CHAPTER FIVE—
DEFENSE WASTES

A long-recognized problem of the U.S. nuclear weapons program is the large amount of nuclear waste it has produced at several large sites across the nation. Defense wastes is a category that includes many types of nuclear waste of varying degrees of hazard. The range of defense wastes includes not only spent-fuel rods and contaminated equipment that are analogous to the wastes resulting from civilian light-water power reactors, although usually different chemically, but also a wide variety of waste types that have no counterpart in the present civilian program (e.g., the residues from fabricating plutonium components for nuclear weapons).

THE HANFORD TANKS

The Hanford site that produced weapons plutonium has the largest volume of wastes among U.S. sites and is widely believed to pose serious health and safety problems. Many of the conditions at Hanford can be found in varying degrees at the other material production and processing sites for nuclear weapons systems. A considerable amount of radioactivity remains on the site, more than half of which is in the form of aqueous solutions and solid residues in 177 underground mild steel tanks of high-level waste (HLW). A map of the Hanford site tanks and evaporators is shown in Figure 5-1; the 177 tanks are clustered in various locations within the "200 area" processing complex. The tanks are of two general types: 149 single-shell (SS) tanks constructed over the period 1943 to 1964 and 28 double-shell (DS) tanks built between 1968 and 1986. The latter have two steel walls with a void space between that provides leak detectability and additional containment in the event of a leak of the primary shell. The void space also provides access that permits some nonintrusive monitoring of the tank contents. A cutaway view of a typical SS tank is shown in Figure 5-2. The older SS storage tanks are well beyond their design lives, and 67 of them are assumed to be leaking, which resulted in draining and evaporation of much of the water from the solutions stored in them. The more recently built DS tanks are still sound and none have been found to leak.

Apart from the tank wastes, the second largest source of radioactivity at the Hanford site is the encapsulated radioactive sources. These consist of two types: sealed capsules of 90Sr and 137Cs, and unprocessed irradiated reactor slugs used as fuels from an onsite reactor (the N-reactor). The sealed strontium and cesium capsules are the result of a concentration process in which part of the liquid contents of the tanks was processed and the effluents returned to the tanks. The capsules are in sealed containers stored in deep water-filled pools (basins) that cool and shield the radioactive materials. The slugs are also stored in basins, but some are not well contained. Some additional radioactive residues at Hanford are contained in the form of locally contaminated soils used as waste disposal pits (cribs) over many years of low-level waste (LLW) disposal activities at the site.

This chapter deals only with the more contentious issues surrounding the tank wastes. It examines separations options that could be considered for remediation of these wastes and is meant to be suggestive of issues and solutions to be considered for the clean-up of other defense high-level wastes.

The 66-million-gallon total volume of waste in the Hanford tanks, and the debate over the appropriate treatment for its conversion to a suitable permanent waste form, make Hanford tank waste a primary focus of current DOE remediation efforts. A major concern about the Hanford tanks has been the possibility that chemical instability of the contents of some tanks could conceivably lead to uncontrolled chemical (not nuclear) reactions, with consequent



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 87
Nuclear Wastes: Technologies for Separations and Transmutation CHAPTER FIVE— DEFENSE WASTES A long-recognized problem of the U.S. nuclear weapons program is the large amount of nuclear waste it has produced at several large sites across the nation. Defense wastes is a category that includes many types of nuclear waste of varying degrees of hazard. The range of defense wastes includes not only spent-fuel rods and contaminated equipment that are analogous to the wastes resulting from civilian light-water power reactors, although usually different chemically, but also a wide variety of waste types that have no counterpart in the present civilian program (e.g., the residues from fabricating plutonium components for nuclear weapons). THE HANFORD TANKS The Hanford site that produced weapons plutonium has the largest volume of wastes among U.S. sites and is widely believed to pose serious health and safety problems. Many of the conditions at Hanford can be found in varying degrees at the other material production and processing sites for nuclear weapons systems. A considerable amount of radioactivity remains on the site, more than half of which is in the form of aqueous solutions and solid residues in 177 underground mild steel tanks of high-level waste (HLW). A map of the Hanford site tanks and evaporators is shown in Figure 5-1; the 177 tanks are clustered in various locations within the "200 area" processing complex. The tanks are of two general types: 149 single-shell (SS) tanks constructed over the period 1943 to 1964 and 28 double-shell (DS) tanks built between 1968 and 1986. The latter have two steel walls with a void space between that provides leak detectability and additional containment in the event of a leak of the primary shell. The void space also provides access that permits some nonintrusive monitoring of the tank contents. A cutaway view of a typical SS tank is shown in Figure 5-2. The older SS storage tanks are well beyond their design lives, and 67 of them are assumed to be leaking, which resulted in draining and evaporation of much of the water from the solutions stored in them. The more recently built DS tanks are still sound and none have been found to leak. Apart from the tank wastes, the second largest source of radioactivity at the Hanford site is the encapsulated radioactive sources. These consist of two types: sealed capsules of 90Sr and 137Cs, and unprocessed irradiated reactor slugs used as fuels from an onsite reactor (the N-reactor). The sealed strontium and cesium capsules are the result of a concentration process in which part of the liquid contents of the tanks was processed and the effluents returned to the tanks. The capsules are in sealed containers stored in deep water-filled pools (basins) that cool and shield the radioactive materials. The slugs are also stored in basins, but some are not well contained. Some additional radioactive residues at Hanford are contained in the form of locally contaminated soils used as waste disposal pits (cribs) over many years of low-level waste (LLW) disposal activities at the site. This chapter deals only with the more contentious issues surrounding the tank wastes. It examines separations options that could be considered for remediation of these wastes and is meant to be suggestive of issues and solutions to be considered for the clean-up of other defense high-level wastes. The 66-million-gallon total volume of waste in the Hanford tanks, and the debate over the appropriate treatment for its conversion to a suitable permanent waste form, make Hanford tank waste a primary focus of current DOE remediation efforts. A major concern about the Hanford tanks has been the possibility that chemical instability of the contents of some tanks could conceivably lead to uncontrolled chemical (not nuclear) reactions, with consequent

OCR for page 87
Nuclear Wastes: Technologies for Separations and Transmutation FIGURE 5-1 Hanford site tanks and evaporators. tank rupture and release of radioactivity into the environment. After several years of evaluation and study, there is growing consensus that the likelihood of chemical instability is much less probable than initially believed. As a result, the emphasis is shifting from mitigation—elimination of safety hazards, such as the slight possibility of disruptive chemical reactions—to remediation, the resolution of underlying problems via the removal, separation, and disposal of the radioactive materials in safely storable waste forms. FIGURE 5-2 A cutaway view of a typical single-shell tank. The Hanford tank wastes pose a challenge for separations (and possible transmutation), the study of which may supply useful information for the remediation of other defense sites and wastes. The development and successful execution of procedures for remediation of the Hanford site

OCR for page 87
Nuclear Wastes: Technologies for Separations and Transmutation would be perceived as major progress for the whole defense waste activity and could serve as models for clean-up efforts at the other defense sites in the United States, including the Feed Materials Processing Center (Fernald), Y-12 Plant, Rocky Flats Plant, Mound Laboratory, K-25 Plant, Los Alamos, and the Savannah River Site, among others. (A summary of recent activity and plans at the Savannah River Site is presented in Appendix E.) All of these sites can be remediated, except probably contaminated soils, with known techniques; however, further development of the techniques is desirable to achieve optimum clean-up at a minimum cost in conformity with remediation standards. Inventory and Characterization of Stored Waste The combined waste volume in all Hanford tanks is estimated to be 250,000 cubic meters (66 million gallons), of which about 190,000 cubic meters (50 million gallons) are contained in the SS tanks. These hold approximately 155 megacuries (MCi) of radioactivity, while the DS tanks contain 110 MCi. A summary of the chemical and radionuclide inventory of the tanks obtained from records of past operational data is presented in Table 5-1. At least 99% of the current radioactivity (in curies) is due to 137Cs and 90Sr, even though during the 1967–1984 period, approximately half of the quantities of these fission products in the SS tanks were removed and placed in capsules. Because the Hanford tanks were made from carbon steel rather than stainless steel, the wastes were made strongly alkaline with sodium hydroxide to minimize corrosion, resulting in the high current content of sodium nitrate and sodium nitrite salts in the tanks. There is also a waste component known as ''organic complexant concentrate," which was generated during the processing of many types of onsite residue fractions, as well as some offsite residues. In the DS tanks, this variable-composition fraction constitutes about one-third of the waste volume; the balance comes from both ion-exchange processing and tributyl phosphate (TBP) extractions that were routine production operations at Hanford's Plutonium Finishing Plant. The composition of the material in the Hanford tanks is heterogeneous in all phases, both within a given tank and also among different tanks. A typical tank contains insoluble residues plus solid excess salt, precipitated residues (sludges) at the bottom of the tank, an intermediate layer of residual liquid that is saturated with a variety of salts and contains salt crystals in suspension, and a crust of low-density salts floating on the central liquid layer. (The extent of horizontal heterogeneity within each layer is unknown but can be expected to increase with the viscosity of the tank contents.) These waste materials have been generated over the past 50 years by three different chemical processes (described in Chapter 3) used to recover plutonium from irradiated uranium. Some of the differences in tank composition can be attributed to a change in the target cladding used for the reactor fuel charges, i.e., the change from the aluminum jackets used in the earlier reactors to the zircalloy cladding used for the N-reactor. Two other reasons are the recycling of the wastes from early bismuth phosphate plutonium recovery and purification processes to recover the uranium content, and the Irradiation Source Program that recovered 137Cs and 90Sr from much of the old waste to make intense gamma irradiation sources for industrial uses (such as sterilization of medical supplies and preservation of food for the military). As mentioned earlier, the total inventory (Table 5-1) of significant species in the Hanford underground tanks is fairly well known, although the specific contents of some tanks and their layers are not. The sludges in the SS tanks are known to contain substantial amounts of mineral-like nonradioactive aluminosilicate solids known as cancrinites, presumably formed by reactions of aluminum salts with various silicate residues. Similarly, a significant portion of the insoluble residues in the DS tanks consists of hydrated zirconium oxides, hydroxides, and complex precipitates containing zirconium, silicon, aluminum, iron, and/or other minor elements. The uranium and transuranic (TRU) elements found in the tank wastes typically are precipitated under conditions such as those existing in the tanks, and some radionuclides may have been incorporated into the complex chemical structure of these mineral-like residues formed from nonradioactive elements. The amount of individual radioactive fission products present in the total contents of all the tanks and cribs can be estimated from historical operational power levels of the reactors and the decay-rate constants for each isotopic species. The bulk amount of salts in the tanks can be estimated from records of the volume and composition of acid solution waste discharged to the tanks and the amount of sodium hydroxide added to neutralize the excess acid. The Hanford site's program for core analysis of the tanks has been driven by regulatory and safety concerns. Continued experimental and modeling studies increasingly indicate that there is an extremely low probability of encountering serious chemical instability problems even in those tanks of special concern, i.e., the tanks emitting radiolytic hydrogen and other decomposition products in significant quantities on a periodic basis and those containing potentially energetic compounds in contact with oxidizing salts (the organic compounds and ferricyanide-nitrate salt reaction issue). While sampling of the gas phase above the residues and analysis of one or two cores of residues per tank is useful to satisfy questions relating to possible safety issues, it is of little value in designing chemical remediation processes, particularly if the horizontal heterogeneity is extensive. The information obtained by this core sampling will

OCR for page 87
Nuclear Wastes: Technologies for Separations and Transmutation TABLE 5-1 Summary of the Chemical and Radiochemical Inventory of the Hanford Tanks Material Type Double-Shell Tank Waste Single-Shell Tank Waste Contaminated Soil Total Total Volume (m3) 7.8E+04 1.4E+05 1.0E+05 4.9E+05 Density (g/cc) 1.5 1.6 1.6 1.9 Total Mass (metric tons) 1.2E+05 2.2E+05 1.6E+05 9.2E+05 Chemical Constituents (kg)         Ag 4.4E+02     4.4E+02 Al 2.2E+06 2.2E+06 2.1E+0S 4.6E+06 As 8.4E+02     8.4E+02 B 2.6E+03     2.6E+03 Ba 1.8E+03     1.8E+03 Be 9.9E+00     9.9E+00 Bi   2.6E+05   2.6E+05 Ca 1.7E+04 1.3E+05   1.5E+05 Cd 6.7E+03 4.0E+03   1.1E+04 Ce 3.8E-02 2.3E+05   2.3E+05 Co 8.1E+02     8.1E+02 Cr 6.8E+04 9.6E+04   1.6E+0S Cs 3.9E+02     3.9E+02 Cu 9.8E+02     9.8E+02 Fe 1.0E+05 6.3E+05   7.3E+05 Hg 2.8E+02 1.0E+03   1.3E+03 K 1.4E+06     1.4E+06 Mg 5.4E+03     5.4E+03 Mn 1.4E+04 1.2E+0S   1.3E+0S Mo 8.4E+03     8.4E+03 NH3       0.0E+00 Na 1.7E+07 5.1E+07 1.0E+06 6.9E+07 Nb 5.4E+00     5.4E+00 Ni 1.1E+04 1.8E+05   1.9E+05 Pb 3.7E+03     3.7E+03 Rare Earths 1.2E+04     1.2E+04 Rh 4.4E+02     4.4E+02 Ru 5.5E+02     5.5E+02 Sb 1.9E+03     1.9E+03 Se 1.4E+03     1.4E+03 Si 6.0E+04     6.0E+04 Sm 5.7E+00 2.5E+01 8.3E-03 3.0E+01 Sr 2.5E+02 3.6E+04   3.6E+04 Tc 1.4E+03 9.4E+02 1.8E+01 2.4E+03 Th 8.3E+02     8.3E+02 Ti 4.5E+02     4.5E+02 Zn 2.3E+03     2.3E+03 Zr 3.1E+05 2.5E+05   5.6E+05 Anions         F 3.8E+05 8.1E+05 2.2E+03 1.2E+06 Cl 4.5E+05 4.0E+04   4.9E+05 CO3 1.1E+06 1.7E+06 1.8E+04 2.8E+06 Fe(CN)6 3.5E+03 3.2E+05   3.2E+0S I 1.5E+02 1.4E+02 4.8E+00 2.9E+02

OCR for page 87
Nuclear Wastes: Technologies for Separations and Transmutation Material Type Double-Shell Tank Waste Single-Shell Tank Waste Contaminated Soil Total NO2 5.0E+06 4.8E+06 5.7E+05 1.0E+07 NO3 1.4E+07 9.7E+07 7.9E+05 1.1E+06 PO4 2.2E+05 8.8E+06 7.0E+04 9.0E+06 SO4 3.8E+05 1.6E+06   2.0E+06 OH 1.0E+07 5.3E+06 1.4E+05 1.6E+07 Total Organic Carbon 6. 1E+05 2.0E+05 8.8E+04 9.0E+05 Cancrinite   2.7E+06   2.7E+06 Diatomaceous Earth   3.5E+05   3.5E+05 Portland Cernent/Concrete   5.7E+04   3.9E+08 Soil     1.6E+08 1.6E+08 Water 6.6E+07 4.5E+07 2.1E+06 1.1E+08 Selected Actinides (1cg)         U 3.6E+04 1.4E+06 2.4E+03 1.4E+06 Np 4.2E+01 4.6E+01 1.6E-01 8.8E+01 Pu 1.6E+02 3.8E+02 2.6E-01 5.4E+02 Am 2.9E+01 1.4E+01 4.7E-02 4.2E+01 Cm 1.9E-02 8.6E-04 2.9E-07 2.0E-02 Radionuclides (Ci) Approximate Decay Date: Jan. 1996     3H 5.3E+02     5.3E+02 14C 2.3E+03 3.0E+03 1.0E+02 5.4E+03 60Co 4.7E+03 9.4E+04   9.9E+04 63Ni 2.4E+03 2.9E+05 9.9E+02 2.9E+05 79Se 8.1E+02     8.1E+02 90Sr 1.1E+07 5.0E+07 1.7E+04 6.2E+07 90Y 1.1E+07 5.0E+07 1.7E+04 6.2E+07 93Zr   4.3E+03 7.5E+00 4.3E+03 95Zr 2.6E-04     2.6E-04 95Nb 4.8E-04     4.8E-04 99Tc 2.4E+04 1.6E+04 2.7E+02 4.0E+04 106Rb 1.9E+04 7.4E-01 1.3E-02 1.9E+04 106Ru 1.9E+04 7.4E-01 1.3E-02 1.9E+04 125Sb 6.9E+04     8.9E+04 126Sn 0.0E+00 5.7E+02 4.9E+00 5.7E+02 129I 2.7E+01 2.4E+01 8.2E-01 5.2E+01 135Cs   7.3E+01 1.9E+00 7.5E+01 137Cs 3.0E+07 1.6E+07 4.2E+05 4.7E+07 137mBa 2.9E+07 1.SE+07 3.9E+05 4.4E+07 144Ce 2.1E+03     2.1E+03 147Pm 1.1E+06     1.1E+06 151Sm 1.5E+05 6.5E+05 2.2E+02 8.0E+05 154Eu 5.3E+04     5.3E+04 155Eu 4.3E+04     4.3E+04 226Ra   3.2E-15 1.1E-18 3.2E-15

OCR for page 87
Nuclear Wastes: Technologies for Separations and Transmutation Material Type Double-Shell Tank Waste Single-Shell Tank Waste Contaminated Soil Total 235U 7.0E-02 2.0E+01 3.4E-02 2.0E+01 238U 6.6E-01 4.6E+02 7.9E-01 4.6E+02 237Np 3.0E+01 3.2E+01 1.1 E-01 6.2E+01 238Pu 4.2E+02 4.4E+02 3.0E-01 8.6E+02 239Pu 9.1E+03 2.2E+04 1.SE+01 3.1E+04 240Pu 2.6E+03 5.4E+03 3.7E+00 8.0E+03 241Pu 5.0E+03 4.5E+04 3.1E+01 5.0E+04 241Am 9.9E+04 4.6E+04 1.6E+02 1.SE+05 243Am   1.9E+01 6.5E-02 1.9E+01 243Cm       0.0E+00 244Cm 1.5E+03 6.9E+01 2.4E-02 1.6E+03 a Includes tank structures. NOTE: The absence of a component estimate in the above table should generally be taken to imply the unavailability of information on that component that could be referenced, rather than that the component is not present in waste materials. SOURCE: Boomer et al. (1994) provide only very limited information about each tank's total contents. Performing the sampling and the subsequent material analysis needed for tanks that might present safety hazards would be costly, especially if enough samples were taken to compensate for the heterogeneity in the tanks, and could result in additional radiation exposure to workers. An extensive program of sampling of individual tanks for reasons other than resolution of safety questions is unnecessary, and sampling should be minimized to the amount needed to provide data for satisfactory risk assessment. Remediation Processing Considerations The complexity of remediating the Hanford tank wastes results not only from the variety of wastes generated by different processing techniques and reagents, but also from the thermal heat arising from fission-product decay energy. It would not be technically prudent to begin the processing needed to produce a suitable permanent waste form before the large majority of short-lived fission product isotopes had decayed to either stable end-products or long-lived isotopes that emit little decay heat per unit time. The length of storage of the Hanford wastes to date has been technically beneficial. Any plan for converting the wastes stored in the underground tanks to a suitable permanent waste form should take into consideration that the relatively small amount of radioactive material present is highly diluted with nonradioactive solid salts and large volumes of saturated aqueous brines. Both the precipitated salts and the brines contain suspended or dissolved 137Cs, 99Tc, and 90Sr; the TRU elements, which represent only a few parts per million of the total tank wastes, are to be found (usually together with uranium) in the precipitated sludges that constitute the bottom layer of tank residues. The earlier decision to maintain a pH greater than 13 has exacerbated the precipitation of normally soluble salts and greatly increased the bulk of the solid material in the tanks. In order to separate the TRU elements from the largely nonradioactive residues, the deposits must first be removed from the tanks. This may be accomplished by washing the residues with high-pressure hot water; some mechanical techniques may also be required for physically scraping the residue from the tank bottoms. (There are a number of pipes and protrusions for instrumentation that would complicate mechanical removal.) Depending on the degree of clean-up chosen as a goal—a key issue—chemical dissolution techniques involving fluoride containing strong mineral acids may ultimately be needed to remove radioactive materials to the maximum extent possible. Such treatment would certainly corrode the existing steel tank surfaces and might damage the tank structure or open old leaks. A major concern is the extent to which complete removal, whether by chemical or mechanical means, will reopen leaks and release radioactivity to the soil. The inventory totals and distributions from the records are sufficient to permit the selection and development of a basic separations process, but it would be necessary to proceed with pilot-scale operation to obtain the detailed technical and operational information needed to design a suitable remediation process system. Planning for full-scale processing requires recovering representative samples of solutions, washings, and washed residues from the tanks for

OCR for page 87
Nuclear Wastes: Technologies for Separations and Transmutation analysis. The residues could be dissolved in fluoride containing mineral acid (probably nitric acid), and the resulting clarified solutions and residual solids could be analyzed. This would provide the information for defining additional chemical processes for treating the tank sludges or the saturated solutions. A logical next step toward remediation processing would therefore be to demonstrate pilot-scale sludge washing operations on feedstocks from several typical tanks. This would provide a sufficiently representative sample of sludge residue, which could be analyzed chemically to provide the information needed to evaluate additional chemical process options for separating the actinide fraction from the sludges. The composition must be available for the major fractions of the feedstock to ensure the successful operation of any of the proposed processes. The selected process must be demonstrated with representative fractions of actual feed material before any type of large plant for remediation processing is constructed. Both the pilot plant and the production facility should operate on a blended feedstock to reduce the effect of variations in tank feed. While the production plant should be designed to accommodate variations, the preparation of blended feed solutions is expected to be a normal part of any process. Variations in both the washing techniques and the tank contents need to be considered, averaged, and analyzed to ensure a stable average input during normal operation of the separations plant. SCENARIOS FOR HANFORD TANK REMEDIATION Decision Factors Options for the remediation of the Hanford tank wastes can be weighed in light of four primary factors: (1) risk (local, regional or national, and worker safety); (2) technical feasibility; (3) public acceptability; and (4) cost. While the committee focused primarily on the technical feasibility of the various separations options, with some discussion of costs, it was recognized that public acceptability and perceived risk are closely linked. The actual risks cannot accurately be defined until the separation processes are further defined, but a general rule is that risk will increase with the degree of handling and processing. Figure 5-3 shows alternative processing scenarios defined by how far one proceeds through the diagram. The choice of path at each juncture should be based on the four factors noted above, using information from literature reviews, research and development studies, and pilot-plant tests on the actual waste product resulting from the previous steps. At its most basic level, separations processing of the Hanford tank waste will divide the waste into LLW and HLW streams. The present plan for the high-level stream involves vitrification and disposal in a geological repository, while the LLW may be disposed of by onsite burial or indefinite storage onsite.1 A decision not to proceed with further separations at any of the junctures shown on Figure 5-3 would result in the approximate number of HLW canisters for geologic disposal shown at the right of the figure. Each decision to undertake further processing reduces further the number of canisters requiring disposal. The cost of geologic disposal obviously depends on both the number of canisters and the regulatory protocols that are adopted. Early regulations proposed for a repository at Yucca Mountain limit the amount of defense waste to 10% of the total inventory of 70,000 metric tons of heavy metal (MTHM) equivalent. The Yucca Mountain repository cost has been estimated at $22 billion, with an additional $1 billion for a repository capacity increase of 10%; the $22 billion includes development costs of $13 billion. These costs would be assessed as charges against canisters placed in the repository. Some of the charge would depend only on the total quantity of heavy metal equivalent, but other charges would reflect the expense of producing and handling each waste canister. Consequently, the potential savings in repository costs that could result from reducing the number of waste canisters via chemical separations is unclear and nonlinear. The costs of producing a suitable waste form, presumably glass in canisters made in a vitrification plant, can be estimated more accurately. Appendix E shows estimated costs of $0.3 million per canister for production of 38,000 canisters over 30 years, compared to $1.8 million each if only 1,000 canisters are made. Discussions are currently under way that may lead to the use of larger, more economical canisters for Hanford HLW. Repository disposal costs must be added to these estimated production costs. Alternative Scenarios The alternative processing scenarios shown in Figure 5-3 involve trade-offs between the quantity of HLW that would require deep geologic disposal (with its attendant costs and risks) and the amount of processing that must be done (and its attendant costs and possible risks). The scenarios also vary in how much LLW would require disposal. These processing scenarios are discussed sequentially below, from least to most chemical processing. In order to provide a complete description of the full range of processing options, this section includes alternatives that have varying degrees of practicality. Current DOE plans are focused on the sludge-washing option with the possibility of further minimal separations. In order to com- 1    The current reference approach at Hartford is to vitrify low level waste and store the vitrified products onsite.

OCR for page 87
Nuclear Wastes: Technologies for Separations and Transmutation FIGURE 5-3 Alternative processing scenarios for Hanford tank wastes. plete the list of processing options, two other scenarios (which are not indicated in Figure 5-3) are considered briefly as well. These two scenarios represent the extremes of virtually no treatment and of complete removal of all radioactivity from the entire site. Risk estimates for the various scenarios are difficult to specify; to some extent, waste processing and repository storage merely reallocate the risk from the local Hanford environment to the locations affected by the transport and storage of the vitrified radioactive materials and to the workers involved in performing these actions. Total risk to the many people associated with the system may therefore increase. No Separations Option Physical removal of the tank contents by aqueous dissolution, evaporation of the water, calcination and recovery of nitric oxides, and vitrification of the residue has been estimated to result in approximately 220,000 canisters of mostly sodium aluminate glasses. Since the repository capacity is specified in tons of heavy metal equivalent, this quantity of canisters may not seriously affect the rules for eventual disposal in a geological repository (Johnson et al., 1993). However, their large number would surely exacerbate problems related to transporting canisters to the repository and to the repository's physical capacity, which in turn would present challenges to public acceptability. The diversity of feedstocks from the individual tanks would pose serious technical challenges for any type of vitrification plant, so blending of stocks would be essential. The cost of immobilizing and disposing of the high-and low-level waste fractions with this method (including construction of a vitrification plant, glass canister production over the plant's lifetime, and transportation and fees for geologic disposal) is unknown; a preliminary estimate—which seems very low—for the purposes of this discussion is approximately $15 billion (Johnson et al., 1993). Sludge Washing A treatment called sludge washing is a logical first step

OCR for page 87
Nuclear Wastes: Technologies for Separations and Transmutation in any separations process that aims to reduce the volume of HLW requiring geologic disposal. In sludge washing, the saturated liquid is first pumped out of the tanks, and the remaining residue in the tanks is directly dissolved and washed with a hot, alkaline (pH > 10) solution. In most instances there will already be more than enough alkali present in the tanks to yield a pH of 10, and a strict policy of adding a minimum of reagents would be followed. After this washing, insoluble residues would be removed with remotely operated robotic systems, though the degree of removal is ill-defined at present. These hydraulic actions should remove most of the insoluble particulate matter from the tanks as a slurry, although an unknown quantity, perhaps 5%, of insoluble solid would probably remain. This residual material is expected to be intractable, and its removal by mechanical or chemical means could cause significant damage to the existing tank structures. If tank leakage resulted, it would increase releases to the environment. The washing fluids and the soluble and suspended fraction pumped from the tanks should contain more than 90% of the sodium and cesium and 20-40% of the strontium; approximately half of the technetium would be expected to dissolve. The cesium may be more difficult than sodium to wash from the sludge. Removal of the strontium and cesium from these fractions could be accomplished by established separation processes such as zeolite adsorption and/or ion exchange. It may not be necessary to explicitly target the water-soluble portion of the technetium, because the ion exchange process proposed for removal of cesium and strontium from the aqueous solutions may also remove a large fraction of the technetium by adsorption into the selected exchanger. Technetium is rendered insoluble by reduction. Any remaining radionuclides in the solutions of sodium salts and wash fractions would be in low concentration. The residue remaining after evaporation of the recycle wash water used in this solution-removal and salt-washing step would qualify as LLW but would probably require treatment by processes such as calcining or biological decomposition to decompose the nitrate and generate a satisfactory feed for the process used to produce the final LLW storage forms, which may be largely sodium aluminate. Environmental concerns relating to the calcining process could require recovery or decomposition of any nitrogen oxides produced during this step. In addition, calcining would produce effluent gases (from organic complexants and their decomposition products) that might require scrubbing. The dissolution of the water-soluble fraction of the waste by sludge washing as described above would leave the insoluble sludge fraction, composed of particulate matter, insoluble portions of salt cake and the settled sludge. These materials should contain almost all of the alkaline-insoluble fission products, the remaining strontium, and, most important, essentially all of the actinide elements from thorium through curium. A large portion of the actinides in the insoluble fraction would be uranium compounds. These separated sludges could be stored in the available DS tanks until a processing decision is made for the next step. The simple sludge-washing process (including zeolite treatment of the solution phase to remove cesium, technetium, and strontium, but without any further chemical separations among the water insolubles) could reduce the volume of waste requiring vitrification by an order of magnitude if the majority of the aluminum salts can be dissolved in the alkaline wash. The cost of sludge washing, removal of cesium, strontium, and technetium from the aqueous fraction, final evaporation of the water solvent, calcination of the residues, and onsite disposal of LLW has been estimated to be $5 to $10 billion (Boomer et al., 1994). If the HLW resulting from this complete process—consisting of the adsorbed cesium, strontium, and technetium in zeolite and any remaining insoluble sludge removed from the tanks—were vitrified, it would result in approximately 38,000 canisters for geological disposal. The heat generated by the cesium and strontium would be the principal determinant of the number of canisters. The required capacity of the vitrification plant would be reduced relative to that needed for the "no separations option." It also should be possible to produce a feedstock with less compositional variation (and correspondingly fewer operation problems) than would be possible if the untreated tank contents were vitrified directly. Estimates of the total costs, including sludge washing, LLW disposal, construction of a vitrification plant, glass canister production over the plant's lifetime, and geologic disposal (including transportation and fees), are estimated at some $25 billion or more.2 Costs may be lessened significantly if the aluminum salts are soluble in the caustic wash, resulting in less sludge to the glass ingots and therefore a smaller investment in vitrification to make the glass ingots. The LLW resulting from this process would still contain trace levels of 137Cs, 90Sr, 99Tc, 129I, and other radionuclides. For this, concrete grout or a sodium glass may be a suitable (Class C) waste form for onsite disposal. It may be reasonable to consider using the emptied DS tanks for the permanent storage of this LLW, although this would not conform to current DOE agreements. The residues generated after the aqueous fraction has been evaporated to dryness, calcined, and converted to a permanent waste form (e.g., immobilized in glass or concrete) could conceivably be deposited in these tanks. 2    Boomer et al. (1994) gives a range of estimates from $18.27 to $24.87 billion for various sludge-washing options, to which the costs of tank-waste retrieval must be added. The latter is estimated at $5 to $15 billion depending on the aggressiveness of retrieval efforts.

OCR for page 87
Nuclear Wastes: Technologies for Separations and Transmutation Solvent Extraction for Actinide Concentration Additional chemical processing has been proposed for the insoluble material that remains after the sludge washing to reduce further the number of HLW canisters requiring repository disposal. While complete dissolution of the concentrated sludge may not be readily achieved, treatment with strong mineral acids and hydrofluoric acid should take essentially all actinides into solution. The resulting solution could be subjected to one or more of the processes such as TRUEX described in Chapter 3 for solvent extraction of actinides. Such an actinide concentration step should further reduce the number of glass canisters of HLW needing geologic disposal, from 38,000 to approximately 11,000. The probability is high that this process could be technologically successful. This actinide concentration approach is similar to the "solvent extraction (advanced separations)" case originally described by the Westinghouse Hanford Co., which includes separation of strontium and technetium (Johnson et al., 1993). Total costs for that option, including tank processing, LLW disposal, construction of a vitrification plant, glass canister production over the plant's lifetime, and geological disposal (including transportation and fees) were estimated in the range of $25 billion (Boomer et al., 1994). These numbers are very speculative; the cost of additional separations beyond sludge washing was estimated to be offset by the reduction in production and disposal costs associated with the lesser volume of HLW. Extensive Separations Further treatment of the non-TRU radioactive material discharged by TRUEX-like processes has been proposed to reduce the vitrification plant product to 1,000 canisters. The Westinghouse Hanford Co. estimated the cost of this option at $17.35 billion, to which retrieval costs of $5 to $15 billion must again be added (Johnson et al., 1993; Boomer et al., 1994). The combinations of new and existing separation processes proposed to achieve this further reduction in glass canisters was called the extensive separations proposal (Straalsund et al., 1992), and its block flowsheet is given in Figure 5-4. To be economically attractive, the costs of such further processing would have to be offset by the reduction in canister production and disposal expense, but the 10,000-canister reduction from the preceding option is estimated to save only $4 billion, based on the total per-canister costs estimated in Appendix E. It is very doubtful that the extensive separations process could be constructed and operated for $4 billion. None of the separation processes proposed for this option has been demonstrated beyond the laboratory scale, and some have not been tested even at that level. There would be major technological challenges to overcome, particularly as regards compatibility problems in the sequential separations systems required for this option as defined. In addition, the extensive separations option would also produce large secondary and tertiary waste streams that in turn would require concentration, processing, and disposal. In-Place Option The in-place scenario involves no removal of material from the tanks. The tank contents would be solidified in an insoluble matrix with a demonstrated, essentially permanent stability. Although this is the least expensive option, neither its risk nor its technical feasibility has been evaluated thoroughly. Current evidence indicates that this option would be unacceptable to the local public, although the low cost might be favored by the general public as other government costs increase. Using current technology, there is probably not enough space in the present tanks to stabilize all the soluble materials present to the degree that would be considered satisfactory. A few large new containers would have to be added on the site to hold the overflow as stable solids. The containment system would need to last well beyond the 300 years required for cesium and strontium radioactivity to decline to about 1/1,000 of the present level. The extremely long half-life problem of the TRUs and technetium would remain also. Pristine Site Restoration Restoration of the site to pristine condition is the extreme end of the remediation spectrum and would involve removal of all added radioactivity from every area of the entire site. This level of clean-up would require not only the removal of tank waste and other stored waste, but also the removal of the emptied tanks (which would still contain significant quantities of radioactive materials) from the site. It would also be necessary to remove the underlying sand, the contents and soil surrounding many other HLW disposal and spill locations (e.g., cribs and ponds), and to treat the groundwater as necessary to restore it to approximate the original environment. The costs of remediation at this level would be extraordinarily high. Since the entire site should be considered measurably radioactive, some rational standard for "low enough," or "as low as reasonably achievable" activity, would need to be adopted even to calculate the cost of this option. The costs for pristine restoration have been estimated to be between $300 and $600 billion. Moreover, the level of chemical separations called for would be extremely high, with many attendant uncertainties, and a protracted period of research and development would be required to develop the necessary decontamination and restoration technologies.

OCR for page 87
Nuclear Wastes: Technologies for Separations and Transmutation FIGURE 5-4 Flowsheet for the extensive separations option process. Transporting the waste would also be an increased risk to all populations along the routes, adding to both costs and political opposition. TRU Isolation for Transmutation Producing the relatively pure target-element material required as feed to transmutation systems that would destroy the TRU content of the Hanford tanks (a quantity of less than one metric ton) to the degree claimed by the most optimistic proposals would require even more extensive separations and recovery than the options proposed above. Other transmutation approaches might be less costly (and less effective), but the reactor operations alone would be a large expense, and all transmutation systems require an elaborate reprocessing plant. Because the spent nuclear fuel scheduled for the first national repository contains over 600 metric tons of TRUs, there appears to be no justification for performing expensive further separation processes so that a fraction of the defense waste TRUs can be transmuted rather than sent to a repository. Both the IFR project scientists at Argonne and the ATW project scientists at Los Alamos agree that application of their transmutation systems to the Hanford tank TRUs may not be justifiable because the small quantity of actinide isotopes involved even in the relatively impure state that would be recovered from a PUREX-type separation could be disposed of through any of the glassingot options. Therefore, the Hanford TRUs—and by extension those from all defense waste sites—can be disposed of with little additional strain on the proposed national repository in vitrified form at considerably less cost than that of any of the proposed transmutation options.

OCR for page 87
Nuclear Wastes: Technologies for Separations and Transmutation CONCLUSIONS AND RECOMMENDATIONS The Hanford storage tanks contain the defense nuclear waste of principal concern at present. Wastes stored at Savannah River, Idaho Falls, Oak Ridge, and elsewhere present serious problems, but they are smaller in total volume and present somewhat less urgent issues by comparison with the Hanford tank wastes. The primary goals of tank treatment are to mitigate safety concerns and to remove the 137Cs and 90Sr, which are the major sources of radioactivity in the tanks, thus reducing the amount of HLW for disposal. Separations of the Hanford tank wastes would involve removal or destruction of sodium nitrate and sodium nitrite salts added during processing. These salts (along with aluminum salts, sodium carbonate, and sodium hydroxide) comprise the bulk of the waste. The simplest option for tank-waste remediation is sludge washing with salt dissolution to remove the bulk of nonradioactive salts, 137Cs, and part of the 90 Sr from the tanks, reducing the volume of HLW by an order of magnitude. Removal of 137Cs and 90Sr from these bulk waste constituents would produce a LLW that is approximately equivalent to the Nuclear Regulatory Commission Class C waste category. This is the option currently being pursued at Hanford. The committee makes the following recommendations relative to the wastes stored in the Hanford tanks: Dedicated transmutation systems should not be considered an option for the remediation of the Hanford tank wastes, or of the other weapon system reprocessing wastes that exist in lesser amounts and analogous kinds at Savannah River, Idaho Falls, and Oak Ridge. The major problems at all these sites are the cesium and strontium fission products which are not candidates for transmutation. The other radioactive isotopes are small in quantity and can be managed safely by other disposal options. Core sampling at Hanford should be employed principally for those tanks for which chemical instabilities are a special safety concern. Extensive sampling is not required for creating a remediation processing plan, and it would increase the likelihood of worker exposure. Highest priority should be given to pilot-scale demonstration of the process involving dissolution of water-soluble sludge and washing of water-insoluble sludge. This involves recovery of cesium, strontium, and technetium from the wash solutions, storage of the insoluble residues, and disposal of the remaining salts from the solutions as LLW. Actual concentrated residues from the sludge dissolution and washing processes in the recommended pilot plant should be used in development studies to determine whether further processing of the HLW is technically and economically justifiable. Because of uncertainties associated with all the processing options, even with the minimum sludge wash treatment, the types and volumes of HLW requiring vitrification cannot yet be defined. Consequently, construction of a Hanford vitrification plant should ensure that the plant design will be compatible with the type and volume of waste requiring vitrification. The option of pristine site restoration appears to be highly impractical technically. A carefully coordinated national research and development program should be instituted to evaluate chemical separation options for the clean-up of DOE defense sites and residues, with emphasis on technology development and demonstration of candidate processes. This program should focus on demonstrating economically viable, practical processes for application to current problem areas. REFERENCES Boomer, K. D., S. K. Baker, A. L. Boldt, J. D. Galbraith, J. S. Garfield, C. E. Golberg, B. A. Higley, L. J. Johnson, M. J. Kupfer, R. M. Marusich, R. J. Parazin, A. N. Praga, G. W. Reddick, J. A. Reddick, E. J. Slaathaug, L. M. Swanson, T. L. Waldo, and C. E. Worcester. 1994. Tank Waste Technical Options Report. WHC-EP-0616. Richland, Wash.: Westinghouse Hanford Co. Johnson, M. E., M. L. Grygiel, P. A. Bayness, J. P. Bekemeier, B. D. Zimmerman, and M. B. Triplett. 1993. Tank Waste Decision Analysis Report. WHC-EP-0617 (draft). Richland, Wash.: Westinghouse Hanford Co. Straalsund, J. L., J. L. Swanson, E. G. Baker, J. J. Holmes, E. O. Jones, and W. L. Kuhn. 1992. Clean Option: An Alternative Strategy for Hanford Tank Waste Remediation. Vol. 1. PNL-8388. Richland, Wash.: Pacific Northwest Laboratory.