact. NEPA was intended to ensure that environmental issues would be appropriately considered in the process of reaching decisions on preferred options for tasks such as this one. An EIS must consider all reasonable alternatives for accomplishing the objective, and must in particular examine the "no action alternative." DOE is currently preparing a Programmatic Environmental Impact Statement (PEIS) analyzing the options for both plutonium storage and plutonium disposition (as well as storage and disposition of the other surplus fissile materials, such as HEU, U-233, and miscellaneous actinides). A draft is expected to be released for public comment in the fall of 1995, with the final document published in early 1996. Subsequent EISs for specific activities at particular sites may then be necessary.

There is an extensive history in recent years of court challenges to the adequacy of EISs. These challenges have sometimes been successful in delaying projects by several years—which, with changing circumstances over time, has sometimes led to cancellation of the project concerned. Whether the PEIS that DOE is now preparing, or subsequent site-specific EISs, would suffer such a fate is impossible to predict. It should be noted that the EIS is only the first step in this process, not the last: in addition to the EIS and gaining formal license approvals from federal regulatory agencies, for example, many state permits may be required unless Congressional action removes these requirements.

Are New Legislation or Generic Rules Needed?

A major question is whether plutonium disposition (particularly the use of plutonium fuels in U.S. reactors) would require new legislation.

It appears that new legislation is not essential, but that some form of congressional approval would be highly desirable. In 1983, Congress barred the use of commercial fuel for nuclear explosive purposes. This legislation was specifi-

66  

In the fiscal year 1993 and 1994 budgets, for example, advocates of particular reactor options inserted Congressional "earmarks" directing that specific amounts of the money allocated to plutonium disposition be spent on studying particular reactor types, though DOE resisted this earmarking. In the fiscal year 1995 budget, the earmarking is more general, specifying a substantial amount of funding that should be spent only on nuclear reactors, and calling specifically for a fast turn-around study of the so-called "triple-play" reactor options—reactors that would burn plutonium, produce electricity, and produce tritium for the weapons stockpile.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

cally aimed at preventing plutonium from reprocessed civilian spent fuel from being used for nuclear weapons. There is no prohibition, however, on the use of material from weapons for commercial fuel. (Indeed, the HEU purchase agreement, in which LEU produced from HEU from Russian nuclear weapons will be purchased by the United States for commercial use, will have established this principle before plutonium disposition begins.) Nor is there any general legislation forbidding the use of plutonium fuels in U.S. commercial reactors. However, the overall process (including NEPA determinations) would be smoothed if there were explicit Congressional approval of the particular disposition options chosen. Given that Presidential Decision Directive 13 indicates that the United States does not encourage reprocessing of plutonium, 67 Congressional approval would be particularly desirable if reactor options were chosen, to emphasize the national interest in accomplishing the plutonium disposition mission. The panel was informed unequivocally by the CEOs of two nuclear utilities that clear congressional approval would be a requirement before commercial utilities would allow WPu-MOX to be used in their currently operating commercial reactors. (The continued conflict in Nevada concerning opening a repository at Yucca Mountain, however, demonstrates that it is not enough merely to have Congress pass a law stating that an action is in the national interest.)

Another question is whether, in the absence of congressional legislation, a generic rule-making by the Nuclear Regulatory Commission (NRC) would be necessary before plutonium could be used in U.S. reactors. In the mid-1970s, when industry and government predicted that reprocessing and recycle of plutonium in reactors would soon be necessary, the NRC began a lengthy regulatory process to address the associated environmental issues, known as the Generic Environmental Statement on Mixed Oxide (GESMO). The GESMO public hearings were halted by the NRC in 1977 in response to the Carter administration's opposition to reprocessing.68 Since that time, case law has changed so that a hearing would not be required were such a rule-making to be undertaken today. Even more important, it probably would not be necessary to develop a generic rule at all if only a few reactors were involved and the reprocessing issue was not raised. Instead, license amendments for the few reactors involved would be required (see below).

67  

White House (1993). This statement has been widely interpreted as opposing use of plutonium in general, but appears to be specifically directed at use that involves additional reprocessing, which need not be the case here.

68  

Since the NRC is an independent agency, the administration could only request, not direct, the NRC to halt the GESMO hearings. The administration did make clear that there would be no federal funding requested for any aspect of the reprocessing regime. Because the potential reprocessing industry depended on federal funding, including development of breeder reactors, the administration's positions made the GESMO hearing irrelevant.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

Federal and Nonfederal Reactors: Who Regulates?

Any plutonium disposition option is likely to involve construction, modification, and operation of important facilities that must be regulated—particularly fuel fabrication plants and reactors for the reactor options, or vitrification facilities for the disposal option. Technically, the licensing issues related to the reactors themselves are only a matter of degree, and do not appear to pose any major technical obstacles. But the procedural issues—how these various facilities would be regulated, and by whom—will be important.

Reactors owned by the U.S. Department of Energy (either government-owned, government-operated [GO-GO] or government-owned, contractor-operated [GO-CO]) would be a special case—one in which current practices are under discussion and review. Currently, the NRC does not regulate any federally owned nuclear plants used for purposes connected with the defense mission. The Defense Nuclear Facilities Safety Board (DNFSB) was established by Congress in 1988 to provide a form of regulatory oversight for DOE weapons facilities. As currently constituted, however, the DNFSB is a purely advisory body with no authority to issue binding regulations or orders. Congressional legislation has been introduced to provide external regulation of some aspects of DOE. This legislation likely will be reintroduced in the 1995 Congressional session, with possible modifications based on DOE recommendations following advice from an advisory committee established to review all aspects of external regulation of DOE.

Were DOE to build a nuclear plant for this mission, or were DOE to buy and complete or take over an existing commercial nuclear plant, DNFSB could be asked to provide advisory oversight. It is not completely clear that DNFSB would be the appropriate agency, however, because such a plant would be for the purpose of generating electricity (using MOX fuel which happens to be fabricated using weapons material), rather than part of the weapons complex. On the other hand, the NRC currently has no statutory authority to regulate DOE facilities, so further legislation would be required if it were to serve in anything more than an advisory role comparable to that of the DNFSB. Absent such legislation, a full-fledged regulatory regime would not exist for a DOE-owned reactor in either the DNFSB oversight or the NRC advisory case. Ultimately, it appears both likely and desirable that reactors burning WPu-MOX will have to meet the same NRC safety and licensing standards applied to commercial reactors.

There are major differences in the operation of the NRC and the DNFSB, including an order of magnitude difference in the size of their staffs. The DNFSB has no hearing process, does not operate under the Administrative Procedures Act, and has not developed the practice of involving the public in its deliberations (an issue on which it has been taken to court). Under current procedures, the length of time required to license a facility, or to get approval for

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

operation, would probably be shorter under the NRC than the DNFSB, if the review and hearings processes were similar. Certainly this would be the case if the DNFSB's actions were challenged in court.

If an existing licensed commercial reactor were taken over by DOE for plutonium disposition, the issue of transferring its license would be raised. Although licenses have been transferred from one utility to another, a license or a construction permit has never been transferred from a commercial facility to the government.69 If a transfer to DOE were proposed, it is most likely that the NRC would require the utility to apply for a license termination. The NRC's role then would be to make a finding of the effect of such action on the safety of the public.

If a new reactor were to be constructed with DOE ownership, under current procedures the DNFSB would provide oversight. Overseeing construction and operation of such a reactor or reactors would pose considerable challenges for the DNFSB. The time required to build such a reactor, however, would give the DNFSB the time to develop gradually the expertise to oversee operation.

If the NRC were asked to provide advice concerning a DOE-owned reactor (as was done for the Fast Flux Test Facility, for example), there would be no formal NRC licensee, but there is the issue of who would pay for these NRC reviews. The NRC is required by law to recover all its costs from the applicants and licensees, but the NRC has no authority to charge DOE. DOE, however, could provide the funding for such reviews. Or, if DOE could characterize the reactor as a “demonstration" reactor, and it were operated on a commercial grid, the reactor would be licensed by the NRC, and as a licensee, would pay the costs of regulation.

The licensing procedure for a reactor that was privately owned would be more clear-cut. Nonfederal reactors operate under license from the NRC. The operating license review includes consideration of the core. Major changes in the core, such as introduction of MOX fuel, would be reviewed by the NRC for potential effect on the safety of the reactor. Although a claim might be made that a license amendment would not be required,70 it would be prudent to assume that one would be necessary. The license amendment process would offer the opportunity for challenges, leading to public hearings. Under current law, the NRC could decide that no significant hazard would result from the amendment and allow the amendment to take effect, with the hearing held afterwards. This would be somewhat more likely for one-third MOX cores, which have been used overseas, than for full-MOX cores, but would still be a politically difficult decision for the Commissioners.

69  

The only exception is that ownership of uranium-mill tailing piles, after stabilization, is transferred to the federal government.

70  

Such a claim is made in GE (1994, p. 5.2-8).

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

To summarize, the two reactor options being considered can be assessed as follows:

  1. Privately owned reactors would be regulated by the NRC, would require some type of environmental impact document, and, if currently holding licenses, would require license amendments. Barring a major technical safety problem (unforeseen at present), the licensing process should be completed well before the time required to build a MOX fabrication facility and obtain approval for its operation. Congressional endorsement would be highly desirable and may be required by utility owners of current reactors.

  2. Government-owned reactors (either GO-CO or GO-GO) would be overseen by the DNFSB if current procedures were used, but the Secretary of Energy could ask the NRC to provide informal advice or the Secretary could ask Congress to amend the Atomic Energy Act to enable the NRC to regulate these reactors. Congress may act in 1995 or 1996 to provide external regulation of operations now managed by DOE. If the oversight process were handled by the DNFSB, its small size and unfamiliarity with regulation could make the process longer than if handled by the NRC. Nevertheless, the MOX fabrication facility still is likely to be the pacing item.71

Regulating Fuel Fabrication Facilities

For the reactor options, fuel fabrication facilities would also be required, and these too would require regulation. Procedurally, if the fuel fabrication facility were owned by the Department of Energy, or operated under a DOE contract with DOE oversight, the same considerations described above for the case of government-owned reactors would apply (even if it supplied fuel to a nonfederal utility). If the facility were privately owned and the only contractual obligation to the government were to provide fuel, the facility would have to be licensed by the NRC—even if all the fuel it produced were to be used by the federal government (as a precedent, the NRC licensed and regulated the privately owned facilities that made nuclear fuel for the U.S. Navy).

Technically, NRC licensing of a plutonium fuel fabrication facility might be somewhat more difficult than licensing reactors to use MOX. Given the absence of civilian reprocessing and MOX use in the United States, it has been many years since the NRC licensed a large plutonium bulk-handling facility. It may take some time to develop the necessary regulatory expertise. The DNFSB does

71  

However, if use is to be made of a commercial reactor which had been closed rather than being modified to meet new NRC requirements, then DOE would come under great pressure from Congress to make such modifications and get NRC (informal) approval before operation. This could add several years and hundreds of millions of dollars to the process.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

provide oversight of existing federal facilities handling plutonium in bulk forms, but large-scale plutonium processing in the DOE complex has been virtually shut down since before the DNFSB was established.

Regulating Vitrification Facilities

Vitrification would presumably be done in DOE-owned facilities (possibly operated by private contractors), as is currently planned for vitrification of U.S. HLW. Currently, the NRC does not have regulatory oversight of these facilities, which are reviewed by the DNFSB. (The NRC, however, does plan on monitoring vitrification facilities, to assure that the product will meet the NRC deep geological disposal regulations.) DNFSB oversight of plutonium vitrification in these facilities could raise issues similar to those described above in the case of possible DNFSB oversight of DOE reactors. Since the current waste already includes some plutonium, no fundamental regulatory changes would need to be made to review the inclusion of WPu. The principal technical issues would relate to criticality safety, the potential for plutonium releases, and worker exposures, all of which are issues DNFSB has addressed extensively in other areas. Presuming a vitrification facility was designed to provide adequate assurance of safety in these areas, gaining DNFSB approval for its operation should not pose special obstacles.

Repository Regulatory Impacts

Either spent MOX fuel or vitrified HLW glass would ultimately have to be disposed of in a geologic repository licensed by the NRC under regulations written to conform with other regulations written by the EPA. In both cases, the products are sufficiently different from the analogous products now scheduled for disposal—LEU spent fuel and vitrified HLW without significant quantities of plutonium—that it is likely they will have to be independently certified as acceptable waste forms for disposal.

Although EPA has issued final regulations for disposal of wastes in the Waste Isolation Pilot Plant (WIPP), there are as yet no final regulations which a Yucca Mountain repository must meet. A National Research Council committee will recommend a regulatory approach for a Yucca Mountain repository, which EPA must consider. EPA and NRC regulations will follow. Neither the addition of WPu into the vitrification process, nor the use of MOX fuel with its generally higher plutonium content, has been examined for its impact on the repository license. The criticality issues addressed elsewhere in this report appear to be the most important technical questions that would be of concern.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

Licensing and Russian Plutonium Disposition

If the future regulatory process in the United States is somewhat uncertain, that in Russia is even more so. In the wake of the collapse of the Soviet Union, the recently established Russian nuclear regulatory agency, GOSATOMNADZOR (GAN) is still finding its role. At this writing (early 1995), a new Atomic Law is still being debated in the Russian parliament. It is inevitable that the regulatory environment in Russia will evolve substantially between now and when plutonium disposition is actually underway on a large scale, in ways that are difficult to predict. At the moment, GAN is much weaker politically than the Ministry of Atomic Energy, which it is intended to regulate. As a result it appears unlikely that GAN objections to particular plutonium disposition facilities, even were they to arise, would have a major impact on the timing or cost of plutonium disposition in Russia.

Licensing and Plutonium Disposition in Other Countries

If Canadian CANDU reactors were used for plutonium disposition, these would be regulated by Canada's Atomic Energy Control Board. As noted in Chapter 4, the regulatory environment in Canada tends to be less adversarial and involve more co-operation between regulators and utilities than is the case in the United States.72 As in the U.S. LWR case, reactor licensing would probably not be a major source of delay in the case of the CANDU option. Just as in the United States, however, the broader issue of overcoming all the potential political and regulatory barriers would be a difficult one in Canada if there were significant public opposition to MOX use in Canadian reactors.

If U.S. or Russian WPu were used in Europe or Japan, the relevant regulatory agencies and publics would be substantially more familiar with civilian use of plutonium. While the use of weapons-grade plutonium does involve somewhat different issues, there would not appear to be any major problems in the licensing process itself. The more difficult problems would arise from the political issues involved in shipping large quantities of WPu from the United States or Russia to these countries.

72  

Ahearne (1988).

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

REFERENCES

ABB-CE 1993: ABB-Combustion Engineering. Pu Consumption In ALWRs. DOE Plutonium Disposition Study. Windsor, Conn.: ABB-Combustion Engineering. Final Report, May 15, 1993.

AECL 1994: AECL Technologies, Ltd. Plutonium Consumption Program CANDU Reactor Project Report. Contract DE-AC03-945F20218. ??: AECL Technologies, July 3 , 1994.

Aheare 1988: J.F. Ahearne. "A Comparison Between Regulation of Nuclear Power in Canada and the United States." Progress in Nuclear Energy 22(3):215-219.

APS 1975: American Physical Society, Study Group on Light-Water Reactor Safety. "Report to the APS." Reviews of Modern Physics 47, Supplement No. 1, Summer 1975.

APS 1978: American Physical Society, Study Group on Nuclear Fuel Cycles and Waste Management. "Report to the APS." Reviews of Modern Physics 50(1) Part II, January 1978.

Atomic Energy Society of Japan 1992: Proceedings of the International Conference and Design and Safety of Advanced Nuclear Power Plants (October 25-29, 1992, Tokyo). Tokyo: Japan Publications Trading Company, 1993.


Battelle 1993: Battelle Pacific Northwest Laboratory. "Preliminary Estimate of Plutonium Disposition Capability of VVER-1000 Reactors." Unpublished manuscript, May 6, 1993.

Berkhout et al. 1993: Frans Berkhout, Anatoli Diakov, Harold Feiveson, Helen Hunt, Edwin Lyman, Marvin Miller, and Frank von Hippel. "Disposition of Separated Plutonium." Science and Global Security 3:161-213, 1993.

Bowman and Venneri 1995: C. D. Bowman and F. Venneri. "Nuclear Excursions and Eruptions from Plutonium and Other Fissile Material Stored Underground." Report LA-UR 94-4022, Los Alamos, N.Mex.: Los Alamos National Laboratory, 1995 (in press).


Canavan et al. 1995: Gregory H. Canavan, Stirling A. Colgate, O'Dean P. Judd, Albert G. Petschek, and Thomas F. Stratton. Comments on "Nuclear Excursions" and "Criticality Issues" LA-UR: 95 0851. Los Alamos, N.Mex.: Los Alamos National Laboratory, March 7, 1995.

Cowan 1976: G. A. Cowan. "A Natural Fission Reactor." Scientific American 235:(July)36-47, 1976.

Culbreth 1993: William Culbreth. Personal communication to Matthew Bunn, September 21, 1993.


Dahl 1993: Roy E. Dahl. "An Assessment of FMEF Supporting the Isaiah Project-Preliminary Report." Unpublished manuscript, October 1993.

Delene and Hudson 1993: J.G. Delene and C.R. Hudson, II. Cost Estimate Guidelines for Advanced Nuclear Power Technologies, ORNL/TM10071/R3. Oak Ridge, Tenn.: Oak Ridge National Laboratory, May 1993.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

Eisenbud 1973: Merril Eisenbud. Environmental Radioactivity, 2nd ed. New York: Academic Press, 1973.

ERI 1993: Energy Resources International. "Nuclear Fuel Cycle Supply and Price Report." As reported in Nuclear Fuel, June 9, 1993.


Fetter and von Hippel 1990: Steve Fetter and Frank von Hippel. "The Hazard from Plutonium Dispersal by Nuclear-warhead Accidents." Science and Global Security 2:21-41, 1990.

Fischer et al. 1987: L.E. Fischer, C.K. Chou, M.A. Gerhard, C.Y. Kimura, R.W. Martin, R.W. Mansing, M.E. Mount, and M.C. Witte. Shipping Container Response to Severe Highway and Railway Accident Conditions. NUREG/CR-4829. Washington, D.C.: U.S. Nuclear Regulatory Commission, February 1987.


GA 1993: General Atomics. MHTGR Plutonium Consumption Study-Final Report . DE-AC03-89SF-17885. San Diego, Calif.: General Atomics, May 14, 1993.

GA 1994: General Atomics. MHTGR Plutonium Consumption Study, Phase II Final Report. GA/DOE-051-94. San Diego, Calif.: General Atomics, April 29, 1994.

GE 1993: General Electric Nuclear Energy. Study of Pu Consumption in Advanced Light-Water Reactors: Evaluation of GE Advanced Boiling-Water Reactor Plants. DE-AC03- 93SF19681, NEDO-32351. San Francisco, Calif.: General Electric. May 13, 1993.

GE 1994: General Electric Nuclear Energy. Study of Plutonium Disposition Using Existing GE Advanced Boiling Water Reactors. DE-AC03-93SF19681, NEDO-32361. San Francisco, Calif.: General Electric, June 1, 1994.

GE-ANL 1993: GE Nuclear Energy and Argonne National Laboratory. ALMR Plutonium Disposition Study. GEFR-00919. San Francisco, Calif.: General Electric, May 1993.


Hohenemser 1988: Christoph Hohenemser. "The Accident at Chernobyl: Health and Environmental Consequences and the Implications for Risk Management." Annual Review of Energy 13:383-428, 1988.

Holdren 1992: John P. Holdren. "Radioactive Waste Management in the United States: Evolving Policy Prospects and Dilemmas." Annual Review of Energy 17:235-259, 1992.

Holdren et al. 1989: J.P. Holdren, D.H. Berwald, R.J. Budnitz, J.G. Crocker, J.G. Delene, R.D. Endicott, M.S. Kazimi, R.A. Krakowski, B.G. Logan, and K.R. Schultz. Report of the Senior Committee on Environmental, Safety, and Economic Aspects of Magnetic Fusion Energy. UCRL-53766. Livermore, Calif.: Lawrence Livermore National Laboratory, September 1989.


IAEA 1986a: International Atomic Energy Agency. Decay Data of the Transactinium Nuclides. Technical Reports Series No. 261. Vienna: IAEA, 1986.

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

IAEA 1986b: International Atomic Energy Agency. Definition and Recommendations for the Convention on the Prevention of Marine Pollution by Dumping of Wastes and Other Matter in the Ocean, 1972. Safety Series No. 78. Vienna: IAEA, 1986.

IPPNW/IEER 1992: International Physicians for the Prevention of Nuclear War and Institute for Energy and Environmental Research. Plutonium: Deadly Gold of the Nuclear Age. Cambridge, Mass.: International Physicians Press, 1992.


Kessler et al. 1992: G. Kessler, D. Faude, R. Kroebel, and H. Wiese. "Direct Disposal Versus Multiple Recycling of Plutonium." Kernforschungszentrum Karlsruhe. Paper presented at a conference in Tokyo, November 1992.

Kudriavtsev 1993: E.G. Kudriavtsev. "Russian Prospects for Plutonium Accumulation and Utilization." Paper presented to International Atomic Energy Agency meeting on Problems of Separated Plutonium, April 1993.


Lahs 1987: William R. Lahs. Transporting Spent Fuel: Protection Provided Against Severe Highway and Railroad Accidents. Washington, D.C.: U.S. Nuclear Regulatory Commission, March 1987.

Lange and Hanson 1993: Larry Lange and Christopher Hanson. "A New Nuclear Proposal." Seattle P-I, p. 1, August 26, 1993.

Levina et al. 1994: I.K. Levina, V.V. Saprykin, and A.G. Morozov. "The Safety Criteria and VVER Core Modification For Weapon Plutonium Utilization." Paper presented at NATO Advanced Research Workshop on "Mixed Oxide Fuel (MOX) Exploitation and Destruction in Power Reactors," Obninsk, Russia, October 16-19, 1994.

Lucoff 1989: D. M. Lucoff. "Passive Safety Testing at the Fast Flux Test Facility." Nuclear Technology 88:21-29, 1989.


McKibben et al. 1993: J.M. McKibben, R.W. Benjamin, D.F. Bickford, L.P. Fernandez, W.N. Jackson, W.R. McDonell, E.N. Moore, P.B. Parks, M.J. Plodinec, W.M. Rajczak, S.K. Skiles, and G.G. Wicks. Vitrification of Excess Plutonium. WSRC-RP-93-755. Aiken, S.C.: Westinghouse Savannah River Company, 1993.

Mikhailov et al. 1994: V.N. Mikhailov, V.V. Bogdan, V.M. Murogov, V.S. Kagramanian, N.S. Rabatnov, V. Ya. Rudneva, and M.F. Troyanov . "Utilization of Plutonium in Nuclear Power Industry of Russia." Paper presented to the "International Policy Forum: Management and Disposition of Nuclear Weapon Materials," Leesburg, Virginia, March 8-11, 1994.

Murogov et al. 1994: V.M. Murogov, V.S. Kagramanian, and N.S. Rabatnov. "The Use of Weapon and Reactor Plutonium in WWER and BN Type Reactors." Paper presented at NATO Advanced Workshop on "Managing the Plutonium Surplus: Applications and Options," Royal Institute of International Affairs, London, January 24-25, 1994.


NAS 1979: National Academy of Sciences, Committee on Science and Public Policy and Committee on Literature Survey of Risks Associated with Nu

Suggested Citation:"Chapter 6: Comparing the Options." National Academy of Sciences. 1995. Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options. Washington, DC: The National Academies Press. doi: 10.17226/4754.
×

clear Power. Risks Associated with Nuclear Power: A Critical Review of the Literature. Summary and Synthesis Chapter. Washington, D.C.: The Academy, 1979.

NAS 1994: National Academy of Sciences, Committee on International Security and Arms Control. Management and Disposition of Excess Weapons Plutonium. Washington, D.C.: National Academy Press, 1994.

National Research Council 1983: National Research Council, Board on Radioactive Waste Management. A Study of the Isolation System for Geologic Disposal of Radioactive Wastes. Washington, D.C.: National Academy Press, 1983.

National Research Council 1988: National Research Council, Committee on Biological Effects of Ionizing Radiation. Health Risks of Radon and Other Internally Deposited Alpha Emitters (BEIR IV), Washington, D.C.: National Academy Press, 1988.

National Research Council 1989: National Research Council, Committee to Provide Interim Oversight of the DOE Nuclear Weapons Complex. The Nuclear Weapons Complex: Management for Health, Safety, and the Environment . Washington, D.C.: National Academy Press, 1989.

National Research Council forthcoming: National Research Council, Committee on Separations Technology and Transmutation Systems. Report of the Committee on Separations Technology and Transmutation Systems. Washington, D.C.: National Academy Press, forthcoming.

Novikov et al. 1994: A.N. Novikov, V.V. Saprykin, A.A. Suslov, and A.P. Lazarenko. "Use of MOX (R-Pu and W-Pu) Fuel in VVER-1000 (Neutron-Physical Aspects of Possibilities)." Paper presented at NATO Advanced Workshop on "Managing the Plutonium Surplus: Applications and Options," Royal Institute of International Affairs, London, January 24-25, 1994.

Nuclear Fuel 1992: Mark Hibbs. "German Utilities Bracing for MOX Fuel Cost Increases." Nuclear Fuel, January 6, 1992.

Nuclear Fuel 1993: Mark Hibbs. "Court Says Hesse Must Pay Siemens for Costs of Shutting MOX Plant." Nuclear Fuel, April 26, 1993.


OECD 1990: Organization for Economic Co-Operation and Development, Nuclear Energy Agency. Uranium: Resources, Production, and Demand. Paris: OECD, 1990.

OECD 1992: Organization for Economic Co-Operation and Development, Nuclear Energy Agency, Committee for Technical and Economic Studies on Nuclear Energy Development and the Fuel Cycle. The Economics of the Nuclear Fuel Cycle. OECD NEA/EFC/DOC(92)5. Paris: OECD, December 1992.

OFR 1992: Office of the Federal Register, U.S. Code Of Federal Regulations, Title 10 (Energy) , Washington, D.C.: U.S. Government Printing Office, 1992.

Page 394 <