Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
2 Why Consider Flexibility in Disposal Options? Why should the nation consider flexibility in disposal options for some waste currently classified as transuranic (TRU) or high-level waste (HLW)? In this chapter, the committee explains that some waste cur- rently classified as TRU or HLW may not warrant disposal in a deep geologic repository because the effort, exposures, and expense associated with retrieval, immobilization, and shipment to a repository may be out of proportion with the reduction in human health risk achieved, if any. The committee identifies three waste types, each of which contains some wastes that merit consideration by the Department of Energy (DOE) and others for alternative disposal. These waste types are described in detail as case studies. This chapter establishes the basis for the committeeâs consideration of a process that could be applied to DOEâs request for alternative disposition of some HLW and TRU waste by discussing the difficulties caused by the current definitions of HLW and TRU, and by describing three waste types containing waste streams that could be candidates for alternative disposal. The committeeâs approach toward greater flexibility was the result of many factors, including gaps and uncertainties in the definitions in Chapter 1, recent litigation, congressional action targeting HLW disposal, along with the testimony from DOE, its contractors, 37
38 RISK AND DECISIONS ABOUT TRU AND HLW stakeholders, and citizen groups on management strategies for HLW and TRU waste. Finding 1: Deep geologic disposal is the default disposition option for HLW and TRU waste. There is a long history of studies supporting deep geologic disposal of long-lived radioactive wastes. Deep geologic disposal remains the na- tionâs approach for disposal of TRU and HLW. 2.1 SOURCE-BASED DEFINITIONSâ WIDELY VARIED WASTE Until the October 2004 legislation, all reprocessing waste that met the old statutory definition of HLW was required to be disposed of in a permanent geologic repository. If the waste were exempted from the definition or reclassified, then it could be disposed of elsewhere. How- ever, the general thrust of this definition is inclusion, not exclusion; that is, it offers opportunities in the second clause of paragraph (A) and in paragraph (B) for an agency (the U.S. Nuclear Regulatory Commission [U.S. NRC]) to add to what is considered HLW, not to create exemptions (see the definition repeated below). The inability to create exemptions may help to limit potential abuses by preventing loopholes, but it also prevents the consideration of reasonable alternatives when they make sense. High-Level Waste is (A) the highly radioactive waste material resulting from the reprocessing of spent nuclear fuel, including liquid waste produced directly in reprocessing and any solid material derived from such liquid waste that con- tains fission products in sufficient concentrations; and (B) other highly radioac- tive material that the Commission, consistent with existing law, determines by rule to require permanent isolation. (U.S. Code, Title 42, Section 10101) The definition states that any solid material derived from liquid HLW containing fission products in sufficient concentrations is HLW. This implies a concentration-based standard, but it is indeterminate: the definition seems to establish that certain solid wastes derived from high- level liquid waste could contain fission products below some undefined âsufficient concentrationâ that would fall outside the HLW definition
WHY CONSIDER FLEXIBILITY IN DISPOSAL OPTIONS? 39 (Burket, 2004; Mears and Ruple, 2004).1 However, liquid reprocessing wastes have no statutory exemptions. While there may also be an implicit de minimis exception to the definition for both liquid and solid reproc- essing wastes (the last few grams in a tank or a kilogram or two of con- taminated soil?) this has not been claimed by the relevant agencies or 2 tested in the courts. It would also be possible to elaborate on the mean- ing of âhighly radioactiveâ to exclude certain fractions (however de- rived) of the initial reprocessing waste stream. This is a logical possibil- ity, but it too is undefined. Following the Idaho court decision that re- jected DOEâs process for exempting wastes (see Sidebar 1.3), there were no exceptions to the characterization of wastes from reprocessing as HLW that are agreed upon as valid. The recent legislation created ex- emption criteria for waste in South Carolina and Idaho based on the U.S. NRCâs concentration limits and performance objectives for near-surface disposal of low-level radioactive waste and requirements that the waste be treated to remove highly radioactive radionuclides âto the maximum extent practicalâ (see Sidebar 1.4). The definition of TRU waste is, by exclusion of HLW, also a source- based definition. Most TRU waste is indeed quite different from HLW, but some is substantially similar. In contrast to HLW, 98 percent of TRU waste is contact handled: it has relatively low concentrations of the shorter-lived fission products and, thus, emits less radiation and gener- ates less heat, but the long half-lives of the transuranic isotopes and their decay products mean that the hazard they pose does not diminish sub- 1 Technically, this is not an exemption in the sense of shifting the burden to the regulated entity to obtain a deviation from the general rule. Read literally, the definition suggests that derived solids are HLW only if they contain suffi- cient concentrations of radionuclides; otherwise, apparently, derived solids gen- erally are not HLW. As a practical matter, however, both the solid and the liquid fractions of the tank waste begin their existence as a single, highly radioactive primarily liquid waste; therefore, it is usually a matter of treating the solid de- rived wastes to reduce radioactivity, rather than deciding whether they contain enough radioactivity. 2 Such an interpretation would not be unprecedented. Courts have interpreted certain parts of the Clean Air Act to include a de minimis exception, Alabama Power Co. v. Costle, 636 F.2d 323, 360 (D.C. Cir. 1979) (âCategorical exemp- tions may also be permissible as an exercise of agency power, inherent in most statutory schemes, to overlook circumstances that in context may fairly be con- sidered de minimisâ), but in other cases have declined to do so because the stat- ute clearly precluded it; (see, e.g., Public Citizen v. Young, 831 F.2d 1108 D.C. Cir. 1987; Delaney Clause of the Federal Food, Drug, and Cosmetic Act).
40 RISK AND DECISIONS ABOUT TRU AND HLW stantially for millennia.3 Like HLW, remote-handled TRU waste can con- 4 tain high concentrations of radionuclides that emit penetrating radiation, and thus are similar to HLW in terms of its requirements during waste management and disposal. Remote-handled TRU waste constitutes only about 2 percent by volume and 3-4 percent of the radioactivity of the to- tal inventory of TRU waste. The definition of TRU waste provides administrative mechanisms for removing waste from the TRU waste classification. In this way it contrasts sharply with the definition of HLW, which does not contain parallel language. However, which TRU waste should be managed by means other than permanent geologic disposal is ill-defined, and DOE and its regulators have not made much use of alternative disposal provi- sions for TRU waste. As far as the committee is aware, only one excep- tion has been granted for TRU waste. In this exception, DOE disposed of some TRU waste near the surface and some at intermediate depth. Spe- cifically, DOE disposed of about 60 metric tons of classified TRU waste containing around 330 curies (Ci) of plutonium-239 in four boreholes approximately 35 meters deep at the Nevada Test Site between 1984 and 1989 (SNL, 2004). Sandia National Laboratories carried out a perform- ance assessment to demonstrate that this disposal meets the Environ- mental Protection Agencyâs (U.S. EPAâs) requirements for disposal of TRU waste (40, CFR 191). DOE and U.S. EPA agreed that disposal of this waste in this manner was satisfactory. As noted in Finding 1, the committee recognizes the necessity and appropriateness of deep geologic disposal for HLW and TRU waste. However, as explained more fully below, the evolution of HLW man- agement and treatment (among other things) has led to the creation of a series of different waste streams. Changes in treatment technology and other factors suggest to the committee that in certain limited cases a process could be considered for reclassification and disposal of HLW and TRU waste. 3 Transuranic isotopes have very long half-lives (i.e., plutonium-239 has a half-life of 24,400 years; neptunium-237 has a half-life of 2 million years) or decay into isotopes that have long half-lives, which can potentially give rise to long-term management problems and uncertainty in exposure scenarios. 4 Penetrating radiation can come from fission products or transuranic iso- topes. Some transuranic isotopes, like americium-241 (a decay product of the relatively short-lived plutonium-241) emit gamma rays and others, like pluto- nium-238 and -240, cause neutron emissions through (alpha,n) reactions in lighter nuclei or through spontaneous fission.
WHY CONSIDER FLEXIBILITY IN DISPOSAL OPTIONS? 41 Cs & Sr capsules at Hanford Vitrified HLW Spent Fuel at SRS Concentration of Short-Lived Radionuclides of Concern High-Level Waste Remote-Handled Grouted HLW tank TRU heels at SRS Class C Limit Saltstone disposed of at SRS Contact-Handled TRU Concentration of Long-Lived Radionuclides of Concern FIGURE 2.1 Chart of the concentrations of short-lived and long-lived ra- dionuclides for waste that might be considered HLW along with TRU waste and spent nuclear fuel. The boundary of each waste class is meant to surround the various waste streams and does not represent quantities. Class C limit demarcations represent radionuclide concentra- tions in low-level waste below which near-surface disposal is permitted. 5 Note that Saltstone would not now be considered HLW. See extensive discussion of this diagram in the main text. The range of variation within the different types of waste is repre- sented qualitatively in Figure 2.1, which illustrates the long-lived and short-lived radionuclide composition of HLW, TRU waste, and spent nuclear fuel. In this figure, adapted from one by Fehringer and Boyle (1987), moving to the right reflects an increase in the concentration of long-lived radionuclides of concern (e.g., americium-243) and moving 5 Saltstone is DOEâs name for the cementitious waste form used at the Savannah River Site to immobilize liquid waste from processing HLW that is being sent to the vitrification plant (see discussion on Waste in HLW Tanks at the Savannah River Site below).
42 RISK AND DECISIONS ABOUT TRU AND HLW up reflects an increase in the concentration of shorter-lived radionuclides of concern (e.g., cesium-137). Radionuclide concentration limits that are generally acceptable for near-surface burial of low-level wastes (Class C limits contained in 10 CFR 61) are represented by a vertical line and a horizontal line within the chart. The regulation contains a table of con- centration limits for short-lived radionuclides and a table of concentra- tion limits for long-lived radionuclides, so these boundaries are simply notional demarcations indicating that low-level waste in the lower, left quadrant is generally acceptable for near-surface disposal under the regu- 6 lations. Congress used the Class C limits and the performance objectives from 10 CFR 61 as part of the new lawâs criteria for determining what reprocessing waste is not HLW. The inclusion of Class C limits should not be construed to imply that the committee has determined that waste that sits below those limits is suitable for near-surface disposal. The lim- its are included as reference levels only. The figure illustrates the varied nature of waste and material that might be considered HLW. In the upper left corner is the most concen- trated radioactive material in the DOE complex: the cesium and stron- tium capsules at Hanford. Low-activity waste from the treatment of HLW at the Savannah River Site already disposed of on-siteâthe Salt- 7 stoneâis at the lower left portion of the HLW boundary. Vitrified HLW and calcined HLW are near the upper right. The waste grouted in two tanks that were declared closed at the Savannah River Site (tanks 17 and 20) is displayed straddling the Class C limit because the waste is below the Class C limit if one averages the concentration over the grout in the tank, but the Natural Resources Defense Council (NRDC; Cochran, 2003) has argued that there is not substantial mixing, and the concentra- tion of the waste itself remains above the Class C limit. It is possible that the quantities and concentrations of the heels in tanks 17 and 20 may be lower than those of other tanks (dâEntremont and Thomas, 2002), so the waste depicted on the figure might not be representative of future grouted tanks. By definition, TRU waste has relatively high concentra- 6 Greater-Than-Class-C low-level waste is not deemed generally acceptable for near-surface disposal, currently is stored, and has no ultimate disposition path under development. In its FY2005 budget request, DOE sought to create a program within a new Office of Future Liabilities to take responsibility for dis- posal of Greater-Than-Class-C waste (DOE, 2004). 7 This waste was not considered HLW when it was disposed of and may not be considered HLW under Section 3116 of the Defense Authorization Act of 2005.
WHY CONSIDER FLEXIBILITY IN DISPOSAL OPTIONS? 43 tions of long-lived radionuclides or of radionuclides that will decay into 8 long-lived radionuclides. Remote-handled TRU waste has higher con- centrations of short-lived fission products and so appears higher in the figure. The concentration of radionuclides in spent nuclear fuel depends on the burnup of the fuel (i.e., how many fissions have occurred per unit fuel). Lightly irradiated fuel has relatively low concentrations of radioac- tivity. 2.2 WASTE STREAMS THAT MAY NOT WARRANT DEEP GEOLOGIC DISPOSAL Finding 2: Some waste currently classified as TRU or HLW may not warrant disposal in a deep geologic repository, either because (1) it is infeasible to recover and dispose of every last bit of waste that might conceivably be classified as TRU or HLW or (2) because the effort, exposures, and expense associated with retrieval, immobilization, and disposition in a repository may be out of proportion with the risk reduction achieved, if any. As Chapter 1 and the discussion above show, HLW and TRU waste classification schemes define waste based primarily on their source. Source-based schemes are often the best way to manage classification for a number of reasons. They tend to be simple and easy to apply. Waste classification systems are normally established without knowing the spe- cific characteristics of all of the waste streams that will be produced. A source-based classification system is thus useful because it avoids prob- lems with some other systems that require revision of the waste classifi- cation system for each new waste stream or advance in treatment tech- nology. In addition, it can provide direction to the implementing agency or agencies so that plans for treatment and disposal, and allocation of costs and responsibilities, can be made at the wasteâs source. Source-based classification systems have certain disadvantages, too. One potential drawback is that such systems lack flexibility to reclassify waste or to treat waste in such a way that it minimizes impacts to human 8 The New Mexico Environment Department fined DOE for shipping TRU waste that had failed to meet waste acceptance criteria due to low radionuclide concentrations from the Idaho National Engineering and Environmental Labora- tory to the Waste Isolation Pilot Plant.
44 RISK AND DECISIONS ABOUT TRU AND HLW health and the environment. In the United States disposal options are dic- tated by the waste class (i.e., waste of a given class may be disposed of only in the manner designated for that class), and when the class is based on the source rather than on the measurable characteristics of the waste, the disposition options may not match the risks posed by the waste. Some source-based definitions accommodate flexibility by providing for exceptions or exemptions of waste, based on meeting a very stringent set of health or risk-based criteria. Chapter 3 contains a more detailed dis- cussion of these classification schemes and an analysis of the âover- and underinclusiveâ problem of defining waste. Chapter 3 also sets out the basis for the committeeâs transparent and constrained risk-informed process for considering alternative disposal. As case studies, the committee selected three waste types that are il- lustrative of the reasons for considering alternatives to disposition in a deep geologic repository for some HLW and TRU: 1. They appear to include wastes that are relatively low in radioac- tivity and/or hazard compared to other HLW and TRU waste that DOE manages, and perhaps could be managed in some manner other than disposition in a deep geologic repository. 2. For some of these wastes it is infeasible to recover and dispose of every last bit of waste that might conceivably be classified as TRU or HLW. 3. The effort, exposures, and expense associated with retrieval, im- mobilization, and disposition in a repository may be out of pro- portion to the risk reduction achieved (if any). For the same reasons, these waste types also contain specific waste streams that DOE could consider as candidates seeking approval for al- ternative disposal if a process for considering such matters were to be put in place. The waste types are 1. HLW remaining in tanks (âheelsâ), 2. low-activity products from treatment of HLW, and, 3. buried TRU waste (not buried in a manner that facilitates re- trieval).
WHY CONSIDER FLEXIBILITY IN DISPOSAL OPTIONS? 45 Finding 3: The committee makes no recommendation whether spe- cific wastes should be approved for alternative disposal, but it has identified three waste types that contain waste streams that merit consideration: (1) HLW remaining in tanks (heels); (2) low-activity products from treatment of HLW; and (3) buried TRU waste (not buried in a manner that facilitates retrieval). The remainder of this chapter provides a discussion of the three waste types. The first two of these are discussed together in the next sec- tion to avoid redundant descriptions of the tank wastes. Note that whether or not the waste streams are indeed of relatively low radioactiv- ity and low risk, and whether the cost of cleanup is indeed disproportion- ate to benefits, have to be evaluated independently as part of any poten- tial exemption process. HLW Remaining in Tanks (Heels) and Low-Activity By-Products from Treatment of HLW DOE is responsible for managing and disposing of wastes from nearly 250 tanks containing HLW at the Idaho National Engineering and Environmental Laboratory, the Savannah River Site, and the Hanford Site. The wastes are diverse, comprising a highly heterogeneous mix of chemicals with the radioactive and non-radioactive constituents in a vari- ety of physical and chemical forms. The plans for these wastes estab- lished in existing compliance agreements at various levels of detail are conceptually similar. Waste in HLW Tanks at Hanford Chemical separation plants at Hanford dissolved the irradiated fuel from on-site plutonium production reactors. The HLW generated from these operations was pumped to underground tanks that were grouped in sets called tank farms. Several different chemical separation processes were used at Hanford at different times, and some waste streams were subjected to further separations to recover residual uranium and to separate cesium and strontium. As generated, the wastes were acidic, but sodium hydroxide was added to reduce the corrosive effects of the waste on the carbon steel used to line the concrete tank structures. This
46 RISK AND DECISIONS ABOUT TRU AND HLW neutralization step increased the waste volume dramatically and resulted in the precipitation of many constituents out of the liquid waste and settling of these constituents to the bottom of the tanks as sludge. Waste varies from tank to tank, but in many tanks the waste can be understood as existing in three roughly defined phases: the sludge, which contains most of the actinides and strontium; a soluble crystalline solid called saltcake; and a liquid supernate. The latter two phases contain some strontium and most of the cesium, iodine, and technetium. Complicating these waste streams are other materials that were added to the tanks, such as debris, cement, diatomaceous earth, and broken or obsolete contaminated equipment. Until 1964, all of the tanks at Hanford were constructed with no sec- ond liner to contain waste in case of a failure of the primary tank liner. 9 There are four different designs of these 149 single-shell tanks. At least 3 67 of the tanks have leaked between 2700 and 5400 m (750,000 and 1.5 million gallons) of HLW into the ground (Gephart, 2003). Another 28 tanks were constructed after 1964, all with a secondary liner, so they are called double-shell tanks. Because of leaks in some of the single-shell tanks, the pumpable liquids from those tanks have been pumped into the double-shell tanks. 3 Hanford now has approximately 196,000 m (54 million gallons) of HLW, about 60 percent in single-shell tanks and about 40 percent in double-shell tanks (Wiegman, 2004). The saltcake and sludge (which each constitute about one third of the waste each) in the single shell tanks contain a little over 100 million curies (MCi) of radioactivity. The dou- ble-shell tanks have a little over 90 MCi contained in waste consisting mostly of liquids, but also sludges and salts (Wiegman, 2004). HLW must be retrieved from the tanks and immobilized for eventual disposal in a deep geologic repository. For most tanks, DOE currently plans to use techniques such as dissolution and sluicing followed by pumping of the resulting solutions and slurries from the tank, ultimately using steam jets and vacuum heads to get at the last portions, although some amount of waste (the heel) is expected to be irretrievable using these techniques. DOE estimates that if it immobilized all of the retrieved waste for disposal in a geologic repository without putting the HLW through a separations process, Hanford would generate more than 100,000 canisters of HLW (GAO, 2004). This is lower than a 1993 esti- mate reported in a 1995 report (NRC, 1995a) which put the figure at 9 For a description of the tanks and the wastes, see Gephart (2003).
WHY CONSIDER FLEXIBILITY IN DISPOSAL OPTIONS? 47 220,000 canisters. In either case, the disposal cost of such an approach (estimated at more than $65 billion now, up from $15 billion in 1993; GAO, 2004) was deemed by DOE to be too high compared to the risk 10 reduction achieved relative to alternative approaches. An alternative approach (one referred to as the âbaselineâ approach, which was agreed to by the parties to the existing compliance agree- ments) is to chemically process the retrieved waste to concentrate most of the radioactivity in a high-activity waste stream and concentrate most of the nonradioactive chemicals and relatively small amounts of radionu- clides in a relatively low-activity waste (see Sidebar 2.1). This reduces the volume of high-activity waste and creates a larger waste stream of 11 lower-activity waste, but the latter waste stream is planned for immobi- lization and near-surface on-site disposal. The amounts in each category depend on the details of the approach, but the current plan at Hanford would produce up to 14,500 canisters 3 (15,700 cubic meters [m ]) of vitrified high-activity HLW (DOE, 2002b) 3 and around 270,000 m of low-activity waste for disposal on-site. The overall cost of this approach, which sends all of the waste through the Hanford Waste Treatment Plant, is estimated by the U.S. Government Accountability Office (GAO) to be $26 billion (net present value; GAO, 12 2004). However, because of the large amount of vitrified low-activity waste compared to the glass production rate of the planned low-activity vitrification facility, the baseline plan does not meet the 2028 completion date agreed to in the federal facility agreement for Hanford (Hanford FFA, 2003). The DOE accelerated cleanup effort has proposed cost and schedule savings by sending more than half of the low-activity waste to âsupplemental treatment,â instead of through the Waste Treatment Plant. Supplemental treatment options include bulk vitrification, steam reforming, 10 This view was not shared by some people who spoke before the committee, most notably some of the representatives from American Indian nations. 11 Because of the difficulty of removing them, current treatment plans leave the fission products technetium-99 and iodine-129 in the low-activity waste stream. These radioisotopes are long-lived and mobile in the environment. 12 The values reported here for GAOâs estimates are the mean values. The range on these values is as small as about Â± 7 percent for some estimates and as large as Â± 12 percent for others. These ranges do not affect the general conclu- sions that can be drawn from looking at the mean value. The committee has not examined these cost estimates in detail but observes that past project costs often have differed from estimates by far more than 12 percent, most commonly ex- ceeding estimates.
48 RISK AND DECISIONS ABOUT TRU AND HLW containerized grout (âcaststoneâ), and sulfate removal (to enable more effective processing in the Waste Treatment Plant). GAO estimates that DOEâs accelerated cleanup plans could save $12 billion. Sidebar 2.1: Plans for Tank Wastes The generic description of DOEâs plans for processing and immobi- lizing leads to three wastes requiring disposal: 1. HLW glass logs produced by adding glass-making materials to the high-activity waste, melting this mixture to drive off water and some volatile anions such as nitrates, and pouring the molten glass-waste mixture into stainless steel cylinders where it solidi- fies, these glass logs are to be stored on-site until they can be transported to and emplaced in a geologic repository presumed to be the proposed site at Yucca Mountain, Nevada; 2. an immobilized, solid, low-activity waste that is to be buried in near-surface disposal facilities established at the site; and 3. tanks containing residual waste that are to be filled with grout or other stabilizing and immobilizing materials. It is not technologically possible to remove every gram of HLW from the tanks without removing the tanks themselves. Few are suggest- ing that the tanks be exhumed, decontaminated, and reburied because the costs (estimated by DOE at $6 billion and by GAO at over $67 billion for all of the single shell tanks; GAO, 2004) are likely to be far out of pro- portion to any human health risk reduction. It stands to reason that the risk to workers involved in this effort would be considerable. Yet some HLW residuals will remain in the tanks. The amount, concentration, and form of that tank heel, along with the barriers that should be emplaced to isolate the residual waste, are central issues in the disposition debate. Present plans based on compliance agreements call for DOE to remove either 99 percent of the volume of HLW (the allowable remainder vol- ume weighted according to the capacity of each single-shell tank) or the limit of waste retrieval technology capability, whichever leaves less in the tank. At its site visits, the committee received comments that this standard is, variously, reasonable, unrealistic, arbitrary, not connected to risk, and leaving a lot in the ground.
WHY CONSIDER FLEXIBILITY IN DISPOSAL OPTIONS? 49 The extent of retrieval depends on the physical and chemical form of the waste in the tank, the physical structures within the tank, physical access to the tank, and the technology used to retrieve the waste. In the single-shell tanks at Hanford, the wastes are highly heterogeneous, and the tank heels will range from hard, insoluble crusts to thick, claylike layers and patches to viscous sludge. It is possible that sluicing, dissolu- tion, and more aggressive methods for recovering the wastes from single- shell tanks could lead to further leaks, although this has not been demon- 13 strated to be the case. The physical design of the older tanks makes waste retrieval quite challenging. Tank access ports are limited in num- ber and size, which will impede access to the tank contents, and many of the tank bottoms are uneven and contain internal structures that further complicate waste retrieval. DOEâs current plan, once a regulatory decision has been made de- termining a tank waste-retrieval operation is complete, is to fill the tank 14 with materials to aid in immobilizing the residual tank contents and to provide structural support to avoid future tank collapse that could lead to water ingress (DOE, 1997b). Other engineered barriers would then be placed over and perhaps around the tanks. DOE has not selected the ma- terial that would be used to fill the Hanford tanks. Waste in HLW Tanks at the Savannah River Site 3 The Savannah River Site (SRS) has approximately 135,000 m (37 3 million gallons) of HLW: 123,000 m (34 million gallons) of saltcake 3 and supernate containing 204 million curies (MCi), and 11,000 m (3 million gallons) of sludge containing 215 MCi of radioactivity. All of the 51 HLW tanks at the SRS have a secondary layer of containment, 13 One can reason that mobilizing encrusted radionuclides and spraying high- pressure jets into a tank that has leaked would probably result in further leakage through the same perforations and that leakage might further mobilize wastes that have already leaked into the soil immediately around the tanks. Alterna- tively, one could hypothesize that the perforations only leak when there is a sub- stantial weight load in the million-gallon tank, and that the half-inch thick steel can withstand water jets, even if the steel is somewhat corroded. The behavior has not been tested, but the risk of further leakage might not be worth taking. 14 The NRDC has questioned the effectiveness of DOEâs approach with re- spect to immobilizing the residual waste, arguing that DOE has not properly accounted for the radioactive constituents of the waste (Cochran, 2003).
50 RISK AND DECISIONS ABOUT TRU AND HLW although 24 of the oldest ones have only a secondary pan (sometimes described as a âsaucerâ to the tank âteacupâ). Most of these ânoncompliantâ tanks have a history of cracks or leakage (WSRC, 15 2004), but only one is believed to have leaked a small quantity of waste to the environment (Davis et al., 1977). Compared to the Hanford single- shell tanks, the Savannah River tanks are in relatively good condition. Two of the tanks have been closed and filled with grout, and three more are empty, to the extent DOE has deemed technologically and economically practical. The other 46 tanks contain wastes that are less chemically varied (produced by and subjected to fewer chemical processes and containing less troublesome additives) than those at Hanford and are generally more amenable to retrieval than the Hanford wastes because they are more soluble and access within the tanks is easier. DOE has been retrieving waste at the SRS and piping the sludge in the form of slurry to the Extended Sludge Processing Facility for sludge washing (removing the soluble nonradioactive chemical components) and then to a vitrification facility, called the Defense Waste Processing Facility, which began operations in 1996. So far, SRS has produced more 3 than 1750 canisters of HLW, vitrifying about 2.9 m (800 gallons) of sludge to produce about 2000 kilograms (4400 pounds) of waste glass for each waste canister. The plan, under which DOE has processed its HLW, as approved by the State of South Carolina, would send all of the retriev- 3 able sludge to the Defense Waste Processing Facility. Some 300,000 m of salt waste, made up of the supernate and saltcake dissolved in water added to the waste, containing 207 MCi of radioactivity (including 201 MCi of cesium-137, or nearly 95 percent of the siteâs cesium-137) is to be pumped to a Salt Waste Processing Facility, which is to extract over 3 99.9 percent of the radioactivity, concentrate it in 11,000 m of waste, and send that to the vitrification plant. The remaining radioactivity (75 kCi, including 10 kCi of cesium-137, 20 kCi of technetium-99, 200 Ci of strontium-90, and 500 Ci of actinides) residing in a dilute solution in 3 315,000 m of liquid would go into a low-activity waste form, called Saltstone, for disposal in near-surface vaults on-site. Despite the fact that the Savannah River Siteâs wastes are easier to retrieve than those at Hanford, DOE still has difficulty recovering the last 15 Thirteen tanks are known to have leaked waste out of the primary contain- ment. Most of these leaks were small and some dried before the waste reached the secondary containment pan (WSRC, 2004).
WHY CONSIDER FLEXIBILITY IN DISPOSAL OPTIONS? 51 fraction of the waste from the tanks. The two tanks that are grouted had on the order of a thousand gallons of residual waste (approximately equivalent to one inch in the bottom of the larger tanks) in each tank bot- tom. The approach used to isolate the heels in these tanks is to fill the tanks with successive layers of various types of cementitious grout to inhibit water access to the heel and collapse of the tank due to deteriora- tion. Initial layers of grout are designed to establish reducing conditions in the tank that serve to minimize the solubility and mobility of most ra- dionuclides. Other layers are designed to provide structural support. The nature of the engineered barriers if any, to be considered outside of the tanks has not yet been decided. The Savannah River Site also has had difficulty developing the chemical processes to be used in the future Salt Waste Processing Facil- ity. After it was found that the preferred approach, an in-tank precipita- tion process to remove cesium, unexpectedly generated large amounts of benzene, DOE asked the National Research Council (NRC, 2000a) to make recommendations on how DOE should develop the needed tech- nology and bring the facility on-line. The NRC committee appointed for that study recommended that DOE pursue several technology options until a preferred technology option is proven to work. The committee also noted that the wastes vary dramatically in chemical composition, so DOE may want to use different technologies for different wastesâa tai- lored approach. DOE is still working to develop the technologies and the facility for this processing. Saltstone is DOEâs name for the cementitious waste form used at Sa- vannah River to immobilize liquid waste from processing HLW that is being sent to the vitrification plant. The reference composition of Salt- stone is given in Table 2.1. It is made up of Portland cement, blast- furnace slag, and fly ash mixed almost one-to-one with the waste salt solution.
52 RISK AND DECISIONS ABOUT TRU AND HLW TABLE 2.1 Saltstone Reference Formulation Reference Reference Saltstone Dry Saltstone Ingredient Solids Premix (wt. %) (wt. %) Portland cement, Type I 8 54 Slag, grade 100 46 Fly ash, class F 46 Reference salt solution, 29 wt. NA 46 % salt Water: cementitious solids ratio NA 0.605 Source: WSRC, 2001. The Saltstone Facility is a fairly simple operation that mixes the waste flow with the cement ingredients and pumps the mixture out to disposal vaults. Concentrations of radioactivity in the waste handled to date have been low enough to allow unshielded operation. DOE has al- ready filled one six-cell concrete vault with Saltstone and has begun fill- 3 ing a 12-cell vault. Each cell holds approximately 6600 m of the waste form. Once a vault is filled, the monolith is mounded over to direct water away from the waste and to provide earthen shielding of radiation from the waste. The majority of Saltstone production has not begun because the Savannah River Site so far is processing and immobilizing only the sludge from its HLW tanks, and the vast majority of the salt waste will come from processing the saltcake and supernate from the tanks. DOE has shut down all work at the Saltstone facility pending resolu- tion of the NRDC v. Abraham (2003) decision. The NRDC and the states of Washington, Idaho, and South Carolina have contended that DOE need not have ceased this operation under the terms of the Idaho District court decision in NRDC v. Abraham. DOE has proposed to accelerate processing of HLW by sending less waste through the Salt Waste Processing Facility. The proposed plan, called âtailored salt processing,â would separate the salt waste into three streams by draining the liquid (which will carry most of the cesium), and
WHY CONSIDER FLEXIBILITY IN DISPOSAL OPTIONS? 53 then sending the high-curie, high-actinide salt (including the liquid with the cesium) to the Salt Waste Processing Facility, which would operate as before, with cesium and actinides going to the Defense Waste Process- ing Facility and decontaminated salt solution going to Saltstone. The re- maining salt, from which liquid was drained, is called âlow-curie saltâ and âlow-curie, high-actinide salt.â The former would go directly to Saltstone. The latter would go through an actinide removal facility, send- ing the actinides to the Defense Waste Processing Facility and the rest to Saltstone. This would leave 17.6 MCi of cesium-137, 1.7 MCi of stron- tium-90, 100 kilocuries (kCi) of actinides, and 20 kCi of technetium-99 16 in the Saltstone. Another way of looking at this is that the cesium in Saltstone would increase by a factor of 1750, the strontium by a factor of 8400, and the actinides by a factor of 200, while the technetium content would stay the same. The concentrations would not rise by similar fac- tors because the overall quantity of Saltstone would increase signifi- cantly, but a new Saltstone Facility would be required to allow shielded operation. DOE says this would save $7 billion and enable it to finish shipping canisters to a HLW repository up to 20 years sooner. One of the advan- tages, from DOEâs perspective, is that waste would be removed from the tanks and solidified earlier. Another is that DOE can start processing some wastes now and bring the Salt Waste Processing Facility on-line later when DOE has fully developed the technology to make the facility accomplish its design goals. Waste in HLW tanks at INEEL DOE has 11 underground tanks used for storage of liquid radioactive waste in the Idaho Nuclear Technology and Engineering Center Tank Farm at the Idaho National Engineering and Environmental Laboratory (INEEL). These tanks are among the newest in the DOE system and are much smaller than those at either the Savannah River Site or Hanford. This allows good access to most parts of the tank. In addition, most of the waste is from chemical processing of spent nuclear fuel from the Na- val Reactors Program, which uses highly enriched uranium fuel. The pur- pose of this processing was not to recover plutonium for the weapons 16 This could bring the waste concentrations close to the class C limits found in 10 CFR 61.55.
54 RISK AND DECISIONS ABOUT TRU AND HLW program, but to recover the residual highly enriched uranium. However, some of this highly enriched uranium remained in the waste after separa- tion, and to avoid criticality accidents that could occur if too much ura- nium precipitated in the tanks, the waste was not neutralized. The tanks had to be constructed of stainless steel to contain the acidic waste. Most of the HLW generated at INEEL was sent to a facility that cal- cined the waste, that is, rapidly oxidized the waste into a granular solid, likened to laundry soap, which was stored in large, stainless steel bins. This waste which currently contains about 44 MCi of radioactivity, is to be immobilized in a form suitable for disposal in a HLW repository and then shipped out of the state for disposal. For the roughly 500 kCi of ra- 3 dioactivity in 3600 m of liquid radioactive waste remaining at the Idaho Nuclear Technology and Engineering Center tank farm (INEEL, 2004), DOE estimates that 1-3 percent by volume is liquid waste from the first cycle of reprocessing (what DOE considers HLW) and 20-30 percent is from second and third cycle liquids. The remainder consists of decon- tamination washing solutions from the calcine facility, evaporator bot- toms, and other process wastes. âSodium bearing wasteâ is the term used to describe the mixture of liquid wastes from all of these sources except the first cycle of reprocess- ing (DOE, 2002c). This liquid waste stream has substantial quantities (more than 2 moles per liter) of sodium nitrate salts, resulting from the addition of sodium hydroxide to the washing solution to enhance its ef- fectiveness in removing some residues. The mixture was concentrated through evaporation, and some was sent through the calciner to produce calcine waste, although the high sodium content makes the direct calci- nation process perform poorly, so treatment prior to calcination is needed. DOE is now consolidating all of its wastes from the tanks into 3 three of the roughly 1100 m (300,000 gallon) tanks while it cleans seven other tanks. An eleventh tank is a clean spare. The consolidation opera- tion will further commingle the waste. The vast majority of the waste is in a liquid form, but a small amount of insoluble solids can be found at the bottoms of the tanks. Samples from several tanks were analyzed extensively. In a typical tank, the liq- uid contains a higher concentration of strontium than do the solids, and the solids have a higher concentration of cesium than the liquid. In tank WM-187, for example, strontium and cesium together contribute 97 per- cent of the total radioactivity of 0.22 Ci per liter. Isotopes of plutonium (mostly plutonium-241 and plutonium-238) constitute 1.9 percent (Bar- nes et al., 2004). The classification of this waste has been in dispute and
WHY CONSIDER FLEXIBILITY IN DISPOSAL OPTIONS? 55 is the subject of litigation. The State of Idaho considers this waste to be HLW; DOE considers the waste to be mixed transuranic waste. Waste processing at INEEL has an important difference from that at the other two sites: there are no plans to perform chemical separations on the liquid waste to generate a high-activity fraction and a low-activity fraction. All of the waste is to be converted to a solid waste form, al- though the technology to be used and the waste form have not been final- ized. Buried TRU As noted in Section 1.2, âburied TRUâ is waste that meets the defi- nition of TRU waste, but was disposed of in near-surface pits and trenches prior to the practice of retrievable storage for disposal at the Waste Isolation Pilot Plant (WIPP). Buried TRU at INEEL The Idaho National Engineering and Environmental Laboratory con- tains the Radioactive Waste Management Complex (RWMC) in the southwest portion of the site. The RWMC consists of about 177 acres contained within natural and constructed earthen dikes and was estab- lished in 1952 to handle the testing stationâs waste. It was also the dis- posal location, starting in 1954, for virtually all of the radioactive waste 17 generated by the Rocky Flats Plant in Colorado and shipped off-site. Indeed, practically all (on the order of 95-98 percent) of the waste in the RWMC is from Rocky Flats. The Subsurface Disposal Area, a 97-acre portion of the RWMC, has been used to dispose of 56,920 mÂ³ of radioactive waste from Rocky Flats, INEEL, and a few other sites. This waste includes buried TRU (or- dinary TRU and mixed TRU waste âirretrievablyâ disposed prior to 1970), alpha-LLW (ordinary and mixed low-level waste [LLW] that con- tains TRU isotopes with alpha activity between 10 and 100 nCi/g), and LLW (mixed LLW until 1984 and ordinary LLW to the present). For regulatory purposes, both buried TRU and alpha-LLW, an estimated total 17 Large amounts of waste remained onsite at Rock Flats, mostly in the form of contamination in buildings, in pipes and equipment, and in the soil.
56 RISK AND DECISIONS ABOUT TRU AND HLW 36,800 m3 containing 297,000 Ci of TRU radioactivity (DOE, 2000),18 are being managed by INEEL as TRU waste. Much of the site is slated for cleanup under the Comprehensive Environmental Response, Com- pensation, and Liability Act (CERCLA), and the preliminary plan was to exhume most of the buried TRU waste from the RWMC (about 90 per- cent by volume and about 70 percent of the TRU radioactivity). The re- trieved waste would be treated and shipped for disposal at WIPP. The waste was disposed of in several units: â¢ 20 pits containing contact-handled TRU waste (pits 1-6, 8-12), LLW, and volatile organic compounds; â¢ 58 trenches containing remote-handled TRU waste (trenches 1- 10, 19, 32) and LLW; â¢ 21 soil vault rows (narrow trenches with concrete lining) contain- ing remote-handled TRU and LLW; â¢ Pad A containing LLW and waste contaminated with nitrate salts and uranium; and, â¢ one acid pit containing partially grouted, contaminated soil. The management of TRU waste is difficult in a number of ways: it poses serious radiological hazards (to varying degrees); some (so-called mixed waste) also poses toxicological hazards; a small portion is fissile material, which raises criticality concerns; and it is also possible that classified waste was mistakenly disposed of, which raises security con- cerns, although the Rocky Flats Plant assured INEEL that this is not the case. Waste from the post-1970 period appears to have been well docu- mented when placed in storage, and its present configuration was de- signed to facilitate retrieval. (The process of characterizing it and prepar- ing it for shipment off-site is nevertheless extraordinarily elaborate and expensive.) However, the pre-1970 material in the Subsurface Disposal Area is highly miscellaneousâphotographs of the disposal process show barrels, boxes, trucks, industrial equipment, and debris simply dumped into the disposal unitsâand was not meant to be retrieved. Through an 18 This represents only the radioactivity from isotopes meeting the definition of TRU, decay corrected to 2006.
WHY CONSIDER FLEXIBILITY IN DISPOSAL OPTIONS? 57 effort termed Ancillary Basis for Risk Analysis (Holdren et al., 2002; Zit- nik et al., 2002), INEEL has identified several contaminants of concern for each pathway from the Subsurface Disposal Area. Pit 9 was chosen as a demonstration case for remediation of the op- erable unit that covers the Subsurface Disposal Area under CERCLA. Pit 9 was selected because it is at the western edge of the Subsurface Dis- posal Area, near the road access, so operations there are unlikely to dis- turb other pits and trenches, and because waste in the pit was disposed of more recently, so records were expected to be more useful for characteri- zation. Efforts at carrying out the selected remedy (physical separation, treatment, and stabilization) proved much more difficult than expected, mostly because of contamination of the soil in the pit, which poses inha- lation hazards. After spending tens of millions of dollars on subsurface characterization and after a $200-million failed start with the first attempt at limited retrieval, DOE successfully completed a demonstration project (the glovebox excavator method, or GEM, pilot project) that recovered about 454 barrels (about 77 cubic meters) of waste from Pit 9 at a cost of $79 million (including the cost to build the facility). Buried TRU at Hanford By volume, the Hanford Site has more buried TRU waste than any 3 other site, with a total of 75,800 m . But this waste contains a little over 60,000 curies of TRU activity, or about one fifth as much as INEELâs 3 buried TRU waste (DOE, 2000). In addition, Hanford has over 31,000 m of TRU-contaminated soils containing 32,400 curies of TRU activity as a result of liquid waste discharges in cribs, ditches, and trenches. DOE plans to manage most of these wastes by containing them in place. Two companion burial grounds that DOE plans to exhume (DOE, 2000) likely constitute some of the most difficult technical challenges faced by DOE with respect to TRU waste at Hanford: The 618-10 and 618-11 Burial Grounds. This summary is based on information provided to a workshop of technical experts sharing experience in dealing with buried TRU waste (Hulstrom, 2003) and on information provided during the committeeâs visit to the Hanford Site. The 618 Burial Grounds received waste from the Hanford 300 Area. The 300 Area was used for fuel fabrication, research and development activities (pilot-scale tests) supporting the development of processes used in the 200 Area (e.g., PUREX), and other activities such as those devel- oped in the Plutonium Recycle Test Reactor (PRTR) facility. Exact in-
58 RISK AND DECISIONS ABOUT TRU AND HLW ventory records are limited and often contradictory and, in some cases, based only on interviews with people who operated the facilities. The 618-10 Burial Ground, which operated from 1954 to 1963, oc- cupies approximately 2.3 hectares (5.7 acres) about 3.8 kilometers west 3 of the Columbia River. It contains an estimated 98,000 m of waste, in- 3 cluding 8.4 m of remote-handled transuranic waste (RHâTRU). These 3 8.4 m are the only TRU wastes in the burial ground. Wastes were dis- posed of in 12 trenches and 94 âvertical pipe units.â Most trenches are presumed to contain low-level waste and low-level mixed waste. The vertical pipe units are estimated to contain a mixture of low-level mixed waste and RHâTRU. In 1961, a fire occurred in one trench. During stabi- lization operations in 1983, oil puddled to the surface indicating the breach of a container and the presence of liquids. The 618-11 Burial Ground operated from 1962 to 1967. It spans 8.6 acres and is located 3.6 miles west of the Columbia River. The 618-11 3 3 Burial Ground contains an estimated 78,000 m of waste, with 94 m of 3 RHâTRU and 10,200 m of contact-handled transuranic waste (CHâ TRU). Wastes were disposed of in three trenches, 50 vertical pipe units, and three to five caissons. Similar to the 618-10 Burial Ground, the trenches predominantly contain low-level waste and low-level mixed waste. The vertical pipe units and caissons are estimated to contain mostly RHâTRU. The radiological hazards presented by these burial grounds include cesium, strontium, thorium, uranium, plutonium, americium, curium, and neptunium. Other hazards include beryllium, uranium, and zirconium metals, and sodium-potassium metals (some of which are pyrophoric), petroleum products, and organic chemicals. The wastes believed to be in the burial grounds include spent nuclear fuel, HLW, CHâ and RHâTRU waste (some mixed), and LLW (some mixed). Hanford officials in- formed the committee that radiation levels at the edge of the burial grounds have been measured to be as high as 5 rem per hour and contact doses as high as 500 rem per hour. Although the general practice was to place the higher-activity waste in the vertical pipe units or caissons, some such waste likely exists within the trenches. The waste is both a source of environmental contamination and a hazard to remediation workers and inadvertent intruders. A tritium plume in the vicinity of the 618-11 Burial Ground has concentrations of 8.1 mil- lion pCi/L (400 times the drinking water standard). The estimated time for a contaminant to travel from the burial grounds to the Columbia River is between 3 to 30 years. As noted above, the RHâTRU waste may have contact doses of up to 2500 times the contact-handled limit. Fur-
WHY CONSIDER FLEXIBILITY IN DISPOSAL OPTIONS? 59 ther, some pyrophoric waste (easily ignited materials) may be present in the burial grounds (Hulstrom, 2003). Cleanup of these burial grounds is difficult because the contents are poorly characterized, diverse, reactive, and intensely radioactive. DOE is currently exploring technologies that might be used for cleanup. How- ever, there is no clear plan or schedule for such cleanup, and DOE repre- sentatives expressed the hope that these burial grounds could be closed without retrieval because of the high cost and occupational risk antici- pated in any attempt to retrieve and treat the wastes. 2.3 THE NEED FOR FLEXIBILITY IN DISPOSAL OPTIONS IDENTIFIED The foregoing descriptions of three waste types and plans for manag- ing them illustrate the variability in the composition and condition of each waste type and the costs of retrieval or waste-form production. This discussion does not provide a complete picture: The descriptions of the wastes could be translated into an understanding of the hazards of the wastes, but a description of exposure pathways and scenarios is needed to understand the risks they pose. The analysis of costs is incomplete, with little treatment of uncertainty and no sensitivity analysis. And other impacts, such as those on ecosystems, are not addressed. But while there is not yet sufficient analysis to support a decision on how to manage each waste stream, the information presented indicates a need. Finding 4: The nation needs a way to decide which of the wastes mentioned in Finding 3 should be disposed of in a deep geologic re- pository and which, if any, should be allowed alternative disposal. Litigation over authority and agreements about waste disposition left DOEâs waste disposition program with substantial uncertainty concerning the path forward. Provisions in the Defense Authorization Act of 2005 cre- ate a process for addressing HLW at the Savannah River Site and at the Idaho National Engineering and Environmental Laboratory, although many details remain to be resolved and Hanford is not affected directly by the legislation. TRU waste already had an exemption provision, but dis- putes remain. Given the various disputes and the reality that not all of the waste will or can be recovered and disposed of in a deep geologic reposi- tory, an acceptable process for deciding what wastes require repository disposal is still needed. The recent legislation should remove some of the
60 RISK AND DECISIONS ABOUT TRU AND HLW obstacles to DOEâs working with others in South Carolina and Idaho to implement the approach recommended in this report.