Disposal of Plutonium Without Irradiation
This chapter discusses options for immobilizing weapons plutonium (WPu) without irradiation, in ways that would make it significantly less accessible for reuse in nuclear weapons and also less hazardous to the environment. The chapter begins with an overview of the technology and a discussion of several of the key technical issues facing vitrification. This is followed by an assessment of this approach on the basis of the criteria developed in Chapter 3 and applied to reactor options in Chapter 4—including timing; safeguards and recoverability; environment, safety, and health; and cost. (A more detailed comparison of the options against the key criteria is found in Chapter 6.)
The goal of the immobilization options, enunciated elsewhere in this report, is to make the WPu roughly as inaccessible for use in weapons as the much larger and growing quantity of plutonium in spent fuel worldwide. (Lesser goals may be of interest as interim steps.) While there are a variety of materials into which plutonium could be embedded to achieve this goal, the primary case we examine in this chapter is incorporating the plutonium in borosilicate glass—a process known as vitrification. We focus in particular on glass that also incorporates radioactive high-level waste (HLW) or other fission products, so as to create a major radiation barrier to handling the material.
This case was chosen as the baseline, not as the result of a detailed comparison of alternative waste forms by the panel, but because, after decades of such comparisons for the mission of disposal of HLW, borosilicate glass has been chosen as the waste form of choice for the HLW disposal mission in the
United States and most other countries. While the plutonium disposition mission is different in some important respects, it appears desirable (to minimize costs, delays, difficulties of gaining approvals, and the like) for the plutonium disposition mission to make use of existing processes, approaches, and facilities to the extent practical, a logic that focuses attention on the borosilicate glasses already scheduled to be produced. (A brief discussion of a few of the alternative waste forms that have been proposed for this mission is also provided.) As discussed in this chapter, plutonium could also be incorporated in glass without fission products, but we do not believe this would provide a large enough barrier to reuse in weapons to be satisfactory as a final disposition option.
Much engineering development work has been done over several decades on vitrification of radioactive materials, both in the United States and in other countries. Based on this work, the general features of the technology are well known: several vitrification facilities have been built and operated around the world using different glasses, different radioactive species, and different throughputs. While glasses containing substantial quantities of plutonium have never been produced on a large scale, the key engineering parameters that would govern a large-scale WPu vitrification operation are believed to be understood. In this sense, the technical feasibility of vitrifying WPu has been adequately demonstrated.
Several important technical issues must be resolved before vitrification of WPu could move from a technical possibility to an operational reality, however. Some of these issues stem from the fact that plutonium has never been vitrified on a large scale before. In addition, further work is required to determine the best mix of plutonium, glass, and fission products for this purpose. The principal objective of vitrifying WPu is to deter its potential reextraction for weapons use. To make this deterrent most effective, it would be desirable to have: (1) small amounts of plutonium in each "log" of glass; (2) large and heavy logs that would be difficult to steal; and (3) large quantities of fission products and other contaminants in the logs to make reextraction difficult. But to minimize cost, take maximum advantage of existing vitrification programs, and meet other criteria, there may also be reasons to increase the amount of plutonium in each log, decrease the amount of fission products in each log, or make the logs smaller. Hence the selection of how much WPu to put in logs of what size, with what composition of other contaminants, requires a systematic exploration of the parameter space, taking into account engineering, handling, and cost aspects. For WPu vitrification, such a comprehensive evaluation has not yet been done.
Some of the technical issues associated with this approach are discussed below. In general the technical uncertainties associated with this approach are somewhat greater than are those surrounding the use of mixed-oxide fuel (MOX) in light-water reactors (LWRs). Nevertheless, the panel believes that WPu vitrification represents a feasible technology that could meet the "spent fuel standard," could be available in the relatively near future (within about a
decade), and could potentially immobilize all of the nominal 50 tons of excess WPu in glass in a relatively short time (a few years, very likely less than 10) once the vitrification campaign had begun.
Carrying out a vitrification plan such as this would require the U.S. Department of Energy (DOE) to make a major commitment to it, involving not only financial support but broad institutional support from the highest levels. Among the needed elements of such a commitment would be actions in order to retain the key technical personnel, recruit others, upgrade facilities, provide high-level "protection" from the budgetary and other institutional threats that will inevitably occur, and so on.
Overview of the Technology
The technology of plutonium vitrification is simple to explain even though it is complicated in detail. Basically, the final product is a glass in which plutonium, and other radioactive wastes in the approach examined here, are dissolved or suspended as impurities while the glass is in the melted state, after which the glass is cooled and solidified. The final form is a glass "log" usually weighing in the range of tens to thousands of kilograms, although the glass can also be made in the form of a powder or in small pieces. Once produced, the glass logs incorporating plutonium and fission products would be stored until a nuclear waste repository became available, at which time they would be emplaced in the repository as waste-without making use of the energy value of the plutonium.1 Once the glass is produced, it is well within current technical capabilities to handle and store the plutonium-laden glass safely from the perspectives of worker safety, environmental contamination, and criticality.
The basic waste vitrification device is a melter, into which a glass powder, known as glass frit, is continuously fed, along with whatever is to be dissolved in the glass. The nonglass material can be a liquid slurry or a dry feed. The melter melts the frit and dissolves (or suspends) the nonglass material in the
glass (Marples 1988). The glass is heated and kept in a molten state, usually by joule heating in a ceramic melter (in which a large alternating current is passed through the molten glass itself) or inductive heating in a metallic melter. Glass product is continuously or intermittently extracted in molten form from the melter as more material is fed in, and is poured into a mold where it cools into a solid form—typically a large glass log. It is beyond our scope here to discuss the several different approaches to heating the melter, to feeding in the frit and the impurities, to accomplishing effective dissolution, and to cooling the final glass log.
Melters to produce the requisite glass can be large (several meters in diameter and in height), small (substantially less than a cubic meter), or in between. Technologies exist for small-scale modular melters that could be built and shipped to the sites where the plutonium now resides (including sites in Russia). To incorporate large quantities of fission products (whether in glass or other waste forms), however, would require a remote-handling facility with adequate protection for workers against the intense radiation—meaning that whether the melter itself is large or small, it will be part of a facility which is large, expensive, and complex. Thus it is likely to be highly desirable to make use of existing facilities to the extent practical. In some options, plutonium might be incorporated in HLW glass logs already scheduled to be produced for HLW disposal, using the same facilities (with some important modifications required). In other options, plutonium might be incorporated in glass logs in addition to those previously scheduled to be produced, either using existing vitrification facilities or new ones constructed for this purpose (probably within an existing remote-handling facility such as a reprocessing canyon or vitrification plant). The issue of what facilities might be used for the WPu disposition mission is an important one, discussed later in this chapter.
The answers to several critical technical questions will determine the feasibility of safely and economically vitrifying WPu. The principal issues can be grouped into three main categories:
designing a form that would meet the "spent fuel standard" with tolerable production costs, schedules, and risks, including options for the physical, chemical, and radiological characteristics of the glass;
difficulties of producing the chosen glass form, including plutonium handling and criticality issues in both preprocessing and during the vitrification process itself, and how much plutonium can in fact be loaded into the glass; and
the suitability of the resulting glass forms for deep geologic disposal.
Before discussing each of these categories of issues, it is necessary to discuss briefly borosilicate glass and several of the possible alternatives to it for the WPu vitrification mission.
THE CHOICE OF WASTE FORM
After decades of study of scores of different possible waste forms, most of the international community has settled on borosilicate glass as the waste form for immobilizing high-level radioactive wastes. Borosilicate glass is being used for this purpose, or is planned for use, in the United States, France, Great Britain, Japan, Germany, and other countries. The choice of borosilicate glass is based on several favorable properties (Marples 1988): it can incorporate almost all of the important radioactive fission products dissolved as oxides; it can contain waste at levels as high as 20 or even 25 percent by weight; it is tolerant of widely varying waste compositions; it is reasonably resistant to leaching by water; it is relatively resistant to radiation damage; it can accommodate the chemical changes that occur when the waste impurities decay radioactively; and the production process is relatively simple and reliable, with a reasonably low formation temperature, and with a glass product that is not corrosive to the process equipment, unlike phosphate and lead phosphate glasses. 2
Providing a sufficient radiation barrier to meet the spent fuel standard for 50 tons of plutonium will require tens or hundreds of millions of curies of radioactivity, a small but significant fraction of the total amount of separated radioactive fission products currently stored in the United States. Hence, incorporating these fission products with plutonium into waste forms other than the borosilicate glasses on which the HLW disposal program is now centered would represent a substantial modification of that program, with the attendant potential for delays and uncertainties for both the HLW disposal program and the WPu disposition program. Nevertheless, it is worth briefly considering a few of the most prominent of the alternate waste forms that have been proposed for the WPu disposition mission-some of which are unique to the WPu disposition mission, and some of which have been considered for the HLW disposal mission in the past. DOE's Office of Fissile Materials Disposition is studying a wide array of possible waste forms for immobilization of WPu, including those discussed below and numerous others.
Phosphate glass is used in the Russian HLW vitrification operation at Chelyabinsk, which raises the question of whether this waste form might be appropriate for vitrification of WPu in Russia (if such an option were ever seriously pursued by the Russian government). Russia is essentially alone in the world in choosing phosphate over borosilicate glass for disposal of HLW; the choice was apparently made in part because of the lower formation temperature
and therefore lower cost of production associated with phosphate glasses. Phosphate glass is less appropriate than borosilicate glass for the WPu disposition mission because it is less durable and less resistant to criticality if WPu is embedded in it (it does not include large quantities of neutron-absorbing boron as a basic constituent). Neutron absorbers could be added to the phosphate glass to control the latter problem, but this possibility has not been examined in detail. The panel believes that if WPu were to be vitrified in Russia, switching to a borosilicate glass would probably be a better approach than continuing to rely on the phosphate glass currently produced (Diakov 1992). Although borosilicate glasses have been studied in Russia, the panel is not aware of any Russian plans to switch to borosilicate glasses, or of any estimates of the cost and schedule for modifying the Russian facility to produce borosilicate rather than phosphate glasses.
A synthetic rock known as “synroc" was developed as a possible HLW form in the United States years ago, but was ultimately abandoned in preference for borosilicate glass. Some work on the concept has been pursued in Australia in the intervening years. The choice of borosilicate glass was based on a number of technical issues related to synroc that have not been resolved, including the larger amount of hot-cell processing required to produce the synroc, and the greater flexibility of glass in incorporating a wide range of wastes. The latter concern might not be a serious problem in the case of plutonium disposition, if the radiation barrier were to be provided by fission products such as the cesium-137 stored at Hanford, rather than by HLW combining a range of products. Nevertheless, it does not appear that synroc has any unique advantages for incorporating plutonium compared to borosilicate glass that would suggest that it would be clearly superior for the WPu immobilization mission.
Some authors have proposed a variety of cement compositions for disposal of radioactive wastes. For disposal of low-level wastes (LLW), this approach has considerable promise, but the consensus of the international community is that for containment of HLW, glasses are superior to cements.
Some analysts have proposed that the "pyroprocessing" approach that was to have been used to prepare fuel for the U.S. integral fast reactor (IFR) program be used instead to combine WPu with spent fuel into a waste form that would meet the spent fuel standard. As part of a redirection package in the wake of the
cancellation of the IFR, DOE plans to provide several million dollars for examination of this approach by the Argonne National Laboratory (ANL).
For the IFR, the pyroprocessing approach (based on molten-salt dissolution and precipitation) would have been used both to recycle IFR fuel and to reduce spent oxide fuels from LWRs to a metal form, from which the short-lived fission products would have been separated. This metal, containing substantial quantities of uranium, plutonium, other actinides, and long-lived fission products, then would have been introduced as fuel for the IFR. For WPu disposition, the concept is to use the pyroprocessing approach to reduce a mixture of oxide spent fuel and WPu to a metal form; in this case, for maximum proliferation resistance, the fission products from the spent fuel would not be separated but would remain in the metal product.3
This approach has several disadvantages that the panel believes effectively rule it out as a serious competitor for the near-term plutonium disposition mission:
First, to meet the spent fuel standard would require the plutonium to be mixed with a large amount of material (roughly 1,000 tons for 50 tons of plutonium, if the product was to be 5-percent WPu by weight); this would require building, in effect, a substantial reprocessing plant. The costs, delays, and approval difficulties involved in building such a facility would be substantial.
Second, at the time the IFR was canceled, reduction of LWR oxide spent fuel to metal was the least developed part of the pyroprocessing scheme, with only a few experiments completed. Feasibility at an engineering scale had not been demonstrated (National Research Council forthcoming). Demonstration and validation of this technology would involve additional costs and delays, with no guarantee of success in providing an economical and effective WPu disposition option.
Third, it is doubtful that it would be desirable to process the WPu in this way unless the resulting intensely radioactive product were suitable for geologic disposal, for if it were not, the material would eventually have to be processed yet again to prepare it for such disposal or for use as fuel (a step that would also require a large remote-handling facility). As the output of the pyroprocessing was never intended to be a waste, it remains uncharacterized as a waste form. Characterization and certification of waste forms for radioactive isotopes that will last many thousands of years is a lengthy and painstaking process that would almost certainly introduce additional delays. It appears unlikely that a metal matrix such as that produced by the pyroprocessing would be a suitable waste form for the chemical environment of Yucca Mountain: the metal, once exposed to water, would be expected to undergo both hydration and oxidation reactions, breaking down its structure and releasing the radioactive materials it contained.
For all these reasons, the panel believes this approach is not competitive with either vitrification in borosilicate glass or the use of MOX in existing reactors, both of which would be likely to involve lower costs, lower technical uncertainties, and shorter delays.
Specialists at the Los Alamos National Laboratory (LANL) have proposed that plutonium be combined with beryllium in such a way that the alpha-n reactions, followed by multiplication of the neutrons in the plutonium, would create enough neutron radiation for the material to be "self-protecting" by Nuclear Regulatory Commission (NRC) and International Atomic Energy Agency (IAEA) standards (100 rads/h at 1 meter) (Toevs and Trapp 1994). In effect, what is proposed is to create extremely large plutonium-beryllium neutron sources. The materials envisioned would be between 10 and 30 percent plutonium by weight. (Higher plutonium percentages, combined with compact geometries, would bring the plutonium closer to criticality, resulting in more neutron multiplication and a higher intensity radiation field.)
While such a plutonium-beryllium combination would be more self-protecting than pure plutonium metal or oxide, it appears extremely unlikely that it would be possible to design such materials in a way that would fully meet
the spent fuel standard without coming perilously close to criticality. As currently envisioned, the radiation field from such materials would be far lower than that from spent fuel until the spent fuel is many decades old, while the plutonium content by weight would be far higher.4 Thus these materials would appear to be a substantially more attractive target for theft by a potential proliferator—or for reincorporation into weapons by a weapons state—than plutonium in spent fuel. Moreover, as with the pyroprocessed waste-form concept, it appears unlikely that such a material would be considered an acceptable waste form for ultimate disposal. Combining plutonium and beryllium would result in a mixed waste, creating very difficult regulatory issues in the United States. If these materials were not suitable for disposal, the additional costs and complexities of eventually processing this material to some other form would eventually have to be borne. The cost of producing these materials, however, might be significantly less than the cost of WPu disposition using the vitrification or MOX fuel options.
TECHNICAL ISSUES FACING VITRIFICATION
Options for a Proliferation-Resistant WPu Glass
The principal objective of WPu vitrification is to place barriers in the way of any party wishing to reuse the WPu for nuclear explosives. Nevertheless, just as with the reactor options (except those designed for near-complete elimination) reextraction of the plutonium is not precluded from any of the glass forms under consideration. The ease or difficulty is only a matter of technical skill, access to facilities, money, and time.
As with spent fuel, the chemical processes needed to extract plutonium from glass are not especially difficult or obscure. The primary difficulty arises from coping with the radioactivity of the fission products also embedded in the glass. To meet the spent fuel standard, the amount of radioactivity would have to be sufficient to require remote operations, such as those used in reprocessing plants. The chemical processes for extracting plutonium from glass would be conceptually similar to those for extracting plutonium from spent fuel. Most types of glass can be easily dissolved in suitable acids, after which separating plutonium and the other impurities requires a series of chemical processing steps that are well known. The difficulty of subsequent steps to purify the plutonium itself would depend on what other impurities, radioactive and nonradioactive, were present in the mix (see below). Other options for recovering the plutonium
from the glass are also widely known, including electroprocessing and plasma-arc processing.
The difficulty of diverting the plutonium-laden glass, transporting it to where it could be processed to extract the plutonium, and conducting that processing would depend on the size of the glass logs, the amount of plutonium in each log, and the radiological and other contaminants also incorporated in the glass. As noted above, no comprehensive trade-off analysis among these variables to select the optimum glass form for cost-effective proliferation resistance has yet been done. The difficulty of these various steps, and the barriers that particular features of the product could pose to use in weapons, would also depend importantly on the skills and resources available to the party that wanted to extract the plutonium for weapons-in particular whether that party was a major weapons state like the United States or Russia, or whether it was a nonweapons-state or nonstate group.
There are several different glass-product options that need to be differentiated. Some differ in their chemical or radiological composition and some in their physical size.
Three general classes of physical size have been discussed with the panel by vitrification experts:
Small glass frit, beads, powders.
Small glass logs of 35- to 70-liter size (one liter of glass weighs about 2.8 kilograms; kg), typically produced by a small melter. These 100- to 200-kg logs are heavy enough so a single person cannot carry away a log, but small enough for hijacking with a forklift.
Large logs, about 3 meters (m) long, 60 centimeters (cm) diameter, weighing about 1,700 kg, plus 200 or more kg for the surrounding canister; too large for forklift handling and regularly available transport. This is the glass form currently planned as the output from U.S. HLW vitrification programs.
Of course, other options are available. Obviously, the larger the size the more impediment to theft and easy post-theft handling. The large glass logs to be produced in vitrification operations at the Savannah River and Hanford sites, about 3 m and over 2 tons each, cannot easily be moved. Especially when combined with radioactivity sufficient to require remote handling, size can be a substantial handling problem for at least some potential parties wishing to reuse the WPu. As an example, the large HLW-laden logs to be produced at Savannah River will be handled individually by a specialized vehicle weighing over 100 tons.
It would be desirable if the plutonium concentration in the logs were low enough that more than one log would have to be stolen to recover enough material for a single weapon. This would imply a very large number of logs, however. If the logs were limited to 2 kg of plutonium each, for example, disposing of 50 tons of WPu would require the production of 25,000 plutonium-bearing logs, more than the entire amount scheduled to be produced in the U.S. HLW disposal program. By contrast, if the logs contained some 20 kg of plutonium (roughly 1 percent by weight for a 2-ton log like those to be produced at the Savannah River Site [SRS], comparable to the percentage of plutonium in spent fuel), only 2,500 logs would have to be produced. In that case, incorporating plutonium into a fraction of the logs already scheduled to be produced at SRS (see below) would be sufficient, without requiring production of any additional logs. Higher concentrations of plutonium may also be possible (see below), but do not appear to be necessary for most options.
Three general classes of glass compositions vis-à-vis their radioactive content have been discussed with the panel by vitrification experts.
Glass with plutonium and HLW. This glass, spiked with high-level radioactive waste that is the detritus of defense processing activities, can be made sufficiently radioactive that handling it would be extremely hazardous to life (lethal external dose in minutes). The nominal case for the HLW glass logs scheduled to be produced at SRS is a radiation field of 5,200 rem/h (roentgen-equivalent-man per hour) at the container surface (roughly 900 rem/h at I m), but in fact nearly all of the logs will be less radioactive than this, many of them having roughly half this dose rate (Westinghouse 1994).
Glass with plutonium and specific fission-product spiking, such as spiking with the cesium-137 (Cs-137) now stored at Hanford. This would create a similar radioactive barrier, but being a single chemical constituent might allow the use of simpler chemistry and preprocessing than use of the complex HLW in storage at Hanford and SRS. Sufficient Cs-137 (about 50 million curies) is stored at Hanford to produce 500 two-ton logs (which could incorporate 50 tons of WPu if the concentration were a high 5 percent by weight) with a radiation field of 2,000 rem/h at I m. DOE currently has no definite plans for disposition of this Cs-137.
Glass with plutonium only. This glass is not very radioactive, and the radioactivity presents little if any impediment to theft or post-theft
handling (except for the hazards of the plutonium itself, particularly if released in an aerosol form).
When spiked with either defense HLW or fission products such as Cs-137, the glass can be made so radioactive (hundreds or thousands of rem/h at 1 m) that lethal doses would arise quickly (McKibben et al. 1993). This would then require remote-handling operations, not only in the handling of the logs but in the initial processing stages as well, substantially increasing the proliferation resistance of the product. The radioactive decay rates of both the defense HLW and the Cs-137 spiking seem to follow roughly the 30-year half-life of Cs-137 (Wicks 1993). Thus a 2,000-rem/h dose rate would decay to some 60 rem/h (below current regulatory standards for being "self-protecting") in 150 years. In another 100+ years the radioactive-handling problem would be small.
It is important to note that intense radioactivity is a major barrier not only in the handling of the logs by a potential reextraction team, but also in the initial chemical processing of the glass and of the plutonium-laden chemical product of glass dissolution. The need for complex remote-handling technology, the problems with equipment becoming and remaining intensely radioactive, and the storage difficulties all contribute to substantially greater costs and greater occupational risks. Many of these costs and risks, however, will have to be borne in any case to dispose of the existing fission products at DOE sites; the important questions in this case are the net additional costs and risks that would be associated with adding plutonium to the HLW vitrification campaigns that will be required in any case.
Chemical (Nonradioactive) Spiking
The plutonium-laden glass could be spiked with nonradioactive chemicals similar to plutonium (actinides, lanthanides) that would make chemical extraction of the plutonium more difficult. (See, for example, Makhijani and Makhijani 1994, Simonson et al. 1994.) After discussion of this possibility with a number of experts in plutonium chemistry, however, the panel concludes that the benefit of available chemical spiking techniques would be modest. Well-known chemical processing steps—within the capability of nearly all states and some nonstate groups—are available that could separate the plutonium from essentially any elements that might be added to the glass.
In particular, if the glass had no fission product contaminants to create a radiological barrier, extracting WPu from it would be substantially easier, whatever the mixture of chemical contaminants added, than extracting WPu from highly radioactive glass requiring remote handling. For states such as the United States and Russia, a chemical barrier alone would be insignificant. For others, it is the panel's judgment that most potential proliferators with the technical expertise, personnel, and the organization required to produce an operable weapon
from separated plutonium—a substantial technical task in itself—would also be able to extract plutonium chemically from a glass log not spiked with radioactivity. Having to do so would not substantially increase the overall time and cost of the project of building a weapon.
Thus the principal advantage of vitrifying plutonium without an added radiation barrier would be to put the plutonium into a form that was large and heavy, and therefore more difficult to steal without detection.5 If the plutonium was nevertheless diverted, the fact that it was in vitrified form would be only a modest additional obstacle on the path of its use for production of a nuclear explosive.
Some analysts have pointed out that Russian officials strongly oppose options that would throw away the plutonium's energy value, and have suggested that vitrification in a plutonium-only glass from which the material could be readily reextracted for fuel at a later time might be a more acceptable alternative for Russia. This approach would only be viable, however, if the plutonium-only vitrification provided a major improvement in proliferation resistance, and the panel has not been presented with evidence that would suggest that this is the case. Thus, the panel believes that putting WPu into such a simple glass form would not constitute enough of a barrier for this to be a serious contender as a long-term disposition option. Therefore, this simpler option is not discussed extensively here.
Nevertheless, the panel recognizes that there might be scenarios in which carrying out such an approach might be considered worthwhile as an intermediate step before revitrifying WPu with fission products. For example, substantial quantities of plutonium are now stored in forms that are not suitable either for long-term storage or for transport; vitrifying these forms of plutonium at the sites where they are now stored in order to produce a safe form for transport to another site where they could be revitrified with fission products could be an attractive disposition approach for these materials. Indeed, in some circumstances it might be judged desirable to undertake a relatively rapid campaign to vitrify all of the WPu without any other radioactive species, thereby inexpensively and rapidly transforming the existing metallic WPu pits and ingots into a glass form, but with a second follow-on vitrification step clearly in mind in which the WPu-laden glass would be revitrified later with highly radioactive species included to provide the desired deterrent.
Summary of Proliferation Resistance Issues
Clearly a wide range of factors would affect how difficult it would be for a particular party to acquire the glass logs produced in this process and extract WPu from them for weapons, including characteristics of both the party concerned and the logs themselves. The intense radioactivity of the logs would pose the greatest barrier. This barrier does not scale linearly with the dose rate: at a certain dose rate (not well defined), remote-handling operations become necessary to process the glass, drastically increasing the cost and complexity of the operation. Beyond that point, increasing radioactivity would continue to pose additional difficulties, increasing the difficulty of acquiring and transporting the logs and so on. To meet the spent fuel standard, enough radioactivity would have to be incorporated into the glass to require remote operations-but how much benefit would be derived from specific further increases in dose rate beyond that point requires further study.
It is important to note that the WPu in this option would remain weapons-grade, unlike the plutonium that could be extracted from most types of spent fuel. Potential bomb-makers would prefer weapons-grade material to reactor-grade material. As noted elsewhere in this report and in the report of the parent committee (NAS 1994, pp. 32-33), nuclear weapons that would have assured yields in the range of 1 kiloton can be constructed with reactor-grade material, using technology similar to that required to produce simple weapons with weapons-grade material. With more sophisticated designs, higher assured yields could be achieved.
Overall, the panel judges that the plutonium in glass comparable to that scheduled to be produced for HLW disposal in the United States would be approximately as inaccessible for weapons use as plutonium in commercial spent fuel.6
Difficulties of Producing Glass with WPu and Fission Products
Plutonium-Loading Capacity of the Glass
A significant technical question is how much plutonium and other waste impurities can be placed into the glass. For those options in which WPu glass logs would be produced above and beyond the HLW glass logs already scheduled for production, the overall cost will be quite sensitive to the number of logs that must be produced (determined by the amount of plutonium that can be put in each). By contrast, if the plutonium is to be added to HLW glass logs already
scheduled to be produced for HLW disposal, the cost of the overall operation will be only modestly sensitive to the number of logs that must incorporate plutonium, unless the level of plutonium that can be added to each log is so low that the overall number of logs to be produced must be significantly increased.
The amount of plutonium that can be put into glass is limited by several factors: the solubility of plutonium in glass, the need to prevent formation of a form that could sustain an accidental chain reaction (known as criticality), and the suitability of the glass product for geologic disposal.
Westinghouse Savannah River Company (WSRC) has quoted experimental work by Plodinec and Wiley (1979) in which 7 percent plutonium was successfully dissolved in a bench-scale test in a glass composition somewhat different from that now planned for use at SRS. (All percentage loadings quoted herein will be percent by weight.) SRS personnel and analysts at the Lawrence Livermore National Laboratory studying the issue for DOE's Office of Fissile Material Disposition believe that even higher loadings may be possible. (The HLW-laden glass to be produced at the Defense Waste Processing Facility [DWPF] at SRS will contain 20-30 percent fission products by weight.) Further experimental work would be required to demonstrate that loadings of several percent by weight could be achieved with current glass compositions in large-scale production, while producing a glass meeting acceptable repository performance criteria (see discussion of repository issues, below).
Criticality also does not appear to place restrictive limits on the amount of plutonium that can be dissolved or suspended in the glass because of the large quantity of neutron-absorbing boron in the glass. WSRC has published simple calculations suggesting that if the glass were homogeneous, it would not be critical at plutonium loadings below 15 percent (McKibben et al. 1993). Given the extreme importance of preventing criticality, however, it is essential to take into account possible inhomogeneities in the glass, as well as possible accident scenarios. The issue of possible criticality during preprocessing and in the melter is discussed in more detail below.
In the case of options in which the WPu would be added to already planned vitrification campaigns, concentrations in the range of 1 percent by weight plutonium would be sufficient to allow all of the nominal 50 tons of excess WPu to be incorporated in the logs already scheduled to be produced at a single site (either Hanford or Savannah River), without any increase in the number of logs. Given that lower concentrations are better from the point of view of the proliferation barrier, there does not appear to be any substantial incentive to move in the direction of higher plutonium concentrations for these options. Only in the case of options in which the plutonium would be incorporated in a separate vitrification campaign, where the processing costs would scale directly with the number of logs produced, would there be any incentive to pursue higher plutonium loadings. Even in that case, it is likely that for reasons of both proliferation resistance and repository performance, loading in the range of one to a few per-
cent will be desirable. Hence, it appears that neither solubility nor initial criticality are likely to be major constraints on achieving the plutonium loadings of most interest.
Handling Plutonium in Upstream Processing and in the Melter
The chemical processing involved in preparing HLW for vitrification is quite complex, and the intense radioactivity of the HLW complicates the problem further. (Indeed, discovery of a series of difficulties with the flow-sheet for chemical preprocessing of wastes at Savannah River has delayed the DWPF project there by several years.) Adding plutonium to this process would raise additional complications, both because of the environment, safety, and health (ES&H) hazards if plutonium were to be accidentally released in aerosol form, and, even more important, the need to design all the relevant processes to ensure against possible criticality.
Plutonium metal would in most cases have to be converted to other forms to prepare it for vitrification.7 Two principal approaches are possible: (1) oxidation (roasting, calcining) of plutonium into an oxide that is fed dry into the melter, or (2) creation of an acidic plutonium solution (McKibben et al. 1993). The oxide form has a significant biological hazard potential if released in aerosol form. The dry-feed approach associated with the oxide form, however, makes it easier to transport the material. The panel believes that the pros and cons of each approach need to be explored.
There are also several approaches for how the WPu might be combined with the HLW and with the glass. The WPu (in oxide or acid solution form) might be mixed with the liquid HLW solution prior to its introduction into the melter, or the two might both be added to the melter separately, to be combined as they each dissolve in the glass. Alternatively, the WPu might be “previtrified" without HLW, and the resulting plutonium-bearing glass added to the melter as a frit. Whatever the approach, detailed engineering work will be required to ensure against criticality, and against aerosolization of the plutonium, throughout the process.
Another unanswered question is whether the design specification for the plutonium-laden glass will require total plutonium dissolution in the glass or can accept some fraction of undissolved plutonium. In some cases, allowing some undissolved plutonium to remain could simplify mixing, and upstream process-
ing, allow melter operation at lower temperatures, and potentially increase throughput.
These issues in the upstream processing, particularly criticality, require careful engineering that has not yet been done. But the panel believes that resolving these issues is within current technical capabilities.
It is also essential to avoid criticality in the melter as the WPu glass is being produced. This depends on the glass composition (including the fraction of plutonium being added to the glass), the melter design, and possible inhomogeneities in the glass.
A variety of different factors in melter design affect the criticality risks that could arise from vitrifying substantial quantities of WPu. Principal among these are the size and shape of the melter and what provisions are made for ensuring the glass in the melter is well mixed (particularly whether the glass is stirred or unstirred).
Currently both the DWPF and, apparently, Hanford's not-yet-built Hanford Waste Vitrification Project, plan to use very large, unstirred melters (Omberg 1993). Most of the major U.S. engineering work on melter technology in recent years has gone into designing and building the DWPF facility at SRS, where a major (20- to 30-year) campaign to vitrify defense-produced HLW is scheduled to begin in 1996 (McKibben et al. 1993). The DWPF melter, remote-handling facilities, and all the supporting auxiliary equipment are designed for this application. Based on material presented to the panel (McKibben and Wicks 1993, 1994; Gray 1994), it appears that the existing first-generation melter at DWPF is not well suited to vitrifying plutonium for a variety of reasons including both criticality concerns and offgas-system questions, issues which we will address in turn.
Large melters like the DWPF have the advantage of large throughputs for a single melter and single feed stream. They have the disadvantage, for the HLW disposal mission, that if a problem arises with the single large melter, production is stopped completely. The amount of glass in the melter at any one time is very large, moreover, and the residence time in the melter is very long, increasing the potential for inhomogeneities in the glass (such as settling of heavier elements or precipitation of some chemical species). In France, by contrast, several smaller melters are used in parallel. The consensus of the vitrification community appears to be moving in the direction of smaller melters. One key process advantage seems to be that these smaller melters can be shut down, cleaned out, and restarted in a matter of a day or two, whereas the shutdown-restart time for the larger DWPF melter system is much longer (Wicks 1993).
For the WPu disposition mission, large melters have the additional major disadvantage that a very large amount of plutonium would be in the melter at any one time, increasing criticality concerns. In the DWPF, for example, if the plutonium were 1 percent by weight in the glass, the melter would contain well over 100 kg of plutonium at any one time. In addition, the DWPF is an unstirred
melter. In such a melter, the potential exists for several problems such as multiple-glass-phase separations, glass-impurity separations, physical stratification (within the glass, as well as heavy material collecting by precipitation at the bottom), and so on—though it is expected that convection in the molten glass will provide significant mixing. Although some information exists on stratification of other heavy species, the panel learned that almost nothing is now known about potential plutonium stratification in any of the melter designs now under active exploration or already in use. Finally, the DWPF is also a "siphon-pour" melter, rather than a "bottom-pour" melter: the glass leaves the melter through a pipe roughly a meter above the bottom of the melter, rather than through a hole in the bottom, so that there will be a meter of glass at the bottom within which heavier species such as plutonium might build up and be drawn off only partially. In short, it appears that the current DWPF melter would not be appropriate for the plutonium vitrification mission.
Each DWPF melter, however, will have a useful life of only a few years and must then be replaced. A second melter similar to the first has been built for this purpose. If a requirement were developed to design a subsequent melter (the third or fourth for DWPF) for vitrifying plutonium, the panel believes that such a design would be feasible. The issues in developing such a design are technically challenging and have only begun to be explored, but there do not appear to be any inherent technical obstacles.8
A melter designed for WPu vitrification would probably be smaller (perhaps with more than one melter operating in parallel) and designed specifically to ensure criticality safety (perhaps with an "inherently safe" geometry that would prevent criticality). A WPu vitrification melter might also incorporate stirring of the molten glass. Stirred melters would be expected to have less serious difficulty with criticality via stratification than unstirred melters. The common practice in criticality assurance, however, is to avoid relying principally on mechanical devices such as the mechanical stirring itself as the means for avoiding criticality; indeed, if a mechanical stirrer that could fail were the only barrier to criticality in the melter, the design would not provide sufficient assurance of criticality safety, and it is the panel's judgment that it would not pass regulatory review.9 Stirred-melter technology is relatively new, and several issues require additional work before it could be proposed for actual production use for WPu vitrification. Small stirred melters, which have been demonstrated at smaller scales under test conditions (although not with plutonium), can operate in a glove-box-type environment with smaller throughputs, about 30 kg/h compared to about 100 kg/h for the larger first-generation DWPF melter. One
important advantage for plutonium vitrification would be that, if plutonium is to be incorporated in the glass without any other radioactive species in an interim step, a glove-box-type operation could be used without the complication of the remote-handling equipment that makes the DWPF operation so complex.
The stirred-unstirred trade-off issues (and the large-or-small melter tradeoff issues) are complex and require further investigation, but there is a basis for confidence that some form of melter technology can be made available that can provide options for critically safe vitrification of WPu. Four small melters in parallel would provide the capacity of the proposed large melter, would probably be cheaper and quicker to develop and perfect, and might, by virtue of modularity, be less costly in operation. We recommend, therefore, that small melters be pursued for future replacement of the large melter now in place at DWPF, particularly if plutonium vitrification is to be carried out in that facility.
Another criticality concern arises in the system that filters the gases released from the melter—known as the "offgas" system—where the potential exists for the accumulation of plutonium in the ducting, filters, scrub solution, etc. The volume and character of the offgas depend critically on whether the melter feed is dry or a slurry: with dry feed the offgas volume is much lower and the problem therefore likely to be easier to resolve (McKibben et al. 1993). The design specification for the melter offgas system for DWPF, assuming the current plan for a DWPF campaign to vitrify HLW (but not plutonium), has concentrated on Cs-137 in the offgas because it is believed by the designers to be the most difficult species that drives the design (Sullivan 1993). The panel believes that a major design and testing effort for a modified offgas system is likely to be required before plutonium vitrification could proceed. This will include developing information on plutonium volatility as a function of various parameters.
As of today, no waste-form criterion or standard exists that would govern the properties of a plutonium-laden glass destined for deep geologic disposal in the proposed Yucca Mountain repository.10 Some (interim) DOE agreements exist between different DOE offices, but the panel believes that nothing now exists that is well founded on a thorough examination of all of the issues. Several issues are involved, including waste leachability aspects, waste physical form, heat generation rates, chemical properties of the waste, and criticality issues.
All of these issues must be, and are being, addressed for the spent fuel that would be the Yucca Mountain repository's principal waste form, including criticality. In that sense, there is nothing unique about plutonium-laden glass. But
plutonium-laden glass has never previously been studied. Serious criticality issues could arise for plutonium-laden borosilicate glass, in part because the boron—needed for criticality control in the glass—leaches out much more rapidly than the plutonium (see Plodinec and Wiley 1979, Bibler et al. 1985), leaving significant potential for plutonium criticality in the waste over subsequent millennia even for modest plutonium loadings in the glass.11 The idea of mixing into the glass an effective neutron absorber with a very low leach rate, such as gadolinium, to assure criticality control even in the absence of boron, has been proposed but not yet explored in detail (McKibben 1993, Omberg 1993, Simonson et al. 1994). The repository criticality issue is described in more detail in Chapter 6.
The panel believes that other glass leachability issues besides criticality are also important and require research. As mentioned above, it is likely that the migration of plutonium from the glass in the repository into underground water is solubility-limited for both steps of what is thought to be a two-step process: first, plutonium leaching from the glass to form a separate precipitate, and second, the dissolution of that precipitate in the groundwater. The various rates are not yet well enough understood, however. The migration of plutonium into the environment may not be a problem, for example, if the overall transport time is longer than several times the 24,000-year half-life of Pu-239.
If the glass is loaded with other radioactive species (HLW or Cs-137, for example), the process governing the leachability of each important species needs to be understood. The panel believes that there may be glass-design approaches, to modify these leach rates if necessary, that remain unexplored. (The answers to these questions will vary, perhaps significantly, depending on the specific glass composition.) As noted in Chapter 6, there are reasons to believe that the addition of plutonium, which appears to act chemically to strengthen the chemical network of the glass, may even modestly improve glass performance in containing other radioactive species.
Questions exist as to the design and performance specifications for the outside canister that would surround the glass in the repository. Because consideration of WPu-laden glass is only recent, no final canister design has been specified yet. Indeed, the standards that such a canister must meet are also in a state of flux now due to the current reevaluation of the Environmental Protection Agency (EPA) standards for Yucca Mountain (see Section 801 of the Energy Policy Act of 1992).
An issue that has not yet been sufficiently explored is whether the alpha-loading of heavily plutonium-laden glass might compromise the glass's physical
integrity: helium from alpha decay can, at sufficient concentrations, cause internal pressure buildup, cracking, physical expansion, etc. Alpha-loading experiments to date (Weber 1991) seem to be limited to alpha-loadings corresponding to plutonium loadings only in the 0.1-percent range—an order of magnitude or more below the range of interest for WPu disposition. Even if alpha-loading cracking occurs for plutonium loadings of interest, it is not known whether this issue will matter if the glass is destined for a deep geological repository like Yucca Mountain. Plutonium migration from the glass to surrounding groundwater is likely to be solubility-limited, in which case the additional surface area provided by severely cracked glass may not significantly affect the total plutonium leaching rate into the water (see discussion below). Migration of other radioactive species might be affected, however. Some dose-rate effects of the alpha-loading may also exist and would require exploration.
If it were ultimately determined that substantial modifications to the glass composition were required for a WPu-HLW glass to be an acceptable waste form for geologic disposal—which the panel does not believe is likely— modifying the vitrification process could require unknown extra resources and time. Even with approximately the current glass composition, the addition of substantial quantities of plutonium (and possibly new neutron absorbers) will inevitably require laboratory tests and engineering analyses to validate the glass's repository performance.
Applicability to Other Forms of Plutonium
Both the United States and Russia face severe problems with plutonium in forms such as scrap and residues that are not suitable for either long-term storage or transport to appropriate processing sites. Many tons of plutonium exist in these forms. In addition to handling excess pits from dismantled weapons, vitrification may have an important role to play in immobilizing these unstable forms of plutonium for disposal. Small melters that could be set up onsite to vitrify these scraps and residues—and thereby both stabilize them to reduce the hazards of near-term storage and prepare them for ultimate disposal—deserve consideration. Once vitrified onsite, this plutonium would be in a form sufficiently stable to ship elsewhere for revitrification with fission products if desired.
ASSESSMENT BY KEY CRITERIA
Facility Options and Schedule
As noted elsewhere in this report, the schedule for WPu disposition— including both when a campaign could begin and when it could be finished—is a major component of the security criterion for comparing options. In the case
of WPu vitrification, the critical-path pacing items are likely to be (1) the design and its confirmatory testing, (2) the regulatory and political approval for the pretreatment and melter steps, (3) obtaining government financial support from Congress, and (4) construction. Although obtaining regulatory and political approvals could delay everything significantly, the major outstanding technical issues should be susceptible to resolution within a few years.
The likely schedule for vitrification cannot be assessed in more detail without simultaneously considering what facilities would be employed to carry out the mission. There are two broad options in this regard: WPu vitrification could rely on modification of vitrification facilities already in existence or planned, or new facilities could be built specifically for WPu vitrification. In the latter case, the facilities might still take advantage of existing infrastructure, for example by incorporating the melter within an existing remote-processing facility (such as the reprocessing canyons or vitrification buildings at Savannah River or Hanford). In both the modification and new construction cases, a complex approval and licensing process (including analyses under the National Environmental Policy Act) will be required that will involve delays and uncertainties of unquantifiable magnitude.
Two vitrification facilities exist in the United States today, both operated by DOE, one at Savannah River, South Carolina, and one at West Valley, New York. A third large DOE facility is planned at Hanford, Washington, but has been put on hold pending a reevaluation of previous plans. We discuss below each of these facilities, as well as foreign vitrification facilities, and consider how the various choices would affect the schedule for WPu vitrification.
Savannah River Facility
Vitrification technology is highly developed at DOE's Savannah River Site, and an advanced facility, the Defense Waste Processing Facility, has been under construction for several years there with the goal of vitrifying much of the high-level radioactive waste in the 30 million gallons currently stored there in tanks (McKibben et al. 1993). The plan includes a complex pretreatment process for extracting the bulk of the long-lived radionuclides from this waste for vitrification. The HLW logs from the DWPF will be stored onsite pending ultimate disposal at Yucca Mountain or another DOE HLW disposal site. The remaining LLW waste from the pretreatment process is to be disposed of in near-surface LLW disposal facilities onsite at SRS.
SRS has a full suite of special facilities capable (in some cases with significant modifications) of carrying out the many other steps in an overall WPu vitrification process, including facilities for storing WPu pits, carrying out the various pretreatment steps, and downstream storage and handling of the glass product canisters.
In the DWPF process as currently planned, the pretreated intensely radioactive waste is fed to the DWPF melter as a solid-liquid mix. The melter's product is a glass-filled canister containing about 600 liters (about 1,700 kg) of borosilicate glass, and weighing just over 2 tons (about 80 percent glass, the rest being the weight of the stainless steel canister itself), standing about 3 m tall and with a 60-cm diameter.
The melter capacity is about 100 kg/h of glass, so the melter can fill one canister in about 17 hours. The plans call for HLW loading in the range of about 20 percent by weight in the glass. Approximately 6,100 glass logs are scheduled to be produced (McKibben et al. 1993). At 17 hours/canister, the entire campaign would take about 15 years if it operated nearly 24 hours a day with no problems or interruptions. However, a longer campaign over two to three decades is currently planned.
The existing waste contains small amounts of plutonium, which will amount on average to about 0.25 kg/canister, or about 0.015 percent of the glass by weight. To add 50,000 kg of WPu to one-third of the canisters in the currently planned campaign would require about 23 kg of WPu per canister, which is about 1.3 percent additional plutonium by weight added to the glass. This appears to be within existing technical limits, as described above.
The existing DWPF has suffered years of delays and substantial cost increases, and is still in its final stages of construction and commissioning. As of early 1993, the planned startup date was in 1994 (GAO 1992). At that time, an accidental flooding of the melter led to a prolonged assessment of plans and operations. The current schedule calls for full-scale production with HLW in 1996, with a campaign lasting until 2018. The project has been held up principally due to technical concerns about the upstream preprocessing of the wastes. Current practices at Savannah River, in the panel's view, indicate a lack of institutional commitment to making progress towards operation. Furthermore, the current operational design and training have indicated potentially serious weaknesses (such as the flooding of the melter just mentioned). Thus some skepticism is in order about the program's ability to meet requirements, even in the absence of WPu.
Experts from SRS have quoted a time for startup of the DWPF in a plutonium vitrification mode as about mid-2005 (McKibben and Wicks 1993). Given that two years have passed since this estimate was made without major activities being undertaken, this estimate would presumably slip to at least 2006. This estimate was made when the planned DWPF startup was still in 1994, but the initial steps are primarily analysis, design, and regulatory ones in which the status of the DWPF would not play a major role. If DWPF was still not operating in the mid-2000s, however, a WPu disposition mission relying on that facility could be delayed.
Based on an assumed 1.3-percent plutonium loading by weight, the plutonium part of the campaign would take eight years (through 2013), and would
not require any additional logs to be vitrified—that is, the nominal 50 tons of WPu would go into the logs currently planned for the eight years between 2005 and 2013 without any new logs being required. Of course, if one could add as much as 4 percent WPu by weight to each canister, it would require only about 700 canisters to cope with the nominal 50 tons surplus WPu. At this higher loading, the WPu part of the campaign could be concluded in less than two years from when it began.
These estimates are those of the SRS team, and assume that vitrification would not proceed until the issues of melter and repository criticality, upstream processing, and offgas system modification had been satisfactorily resolved. The panel has not had either the time or the resources to develop independent estimates. The SRS estimates appear reasonable, assuming a high-priority decision to accomplish the mission, and assuming that the DWPF program progresses without major problems. The estimates have not been examined either for innovative ways to accomplish particular steps more quickly or to root out excessive optimism. They must therefore be considered uncertain. As with other disposition options, achieving success in a timely way would require overcoming the institutional barriers noted at the outset of the chapter, which would require high-level DOE commitment.
West Valley Facility
In West Valley, New York, the site of the only commercial reactor-fuel-reprocessing facility that ever operated in the United States, there is a vitrification facility under construction called the West Valley Demonstration Plant (WVDP). It has been developed by DOE with the goal of vitrifying over 2,000 cubic meters of HLW into glass canisters about the same size as those from Savannah River's DWPF (Jain and Barnes 1993). The waste was produced about two decades ago when the fuel-reprocessing plant operated, and is mostly liquid fission-product waste in alkaline solution. The schedule calls for a startup date slightly later than that at Savannah River, and the planned vitrification campaign will produce about 300 canisters in an 18- to 24-month period. The WVDP melter design has about half of the capacity of the DWPF melter and its technology is similar.
It would be possible to vitrify some part of the WPu at West Valley with the fission products located there. These fission products would be adequate for only a small fraction of the plutonium, however. If all of the excess WPu, or a substantial fraction of it, were to be vitrified at West Valley, fission products would have to be imported to that site—such as the encapsulated Cs-137 now at Hanford. In this case, either the upstream plutonium processing steps would have to be accomplished at other sites, with the resulting plutonium form shipped to West Valley, or the upstream plutonium-handling and plutonium-processing facilities that already exist at sites such as Hanford and SRS would
have to be duplicated at West Valley, with the attendant costs. Since this site is less prepared for this option than Savannah River, the overall schedule would presumably be somewhat longer. Moreover, the involved local community, as well as the West Valley operators, believe the site will only be used for vitrification of local waste. The site has been the subject of considerable local concern, and there has been no mention by DOE of the possibility of using the West Valley plant for WPu vitrification.
Hanford Waste Vitrification Project
DOE had planned a major vitrification facility at Hanford, Washington, known as the Hanford Waste Vitrification Project (HWVP), with a startup date in 1998, but the project has been put on hold pending a review of the plans.12 The plans for this facility would have employed technologies very similar to those in Savannah River's DWPF, and is not discussed further here. Clearly, if this facility is ever built it could also be used for vitrifying WPu, if a decision to do so were taken with enough lead time. In fact, because of the deferral, it might be easier and cheaper (although probably not faster) to design a WPu capability into this facility from the start than to adapt the already built DWPF. The schedule for beginning vitrification of WPu at HWVP is indeterminate, given that it is as yet unknown when HWVP will begin operations. If HWVP could be designed from the outset for WPu, and began operations in the mid-2000s time frame envisioned for the start of WPu vitrification at Savannah River, waiting for HWVP might impose only a modest delay on initiating WPu vitrification operations.
Foreign Vitrification Facilities
Significant vitrification capability exists abroad (Odell 1992), the most impressive being that in France. For decades, the French have carried out research into various glass technologies, and at Marcoule they have operated a sophisticated facility since 1978. A second and more advanced French facility at Cap de la Hague started operations in 1989. In Belgium, a facility at Mol that is jointly supported by the German government operated from 1985-1991, and is now in the process of upgrading. In 1987, a large facility began operations with phosphate glass at Chelyabinsk in the Soviet Union (now Russia), and a British facility with technology similar to the advanced French approach started up at Windscale in 1990. A facility in Tarapur, India, has also operated for some years.
These are the only facilities worldwide that have significant vitrification capacity, although several research-scale operations exist in various institutions, and plans for large facilities are now developing in a few other countries, most notably in Japan and China.
These foreign processes differ in their technical approaches: for example, some use calcining in the pretreatment stage while others do not. The melter in Japan operates on porous-glass preforms saturated with fission-product solution rather than on glass frit. The waste streams feeding into the melters differ as well, some vitrifying fuel-reprocessing liquid wastes, some military-processing wastes, and some solid wastes. Much of the glass product from these foreign facilities is intended for permanent (or at least very long-term) onsite storage, usually with either forced- or natural-convection air cooling.
Taken all together, the various plants have a very large capability—several times greater than that of Savannah River's DWPF—that is diverse in both technology and operating philosophy (Odell 1992). Almost all of this foreign capability would be technically suitable or adaptable for vitrifying WPu in one form or another, if institutional barriers could be overcome. The schedule to do so would be dependent both on technical issues similar to those described above for U.S. facilities, and the institutional issues involved in shipping WPu overseas—and would also require convincing major reprocessors whose livelihoods depend on separating plutonium from fission products that the reverse operation should be performed on excess WPu.
As disposition of Russian WPu is a major issue, the Russian vitrification operation is of particular interest. A waste vitrification facility with a nominal output of 1 ton of glass per day is in operation at the Chelyabinsk-65 site in Russia and, by September 1993, was reported to have processed 150 million curies of radioactive waste, at a loading of between 150,000 and 200,000 curies per ton. The glass produced has somewhat higher loadings of radioactivity than are planned at Savannah River. Nearly 700 million curies of HLW remain in waste tanks at this site, similar to the holdings at Savannah River and somewhat more than the amount at Hanford.13 As noted earlier, the phosphate-glass composition employed at this facility is less appropriate for WPu disposition than borosilicate glass. Alternate melters could be used at this facility, however, to produce borosilicate glass if a decision were taken to do so.
Some of the small melters developed in the U.S. vitrification program, in particular, are relatively low in cost and transportable, and could therefore be shipped to Russia for a vitrification campaign there, if modification of existing Russian melters proved too costly. Russia has operational remote-handling facilities that could be used to operate such melters while incorporating HLW or cesium capsules in the product to create a radioactive barrier. Such small
melters could be used to produce either small glass logs (which would pose a somewhat lower barrier to theft) or large glass logs like those produced in larger melters. The net cost of this approach depends on whether it is seen as an alternate way of handling the HLW vitrification campaigns already planned (in which case much of the cost might be offset by reductions in other vitrification costs) or as a separate campaign for disposing of WPu. Assuming the former case, the basic contributors to cost would be similar to those in the United States, though in Russia at present, both capital and labor costs are substantially lower than in the United States.
In general, Russian authorities have objected to WPu disposition options that would “throw away" the plutonium without generating electricity. Given the environmental legacy of past handling of plutonium and the widespread public distrust of government safety assurances, moreover, gaining public acceptance and licenses for a plan to bury plutonium in a repository in Russia might be difficult. The Russian Ministry of Atomic Energy (MINATOM) itself has recently emphasized the environmental dangers of burying long-lived actinides such as plutonium, as part of its advocacy of a closed fuel cycle in which plutonium would be reprocessed and reused. As in the case of spent fuel, however, the ease of storing and safeguarding the vitrified logs would make it possible for Russia to defer decisions on committing them to geologic disposal for a substantial period.
Specially Constructed WPu Vitrification Facilities
It is also possible to construct a new, dedicated facility for WPu vitrification, rather than using existing facilities. Given the complexities of systems capable of handling both fission products and plutonium, a decision to construct a new facility would be expected to mean a considerable stretch-out of the schedule. It is probable that even with a major commitment, the design and approval phase could take 4-6 years longer than using the DWPF. This approach would allow an optimization of the facility for this application, however, and would impose less disruption on ongoing HLW disposal programs.
Although construction of a special-purpose facility might simplify the task of vitrifying the plutonium, the total costs would be higher, because all the costs of production, handling, and disposal of this waste form (including the potentially substantial costs of providing and operating facilities capable of handling the highly radioactive materials that might be added to it) would have to be charged to the plutonium disposition mission, rather than only the net additional costs of adding plutonium to a previously planned HLW vitrification campaign.
Careful study is required, however, of how much the costs and delays of this approach would exceed those of modifying an existing or planned facility. The modifications to existing facilities needed for WPu vitrification may be substantial. In the case of the DWPF, for example, necessary modifications
would probably include a new melter; a modified offgas system; modified, critically safe systems for feeding plutonium into the melter; and installation of a complete safeguards and security system (there is currently no safeguards system for the Savannah River waste operations, since they are not handling material of proliferation concern).
Costs and delays involved in building a new facility could be reduced by making use of existing remote-handling facilities. Given the physical arrangements at DWPF, for example, it seems quite possible that a small additional melter could be built within the same building; alternatively, this could be done at the building that will house the HWVP, or in the reprocessing canyons at SRS or Hanford. Such a separate melter could vitrify plutonium with HLW, or the Cs-137 stored at Hanford could be used to provide a comparable radiation barrier. If an independent melting operation was going to be undertaken in any case, using this Cs-137 might be simpler in important respects, given the chemical complexities of the wastes at both Hanford and Savannah River.
It should be noted, however, that producing additional logs beyond those already planned would involve additional costs. All of the planned capacity in the Yucca Mountain repository will be filled by wastes already scheduled to be produced. Therefore production of additional waste products specifically for WPu disposition (rather than piggy-backing on planned HLW vitrification campaigns) would require either displacing other wastes now scheduled to go into Yucca Mountain, expanding that repository's capacity, or waiting for an indeterminate time until a second repository became available. A significant cost is associated with the disposal of each additional log that would be produced. (The same is true for spent fuel, if the reactor used for plutonium disposition would not otherwise have operated and produced this waste.)
In a new facility, the time for the full campaign of vitrifying 50 tons of WPu would depend on choices as to the capacity of the new facility. Such a facility would presumably be designed to accomplish the campaign expeditiously; even a 1- to 2-year campaign time could be fully feasible if desired.
Vitrification Without Fission Products
If WPu were to be vitrified without fission products—as an interim step before later "revitrification" as described above—a remote-handling facility would not be necessary, and the schedule for this initial step might be significantly compressed. The Savannah River experts estimated that small glove-box melters could begin production-scale operations with WPu glass in roughly nine years (most of that time being involved in approvals, installation of equipment, and preparing for large-scale conversion of various forms of plutonium to other forms). In particular, the small-stirred melter technology might be particularly attractive for this approach (McKibben et al. 1993). One such advanced melter is now onsite at SRS. This new design is smaller and less expensive, has a
throughput of about one-fourth to one-third of the larger unstirred DWPF melter capacity, and could be deployed relatively quickly, especially for use in vitrifying WPu alone without spiking. Also, its smaller size means that constructing additional similar melters could be feasible for accelerating the WPu vitrification throughput, either at SRS or elsewhere.
Without fission products, a remote-handling facility would not be required. This would eliminate a number of time-consuming and expensive steps in the process, such as designing the optimal scheme for the pretreatment of the mix of WPu with the other (highly radioactive) contaminants, and would simplify safety analyses.
The panel believes that for the vast majority of the WPu that is in the form of pits, there is little to be gained in terms of security from vitrification without fission products. A small fraction (but still a significant absolute amount) of the WPu in both the United States and Russia exists in other forms, however, such as partially machined pits, metal scraps, and so on. For this material, there would likely be an increase in both its safety and security if it were vitrified even without fission products, because it would all be converted to a physical form that can be easily accounted for and safely stored. Whether these advantages outweigh the financial costs, the administrative burdens, and the ES&H impacts of such an operation is both a technical and a policy question that the panel recommends should be seriously examined. Vitrifying the WPu pits themselves without fission products does not seem to the panel to be worthwhile; however, neither the security nor the safety benefits seem to be substantial enough to justify the costs and risks, except perhaps as a step in the process of vitrifying the material in a form incorporating fission products.
Late in its deliberations, the panel learned that options are being considered to vitrify actinides now stored in solution at the F-canyon at Savannah River, as part of the clean-out of that facility. Preliminary estimates suggest that the WPu stored in solution there (as well as solutions of some other actinides) could be vitrified in a glass comparable to those discussed here for only modest additional cost compared to the investment that must be made to clean out this facility in any case. This would offer the opportunity for an early production-scale demonstration of WPu vitrification. Such a demonstration would be very valuable, as the lack of such a clear technology demonstration is currently one of the weaknesses of the vitrification approach when compared to the MOX options. The panel therefore believes that the possible synergy of combining clean-out of the F-canyon solutions with an early demonstration of WPu vitrification should be seriously examined, and, if the approach is found to be technically and economically viable, seriously evaluated in the context of the overall WPu effort and, if appropriate, implemented. The panel also believes that the feasibility of incorporating fission products in this demonstration should also be explored in order to achieve a full demonstration of the entire approach described in this chapter at an early date.
Approvals and Licenses
As noted above, gaining regulatory approval for the various plutonium processes required for vitrification would be a major and uncertain component of the schedule for carrying out vitrification operations. (This component is included in the estimates quoted above, with the assumption that a high national priority was assigned to accomplishing the WPu disposition mission.) Certifying the safety of the additional processes needed to add plutonium to currently scheduled HLW vitrification campaigns would take several years. Careful attention would have to be paid to melter design to ensure against criticality and to the offgas system that must prevent release of plutonium into the environment and accumulation of plutonium within the offgas system itself. These engineering issues, while challenging, appear resolvable. Gaining public acceptance at the relevant sites may be more difficult, but if (1) the public is included in the decision-making process, (2) the association with arms reductions is made clear, and (3) a plausible case can be made that once processed, the plutonium will eventually be shipped elsewhere for burial in a geologic repository, then public approval should be achievable. Overall, licensing and approval for this approach would probably be easier than for MOX, at least in the United States. Siting approval and licensing for a vitrification facility dedicated solely to plutonium disposition would probably be more protracted than for an approach piggybacking on already scheduled HLW vitrification campaigns.
Certification of the plutonium-bearing glass as a suitable waste form for emplacement in a geological repository, including resolution of the long-term criticality issue, would be the highest hurdle.
Safeguards, Security, and Recoverability
As noted earlier, the difficulty of extracting plutonium from the glass logs would be generally comparable to the difficulty of extracting plutonium from spent fuel, given the intensity of the radiation fields with which anyone handling the logs would have to cope.
As for the opportunities for diversion or theft of the materials, it is important that all necessary plutonium operations for the vitrification option—both pit processing and production of the plutonium-bearing glass—could be carried out at a single nuclear weapons complex site with extensive safeguards and security. Thus the number of required transportation and storage steps, and the associated opportunities for theft, would be less than in those reactor options requiring more than one site.
Fabrication of HLW logs would also be easier to safeguard than fabrication of MOX fuel bundles (Shea 1993). Monitors would have to confirm only the single step of mixing the plutonium with the HLW. Once that step had taken place, the plutonium would be in an intensely radioactive mix and very difficult to divert. There would be no capability within the vitrification facility for re-
separating the plutonium from the HLW. MOX fabrication, by contrast, requires many steps involving large-scale bulk handling of plutonium with inherent accounting uncertainties, and at each step of the process the plutonium remains in a form from which it could be readily reseparated.
For the glass operation, however, once the plutonium had been mixed with the HLW and incorporated in glass, the very high radioactivity and strong neutron absorption of the glass log would make accurate nondestructive assays of the amount of plutonium in the glass difficult. Thus, the traditional material-accounting approach of detailed measurement of the inputs and outputs of the plant might have to be modified, with safeguards relying more on confirming that the plutonium was mixed with HLW, and on containment, surveillance, and security measures to ensure that no plutonium was removed from the processing area or from the site without authorization. Although this would be an engineering challenge, adequate technologies exist to safeguard the glass production process, particularly given the relative simplicity of safeguarding the glass production process as compared with safeguarding the MOX fabrication process.
As vitrification operations do not normally include fissile materials, the types of security required for handling such materials have not to date been provided for facilities such as the DWPF. Setting up the requisite security system and procedures would be one of the significant modifications required for the DWPF if WPu were to be vitrified there. Once the logs had been produced, they could be stored and safeguarded relatively cheaply until repositories were ready to accept them, in facilities already planned, just as in the case of spent fuel.
Indirect Impact on Civilian Fuel-Cycle Risks
Treating pure weapons-grade plutonium as a waste to be disposed of would support the present U.S. administration's policy of generally discouraging the use of separated plutonium reactor fuels.
The schedule and cost for achieving large-scale WPu vitrification depends on a wide range of factors that are not yet known in detail, including such fundamental matters as the facility to be used, the number of logs to be produced, and the like. Only the roughest estimates are available at present. For those options that would incorporate WPu in HLW glass that would be produced in any case, it is important to focus on the net additional cost of adding the WPu, as in those cases the total cost of the vitrification operation cannot be charged to the WPu disposition mission. For all cases, it is important to separate the various preprocessing costs before vitrification begins from the costs of vitrification itself, as these preprocessing costs are similar to those that must be borne by other options as well.
Vitrification costs will depend on several issues, including: (1) whether existing facilities can be used in whole or in part, including taking advantage of existing facilities even if a new vitrification facility were to be built; (2) whether the duration of the WPu vitrification campaign can efficiently use the chosen facilities; and (3) whether the use of these facilities displaces other necessary or desirable activities. Also, the WPu vitrification cost will depend almost inversely proportionally on how much WPu loading by weight can be safely and economically added to the vitrified glass, unless the WPu campaign fits well into a previously planned campaign to vitrify HLW, such as the currently planned DWPF campaign to vitrify HLW at SRS.
The only detailed cost estimates that have been available to the panel were prepared by Westinghouse Savannah River Company, for vitrification at Savannah River Site.14 The estimated cost for vitrification with HLW in the DWPF is approximately $600 million, plus approximately $400 million to carry out the preliminary steps, including pit processing (which would also be required for the reactor options). The SRS team estimates the cost of vitrification without HLW at less than $200 million (plus the same $400 million preprocessing costs). These estimates are uncertain by at least a factor of two. As noted above, the cost of a separate plutonium vitrification campaign that incorporated radioactive materials such as Cs-137 would be higher, because the high costs of processing highly radioactive glass would then have to be borne entirely by the WPu disposition mission, rather than being shared by HLW disposal operations already planned.
These rough cost estimates are based on carrying out the vitrification at SRS. Certain economies could be realized by designing for WPu vitrification from the beginning in the currently deferred HWVP, but it is not possible now to estimate these very well.
Extensive engineering effort has been necessary to assure that the DWPF at Savannah River, the most advanced vitrification facility that now exists in the United States, can meet all applicable environmental, safety, and health regulations. Although it will be a challenge to provide a comparable level of assurance for a facility to vitrify WPu (whether a modified DWPF or another facility), achieving adequate compliance should be within current technological capabilities. Neither cost nor schedule difficulties should be affected overwhelmingly by problems in these areas.
McKibben et al. (1993). This is an undiscounted estimate; discounting by 7 percent per year (see Chapter 3) would reduce the billion-dollar figure by roughly half, for comparison to other options. These estimates also include previtrification in a plutonium-only glass; eliminating this step would probably lower costs somewhat.
One benefit of the vitrification process is that it can accept either WPu or other plutonium forms, and can immobilize essentially all of the plutonium feedstock, so that the need for handling or disposing of subsequent plutonium-contaminated radioactive waste beyond the WPu-laden glass itself can be substantially reduced. (Waste from production is recycled into the melter to produce new glass.) Of course, some plutonium-contaminated waste streams, including contaminated equipment, will require subsequent LLW handling and disposal. These waste streams (besides the glass product itself) should be manageable within applicable regulations. These ES&H issues are addressed in more detail in Chapter 6.
Argonne 1994: Personal communication of Argonne National Laboratory researchers to Dr. Frank von Hippel, 1994.
Bibler et al. 1985: N.E. Bibler, G.G. Wicks, and V.O. Oversby. "Leaching of SRP Nuclear Waste Glass in a Saturated Tuff Environment." Scientific Basis for Nuclear Waste Management VIII 44:247-256, Materials Research Society, 1985.
Bradley 1992: Donald J. Bradley. "Radioactive Waste Management in the Former USSR," Vol. III. PNL-8074. Hanford, Wash.: Pacific Northwest Laboratory, June 1992.
Bradley 1993: Donald Bradley, Pacific Northwest Laboratory. Personal communication, October 1993.
Diakov 1992: Anatoli S. Diakov. "Imbedding Russian Weapon-Grade Plutonium Into High-Level-Waste Glass: Technical Considerations." Paper presented at International Workshop on the Future of Reprocessing, and Arrangements for the Storage and Disposition of Already-Separated Plutonium, Moscow, December 14-16, 1992.
Forsberg et al. 1994: C.W. Forsberg et al. "Direct Conversion of Radioactive and Chemical Waste Containing Metal, Ceramics, Amorphous Solids, and Organics to Glass." Paper presented to "Spectrum '94: Nuclear and Hazardous Waste Management International Topical Meeting," August 14-18, 1994.
GAO 1992: General Accounting Office. Nuclear Waste: Defense Waste Processing Facility-Cost, Schedule, and Technical Issues. GAO/RCED-92-183. Washington, D.C.: General Accounting Office, June 1992.
Gray 1994: Leonard Gray, Lawrence Livermore National Laboratory. Personal communication, 1994.
Jain and Barnes 1993: V. Jain and S.M. Barnes. "Radioactive Waste Solidification at the West Valley Demonstration Project." In "Proceedings of the Third International Conference on Advances in Fusion and Processing of Glass," Ceramic Transactions 29, American Ceramics Society, 1993.
Lyman 1994: Edwin S. Lyman. Interim Storage Matrices for Excess Plutonium: Approaching the "Spent Fuel Standard" Without the Use of Reactors. PU/CEES Report No. 286. Princeton, N.J.: Center for Energy and Environmental Studies, Princeton University, August 1994
Makhijani and Makhijani 1994: Arjun Makhijani and Annie Makhijani. Fissile Materials in a Glass, Darkly: Technical and Policy Aspects of the Disposition of Plutonium and Highly Enriched Uranium. Takoma Park, Md.: Institute for Energy and Environmental Research, November 1994.
Marples 1988: J.A.C. Marples. "The Preparation, Properties, and Disposal of Vitrified High-Level Waste From Nuclear Fuel Reprocessing." Glass Technology 29(6):230, 1988.
McKibben 1993: J.M. McKibben, Westinghouse Savannah River Company. Personal communication, 1993.
McKibben and Wicks 1993: J.M. McKibben and G. Wicks, Westinghouse Savannah River Company. Personal communication, 1993.
McKibben and Wicks 1994: J.M. McKibben and G. Wicks, Westinghouse Savannah River Company. Personal communication, 1994.
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," Report WSRC-RP-93-755. Aiken, S.C.: Westinghouse Savannah River Company, 1993.
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 (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.
Odell 1992: M. Odell. "Vitrification—World Review." Nuclear Engineering International, pp. 51-53, June 1992.
OTA 1991: Office of Technology Assessment. Long-Lived Legacy: Managing High-Level and Transuranic Waste at the DOE Nuclear Weapons Complex. Washington, D.C.: U.S. Government Printing Office, May 1991.
Omberg 1993: R. Omberg, Westinghouse Hanford Company. Personal communication, 1993.
Plodinec and Wiley 1979: M.J. Plodinec and J.R. Wiley. "Evaluation of Glass as a Matrix for Solidifying Savannah River Plant Waste: Properties of Glasses Containing Li2O." DP-1498. Aiken, S.C.: Savannah River Laboratory, February 1979 (cited in McKibben et al. 1988).
Shea 1993: Thomas Shea, International Atomic Energy Agency Safeguards Division. Personal communication, August 1993.
Simonson et al. 1994: Scott A. Simonson, Kory Sylvester, and Marvin Miller. Vitrification as an Option for Plutonium Disposition. Annual Report September 1993-August 1994. Cambridge, Mass.: Department of Engineering, Massachusetts Institute of Technology.
Sullivan 1993: K. Sullivan, Westinghouse Savannah River Company. Personal communication, 1993.
Toevs and Trapp 1994: James W. Toevs and T.J. Trapp, Los Alamos National Laboratory. "The Radiation Barrier Alloy, Pit Disassembly, and Impact on Pu Disposition Time Scales." Briefing to Frank von Hippel and Matthew Bunn, July 28, 1994.
USDOE 1993: Washington State Department of Ecology, U.S. Environmental Protection Agency, and U.S. Department of Energy. "Hanford Federal Fa
cility Agreement and Consent Order." Document 89-10. Available from U.S. Department of Energy Richland Operations Office, Richland, Wash.
Weber 1991: W.J. Weber. "The Effect of Radiation on Nuclear Waste Forms." Journal of the Minerals, Metals, and Materials Society 43(7):35-39, 1991.
Westinghouse 1994: Westinghouse Savannah River Site. Briefing to Frank von Hippel and Matthew Bunn, August 18, 1994.
Wicks 1993: G. Wicks, Westinghouse Savannah River Company. Personal communication, 1993.