V
Tank Grouting and Closure

Cementitious materials are used worldwide to immobilize low- and intermediate-level radioactive wastes (IAEA, 1999, 2000, 2004).1 Historically, grout has been one of the most commonly used materials for solidifying and stabilizing these wastes, and its technology is at a mature stage of development. Grout stabilization of Resource Conservation and Recovery Act (RCRA) heavy metals (e.g., chromium, lead, mercury) is standard technology for producing waste forms that meet U.S. Environmental Protection Agency (EPA) requirements (NRC, 1999c). The committee agrees in principle with the Department of Energy’s (DOE’s) choice of these Portland-cement based grouts for immobilizing tank waste residues by filling emptied tanks with grout, recognizing that this is an essentially irreversible action. However, there are numerous caveats that arise from DOE’s unusual applications of these materials that are outside the construction industry’s experience. The committee discusses these issues in the first half of this chapter.

The committee recommended in its interim report that DOE “decouple” tank waste removal and tank closure actions where there are indications that significant amounts of radioactive material are present in the tank after cleanout operations have ended. The committee also recommended that DOE work with the State of South Carolina to revise closure milestones, if necessary (see Appendix E, Recommendation 1). In this chapter, the committee reiterates this finding and recommendation and extends them to the Hanford and Idaho sites.

Decoupling does not imply delaying a site’s tank closure program. Decoupling means that for a given tank, once the planned waste removal program has been completed, there is an objective evaluation of the result. Only after this evaluation is a decision made to proceed with the essentially irreversible step of tank grouting or to execute additional waste removal operations. Additional removal operations would likely employ new approaches according to lessons learned from the previous operations. Decoupling is discussed in the second part of this chapter.

USE OF GROUT FOR TANK CLOSURES

By immobilizing the waste and acting as a barrier around it, grouting can reduce the likelihood that the waste will cause harm (see Chapter VI). However, grouting does nothing to reduce the hazard of the waste itself and can be viewed as “treating the symptom rather than the disease.” Grouting sludge heels or other tank wastes on-site is not a substitute for removing radioactive materials to the maximum extent practical, as discussed elsewhere in this report, and disposing of them in a geologic repository.

Freshly prepared grout can be mixed intimately with waste to be stabilized (e.g., salt waste at the Savannah River Site) and the mixture pumped into its final containment where it solidifies (such as the Savannah River Site’s saltstone vaults; see Chapter IV). Alternatively, especially for stabilizing low-level solid wastes, the grout is often poured in and around objects in a drum or larger container, and the containers themselves can be embedded in grout. As will be discussed in this section, DOE’s plans to grout tank waste residues are a hybrid of both practices.

Experience with concrete in the construction and oil industries is extensive, and the materials are relatively inexpensive. Like any materials used in engineering and construction, these materials have well-defined operating

1

Portland cement is a mixture of silicates and aluminates of calcium obtained by firing (usually) a mixture of limestone and clay in a rotary kiln. The solid material (clinker) formed on cooling this molten mass is ground with a small quantity of gypsum. Portland cement reacts with water to form a solid that is stable in water. “Concrete” is a solid product that results from mixing cement, water, aggregate (sand and gravel or crushed stone), and admixtures that may affect its chemical or physical properties. “Grout” is a mixture of cement and water with or without aggregate, proportioned to produce a pourable consistency. Waste materials (sludge, salt) generally behave as additives and become chemically or physically incorporated into the solid material.



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 62
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report V Tank Grouting and Closure Cementitious materials are used worldwide to immobilize low- and intermediate-level radioactive wastes (IAEA, 1999, 2000, 2004).1 Historically, grout has been one of the most commonly used materials for solidifying and stabilizing these wastes, and its technology is at a mature stage of development. Grout stabilization of Resource Conservation and Recovery Act (RCRA) heavy metals (e.g., chromium, lead, mercury) is standard technology for producing waste forms that meet U.S. Environmental Protection Agency (EPA) requirements (NRC, 1999c). The committee agrees in principle with the Department of Energy’s (DOE’s) choice of these Portland-cement based grouts for immobilizing tank waste residues by filling emptied tanks with grout, recognizing that this is an essentially irreversible action. However, there are numerous caveats that arise from DOE’s unusual applications of these materials that are outside the construction industry’s experience. The committee discusses these issues in the first half of this chapter. The committee recommended in its interim report that DOE “decouple” tank waste removal and tank closure actions where there are indications that significant amounts of radioactive material are present in the tank after cleanout operations have ended. The committee also recommended that DOE work with the State of South Carolina to revise closure milestones, if necessary (see Appendix E, Recommendation 1). In this chapter, the committee reiterates this finding and recommendation and extends them to the Hanford and Idaho sites. Decoupling does not imply delaying a site’s tank closure program. Decoupling means that for a given tank, once the planned waste removal program has been completed, there is an objective evaluation of the result. Only after this evaluation is a decision made to proceed with the essentially irreversible step of tank grouting or to execute additional waste removal operations. Additional removal operations would likely employ new approaches according to lessons learned from the previous operations. Decoupling is discussed in the second part of this chapter. USE OF GROUT FOR TANK CLOSURES By immobilizing the waste and acting as a barrier around it, grouting can reduce the likelihood that the waste will cause harm (see Chapter VI). However, grouting does nothing to reduce the hazard of the waste itself and can be viewed as “treating the symptom rather than the disease.” Grouting sludge heels or other tank wastes on-site is not a substitute for removing radioactive materials to the maximum extent practical, as discussed elsewhere in this report, and disposing of them in a geologic repository. Freshly prepared grout can be mixed intimately with waste to be stabilized (e.g., salt waste at the Savannah River Site) and the mixture pumped into its final containment where it solidifies (such as the Savannah River Site’s saltstone vaults; see Chapter IV). Alternatively, especially for stabilizing low-level solid wastes, the grout is often poured in and around objects in a drum or larger container, and the containers themselves can be embedded in grout. As will be discussed in this section, DOE’s plans to grout tank waste residues are a hybrid of both practices. Experience with concrete in the construction and oil industries is extensive, and the materials are relatively inexpensive. Like any materials used in engineering and construction, these materials have well-defined operating 1 Portland cement is a mixture of silicates and aluminates of calcium obtained by firing (usually) a mixture of limestone and clay in a rotary kiln. The solid material (clinker) formed on cooling this molten mass is ground with a small quantity of gypsum. Portland cement reacts with water to form a solid that is stable in water. “Concrete” is a solid product that results from mixing cement, water, aggregate (sand and gravel or crushed stone), and admixtures that may affect its chemical or physical properties. “Grout” is a mixture of cement and water with or without aggregate, proportioned to produce a pourable consistency. Waste materials (sludge, salt) generally behave as additives and become chemically or physically incorporated into the solid material.

OCR for page 62
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report FIGURE V-1 Engineered barrier system to close tank at the Hanford site. A similar plan of engineered grout layers plus a cap system is adopted at the Savannah River Site and the Idaho National Laboratory. SOURCE: Sams, 2005. envelopes in which they can be used with confidence, but outside these limits their performance may be uncertain. The use of grout for the long periods involved in the disposal of radioactive waste (more than 100 years) is outside the general operating envelope for cementitious materials in industrial applications. There are no good precedents for cementitious materials to maintain very low permeability to water and other properties necessary to retain radionuclides for the very long times required by DOE. After a tank has been emptied, it must be filled with a solid material to prevent potential collapse of the roof and walls due to the weight of the overburden and the lateral pressure from the surrounding soil. Such collapse would not occur immediately after emptying the tanks but would be the result of corrosion and aging of the tank structure. The potential collapse of the structures could cause a subsidence of the ground surface (final tank farm closure grading), affecting surface water drainage. Filling the tank with solid material limits such a collapse. DOE plans to close emptied tanks by placing one or more layers of engineered grout in them to provide the structural support described above, encapsulate and stabilize the tank heel, and act as a physical barrier that inhibits the flow of water through the residual waste. Some tanks would have a high-strength layer of grout that would serve as an intruder barrier. Engineered covers to retard infiltration to the tanks after closure are also under consideration at the three sites (see Figure V-1). Engineered Grouts for Tank Closure Specially formulated grouts have been developed to backfill the tanks after the waste retrieval is deemed complete. These grouts are being used for tank closures at the Savannah River Site and are planned for tank closures at the Idaho National Laboratory, as discussed in the sections that follow. A layered system of different types of grout is part of the engineered barrier system designed to reduce groundwater infiltration into the radioactive sludge layer. The main requirements of the engineered grouts used to immobilize radioactive waste follow: They must be suitable for pumping into the tanks, typically through long pipes or “tremies” for placement in a tank without segregation throughout; They must provide near- and long-term high pH and chemically reducing capabilities to maintain the radionuclides and toxic heavy metals, such as technetium and neptunium, in their least mobile chemical forms (i.e., low-oxidation state or reduced form) (Buice et al., 2005); and They must minimize the flow of water through the material (and the consequent leaching of radionuclides and metals from the grout). The cementitious materials to be used to fill the tanks are a mix of Portland cement, ground granulated blast furnace slag, and fly ash. Portland cement enables the grout to set

OCR for page 62
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report (solidify) and gain strength in a reasonable amount of time. It also gives the grout a high pH (approximately 12, which is highly alkaline). Slag (a byproduct of the steel industry) gives the grout a low reduction-oxidation potential, or Eh (i.e., a “reducing” grout). Fly ash (a by-product of coal-fired power plants) helps minimize thermal cracking by limiting the heat generated by the grout during the curing process. The high pH and low Eh should reduce the solubility and mobility of many radionuclides if they are well mixed with the cement matrix or if contacted by water that has been sufficiently altered by the cement matrix. The specific proportions of cementitious materials in the grout are modified to optimize its ability to immobilize the waste, based on an analysis of the waste. At the Savannah River Site, DOE also intends for the grout to serve as a barrier to inadvertent intrusion by burrowing animals or humans drilling or excavating, because it would be clearly different from the native soil. Uncertainties in DOE Tank Grouting The committee agrees with DOE’s selection of grout for tank closures because of the extensive experience base and relatively low cost for DOE’s near-term (approximately 20 year), large-scale needs to immobilize waste. A previous review of materials for stabilizing waste (NRC, 1999c) did not identify any promising material that might be superior to grout for DOE’s tank closures. The committee does not foresee the development of better alternatives and neither DOE nor the committee judges it necessary to explore alternative tank fill materials. However, as noted previously, the use of grout for tank closure is unique both in the basic construction challenges it presents and in DOE’s use of the material to encapsulate tank residues for very long periods of time. In reviewing DOE’s plans for tank grouting, which are detailed in the following section, the committee developed two sets of concerns (see Sidebars V-1 and V-2). These concerns highlight and summarize lacunae in present knowledge, that DOE must address—many on a tank-by-tank basis—to ensure effective radionuclide immobilization. Simply pumping grout into mostly emptied tanks may not fulfill DOE’s responsibilities to its regulators, public stakeholders, or Congress under Section 3116 of the 2005 National Defense Authorization Act (NDAA). Savannah River Site staff is doing extensive work in developing grout formulations for tank wastes and estimating how these grouts might perform, working to address some of the concerns discussed in Sidebars V-1 and V-2. Recent studies improve grout production and batching, grout flow, and measurement of the effective diffusion coefficient of technetium-99 in reducing bulk fill grout. An ongoing cooperative program with the Khlopin Radium Institute in Russia is addressing the modeling of technetium-99 stabilization in grout and possible improvements. An evaluation of alternative materials and admixtures for achieving zero-bleed (i.e., no water separation from grout mix), self-leveling grouts is also going on. The Savannah River Site is also continuously updating its knowledge base on radionuclide leaching from grout as new data are generated, and it is continuing research to combine design features of the reducing grout and bulk fill grout (Langton, 2005). Although most of the information gathered by the committee was provided by Savannah River Site research personnel, Hanford and the Idaho National Laboratory have also done work in this area and benefit from the knowledge gathered at the Savannah River Site (Quigley, 2005; Sams, 2005). In a more fundamental approach, researchers at the National Institute of Standards and Technology (NIST) have developed a model that contributes significantly to understanding and predicting changes in the microstructure and transport properties of grout materials over long times (Garboczi et al., 2004). Tank Grouting at the Savannah River Site Tanks 17 and 20 at Savannah River Site were emptied (see Chapter III), and they have been backfilled with three layers of cement grout and closed. Plans are being finalized to grout and close Tanks 18 and 19. The grout materials are designed to reduce the mobility of any radionuclides and toxic heavy metals remaining in the tanks after cleaning and to lend structural stability to the tanks themselves. In Tanks 17 and 20, the bottom layer, which is in contact with the radioactive residual sludge, is an engineered grout called “smart grout.” The middle layer, the thickest of the three, is a low-strength grout (bulk fill), and the top layer is a harder grout intended to serve as a barrier against inadvertent intruders. The smart grout was formulated to generate less heat of hydration than ordinary portland cement grout and was placed in a series of lifts to allow time for some of the heat of hydration to dissipate to minimize cracking. The plan for Tanks 18 and 19, and currently for future tanks, is to have two layers of cement grout: a thick layer similar to the smart grout and a top layer of higher-strength grout to act as a barrier to intruders (DOE-SRS, 2005a). DOE recognizes that there is effectively no mixing of grout with the insoluble waste heel. Having resisted attempts to remove them (see Chapter III), waste heels are likely to be in inaccessible locations and practically immovable. In addition, there are physical limitations on where the grout can be discharged into a tank (tremie2 placement) and differences in density and viscosity between the cementious material and the tank heel (DOE-SRS, 2005a; USNRC, 2005). In Tanks 17 and 20 a series of tremie placements was made around the circumference of the tank to lay down the first grout layer and contain the tank residues rather than displace them 2 A tremie is a pipe used to convey and deposit grout or concrete rather than simply pouring the material from a height.

OCR for page 62
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report SIDEBAR V-1 Construction Challenges Pertaining to Grouting of Tanks The ability of grouted waste and grout-filled waste tanks to provide the long-term radionuclide immobilization that DOE is anticipating, as described throughout this report, depends greatly on success in meeting several challenges, many of which are well known by the construction industry. Each tank will present variations on these challenges based on its limitations for physical access; internal obstructions; and the amount, location, and properties of residual waste deposits. The adequacy of DOE’s tank closures will depend on careful consideration of each of the following on a tank-by-tank basis: Pumping. Because of the limited access to the tank and the large volume of grout that must be placed, it is likely that the fill material (concrete or grout) would be placed by pumping; hence it must remain pumpable during the entire time of placement. Flow characteristics. The grout must flow to the walls of the tank from the point(s) of placement while retaining its integrity—that is, it cannot segregate into its constituent ingredients. Degree of mixing with or encapsulation of waste. Based on the results of mockups and the few tanks that have been grouted, the heel material does not mix with the grout to an appreciable degree. Also, since the grout remains on top of the heel, encapsulation of the waste is incomplete. However, placement techniques may influence the final distribution of heel and grout material, resulting in better encapsulation of the heel. Inhomogeneity of grout. Because the waste liquid remains on top of the heel in some tanks, there may be some mixing or sorption of the liquid into the grout. In some cases, dry cementitious materials are to be placed pneumatically on top of the liquid waste to stabilize the free liquids. The resulting inhomogeneities must be evaluated to determine whether they affect the overall performance of the grouted tank. Effectiveness of grout (filler) in immobilizing waste. If the walls and pipe surfaces of a tank cannot be adequately cleaned, some radioactive waste will remain above the elevation of the grouted bottom layer. Thus, the low-strength “filler” would require some ability to immobilize the waste. If the low-strength filler does not have the required capability, it may be necessary in some cases to use higher-quality grout to fill the entire tank. Heat generation. The hydration reactions of all cementitious materials evolve heat. Because grout does not conduct heat well, the temperature within the grout can rise significantly and lead to cracking. Heat generation must be controlled by proportioning the grout appropriately using materials that generate little heat and then allowing the heat to dissipate in a way that avoids thermal cracking (see cold joints, below). For the use of grout to be acceptable, either the grouting must not result in thermal cracking, or cracking must not result in significant adverse effects on the performance of the grouted tanks. In either case, testing and analysis are needed to verify DOE’s expectations Long-term monitoring. As discussed in detail in Chapter VII, Sect. 3116 of the NDAA requires a post-closure monitoring program. One desirable component of such a program is to monitor the performance of the waste form to ensure it is performing as expected and to provide early detection of radionuclide release. To accomplish this, it would be helpful if grout construction could be designed to allow the desired monitoring to occur, as is recommended in Chapter VII. Cold joints. One means of managing the heat generated is to use a series of grout placements (i.e., lifts) rather than place all the grout at once, allowing the heat to dissipate between each lift. In normal construction, specific measures must be taken to ensure that the concrete behaves as a monolith across such “cold joints” between the lifts. Because of the limited access into the tank and the hostile environment, it may not be possible to take such measures. Thus, the presence of cold joints must be taken into account in assessing the performance of the grouted tanks. Alternative grouting formulations and techniques need to be tested in mockups (as is done routinely in construction projects), which allows the contractor to gain experience, and verify the properties of the grout as placed. In addition to the construction industry, the oil industry has developed a great deal of expertise in grouting of areas with difficult or limited access. The U.S. Army Corps of Engineers also has personnel who have participated in the development of related technology whose expertise could be brought to bear on these problems. toward the walls. Documents indicate that there were small areas of incomplete grout coverage at the intersections of grout deposited by different tremie placements (USNRC, 1997a). The smart grout covered the fixed, insoluble waste particles (the solid heel, containing primarily actinides and strontium) and displaced the liquids. The liquids, which contain technetium, other soluble radionuclides, and some suspended insoluble particles, were largely displaced to the top of the grout. They were absorbed by a second layer of dry grout to provide further immobilization of the waste. After placing more smart grout on top of the dry grout, an improved version of a “controlled low-strength material” was then added above this to fill much of the tank, inhibiting water flow (a hydraulic barrier) and preventing collapse. Finally, a third layer of a higher-strength grout material was used to fill the voids around the risers and to act as an intruder barrier. DOE’s estimates of grout behavior over time do not assume that the waste is mixed in the grout, but they do

OCR for page 62
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report SIDEBAR V-2 Unusual Requirements of DOE Grout Applications DOE has relied heavily on knowledge of the characteristics of grout and its long-term behavior gained by the construction industry. While this approach is appropriate, it is not necessarily sufficient for the purposes of stabilizing radioactive wastes. The following are some topics on which experience from construction applications alone provides an inadequate knowledge base for DOE’s applications. Additional information, most likely from a research and development program, is needed to provide the necessary understanding of the behavior of the fill material (grout or concrete) over the long term so that appropriate grout formulations can be selected and performance assessments can be based on valid assumptions. Compatibility of grout with liquid waste. In most cases, some liquid waste remains on top of the heel. As the grout flows into place, it either mixes with or absorbs water from this liquid. The radionuclides, toxic heavy metals, and other chemical constituents of this liquid may locally affect such characteristics of the grout as its ability to set or its durability.a Exposure to radiation. Experience in the nuclear industry has established that properties of cementitious materials can be affected by high levels of radiation (e.g., in reactor shielding). Lower levels of radiation may or may not affect the properties of the grout material.b Radiation levels from tank wastes are much lower than in reactor applications and decrease with time. However, the effects of persistent radiation from the tank waste on grout performance have not been evaluated. It is possible that the radiation levels are not sufficient to cause deterioration, or that even deteriorated grout is satisfactory in this application, but this has to be established. Deterioration of the tank floor and sides. It is anticipated that the carbon steel tank floors and sides will eventually corrode away. The concrete slabs and vaults in which the tanks sit were not designed as long-term containment. Performance assessments must continue to account for theeffects of the eventual loss of these barriers. Reducing capabilities of grout. The ability of the grout to stabilize radionuclides and toxic heavy metals rests on its high pH and low Eh. High pH is important in the construction industry because it helps protect embedded steel against corrosion. Thus, there is some understanding of the mechanisms of loss of pH over time. However, Eh has no particular relevance in construction; thus, much less is known about its persistence over time. Extremely long service lives. In the construction industry, a typical service life is on the order of 50 to 100 years, and regular maintenance is necessary to achieve it. While examples of ancient concrete still survive today, they are exceptional and have little relation to modern construction materials and techniques or to service conditions that will be encountered in the tanks. DOE seeks to place grout that retains its properties in some form for 500, 1,000, or even 10,000 years. Often the strategies for durability in the construction industry involve postponement or slowing of deterioration rather than prevention. These strategies have to be reconsidered for the extended service lives required in the tanks. High groundwater table. In a few cases, the elevation of the groundwater table is above that of the tank floor. Coupled with the likely deterioration described above, this could result in the tank heel coming in contact with groundwater that has not been substantially altered by the chemically tailored grout atop the heel. The performance assessment of these tanks must include this condition, and DOE may want to consider more thorough cleaning or other means to reduce the risks associated with these tanks.    a A similar issue has been raised in the context of low-activity waste disposal given the potential for interaction between the chemicals in the waste and the grout. While chemical compatibility could be a problem for saltstone, the committee has not examined the issue and has no evidence that it is a problem.    b Although the doses required are high (see, e.g., Utsunomiya et al., 2003), the principal concern would be that penetrating beta rays could cause solid state radiolysis in hydrated phases, such as those present in grout and zeolites. assume that the grout continues to be an intact hydraulic barrier for 500 years and maintains its alkalinity and reducing capability for 10,000 years. Despite the considerable amount of work performed by DOE contractors, the committee received little quantitative (experimental or other) information to support the 500-year and 10,000-year assumptions. Langton and coauthors describe the different needs and challenges for waste tanks at each DOE site, tank fill materials placement requirements, leaching and durability properties, and technology needs to demonstrate tank fill physical and leaching properties (Langton et al., 2001). The committee is aware of a qualitative analysis of the tank waste grout from 1992 (Lokken et al., 1992), but of only one recent experimental study, which is on the leaching characteristics of grout with respect to technetium-99 (Harbour et al., 2004). DOE’s Performance Objectives Demonstration Document for the Closure of Tank 19 and Tank 18 (Buice et al., 2005) contains a set of calculations concerning reducing capabilities of the grout and examines the 10,000-year assumption. These calculations are discussed in Chapter VI. The committee believes that the short- and long-term performance of tank fill materials warrant further research to bridge a knowledge gap (see Chapter IX).

OCR for page 62
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report Tank Grouting at Hanford Hanford has not yet finalized its closure plan for single-shell tanks. An environmental impact statement has been initiated to assess the final closure configuration of the single-shell tanks after retrieval. For planning purposes, Hanford is currently assuming a landfill closure configuration that utilizes grout as the fill material based on research and field experience gained at the Savannah River Site. Gravel, concrete, and other materials have been considered as fill materials, but their performance is inferior to—or their handling is more complex than—that of grout. According to presentations to the committee, the Hanford tanks would be filled in three layers of flowable grout: Layer 1: a 30 to 90 cm (1 to 3 feet) layer of free-flowing grout that will cover waste residuals and debris on the tank bottom and support subsequent fills. Layer 2: grout that will enhance stability of the tank structure and fill the majority of the tank. Layer 3: high-compressive-strength grout placed in the remaining void space to discourage intrusion. Savannah River National Laboratory staff performed scaled testing (lab, bench, and large scale) to develop grouts with properties suitable for Layer 1 in the Hanford tank and waste environments, which differ from conditions at the Savannah River Site (Harbour et al., 2004). This study found that waste particles at the bottom of the tanks would be only partially encapsulated by the grout. However, the grout would be able to penetrate many of the interstitial regions. The study also found that the stabilizing layer should provide a reducing environment, thus decreasing the mobility of contaminants of concern. The layer would also provide a physical barrier to slow the release of these contaminants in the environment. Both multipoint and single-point tremie placements have been evaluated to accommodate various riser configurations, but additional tests are needed (Langton et al., 2003). Hanford is also planning a tank closure demonstration on one of the smaller C-200 series tanks after completion of waste retrieval. The purpose of the demonstration is to verify tank stabilization by core sampling of the grout layer. This work will also include characterizing contaminated soil outside the tank and stabilizing it by impermeable barrier installation; characterizing and stabilizing one diversion box and direct buried pipelines by in situ grouting; and characterizing and isolating in-trench pipelines. DOE also plans to continue grout formulations studies for Hanford-specific applications. Given the early stage of its tank closure plan, Hanford has the opportunity to benefit from continuing dialogue on tank closure with the other DOE sites (including Oak Ridge and West Valley). Tank Grouting at the Idaho National Laboratory Idaho National Laboratory staff demonstrated a method of placing grout onto a tank floor to permit retrieval of additional slurry from the tank using a variable-depth steam jet. Five sequential placements of the grout pushed liquid toward the jet intake, allowing removal of additional liquid from the large-diameter tanks (see Figure V-2). The sequential placement technique was developed when the site did a 1999 mockup test. A mockup of a tank was constructed at an Idaho Falls industrial facility. This was a full-scale horizontal slice of the bottom of a tank, only a few feet in height, but 50 feet in diameter, with cooling coil structures and simulated residual solids. Grout mixtures were tested to validate assumptions of flowability in both the tank and the surrounding vault areas, and the ability of the grout to move the in-tank solids toward the steam jet (INEEL, 1999). The main assumptions about the tank fill material used in the Idaho National Laboratory tank closure performance assessment are that the outer vault grout fails at 100 years, tank and tank grout fail at 500 years, and piping fails at 500 years. The main difference from the other two sites’ plans for tank grouting is that Idaho’s tank closure plan does not include a layer of high-compressive-strength filler material to serve as an intruder barrier. Idaho National Laboratory staff assumes that intruders who might attempt to drill in the tank farm area would expect to encounter basalt flows; therefore, the presence of a high-compressive-strength grout on the top of the tanks would not necessarily prevent drilling. The grout formulation has not yet been finalized at the Idaho site. Appendix C of the Draft 3116 waste determination (DOE-ID, 2005a, Appendix C) reads: The grout planned for use at Idaho is expected to exhibit strongly reducing conditions, as in Hanford and Savannah River tank closure plans. However, current Tank Farm Facility analysis concludes that reducing conditions in the grout are not necessary to demonstrate compliance with performance objectives. Idaho’s Department of Environmental Quality (IDEQ) has approved partial closure plans (i.e., not for the whole tank farm) for the 1136 m3 (300,000-gallon) tanks WM-182, 183, 184, 185, and 186. DOE has submitted, but IDEQ has not yet approved, partial closure plans for Tanks WM-180 and 181 (300,000-gallon tanks) and WM-103, 104, 105, and 106 (30,000-gallon tanks). DOE has not submitted closure plans for the remaining 300,000-gallon tanks or other portions of the tank farm because those tanks store waste that DOE plans to treat by steam reforming and ship for disposal off-site. The site is collaborating closely with the Savannah River and Hanford Sites on grout formulation and placement methods. Idaho National Laboratory is also working with other DOE sites, the Pacific Northwest National Laboratory,

OCR for page 62
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report FIGURE V-2 Plan for sequential placement of grout in tanks at the Idaho National Laboratory. SOURCE: Lockie et al., 2005. and the United Kingdom Atomic Energy Authority on issues related to tank closure. DECOUPLING WASTE REMOVAL FROM TANK CLOSURE In its interim report about the Savannah River Site, the committee found that tank closure milestones make tank waste removal and tank grouting schedules appear “coupled” (i.e., one following the other as soon as possible) for some tanks (see Appendix E, Finding 1b). For example, in the case of the draft Section 3116 waste determination for Tanks 18 and 19, the milestone for closing these two tanks was one of the top three criteria for determining that “waste has been removed to the maximum extent practical” (DOE-SRS, 2005a). In that report the committee recommended that retrieval and closure not necessarily be closely coupled, especially for tanks containing significant amounts of residual radionclides. Subsequent to the interim report, DOE and the State of South Carolina reiterated their preference for closing tanks soon after retrieval is completed (see the section on objections to decoupling, below). The committee remains concerned that DOE is defining what is practical, in at least some cases, by what is required to meet a milestone or by the letter of the law (radiation doses at a far-future time), rather than making decisions based on sound science and engineering judgment. According to information reviewed by the committee, the volume of sludge residues left in a tank after waste retrieval is completed may vary by two orders of magnitude (a hundredfold). This is not necessarily bad; it simply reflects the inherent uncertainty in expected tank cleaning results at this early point in DOE’s program. Reducing waste volume in a million-gallon tank down to 100,000 gallons is 90 percent removal; reducing waste volumes to 10,000 gallons would be 99 percent removal. Table VI-3 shows some very optimistic assumptions in DOE’s environmental impact statement for tank closure at the Savannah River Site. Many or most tanks were assumed to contain only 100 gallons after retrieval was completed. The accompanying discussion in Chapter VI suggests that residues of 5,000 to 10,000 gallons or more are more consistent with experience with the methods used in the most recent residual waste removal campaigns. The committee concurs with DOE that achieving near-term risk reduction by removing 90 to 99 percent of the waste volume can and should be accomplished as soon as possible. However, the committee does not agree with what appears to be a milestone-driven rush to grout a tank essentially permanently and irrevocably even if much more radioactive

OCR for page 62
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report material remains than expected. For a problematic tank, decoupling waste removal from grout closure—that is, allowing opportunity for objective assessment of the results and reassessment of the path forward—is essential. Advantages of Decoupling There are several advantages to decoupling tank closure from tank cleanup. The first advantage is that for tanks that prove difficult to clean (either because the tank configuration obstructs access or because the waste is recalcitrant), options can be kept open in the near term (5 to 10 years) to remove additional waste and/or to use improved immobilizing material to fill the tank. Filling a tank with grout is essentially an irreversible action. The second advantage of decoupling tank closure is to allow periodic reassessment of technology developments and alternatives to reduce long-term risks presented by the tank heels. A third advantage in delaying closure of these tanks is that it allows time to gather operational experience for tanks containing cooling coils and other waste retrieval challenges (see Chapter III). DOE obtained reasonable results in retrieving waste from Tanks 17 and 20, leaving behind very little residual waste. Tanks 18 and 19, which have undergone waste removal, are estimated to have an order of magnitude more radioactivity than Tanks 17 and 20, but the greater challenges lay ahead. DOE started its tank waste removal and closure campaign with Type IV tanks, which are simpler to work with because of the absence of cooling coils. This approach makes sense with respect to retrieval technology, because it allows DOE to learn from the simpler tanks before tackling the more complex ones. Tanks with coils may present an additional challenge because they are likely to have more solids encrusted on the interior surfaces and those solids will be difficult to reach. This is because (a) there is more surface area to which waste material can adhere and (b) there are more obstructions that make retrieval more difficult. DOE has developed operational experience with in-tank activities such as sampling, slurrying, pumping, removing waste heels with water jets (sluicing), and operating other remotely controlled equipment. In some cases, DOE may need more time than is allowed by the Savannah River Site Federal Facility Agreement closure milestone to apply what it has learned, test, identify any new challenges, and evaluate new technologies to maximize the removal of waste and stabilize residual waste in the more difficult tanks (see also Chapter IX). The fourth advantage of delaying closure of these tanks is that it would allow for a focused research and development program to enhance tank waste removal, improve waste immobilization, and improve tank stabilization as recommended in Chapter IX. A previous National Research Council report also recommended further research in waste retrieval and immobilization prior to tank closure (NRC, 2001a). As noted in the committee’s interim report, the long-term performance of tank fill materials appears not to have been established adequately; the committee discusses uncertainties in the long-term performance of these materials in Chapter VI. To lend confidence to the assumptions used in the performance assessment, a delay in tank closure would give DOE more time to evaluate grout formulation and techniques and to conduct studies of projected long-term performance by laboratory and field testing of tank fill materials (see Chapter IX). DOE itself recognizes the potential benefits of decoupling tank cleanup from tank closure. A series of reports requested by the Tanks Focus Area and developed by Pacific North-west National Laboratory describes the concept and applicability of placing a tank and its residual contents into a safe, stable, and minimum maintenance condition pending final closure options, what is defined as “tank lay-up” (Elmore and Henderson, 2001a, 2001b, 2002a). In these documents, tank lay-up is viewed as a potential necessity to bridge the time gap between tank cleanup and final closure, because sometimes the decision to close a tank is not made for many years after the tanks have been emptied (e.g., see the West Valley discussion in Chapter III and Appendix G); in these reports, tank lay-up is assumed to last for up to 20 years. Tank lay-up activities are discussed at five DOE sites (Hanford, Idaho National Laboratory, Oak Ridge National Laboratory, Savannah River Site, and West Valley Demonstration Project). The reports clearly discuss how lay-up depends on the number and physical condition of the tanks; expected lay-up period; uncertainty in closure requirements; perceived risks associated with waste heels; and the regulatory environment. The more recent of the two reports (Elmore and Henderson, 2002a; pp. 2-3) states: Tank lay-up activities are expected to reduce the perceived risks associated with the tanks. Likewise, subsequent hazard/ accident analyses on a tank-by-tank basis could result in the following: Lowering the hazard classification for certain facilities, which could impact conduct of operations, hazardous waste management, emergency preparedness, and training Reduction in the number of safety-class, safety-significant, and defense-in-depth structures, systems, and components, which could reduce the number of required engineered and administrative controls Reduction in the number of technical safety requirements (e.g., safety limits, limiting control settings, limiting conditions for operation) Reduction in monitoring or surveillance frequencies (e.g., liquid/solids levels, waste temperatures, vapor space pressures, leak detection probing, corrosion prevention) Reduction in tank reporting requirements Reduction of maintenance on the tanks and supporting and interfacing systems (e.g., vapor space filtration, liquid level devices, temperature probes, light-duty utility arm [LDUA], core sampling system)

OCR for page 62
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report Reduction in the interface requirements associated with nontank facilities and systems Reduction in configuration management requirements, procedure maintenance, number and depth of assessments, required personnel training, hazardous materials and radiation protection requirements, and other requirements to be determined on a Site and tank basis. The most recent summary report on tank lay-up activities also recommends that DOE share lessons learned on tank closure activities among its sites (Elmore and Henderson, 2002b). The committee does not advocate decoupling the removal and closure schedule based only on the future possibility of discovering better technologies for cleanup and closure without identifiable current prospects. Rather, the committee encourages developing or adapting specific technologies that are at least in the applied research stage and researching a narrow set of questions that, if answered, could enhance tank heel removal and closure effectiveness. The committee selected a time frame (5 to 10 years) that is in reasonable accord with the overall schedule for tank farm closure and would not extend tank closure indefinitely into the future. Concerns with Decoupling The recommendation in the interim report was not well received by DOE and the State of South Carolina (see Appendix E). In interviews with reporters, South Carolina Department of Health and Environmental Control representatives reiterated their commitment to the schedule for closing tanks and disagreed with the committee’s conclusion that delaying filling of tanks with grout would be beneficial from the perspective of risk. These representatives argued that unless previously agreed to milestones for tank closure continue to be met, progress will stall. This concern could be addressed if separate milestones were established for tank waste retrieval and for closure. The committee notes that to delay grouting of specific tanks may not delay the final closure milestone for the entire tank farm, which will take several years.3 If new technologies become available in the near future (i.e., 5 to 10 years), it may be possible to clean up and close tanks faster (possibly leaving less waste behind), thus meeting the final milestone for closing the tank farms. Even if the decoupling did result in some delay, the federal facility compliance agreements could be modified, as they have on many other occasions, provided that the action improves the outcome. In 2002 the General Accounting Office (GAO, now the Government Accountability Office) issued a report on the implications of DOE’s compliance agreements in waste cleanup (GAO, 2002). The GAO found that compliance agreements have not been a barrier to previous DOE management improvement initiatives. Regulators generally supported these initiatives, saying that they support efforts to implement faster, less costly ways to reduce environmental risks at the sites, as long as DOE’s approach did not reduce funding for individual sites (GAO, 2002). The second objection raised against delaying tank closure is that a tank could collapse due to lateral pressure from the surrounding soil, or from the weight of the overburden. In its interim report, the committee recommended that DOE consider the risks from postponing tank closure compared to the risk reductions that could be achieved if the postponement improves heel removal. A qualitative assessment by DOE of the issues associated with aged and abandoned underground structures and vessels includes the potential for roof and side wall collapse; filling with water from runoff (bathtub effect); and internal seepage, which can lead to overflowing, leaking, or leaching; and buoyancy (Langton et al., 2001). However, the committee is not advocating abandoning the empty tanks on-site and has seen no quantitative assessment of the risks of postponing tank grouting. According to DOE, the tanks are not in near-term danger of collapsing after bulk waste retrieval;4 indeed, the structural support provided by the tank fill is not likely to be needed until DOE is ready for ultimate closure of the tank farm. In most cases, postponing closure of tanks that contain significant amounts of residual waste for several years would appear to have essentially no effect on near- or long-term risk, while leaving open the possibility of further risk reduction if more of the waste can be removed. The third objection against delaying tank closure is that once equipment is in place for tank waste removal (e.g., the superstructure for in-tank operations), it is convenient to proceed to use the same equipment for closure, rather than moving it to another tank and reequipping the first tank when it is ready for closure. This may be a valid concern if DOE is using a superstructure that is difficult or costly to move; it is not clear how much of an inconvenience this would impose. Therefore, the committee recommended in its interim report that DOE evaluate advantages and disadvantages for the entire waste management operation at a given site from both a risk and a cost perspective. If DOE can relax other constraints on tank waste removal, such as the tank space problem, delaying tank closure could free up funds planned for closure activities, and those funds could be devoted to 3 At Hanford the closure schedule is 2024 for single-shell tanks and 2032 for double-shell tanks; at the Savannah River Site the closure milestones are 2022 for Type I, II, and IV, and 2028 for Type III tanks; at Idaho the tanks will be closed in six phases from 2005 to 2016; there are no milestones for closing the calcine bins. 4 It is the committee’s understanding that the geometry of the tanks is inherently stable (i.e., resistant to collapse). The emptied tanks, therefore, need not be filled until immediately prior to closure of the entire tank farm and placement of the engineered cap (if used).

OCR for page 62
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report enhancing waste removal, waste processing, and confidence in the near- and long-term performance of the waste immobilization and tank fill materials. Similarly, research and development require funds, but could, if successful, result in lower costs and increased safety overall (see Chapter IX). FINDINGS AND RECOMMENDATIONS Finding V-1: To protect future inhabitants on or near the present DOE sites, the primary objective of DOE’s tank closure program is to remove reprocessing wastes from DOE sites and permanently isolate the radionuclides in a geologic repository, such as Yucca Mountain. Grouting and other technologies (e.g., Hanford’s low-activity waste vitrification) to immobilize the wastes left on-site are secondary lines of defense for protecting future near- or on-site inhabitants. Recommendation V-1: DOE should maintain its primary objective of removing radioactive tank wastes from DOE sites. Immobilization of wastes left on-site cannot be a substitute or justification for not removing tank wastes from the sites to the maximum extent practical (e.g., to meet schedule commitments). Finding V-2: When a tank has a relatively simple configuration (i.e., without a network of cooling coils or other obstacles) and can be cleaned to an acceptable degree, it is reasonable to continue with tank closure soon after retrieval. However, when the residue in a given tank after cleaning still contains significant amounts of radioactive material, proceeding immediately to closure effectively precludes any further removal of residue from the tank. In its interim report, the committee recommended that DOE consider decoupling tank cleanup and closure activities. Recommendation V-2: In cases where significant amounts of radioactive residues remain after tank cleaning, efforts should be directed to emptying and cleanup of other tanks while more effective retrieval techniques are sought. The committee judges that this approach would result in improved risk reduction. This decoupling need not delay the scheduled closure of the overall tank farm. Finding V-3: Some of DOE’s performance assessments for residual wastes in storage tanks incorporate assumptions about the ability of the grout to retain its structural integrity and chemical properties over centuries and even millennia without a firm basis in either empirical data or fundamental scientific principles. In the near term, decisions about the formulation of grouts for tank fill are being made on the basis of experience in very different applications and, in some cases, on data from short-term tests on saltstone. The committee has not seen any reports of long-term testing or more fundamental research directed at the unique aspects of DOE applications, particularly the binding capacity of grouts and changes in various properties over the extended times contemplated by DOE. Recommendation V-3: The committee recommends that DOE initiate a focused research and development program over a 5- to 10-year period, and longer where necessary, to improve fundamental understanding of the long-term performance of tank fill material and tailoring grout formulations to different tanks or group of tanks. The program should involve collaboration among government laboratories, universities, and industry. Further details, findings, and recommendations on research and development can be found in Chapter IX.