II
Background and Overview of Current Situation

In this chapter the committee provides factual background information in support of its analyses, findings, and recommendations presented in this and subsequent chapters.

OVERALL APPROACH

The Department of Energy’s (DOE’s) overall approach for managing its tank wastes is the following: To the maximum extent practical, retrieve the waste from the tanks (and bins in Idaho, see below); separate (process) the recovered waste into high- and low-activity fractions; and dispose of both remaining tank heels and recovered low-activity waste on-site in a manner that protects human health and the environment. Figure II-1 is a simplified illustration of such an approach. The details of this approach are discussed in the following chapters: waste retrieval in Chapter III, waste processing plans in Chapter IV, and tank closure plans in Chapter V.

THE THREE SITES

This section provides background information on the Savannah River Site, the Hanford Site, and the Idaho National Laboratory and their tank waste.

Savannah River Site

The Savannah River Site has 51 underground tanks1 that are used for storing 138,000 m3 (36.4 million gallons)2 of hazardous and radioactive waste.

The Savannah River Site started generating tank waste in 1954 when a large chemical processing facility, called the F Canyon, was brought into service to separate uranium and plutonium from irradiated targets and spent nuclear fuel from on-site reactors to support the U.S. nuclear weapons program. A second chemical processing facility, the H Canyon, was brought on line in 1955.

Each canyon facility piped highly radioactive liquid waste from the chemical processing operations to a set of tanks located in its area: the F Area Tank Farm has 22 tanks and the H Area Tank Farm has 29 tanks.3 The tanks range in size from about 2,850 to 4,900 m3 (750,000 to 1.3 million gallons). They are vertical cylinders, approximately 23 to 26 m (75 to 85 feet) in inner diameter and 7.5 to 11 m (24.5 to 35 feet) in height from the inner tank floor to the ceiling. They are buried at a shallow depth (1 to 3 m below the land surface), mostly above the water table, although four tanks in the H Area Tank Farm are nearly submerged in the saturated zone (that is, the water table reaches nearly to the top of the tanks).

Access to the interior of the tanks is gained through portals, called risers, which rise from the top of the tank through the ground cover to the land surface. The number of risers in each tank ranges from 7 to 40, and the diameters of most of the apertures range from 58 to 107 cm (23 to 42 inches) depending on tank type. Some risers are larger: The center riser of a Type IV tank is approximately 2.7 m (9 feet) (Fogle, 2002). Annulus ports on Type III tanks are as small as 20 cm (8 inches, see below).

Most of the tanks have a carbon steel inner wall and a cylindrical outer vault wall constructed of concrete, with a space between them called the “annulus.”4 Tanks that have

1

Two tanks (Tanks 17 and 20) were filled with grout and closed in 1997, and three tanks (16, 18, and 19) were cleaned and taken out of service, so there are currently 46 tanks in service. DOE plans to close Tanks 18 and 19 next.

2

This report presents quantities in SI units and the equivalent value in English units in parentheses, e.g., 11 liters (2.9 gallons). The only exception is radioactivity, which is reported in curies first with the quantity in becquerels after, because becquerels are very rarely used in discussion of tank wastes.

3

A map of the Savannah River Site and the General Separations Area where the tanks are located can be found in Appendix J.

4

The gap between the primary and secondary containment is about 60 cm (2 feet).



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 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report II Background and Overview of Current Situation In this chapter the committee provides factual background information in support of its analyses, findings, and recommendations presented in this and subsequent chapters. OVERALL APPROACH The Department of Energy’s (DOE’s) overall approach for managing its tank wastes is the following: To the maximum extent practical, retrieve the waste from the tanks (and bins in Idaho, see below); separate (process) the recovered waste into high- and low-activity fractions; and dispose of both remaining tank heels and recovered low-activity waste on-site in a manner that protects human health and the environment. Figure II-1 is a simplified illustration of such an approach. The details of this approach are discussed in the following chapters: waste retrieval in Chapter III, waste processing plans in Chapter IV, and tank closure plans in Chapter V. THE THREE SITES This section provides background information on the Savannah River Site, the Hanford Site, and the Idaho National Laboratory and their tank waste. Savannah River Site The Savannah River Site has 51 underground tanks1 that are used for storing 138,000 m3 (36.4 million gallons)2 of hazardous and radioactive waste. The Savannah River Site started generating tank waste in 1954 when a large chemical processing facility, called the F Canyon, was brought into service to separate uranium and plutonium from irradiated targets and spent nuclear fuel from on-site reactors to support the U.S. nuclear weapons program. A second chemical processing facility, the H Canyon, was brought on line in 1955. Each canyon facility piped highly radioactive liquid waste from the chemical processing operations to a set of tanks located in its area: the F Area Tank Farm has 22 tanks and the H Area Tank Farm has 29 tanks.3 The tanks range in size from about 2,850 to 4,900 m3 (750,000 to 1.3 million gallons). They are vertical cylinders, approximately 23 to 26 m (75 to 85 feet) in inner diameter and 7.5 to 11 m (24.5 to 35 feet) in height from the inner tank floor to the ceiling. They are buried at a shallow depth (1 to 3 m below the land surface), mostly above the water table, although four tanks in the H Area Tank Farm are nearly submerged in the saturated zone (that is, the water table reaches nearly to the top of the tanks). Access to the interior of the tanks is gained through portals, called risers, which rise from the top of the tank through the ground cover to the land surface. The number of risers in each tank ranges from 7 to 40, and the diameters of most of the apertures range from 58 to 107 cm (23 to 42 inches) depending on tank type. Some risers are larger: The center riser of a Type IV tank is approximately 2.7 m (9 feet) (Fogle, 2002). Annulus ports on Type III tanks are as small as 20 cm (8 inches, see below). Most of the tanks have a carbon steel inner wall and a cylindrical outer vault wall constructed of concrete, with a space between them called the “annulus.”4 Tanks that have 1 Two tanks (Tanks 17 and 20) were filled with grout and closed in 1997, and three tanks (16, 18, and 19) were cleaned and taken out of service, so there are currently 46 tanks in service. DOE plans to close Tanks 18 and 19 next. 2 This report presents quantities in SI units and the equivalent value in English units in parentheses, e.g., 11 liters (2.9 gallons). The only exception is radioactivity, which is reported in curies first with the quantity in becquerels after, because becquerels are very rarely used in discussion of tank wastes. 3 A map of the Savannah River Site and the General Separations Area where the tanks are located can be found in Appendix J. 4 The gap between the primary and secondary containment is about 60 cm (2 feet).

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report FIGURE II-1 Simplified flowsheet of DOE’s waste management plan for its tank and bin wastes at the Hanford, Idaho, and Savannah River sites. The waste processing box corresponds to the Waste Treatment and Immobilization Plant and supplemental treatment plant for Hanford; the steam reforming plant for the Idaho sodium-bearing waste; and sludge washing, and the Salt Waste Processing Facility at the Savannah River Site. It is not clear right now whether the Idaho calcine will be accepted in a deep geologic disposal as is or if it requires further processing. Some Hanford tank waste and the steam-reformed waste from the sodium-bearing waste at Idaho will undergo a waste determination to declare it defense transuranic (TRU) waste and ship it to the Waste Isolation Pilot Plant (WIPP). NOTE: INL = Idaho National Laboratory; HLW = high-level waste; LAW = low-activity waste; SBW = sodium-bearing waste; WTP = Hanford Waste Treatment Plant a metal liner on the outer wall are said to have a “secondary containment” (i.e., a tank inside a tank). If the outer liner rises only partway up the outer wall, it provides only partial secondary containment. Eight of the tanks have no annulus or secondary containment (Type IV tanks), 16 have partial secondary containment (Types I and II tanks), and 27 have full secondary containment (Types III and IIIA tanks). Figure II-2 illustrates the four general tank types. Only the Type III and IIIA tanks with full secondary containment are considered “compliant tanks” under the site’s Federal Facility Compliance Agreement and Consent Order (Federal Facility Agreement), the agreement regulating waste under the Resource Conservation and Recovery Act (RCRA) requirements for wastes stored in tanks (40 CFR 264.193 (b)). The “noncompliant” tanks are generally past their 30-year design life, and many (13, at last report; DOE-SRS, 2005c) have a history of cracks or leakage (either from the tank into the annular secondary region or from the surrounding media into the tank or annulus),5 although only one tank is believed to have leaked a small quantity of waste to the environment. Waste levels in the tanks have been lowered below the location of known leaks, and at present DOE believes that there are no active leaks except in Tank 5 from which waste is currently being retrieved. All but the Type IV tanks contain dense networks of vertical and horizontal “cooling coils,” pipes that circulate cooling water. The cooling water removes heat produced from radioactive decay in the waste. Savannah River Site Tanks Inventories Based on information provided by site personnel, the global inventory of chemicals and radionuclides in the tank farms is reasonably well understood. The information is based on analytical data, reactor fuel burnup and discharge records, reprocessing plant processes, flowsheets, and records of chemical purchases and operations. Waste from the canyons contains acids and other chemicals used in the separation processes, chemicals added to neutralize and alkalinize the waste, and radionuclides (fission products, such as cesium-137, and actinides, such as neptunium-237) not separated during recovery of plutonium 5 Leaks are detected by visual inspection or by conductivity probes in the annulus.

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report FIGURE II-2 Diagrams of tank types at the Savannah River Site (not drawn to scale). There are twelve Type I tanks, four Type II tanks, twenty-seven Type III (and IIIA) tanks, and eight Type IV tanks in the tank farms. Risers are not depicted in most of these diagrams but are present on each tank. SOURCE: Adapted from Mahoney, 2005. and uranium. To prevent corrosion of the carbon steel tanks, sodium hydroxide was added to neutralize the acid and make the waste alkaline before it was pumped to the tanks. This caused metals and most radionuclides to precipitate as an insoluble sludge,6 which settled to the bottom of the tanks (see Figure II-3). The liquid remainder, or unconcentrated supernate, contains soluble salts and is referred to as a salt solution. If concentrated by evaporation, much of the salts initially in solution will crystallize to form a solid saltcake. Thus, the wastes in the tanks exist mainly in three physical forms: sludge, supernatant liquid (“supernate”), and saltcake. Together, the supernate and saltcake are referred to as salt waste. To conserve tank space, most of the salt solutions have been processed through an evaporator (a heated tank that evaporates water from waste) to produce saltcake, leaving relatively small volumes of concentrated supernate solution. The total estimated radioactivity in each physical form is shown in Figure II-4 and Table II-1, which also lists the radioactivities of other wastes on the site. Further details are provided below. The supernate contains more than 90 percent of the inventory of soluble radioactive species, mainly cesium-137. 6 The terms “insoluble” and “soluble” are used here to describe chemical species that exist preferentially in the solid phase or the liquid phase, respectively, in the larger medium (the waste in a tank). No species will exist exclusively in one phase. In the cases discussed here, however, all but a very small fraction of the chemical mentioned exists in the preferred phase in that medium. Also, the sludge entrains some soluble radioactive species.

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report FIGURE II-3 Photograph of a Savannah River Site tank sludge sample. SOURCE: Caldwell, 2005a. The saltcake is a solid material composed of more than 99 percent salts, such as sodium nitrate, that contains concentrations of soluble and insoluble radioactive constituents lower by approximately a factor of 10 to 20 than what is in the sludge. The waste in the tanks (see Figure II-4) contains approximately 426 million curies (MCi; 1.58 × 1019 becquerels) of radioactivity; approximately half of the radioactivity is in the sludge and half in the salt waste. Most of the volume is in the salt waste, approximately 128,000 m3 (33.8 million gallons), whereas the sludge represents approximately 9,800m3 (2.6 million gallons). More than 95 percent of the radioactivity in the salt waste comes from cesium-137 (and its short-lived decay product, barium-137m) and strontium-90 (and its short-lived decay product, yttrium-90). Both the cesium and the strontium isotopes have half-lives of approximately 30 years. The cesium poses a particular hazard for people working near the waste because it emits penetrating radiation (gamma rays). Other radioactive constituents in the waste are of concern for other reasons: DOE has concluded that carbon-14, selenium-79, technetium-99, iodine-129, tin-126, and neptunium-237 dominate the long-term risk to the public from disposed waste because of their long half-lives and their mobility in the environment (Cook, 2005). The actinide isotopes, including isotopes of plutonium and americium, decay into a series of other radioactive substances (together referred to as a decay chain) and also constitute long-term hazards, particularly for inadvertent intruders. Although the global tank farm inventory is reasonably understood, individual tank inventories have greater uncertainties (Reboul and Hill, 2005). Sampling individual tanks is difficult because the waste is highly radioactive, and furthermore it is heterogeneous and segregated into different forms and compounds in different portions of the tanks. DOE reported to the committee (Hill, 2006) on three sampling studies:7 a 2002 statistical comparison of slurried sludge samples from eight individual tanks versus characterization predictions for those tanks, a review of seven supernate samples, and a review of six short (3 feet or 1 meter) salt core samples. These reviews found the following: For significant radionuclides,8 the predicted inventory was on average a factor of 1.6 greater than the measured inventory and 95 percent of the predicted inventories were within a factor of 2.5 less than predicted and a factor of 8 more than predicted. All predictions were between a factor of 10 of the measured value. For all elements considered to be significant (i.e., at least 1 weight percent of the total dried solids), the predicted concentration was within a factor of 1.12 less than the measured concentration and 95 percent of the predicted inventories were between a factor of 4 less than predicted and a factor of 2.5 more than predicted. All of the differences were within a factor of 10. DOE believes that uncertainties in the saltcake radionuclide constituents (i.e., all nuclides that are characterized for saltcake predictions) are within a factor of 2. DOE describes the error in predictions of minor chemical constituent concentrations (those for sodium phosphate, sodium chloride, sodium fluoride, sulphate, etc.) in the saltcake as plus or minus 50 percent. Uncertainties in supernate radio- 7 These studies were not yet public when the committee completed its report, so the reviews were only described to the committee. 8 At the time of the sampling study DOE described, significant radionuclides were those that contributed to inhalation dose potential.

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report FIGURE II-4 Aggregated volume and radioactivity distributions among the tank waste phases in all tanks at the Savannah River Site as of December 2004. SOURCE: DOE-SRS, 2005b. TABLE II-1 Inventory of Radioactive Waste by Type at the Savannah River Site. Type of Waste Volume (m3) Radioactivity (Ci) Total waste in the tanks comprisinga 138,000 426 million Sludge` 9,800 203 million Saltcake 62,000 12 million Supernate 66,000 211 million Vitrified high-level wasteb 1,500 10 million Stored transuranic wastec,d 11,000 490,000 Buried transuranic-contaminated waste and soilc 4,500 18,500 Low-level waste storedc 15,276 1.3 milliond Low-level radioactive waste in disposal cellsc 698,000 11 millione Saltstone as of 2005 25,000f 225f Saltstone (DOE projected) 410,000a 3-5 milliona E Area vaults 117,000g 10 milliong Old Burial Ground Unknownh 4.5 millionh Tanks at closure (DOE projected) 140i 0.72 millioni TOTAL > 867,000 446.6 million NOTE: Shaded area lists wastes that are expected to remain on-site. The data are from different sources, are measured or estimated at different times, and did not indicate quantified uncertainties. This table does not include spent fuel from research reactors. 1 Ci = 3.7 × 1010 Bq. a DOE-SRS, 2005b. b As of January 31, 2006, 2044 canisters containing approximately 0.74 m3 per can. c DOE, 2001b. d Assumes same concentration as E Area Vaults (85 Ci/m3). e Not decay corrected, hence an overestimate. f WSRC, 2004a. g E Area Vault Waste Information Tracking System (WITS) provided by DOE (Clark, 2005a). h WSRC, 1997. i Based on assumed residuals in the tanks (DOE-SRS, 2002) and average concentrations (Buice et al., 2005). See Chapter VI of this report for discussion of this inventory.

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report nuclide predictions for cesium-137, technetium-99, and iodine-129 are within 10 percent, other fission products are within a factor of 2, and actinides are within a factor of 10. DOE believes that predictions of minor chemical constituent concentrations in the supernate are within 50 percent (plus or minus 50 percent). The low-activity waste disposed as saltstone will be characterized through sample analysis, and therefore, the uncertainty in this waste stream inventory will be small. The impetuses for sampling waste that is retrieved from the tanks are the Salt Waste Processing Facility and the Defense Waste Processing Facility (DWPF) feed delivery specifications. The Savannah River Site also plans to sample waste tank heels to demonstrate how the tank heels meet the performance objectives in 10 CFR 61. Detailed discussion of tank heel sampling can be found in the performance objective demonstration document for Tanks 18 and 19 (Buice et al., 2005). Savannah River Site Waste Processing DOE’s plan to manage the waste retrieved from the tanks is to separate the radioactive from the nonradioactive components, the latter of which make up nearly the entirety of the waste volume. This processing generates two waste streams: (1) a high-activity, low-volume waste stream, which will be immobilized and disposed off-site in a deep geologic repository, and (2) a low-activity waste stream, which is to be disposed in near-surface vaults on-site. Figure II-5 illustrates the waste flows that DOE has described for tank wastes at the Savannah River Site. The wastes planned for repository disposition are not the subject of this study because they are not planned for on-site disposal. They are included here because the management of tank wastes must be considered as a system of interconnected parts. Sludge Processing For nearly 10 years, DOE has been retrieving sludge from tanks at the Savannah River Site for immobilization in glass. After retrieval from the tank, the sludge is transferred to a dedicated waste tank where it is “washed” to remove soluble salt constituents that will interfere with the glass-forming process and to reduce the volume of material that is sent to the DWPF for vitrification into logs of waste glass. The logs are to be disposed off-site in a high-level waste repository. The wash water and a low-activity liquid waste stream from the DWPF are sent back to the tanks (see Figure II-5). FIGURE II-5 Waste flows in the Savannah River Site waste management plans. Note that the sizes do not necessarily scale with the sizes of the waste flows. Noncompliant tanks are those that do not have full secondary containment (i.e., a tank inside of a tank). NOTE: ARP = Actinide Removal Process; DDA = deliquification, dissolution, and adjustment; MCU = Modular Caustic Side Solvent Extraction Unit.

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report Salt Waste Processing DOE indicated to the committee that the Savannah River Site is facing a “tank space crisis” because of net waste inputs from current waste processing and waste removal operations. To ensure that sludge removal from noncompliant tanks continues apace and DWPF continues to operate at full capacity DOE is proposing to begin processing salt waste as soon as possible. DOE is still developing facilities to process the salt waste (supernate and saltcake) at the Savannah River Site. Three progressively more sophisticated and effective separation processes are to be brought into service for processing different batches of salt wastes: DOE proposes to use two “interim” processes (described in Chapter IV) for what it calls “low-activity salt,” that is, salt waste that contains what DOE considers to be “low concentrations of radionuclides,” relative to the average until the Salt Waste Processing Facility (SWPF) begins operations. The SWPF was scheduled to begin operation in 2009; however, DOE recently announced an estimated 26-month delay in startup operations because of seismic concerns in the building design (Terhune and Kasper, 2005). When the SWPF comes online, it will process the majority of the salt waste, “augmented as necessary by ARP” (DOE, 2006). Tank wastes are to be processed to concentrate the radionuclides into a high-activity waste stream that will be vitrified at the DWPF. The other separated fraction, consisting mainly of the nonradioactive salts and other constituents with low concentrations of radionuclides that make up the less contaminated, low-activity waste stream, is to be immobilized in the Saltstone Production Facility—an operation that mixes liquid waste with grout9 to create a waste form referred to as saltstone, which is disposed on-site as a monolith in concrete vaults. Until now, the Saltstone Production Facility has handled very low activity waste. The higher radioactivity anticipated in the liquid waste that DOE plans to send to the facility prior to SWPF startup has required DOE to reconfigure the equipment and facility as well as add shielding in certain areas. Tank Closure After waste is retrieved from a tank, DOE plans to operationally “close” the tank, i.e., fill the tank with tailored layers of grout, and sever and seal external penetrations. At some point in the future, groups of tanks will be formally closed and engineered barriers (e.g., a cap) will be emplaced. The plans for waste retrieval and tank closure are discussed in Chapters III and V, respectively. The Hanford Site The Hanford Site was the world’s first plutonium production factory. It was designed, constructed, and operated by E.I. du Pont de Nemours and Company to provide the Manhattan Project with material for the cores of some of the first nuclear weapons.10 The Hanford Site occupies 1,517 km2 (585 square miles) of land on the Columbia River in south central Washington State. The plutonium production reactors were built in the “100 Area,” widely spaced along the Columbia River so that the reactors could use river water as coolant. The chemical processing plants were built 10-20 km south of the river in the middle of the site, called the “200 Area,” or the Central Plateau (see Figure II-6). Nine plutonium production reactors were built and operated at Hanford, and all of them have been shut down. When the production facilities were operational, irradiated reactor fuel was transported by rail from the 100 Area reactors to chemical separation plants in the 200 Area. Five chemical reprocessing facilities (T-Plant, B-Plant, U-Plant, REDOX Plant, and PUREX Plant) and a plutonium finishing plant operated over the history of the site. The waste from these facilities was stored in large underground tanks, and during some of the early years of operation, lower-activity (compared to tank waste) liquid waste was discharged into the ground. The last reprocessing facility ceased operations in 1990. The Hanford Site has 149 single-shell tanks and 28 double-shell tanks (Figure II-7). Tanks are grouped into 12 single-shell tank farms and 6 double-shell tank farms in the 200 East and 200 West Areas of the site. The tanks are interconnected by underground pipes and served the five chemical processing facilities mentioned above. The tank farms also have ancillary equipment used to divert and direct waste within each tank farm, such as valve boxes and pump pits, and between tank farms. The 149 single-shell tanks were constructed between 1943 and 1964. These are vertical cylindrical structures that range in size from approximately 200 m3 to nearly 3,800 m3 (55,000 to 1 million gallons)—133 of the 149 are 2,000 m3 (500,000 gallons) or larger—and were constructed of a concrete shell lined with a single layer of carbon steel. A typical tank is 23 m (75 feet) in diameter and 9 to 16 m tall (30 to 54 feet). The top dome is unlined concrete with 2 to 3m (6 to 10 feet) of earthen cover (Elmore and Henderson, 2001a). Most of the tank bottoms are slightly concave (bowl shaped, with the low point in the center). Access to the tank interiors is achieved through risers that range from 10 cm to 1.1 m (4 to 42 inches) in diameter (Elmore and Henderson, 2001a). The number of risers in a single-shell tank ranges from 9 to more than 20. 9 Except where otherwise indicated, the term “grout” is used here to mean a cementitious material used for waste immobilization or tank fill; see Chapter V. 10 Du Pont turned operation of Hanford over to General Electric in 1946. Subsequently, du Pont became involved in the design and construction of the facilities at the Savannah River Site in 1950, and it continued to operate that site until 1989.

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report FIGURE II-6 Map of the Hanford Site indicating the areas related to plutonium production and storage of high-level radioactive waste. NOTE: Figure does not include every facility in the 200 Area. SOURCE: Mann, 2005. FIGURE II-7 Hanford single-shell tank (on the left) and double-shell tank diagram (on the right). SOURCE: www.Hanford.org.

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report Many of the tank farms do not have evaporators, and early volume reductions were achieved by allowing the waste to boil in the tanks (see Footnote 9). None of the tanks has cooling coils. None of the welds on the carbon steel liners of the single-shell tanks was stress relieved. The combination of thermal stresses and exposure to hot, aggressive solutions has resulted in stress-corrosion cracking at the welds of some tanks. All single-shell tanks are regarded as beyond their design lives and sixty-seven of the single-shell tanks have leaked or are assumed to have leaked about 3,800 m3 (1 Mgal) of waste into the environment (see below). Starting in 1968, the site built 28 double-shell tanks. All of the double-shell tanks are similar in size and design to the largest of the single-shell tanks, but they enclose the waste entirely in a double layer of carbon steel separated by an annular space to collect and monitor leaks. The last was built in 1986. All double-shell tanks were stress-relieved during construction to substantially reduce the likelihood of stress-corrosion cracking. None of the double-shell tanks has leaked. Hanford Tank Inventories The form of Hanford tank wastes is similar to that at the Savannah River Site: highly alkaline waste in supernatant, saltcake, and sludge phases. The sodium hydroxide and sodium nitrite used for corrosion control in the tanks formed sodium nitrate cakes and hydrated oxides of radionuclides and other chemicals in the waste, creating a sludge on the floor of the tanks. However, Hanford dealt with a greater variety of fuels and at different times used more chemical processes than did the Savannah River Site, which is reflected in the diversity of tank waste compositions at Hanford. Table II-2 lists the quantities of waste stored, disposed, or discharged to the environment at the Hanford Site. Beginning in the late 1960s, cesium was separated from the supernatant liquid in waste tanks and strontium was removed from the sludge that was dissolved and processed. These separations, which removed approximately 90 percent of the strontium and cesium from the processed waste, produced intensely radioactive halide salts, now stored on-site in capsules (see Sidebar II-1). 11 Hanford processed 97,000 metric tons of irradiated uranium. Between 1944 and 1980, approximately 700,000 m3 (185 million gallons) of liquid radioactive waste was pumped into 149 single-shell tanks. Releases from the tanks and the piping system were first reported in 1956 and 67 tanks now are estimated to have released between 2,200 and 3,800 m3 (580,000 and 1.0 million gallons) of tank waste into the ground (Honeyman, 2005). Prompted by leaks in some of the single-shell tanks, DOE pumped the free and drainable liquids from those tanks into double-shell tanks and implemented a remedial investigation program to determine the nature and extent of past leaks and various interim corrective measures to reduce groundwater impacts. Now, approximately 121,000 m3 (about 32 million gallons) of saltcake, sludge, and interstitial liquid waste remains in the single-shell tanks (Honeyman, 2005). The site inventory in the 177 Hanford tanks today consists of approximately 204,000 m3 (54 million gallons) of radioactive waste containing 193 MCi (7.14 EBq) of radioactivity. In 1989, the defense-related plutonium production mission at Hanford ended and all production reactors and processing plants were shut down. Most of the work at the site now supports the mission of managing the waste and environmental problems at the site. The 204,000 m3 (54 million gallons) of waste is what remains in the tanks from the roughly 2 million cubic meters (525 million gallons) of tank wastes generated between 1944 and 1988. The balance of the waste was evaporated (71.3 percent),12 disposed to the ground after some radionuclide removal (28.5 percent), or leaked directly to the ground (about 0.25 percent). The saltcake and sludge in the single shell-tanks contain a little over 98 MCi (3.6 × 1018 Bq) of radioactivity. The double-shell tanks have about 95 MCi (3.5 × 1018 Bq) contained in waste consisting mostly (80 percent) of liquids, but also of sludges and salts (Wiegman, 2004; Honeyman, 2005). A visual inventory of the radioactivity in Hanford tank waste is shown in Figure II-8. Aside from concerns about leaks, safety concerns related to the tanks arose as a result of hypothesized and observed chemical reactions, excessive heating, and the possibility of nuclear reactions (criticality) in some of the waste. Of greatest concern was the observed buildup of hydrogen gas generated by chemical and radiolytic reactions in one tank (later found in others). The potential for a flammable mixture to be ignited prompted DOE to install a mixing pump in Tank SY-101, which is located in the 200 West Area, to prevent local buildup of flammable concentrations and to institute a set of operational controls utilizing flammablegas monitors. Heat generation due to radioactive decay and 11 While DOE considers the cesium and strontium capsules to be nuclear materials rather than tank wastes, the committee has examined them because they are highly radioactive materials extracted from the tank wastes and DOE says it plans either to dispose of them on-site or to combine them with the high-activity waste stream to be vitrified and sent to geologic disposal. In essence, DOE faces the same decision about the capsules that it faces when deciding what to do with radioactive material separated in the Waste Treatment and Immobilization Plant. The main difference is when the separations were carried out. 12 Different techniques have been used to reduce the volume of wastes through evaporation. Early techniques included allowing the wastes to self-boil because of the decay heat they generated. Temperatures in many of the tanks routinely were above 150C (300F), and one tank got above 310C (590F). Other early techniques included in-tank evaporation, either by inserting an electric heater into the waste or by circulating hot air into the tanks. Large-scale evaporation began in the 1970s by operating Evaporator-Crystallizers in the 200 West and 200 East Areas (Gephart, 2003).

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report TABLE II-2 Inventory of Radioactive Waste by Type at the Hanford Site. Type of Waste Volume (m3) Radioactivity (Ci) Total tank wastesa,b 204,000 193 million Single-shell tanks 121,000 98 million Double-shell tanks 83,000 95 million Radionuclides separated from tank wastes     Cs and Sr capsules ~4 125 million German logs 34 logs 11 million Waste leaked into environment from tanksc 2,200-3,800 0.3 million Early tank waste intentionally discharged to soilc 454,000-492,000 65,000-4.7 million Evaporator condensate released to groundc 1 million ~3,000 Stored transuranic (TRU) wasted,e 48,000d 246,000d   46,000g 300,000g Buried TRU-contaminated waste and soild,e 107,000 92,000 Low-level waste (including mixed) stored 187,000g 5.5 milliong   9,300d Not availabled Low-level radioactive waste in disposal cells 1.2 milliond 12 milliond   283,000g 11 milliong Low-level radioactive waste at U.S. Ecology commercial disposal site (not DOE)h 382,000 3.9 million TOTAL >3.58 million >350 million NOTE: Shaded entries are wastes that ultimately are expected to remain on-site. The data are from different sources, are measured or estimated at different times, and did not indicate quantified uncertainties. 1 Ci = 3.7 × 1010 Bq. a Schepens, 2005. For more information on cesium and strontium capsules and the German logs, see Appendix K. b This table does not include spent fuel and sludge from the K-Basins or spent nuclear fuel from the Fast Flux Test Facility (FFTF) and the Shippingport Pressurized Water Reactor. c NRC, 2001a, decay corrected to mid-1990s. The lower-bound number for intentionally discharged radioactivity accounts only for cesium-137 and strontium-90. d DOE, 2001b. e As of 1996. f Not decay corrected, hence an overestimate. g DOE-RL, 2004a. h Quantities as of January 1, 2000 (Washington State, 2000). SIDEBAR II-1 Cesium and Strontium Capsules Some of the tank waste at Hanford was processed to remove cesium and strontium. Cesium was removed by ion-exchange columns. Strontium was removed by a solvent extraction method. This method produced liquid and solid alkaline waste containing high concentrations of organic complexants that retain some radioactive elements in solution. These separations, which removed approximately 90 percent of the strontium and cesium from the waste processed, were performed in the B Plant and produced intensely radioactive halide salts (cesium chloride and strontium fluoride). These salts were encapsulated in 2,217 metal cylinders, some of which were used both on-site and off-site as radiation sources. The cylinders are approximately 7 cm (2.75 inches) in diameter and 50 cm (approximately 20 inches) long. The off-site applications never developed as expected and ceased entirely in 1988 after one capsule being used in the commercial sector was found to be leaking (USNRC, 1989). All capsules were returned to Hanford by 1996. As of 1997, nearly 300 capsules had been dismantled and their contents repackaged. Currently, 1,936 capsulesa are stored at the Waste Encapsulation and Storage Facility at Hanford. Although 23 of these had to be overpacked (i.e., sealed in a larger stainless steel container) due to swelling, the capsules are generally considered to be in good condition (DOE-RL, 2002). The total volume occupied by these capsules is about 4 m3 (150 cubic feet), but they account for more than 40 percent of the tank waste radioactivity at Hanford. Off-site disposal at a geologic repository by 2020 is the reference disposition option for these materials. A previous National Research Council report describes the technical challenges that these cesium and strontium capsules present for continued storage and eventual disposal (NRC, 2003a).    a1,335 containing about 83.5 MCi (3.1 EBq) of cesium and its barium decay product, and 601 containing about 36.5 MCi (1.4 EBq) of strontium and its decay product. These are decay corrected to July 2005 from DOE-RL (2003).

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report FIGURE II-8 Radioactivity in Hanford tank wastes, decay corrected to January 2004. SOURCE: Honeyman, 2005. exothermic chemical reactions in the waste (ferrocyanide reacting with nitrate and nitrite mixtures; organic complexants and organic solvents) raised concerns that temperatures in some single-shell tanks could exceed the tanks’ structural limits. One tank was termed a “high-heat tank” due to heat from radioactive decay. DOE determined that the waste should be transferred to a double-shell tank, which is better equipped to take the heat load. Investigation of the chemical component of this potential problem revealed that radiolysis had diminished the concentrations of the relevant species and that they were diluted enough by other components of the waste to prevent significant further reaction rates. Based on information provided by site personnel, the committee judges that the global inventory of chemicals and radionuclides in the tank farms at Hanford is reasonably well understood and is derived in a manner similar to that employed at the Savannah River Site. Individual tank waste inventories have greater uncertainty (Honeyman, 2005). The composition of the waste in Hanford tanks is not fully known because of poor record keeping concerning waste inputs to particular tanks and transfers among the tanks, and the difficulty and high cost of sampling and assay of samples. As of the end of 2005, 86 single-shell tanks and 17 double-shell tanks had been core-sampled. Most of the remaining 74 tanks have been sampled via grab samples13 or auger samples. There are 32 single-shell tanks that have not been sampled since 1986. Like the tanks at the Savannah River Site, many Hanford tanks have a bottom layer of sludge containing strontium 13 The preferred method to estimate inventory is sampling, supplemented by process knowledge. Core, liquid “grab” (using the “bottle-on-a-string” method), and vapor-phase sampling are the methods currently used to characterize waste in the tanks prior to waste retrieval. Auger samples are samples obtained with an “auger tip” (similar to a drill bit), a solid or tubular drill rod, and a “T” handle. The auger tip drills into the waste as the handle is rotated, and material retained on the auger tip is brought to the surface.

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report and transuranic elements along with hydroxides, oxides, and phosphates of the various nonradioactive metals present in the waste. Above the sludge layer sits one of two different salt media: in the single-shell tanks there are sodium nitrates, nitrites, phosphates, aluminates, carbonates, and sulfates in the form of a saltcake with interstitial liquids containing radioisotopes of cesium, technetium, and iodine. Double-shell tanks have a slurry of the aforementioned salts topped by supernatant liquid composed of the same materials, but bearing trace concentrations of strontium and transuranic radionuclides. Finally, a vapor resides above the liquid and solid contents of the tanks. The vapor is mostly air with small amounts of hydrogen, nitrous oxide, ammonia, trace organic chemicals, and water vapor. Some of the tanks contain other materials, such as debris, cement, diatomaceous earth, and broken or obsolete contaminated equipment (see Chapter III). Because of the risks to workers and the high cost of data acquisition, the tanks are not sampled for general characterization purposes (i.e., just to know everything in a tank); however, they are sampled for specific data needs in accordance with the site’s Data Quality Objectives or if requested by oversight groups. For example, the Defense Nuclear Facilities Safety Board (DNFSB, 1993) directed the Hanford Site to sample the tanks for flammable gas. Sampling at Hanford is currently driven by waste compatibility and chemistry control for corrosion mitigation, Waste Treatment and Immobilization Plant feed delivery needs, and single-shell tank retrieval actions to support tank closure. Hanford Tank Waste Processing As mentioned above, Hanford tank waste consists of highly alkaline sludge, saltcake, and supernate. Cesium and strontium isotopes and their decay products comprise most of the radioactivity in the waste. Current planning is to retrieve all waste in the single- and double-shell tanks, separate low-activity waste from the high-level waste, operate the Waste Treatment and Immobilization Plant and supplemental treatment systems to immobilize in glass most of the waste, and package remaining transuranic waste determined not to be from reprocessing of spent fuel for shipment to WIPP for disposal. Waste in the 28 double-shell tanks is to be retrieved and immobilized in glass in the Waste Processing and Immobilization Plant. Plans for the highly radioactive waste processed from the contents of the 149 single-shell tanks are still being developed, but it too is expected to be treated and immobilized in the proposed vitrification facility. Details and status of waste retrieval, processing, and tank closure plans are described in Chapters III, IV, and V, respectively. The Idaho National Laboratory The Idaho National Laboratory was established in 1949 as the National Reactor Testing Station on what had previously been a bombing and artillery range for the U.S. Navy. The site occupies 2,303 km2 (890 square miles) of land in southeast Idaho, approximately 40 km (25 miles) west of Idaho Falls. Nine primary facility areas scattered mostly across the southern half of the site support missions related to naval nuclear propulsion and civilian and military nuclear applications. Some of these missions are ongoing, but others have ended. The spent nuclear fuel processing facilities are located at the Idaho Nuclear Technology and Engineering Center (INTEC), formerly the Idaho Chemical Processing Plant. Between 1953 and 1992, the Idaho Chemical Processing Plant reprocessed 44 metric tons of heavy metal of U.S. government spent nuclear fuel primarily to recover highly enriched uranium (NRC, 1999b). The processing of spent fuel at the Idaho National Laboratory was similar to processing carried out at the Savannah River Site and the PUREX facility at Hanford. Spent nuclear fuel was dissolved in nitric acid and other strong mineral acids and then sent through further processing steps to recover uranium, neptunium, krypton, barium, and xenon. The highly radioactive waste from the first cycle of the solvent extraction system, containing most of the fission products, was piped to the underground tanks. The 11 stainless steel tanks, each with a typical capacity of 1,136 m3 (300,000 gallons), are located within concrete vaults. There are annular spaces between the outside of the tanks and the vault walls. Some of the tanks have cooling coils along their bottoms and walls. Three of the tanks were designed for use with less radioactive wastes and, thus, did not receive first-cycle waste from the reprocessing facilities. Unlike waste at Hanford and the Savannah River Site, sodium hydroxide was not added to the liquid tank waste, thus reducing the volume of storage space needed. Because the tank farm components were made of stainless steel that was compatible with the waste, there was no need to neutralize the waste streams. In fact, the waste streams sent to the tank farm were purposely kept acidic to minimize waste precipitation, to simplify later waste retrieval, transfers, and processing. Maintaining acidic waste streams also reduces the possibility of accidental nuclear criticality in the tank farm. Most of these wastes were removed from the tanks and sent to the calciner for processing. Calcine Waste in Bins The Waste Calcining Facility, which operated from 1963 until 1981, and the New Waste Calcining Facility, which operated between 1982 and 2000, were designed and constructed to calcine (i.e., rapidly evaporate and decompose anions such as nitrate and carbonate to yield a granular solid) aqueous wastes generated from spent fuel reprocessing. The calciners are fluidized-bed units: the liquid waste is sprayed into a vessel containing an air-fluidized bed of granular (200-500 mm diameter) particles and heated to 400-600C. The

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report FIGURE II-9 Calcination process (top) and calcine simulant (bottom). original design used an internal heat exchanger to heat the bed. Later, the system was modified, and kerosene and oxygen were injected into the vessel (see Figure II-9). Kerosene burns in the hot vessel, and its combustion provides the necessary process heat to vaporize the waste. The liquid and volatile portion of the anions evaporate, and most of the remaining constituents of the waste adhere to the granular bed particles. This process reduces the volume of the liquid waste by a factor of between 2 and 10. The granular calcine waste was piped pneumatically into tall, stainless steel bins, (see Figure II-10). The bins are contained in concrete vaults called Calcined Solids Storage Facilities (also known as “bin sets”). The bin sets were built from the late 1950s to late 1980s and were designed to last 500 years. There are seven bin sets that contain from three to twelve bins each. The bin sets are located below (or partially below) the ground surface. The bins are different in size, sometimes even within the same bin set. Bin heights range from 6 m (20 feet) to 21 m (68 feet); their (outer) diameters range from 0.9 meters (3 feet) to 4.1 m (13.5 feet). In the case of annular bins, the space between the outer cylinder and the inner cylinder varies from 0.6 m (2 feet) to 1.9 m (6.25 feet). Storage volumes in each bin set range from approximately 226 m3 (8,000 cubic feet) to approximately 1,506 m3 (53,200 cubic feet). All of the bins can be accessed from the top through installed risers, except for those in bin set I, which has no access risers. The number of risers varies from one to five per bin. In general, the annular bins have more access risers than do the cylindrical bins. Bin set I is expected to be the most challenging for waste retrieval because there is no installed retrieval access. The bins also contain numerous internal obstructions, such as internally mounted wall stiffeners and bottom braces, which could hinder waste retrieval operations (Steiger and Swenson, 2005). The largest bin set (bin set VII) is empty. Bin set I consists of four sets of three concentric units; bin sets II and III are composed of seven cylindrical units, while bin set IV is composed of three cylindrical units; bin sets V, VI, and VII are composed of seven annular units (see Figure II-10). Calcine Bin Inventory Approximately 41 MCi (1.52 × 1018 Bq) of waste,14 nearly all of the liquid waste from reprocessing of spent fuel at the Idaho National Laboratory, had been calcined by May 2000, when the calciner was shut down to comply with a 1999 modification to a notice of noncompliance consent order with the State of Idaho. The composition of the calcine varies depending on the composition of the fuel and its cladding, as well as any chemicals added during reprocessing and calcination. Aluminum- and zirconium-clad fuels yield calcines containing alumina (Al2O3) and zirconia (ZrO2) as major constituents. Hydrofluoric acid was used to dissolve zirconium-clad fuel. Aluminum nitrate was added to the liquid waste to complex the fluoride. Calcium nitrate was added to the waste at the calcining facility (not in the tank farm) to prevent fluoride volatility in the calcination process. Thus, calcium fluoride (CaF2) and alumina are also major constituents of zirconia calcine. Boron, sodium, chromium, iron, lead, mercury, and other trace metals are present in the calcine waste as oxides; some sodium and potassium are present as nitrates; and some magnesium and calcium carbonate from the dolomite startup bed is also present. 14 This value is decay corrected to 2006.

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report FIGURE II-10 Calcine solids storage facilities. The horizontal line outside the bin shows the location of the land surface. SOURCE: Patterson, 2005. Very few characterization data are available on the calcine at the Idaho National Laboratory because of the high dose levels and the difficulty of reaching the material in the bins. Some of the information that is of current interest, particularly the concentration of long-lived radioactive nuclides and RCRA metals, was not routinely collected at the time of waste generation. Information gaps were filled using process knowledge. The relative error bound for calcine inventory is 14 percent at a 95 percent confidence level (Steiger and Swenson, 2005). A previous National Research Council report (NRC, 1999b) discusses calcine characterization in Idaho bins and points out both the discrepancies among characterization information and the heterogeneity of calcine properties potentially existing in the bins. Moreover, what is known today about the calcine appears to be based on pilot tests with cold surrogates and not on sampling information. For example, the only samples of actual calcine that have been retrieved from bins consist of two core samples collected from the second bin set in 1979. Another calcine sample from the output of the calciner was collected in 1993. The calcination heat source was changed from indirect liquid-metal heating to in-bed combustion just around the time this waste was calcined. In-bed combustion generates both oxidizing and reducing chemical environments in different regions, which could affect calcine properties. To date, no samples of calcine produced by in-bed combustion have been retrieved from bins. Sodium-Bearing Waste in Tanks The roughly 500 kCi (1.85 × 1016 Bq) of radioactivity in the liquid radioactive waste remaining at the INTEC tank farm is called sodium-bearing waste. DOE describes sodium-bearing waste as “a liquid mixed radioactive waste produced from the second and third cycles of spent nuclear fuel reprocessing and waste calcination, liquid wastes from INTEC closure activities stored in the Tank Farm, solids in the bottom of the tanks, and trace contamination from first cycle reprocessing extraction waste” (DOE-ID, 2002). Sodium bearing waste has high concentrations (more than 2 moles per liter) of sodium nitrate salts, resulting from the addition of sodium hydroxide to the washing solution to enhance its effectiveness in removing some residues. Some

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report FIGURE II-11 Idaho National Laboratory tank farm volumes. SOURCE: Lockie et al., 2005. of this liquid waste was sent through the calciner to produce calcine waste. Figure II-11 shows how much waste is stored in each of the roughly 1,100 m3 (300,000 gallon) tanks. DOE is now cleaning seven tanks, and one tank is a clean spare. Idaho Tank Inventories The vast majority of Idaho tank waste is in a liquid form, but a small amount of insoluble solids can be found at the bottoms of the tanks. DOE has sampled waste from several tanks. In a typical tank, the cesium-137 and strontium-90 (and their short-lived decay products borium-137m and yttrium-90) account for most of the radioactivity in both the solids and the liquids. In tank WM-187, for example, strontium and cesium together contribute 97 percent of the total radioactivity of 0.22 Ci per liter (8.1 × 109 Bq per liter). Isotopes of plutonium (mostly plutonium-241 and plutonium-238) constitute 1.9 percent (Barnes et al., 2004). The inventory of radioactive waste in the INTEC Tank Farm is listed in Table II-3. More than 187 samples have been retrieved from the tanks at the Idaho National Laboratory since 1987. Most of the samples are obtained via steam jet pump from the tanks. Some samples were obtained directly via a Light Duty Utility Arm (Olson, 2005). The uncertainty in quantities is generally within 10 percent for most chemical constituents; less than 20 percent for most radionuclides; and about 30 percent for solids; however, many organics were not able to be detected during sample analysis. The tanks are currently sampled for steam reforming15 processing needs. Idaho Tank Waste Processing and Tank Closure Waste processing at the Idaho National Laboratory is different from that at the other two sites: There are no plans to perform chemical separations on the liquid waste or calcine to generate a high-activity fraction and a low-activity fraction. DOE has recently selected steam reforming as the technology to convert sodium-bearing waste into a solid form (DOE-ID, 2005b). The calcine waste will be put in a form suitable for disposal in a monitored geologic repository and will be ready for shipment out of Idaho by 2035. DOE plans to ship its solidified sodium-bearing waste to the WIPP 15 In a typical steam reforming process, superheated steam, along with the material to be treated and co-reactants, is introduced into a fluidized bed reactor where water evaporates, organic materials are destroyed, and the waste constituents are converted to a granular, leach-resistant solid (NRC, 2005b; see also, DOE-EM, 2005 and DOE-ID, 2002).

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report TABLE II-3 Inventory of Radioactive Waste by Type at the Idaho National Laboratory Type of Waste Volume (m3) Radioactivity (Ci) Total tank and bin wastea ~5,000 35-36 million Comprising     Treated Sodium-bearing waste in tanks ~ 500-800 ~520,000 Calcine waste in bins 4,400 35 million Waste leaked into environment from pipes and valvesb 107 37,000 Service wastewater injected to aquiferc 45 million 22,000 Stored transuranic wasted,e 65,000 343,000 Buried transuranic-contaminated waste and soild,e 37,000 297,000 Low-level waste (including mixed) stored 2,200 Not available Low-level radioactive waste in disposal cells d 158,000 12 million TOTAL >45 million >49 million NOTE: Shaded entries are wastes that ultimately are expected to remain on-site. These data are from different sources, are measured or estimated at different times, and did not indicate quantified uncertainties. This table does not include spent nuclear fuel stored on-site (i.e., from naval and test reactors as well as from Fort St. Vrain and Three Mile Island) or contaminated soil at the evaporation ponds, which have been remediated. 1 Ci = 3.7 × 1010 Bq. a Lockie, 2005a. Decay corrected to 2012. b Cahn, 2005. Not decay corrected, therefore an overestimate. c DOE-ID, 2003a. Radioactivity not decay corrected. 99.8% of the radioactivity is tritium. d DOE, 2001b. e As of 1996. facility in New Mexico. DOE’s plans for and status of waste retrieval, processing, and tank closure at Idaho are described in Chapters III, IV, and V respectively. POLICY BACKGROUND The 1954 Atomic Energy Act (AEA) gave the Atomic Energy Commission (the predecessor agency of both DOE and USNRC) the authority to manage nuclear waste generated from both defense and commercial nuclear fuel cycle activities. The 1982 Nuclear Waste Policy Act (NWPA) defined the term “high-level waste” (HLW) and officially adopted deep geologic disposal as the nation’s long-term strategy for managing this waste. The definition of HLW, as set out in the Nuclear Waste Policy Act (42 U.S.C. Section 10101), 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 contains fission products in sufficient concentrations; and (B) other highly radioactive material that the [Nuclear Regulatory] Commission, consistent with existing law, determines by rule to require permanent isolation. It is apparent from this text that Congress defined HLW in the AEA and the NWPA in terms of its source. Section 3116 of the NDAA provides an exception to this definition at the sites in South Carolina and Idaho. DOE Order 435.1 still applies to waste determinations at Hanford and potentially to other wastes at the Savannah River and Idaho sites to which Sect. 3116 does not apply. In 1993, the USNRC first set out criteria to determine which portions of certain Hanford nuclear fuel reprocessing waste are not HLW (the waste so determined is also called “waste incidental to reprocessing” in some documents).16 DOE, which regulates itself on most matters related to radioactive waste, developed Order 435.1 which contains provisions for determining that some wastes are not HLW and, thus, can be managed as low-level waste or transuranic waste (DOE, 1999a; 1990b; 2001b). According to DOE Order 435.1, waste can be determined to be incidental to reprocessing by two methods, “citation” or “evaluation.” The citation method simply lists certain wastes, such as resins and clothing, that DOE identifies as incidental to reprocessing. The evaluation method is based on three criteria provided to DOE by USNRC in 1993 in its denial of a petition for proposed rulemaking concerning the definition of HLW (Bernero, 1993). The Commission…has indicated…it would regard the residual fraction as “incidental” waste, based on the 16 The first official document referring to “waste incidental to reprocessing” is the provisions of DOE Manual 435.1 concerning determining whether DOE tank waste is not HLW. “Incidental” waste is mentioned in a March 4, 1993 Federal Register Notice in which the USNRC set forth criteria for determining that waste from Hanford double-shell tanks disposed of in a grout facility would not be HLW USNRC found that the principles for waste classification are well established, endorsing the criteria DOE later used in Order 435.1 (NRC, 2005b; see also Bernero, 1993).

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report Commission’s understanding that DOE will assure that the waste: (1) has been processed (or will be further processed) to remove key radionuclides to the maximum extent that is technically and economically practical; (2) will be incorporated in a solid physical form at a concentration that does not exceed the applicable concentration limits for Class C low-level waste as set out in 10 CFR Part 61; and (3) will be managed, pursuant to the Atomic Energy Act, so that safety requirements comparable to the performance objectives set out in 10 CFR Part 61 are satisfied. (Bernero, 1993) Thus, the tank waste that USNRC reviewed and that was destined for disposal on-site would not be considered HLW if it met the criteria. On this basis, DOE Manual 435.1 created in effect three implicit subcategories of waste: (1) high-level waste, (2) non-high-level waste that is managed as low-level waste, and (3) non-high-level waste that is managed as transuranic waste (see Appendix C, Table C-1). Using the provisions of DOE Manual 435.1, DOE proposed to determine that certain wastes at the three DOE sites that are the subject of this report are not HLW, a step needed for DOE to carry out its separation strategy (high-activity and low-activity) for the tank wastes. This process came to an abrupt halt in 2003 when DOE was sued in Idaho by the Natural Resources Defense Council, Snake River Alliance, Confederated Tribes and Bands of the Yakama Nation, and the Shoshone Bannock Tribes. The plaintiffs argued that Order 435.1 exceeded DOE’s authority under the AEA and the NWPA. In 2004, the court found that the standards DOE established by rule were too discretionary and offered no effective limitation on the agency’s ability to determine which waste could be managed as low-level waste and disposed on-site. The federal district court in Idaho ruled in favor of the plaintiffs, finding that DOE could not continue with its management activities in reliance on Order 435.1.17 DOE appealed the district court’s decision. The U.S. Court of Appeals for the Ninth Circuit did not rule on the legal merits of the district court’s ruling. It reversed the district court on the procedural ground that the case was not yet “ripe” for judicial determination.18 In other words, the Ninth Circuit expressed no opinion on the legality of Order 435.1, but put off the question for a later time, when DOE actually takes action under the authority of Order 435.1. Although the decision that struck down Order 435.1 was vacated, the Order could be contested at its first use. This leaves Order 435.1 in some degree of legal limbo in Idaho, where the only existing opinion (albeit vacated) is negative and in Washington state, which is also in the Ninth Circuit. DOE saw the rulings as a major impediment to its pursuit of a separation strategy at the Hanford and Savannah River Sites and to tank closure at all three sites. So, even before the Ninth Circuit rendered its decision on the appeal, DOE sought a statutory remedy from Congress. In Section 3116 of the Ronald Reagan National Defense Authorization Act of 2005, Congress established criteria for determining that some waste from spent fuel reprocessing is not high-level waste and may be disposed of on-site at the Savannah River Site and the Idaho National Laboratory. The Hanford Site, however, was not included in the provisions of Section 3116 because the state of Washington explicitly is not covered or bound by the section. In its criteria, Congress implicitly divided the non-high-level waste from spent fuel reprocessing destined for on-site disposal into two subclasses, depending on the concentrations of radionuclides in the waste in relation to Class C concentration limits in 10 CFR 61.55 (see Appendix C) although the differences are only procedural (NRC, 2005a). Therefore, under Section 3116, at the Savannah River Site and the Idaho National Laboratory (but not Hanford), there are essentially three subclasses or categories of tank waste from reprocessing: HLW, non-HLW Class C or less, and non-HLW greater than Class C. Section 3116 is similar to Order 435.1 in many ways, most importantly in the standard for removal of radionuclides to the maximum extent practical and in the use of the performance objectives in 10 CFR 61 as benchmark criteria for on-site disposal. However, there are some critical differences. First, Section 3116 addresses only wastes that are to be disposed of on-site and which are subject to a state compliance agreement whereas the provisions of DOE Manual 435.1 could encompass any waste and its planned destination. Section 3116 does not say that waste disposed on-site is low-level waste, although it is implied that such wastes will be managed by near-surface disposal like other low-level waste disposed on-site. Section 3116 was intended to resolve the legality of the overall separation strategy at the Savannah River Site and of tank closures at the Savannah River Site and Idaho (but not, of course, at Hanford). Unlike Order 435.1, however, Section 3116 does not provide authority or guidance on tank waste determinations for retrieved non-high-level waste to be managed as transuranic waste, probably because defense transuranic waste is slated for geologic disposal at the Waste Isolation Pilot Plant in New Mexico and Section 3116 only applies to waste that stays on-site.19 Second, Section 3116 sets out roles for the host states and USNRC, which are absent from Order 435.1. DOE requested informal USNRC input on waste determinations performed before 2004 under Order 435.1 (Camper, 2005; Flanders, 2005). However, the USNRC did not have any official regulatory role in that capacity and provided general and nonbinding comments on DOE’s waste determinations. The 17 NRDC, Inc. et al. v. Abraham, 271 F. Supp. 2d 1260 (D. Idaho 2003). 18 That is, DOE had not yet actually applied Order 435.1 in the Idaho case. NRDC, Inc. et al. v. Abraham, 388 F.3d 701 (9th Cir. 2004). 19 See definition of transuranic waste in Appendix K.

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report situation at the Hanford Site is somewhat different because the Federal Facility Compliance Agreement and Consent Order for this site formally requires USNRC input on the effectiveness of DOE tank waste retrieval. Third, Section 3116 and Order 435.1 differ in their description of the degree of removal of the highly radioactive fraction: Section 3116(a)(2): “has had highly radioactive radionuclides removed to the maximum extent practical” Manual 435.1-1 (p. II-1): “has been processed, or will be processed, to remove key radionuclides to the maximum extent that is technically and economically practical” The meaning of the unmodified term “practical” in Section 3116 requires some interpretation, i.e., whether it is the same as the “technically and economically practical” found in Manual 435.1-1 or something different (including more, fewer, or other considerations). Fourth, a time of compliance is not mentioned in 10 CFR Part 61. A time of compliance of 10,000 years has been recommended by Nuclear Regulatory Commission staff in its guidance on performance assessments, but this recommendation has not been approved by the commissioners, so it does not constitute official agency policy. However, DOE Order 435.1 specifies a time of compliance of 1,000 years for low-level waste disposal facilities, which complicates the issue. This difference may need to be resolved, however, DOE has made it a practice to carry performance-assessment calculations out to the peak dose within 10,000 years, possibly making the difference irrelevant. The time of compliance is discussed further in Chapters VI and VIII. Multiple Legal Drivers and Decision-Making Authorities The cornerstones of DOE’s authority to manage radioactive waste are the AEA and the NWPA. However, the AEA and the NWPA are not the only applicable federal statutes. Federal legislation such as the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA), Resource Conservation and Recovery Act (RCRA), Clean Air Act, Clean Water Act, Safe Drinking Water Act, National Environmental Policy Act (NEPA), and correlative state laws all have a part to play. The relevant considerations under these statutes go well beyond, and often adopt different approaches than, the AEA, NWPA, and Section 3116 of the NDAA. Moreover, these other laws are not administered by DOE, but by the Environmental Protection Agency (EPA) and, through delegated authority, the states. Through Order 435.1, DOE regulates storage and treatment as it relates to radiological components of waste at the three sites. Order 435.1 could be applied to waste determinations at Hanford, subject to resumption of the legal challenges brought previously in Idaho which were suspended on the basis of ripeness. Disposal actions for wastes that DOE determines under Section 3116 not to be high-level waste are to be monitored by the USNRC in coordination with the host state. In addition, all sites have entered into Federal Facility Agreements and Consent Orders on behalf of DOE with each host state and EPA (and the U.S. Navy at Idaho) (Idaho FFACO, 1991; Hanford FFACO, 2003; SRS FFA, 1993; Idaho SACO, 1995). These agreements establish the operational goals and milestones for DOE’s site cleanup operations. They provide authoritative interpretation of DOE’s statutory and regulatory obligations, and they add requirements, such as milestones and technical performance specifications; however, Federal Facility Agreements cannot establish requirements that are contrary to existing federal laws. DOE’s plans for Savannah River Site tank waste disposition are subject to the approval of the South Carolina Department of Health and Environmental Control under the Savannah River Site Federal Facility Agreement (SRS FFA, 1993). The state regulates the hazardous component of the waste through the South Carolina Hazardous Waste Management Act while the tanks are closed under wastewater treatment and hazardous waste regulations (SCDHEC, 2004a, 2004b). The closure milestones in the Federal Facility Agreement for this site are 2022 for Type I, II, and IV and 2028 for Type III tanks. Plans for the Idaho National Laboratory tank waste disposition are subject to the approval of the Idaho Department of Environmental Quality under the Idaho Federal Facility Agreement and Consent Order and the Settlement Agreement and Consent Order (Idaho FFACO, 1991; Idaho SACO, 1995). The 1995 court settlement (called the Settlement Agreement) among DOE, the U.S. Navy, and the State of Idaho requires that sodium-bearing waste be solidified and made ready for disposal outside Idaho by 2009. The 1995 court settlement also requires all high-level waste to be prepared for removal from Idaho by 2035. Through the Idaho Hazardous Waste Management Act (HWMA, modeled on RCRA), the state regulates the treatment and storage of the hazardous components of the waste (sodium-bearing waste, tanks, and calcine). The waste remaining in the tanks following closure may be subject to continued RCRA-HWMA regulation. A 1991 Notice of Noncompliance consent order signed by both the EPA and Idaho established a schedule for DOE to cease use of the tank farm and perform closure activities. The tanks should be closed in six phases from 2005 to 2016. DOE Order 435.1 regulates storage and treatment of radiological components of the waste (sodium-bearing waste, tanks, and calcine). Section 3116 does not apply to waste determination at Idaho if the waste is not slated for onsite disposal. According to a state regulator, potential disposal at the Waste Isolation Pilot Plant as transuranic waste would be regulated by EPA and the State of New Mexico; otherwise, the NWPA applies (Trever, 2005). DOE’s plans for Hanford Site tank waste disposition are subject to approval of the State of Washington, Department

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report of Ecology under the Hanford Federal Facility Agreement and Consent Order, also known as the Tri-Party Agreement (Hanford FFACO, 2003). The RCRA closure plan (under the Washington Hazardous Waste Management Act) applies to tank farms, while the AEA and NWPA apply to tank residuals and waste left in pipes. Tank farm closure is performed under CERCLA. The Hanford closure schedule milestones in the Tri-Party Agreement are the year 2024 for single-shell tanks and 2032 for double-shell tanks. Consideration of various other legal drivers for the tank wastes (such as the Clean Air Act and the Safe Drinking Water Act) illustrates another dimension of complexity. The numerous waste types related to the tanks are governed by many regulations, and in some cases the regulatory framework differs for the same type of waste stream depending on the site. Waste types in the tanks or related to tank waste are the following: Bulk tank waste; Tank residuals (including heel); Tank itself and interior equipment (e.g., cooling coils, nonretrievable cleaning equipment); Cesium and strontium capsule contents and containers (at Hanford only); Piping and valve boxes connecting tanks and between tank farms; Contaminated soil below tanks and pipes and in tank farm areas; and Used waste management equipment (e.g., HLW melters). Different combinations of legal and regulatory standards apply to each of these. Chapter VIII contains findings and recommendations relevant to such a complex legal and regulatory framework. DIFFERENCES AMONG SITES Tables II-4 and II-5 show that there are major differences among the three sites in terms of the geology, hydrology, climate, physical and chemical composition of the waste, and tank-system designs. These differences are described in this section. Differences in Natural and Man-Made Conditions at Each Site The natural features at the sites differ, from elevation to rainfall, although the two western sites have more natural conditions in common with each other than with the Savannah River Site. It can also be seen that in the manmade features, particularly the fuel reprocessing methods TABLE II-4 Differences in the Natural Features from Site to Site   Savannah River Site Hanford Site Idaho National Laboratory Average seasonal low/high temperature C (F) 2/33 (36/92) 0/24 (32/76) −7/18 (19/65) Extremes low/high temperature C (F) −19/42 (−3/107) −32/33 (−25/92a) −45/39 (−49/103) Distance from tank farm to nearest surface water       by land; ~0.9 km 15 km (downgradient) 61 m (ephemeral stream) by groundwater 1.85 km 30 km 200 km Natural flow of nearest surface water average, maximum (cubic meters per second) 0.5, 23.5 (rates measured in Four Mile Branch) 3,360, 19,500 (rates measured in the Columbia River) Intermittent, 82.4 (estimated 100-year flood of Big Lost River) Subsurface medium Loamy sand, sandy clay, clay, silty clay Unconsolidated glaciofluvial sands and gravels, fluvial- lacustrine sediments, basalt Basalt overlain by alluvial deposits of gravel-sand-silt with silt and clay interbeds Average depth from ground surface to water table 9.75 m 90-100 m 143 m Average annual precipitation 124.4 cm 16 cm 22.1 cm Average annual soil infiltration 40 cm 0.4-1.0 cm 0.36-1.1 cm Complications for monitoring and modeling contaminant transport Clay lenses; multiple aquifers with different flow rates and directions; large infiltration rate Deep vadose zone, highly variable conductivity and sorption values, varied stratigraphy Deep vadose zone, highly variable conductivity and sorption values (ranging several orders of magnitude); perched aquifers a Hoitink et al., 2005.

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report TABLE II-5 Differences in the Man-Made Features from Site to Site       Idaho National Laboratory Tank Specification Hanford Site Savannah River Site SBW Tanks Calcine Number of tanks/areas to close 177/18 tank farms 51/2 tank farms 11a/1 tank farm 7 calcine bin sets Tank types 2 (149 SST and 28 DST) 4 (Type I, II, III, IV) 1 2 (annular and cylindrical bins) Tank sizes, 103 gal 55-1,160 750-1,300 30-318 60-471 Construction periods (or years when in service) 1943-1964 SSTs Type I: 1954-1965 1953-1966 1960s 1968-1986 DSTs Type II: 1956-1960         Type III: 1971-1992         Type IV: 1959-1965     Construction material Carbon steel Carbon steel Stainless steel Stainless steel Tank maximum ages in years at closure More than 75 More than 75 More than 60 More than 40 Tank conditions 67 confirmed and assumed leakers, estimated 1 million gallons to soil 11 leakers, 1 to soil No leakers No leakers Tank depth relative to water table Well above water table Some tank bottoms in water table Well above water table Above surface Extent of obstruction in tanks Abandoned equipment, debris Severe obstructions due to vertical cooling coils in most tanks Little or no obstructions (cooling coils on the bottom and walls)   Waste types Viscous, alkaline liquid, sludge, saltcake, diatomaceous eartha Viscous, alkaline liquid, sludge, saltcake, zeoliteb Acidic, liquid sodium waste, and small amount of sludge Calcined powder Waste volume, 106gallons 54 33 1.4 1 Waste radioactivity, 106Ci (3.7 × 1016 Bq) 193 in tanks 136 in capsules and German logsc 426 0.52 24 Retrieval schedule SSTs complete by 2018d and DSTs by 2028a 2019 for Type I, II, and IV; 2024 for Type III HLW retrieval complete by 1998; remaining liquid waste by 2012 Road-ready by 2035 Closure schedule SSTs by 2024d and DSTs by 2032a 2022 for Type I, II, & IV; Type III by 2028; In six phases from 2005 to 2016 Not yet determined NOTE: DST = double-shell tank; HLW = high-level waste; SST = single-shell tank. a Diatomaceous earth was used as waste sorbent material to immobilize residual supernatant liquid in tanks where liquid removal by pumping was not feasible (see Appendix K). b Zeolites were used to remove cesium from the condensed steam recovered from an evaporator. Zeolite particles contain “trapped” cesium ions (along with other ions) and are difficult to retrieve by pumping because of their high settling rate (see Appendix K). c Cs and Sr capsules = cesium and strontium capsules (see Appendix K). d Currently reevaluating retrieval and closure schedules. SOURCE: Adapted and elaborated from CH2M Hill Hanford Group, Inc., 2003.

OCR for page 13
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report and storage, the Savannah River and Hanford Sites have much more in common with each other than they do with Idaho. The types of waste at the Idaho site are really distinct from those at Hanford and the Savannah River Site because only a single reprocessing technology was used and the wastes were kept in acidic condition and then converted into solids. This has made retrieval of the sodium-bearing wastes from their tanks much easier. DOE anticipates that the same will be true for the calcined solids. Differences in Readiness at Each Site Table F-1 in Appendix F shows that the Savannah River Site, the Hanford Site, and the Idaho National Laboratory are at different points in their removal and stabilization of tank wastes. As a result of these differences, the amount of detail in the committee’s discussions of the processes and recommended changes at these sites varies greatly. The experience and progress toward final tank farm closures are different at each site. The Idaho National Laboratory, for example, has retrieved the waste from the majority of its tanks, 7 of the 11 1136 m3 (300,000 gallon) tanks and all 4 of its 114 m3 (30,000 gallon) tanks. Hanford has completed waste retrieval from only 4 of its 177 tanks, while the Savannah River Site has retrieved the waste from 4 of its 51 tanks—in each case a small percentage of the number of tanks to be emptied. Calcine retrieval has not yet been tested with the radioactive solids in the bins at the Idaho site. In some cases, the tanks from which the wastes have been retrieved were chosen because it was thought that they would be the easiest to clean. DOE has put very few tanks (and their wastes) through the other major steps, namely separation, treatment, and disposal of the retrieved wastes and closure of the tanks. Therefore, there is little operational experience so far on tank remediation and closure. To fill this lack of operational experience, move the program forward, and try to encompass all of the likely variants, some experimental work at the bench scale, less experimental work at a pilot scale, and computer programs used to model facility performance have been utilized at the sites. However, the sites acknowledge that there is no assurance that the tank remediation process will move forward seamlessly in the future with more than 240 tanks and bins to close.20 This section shows that there are varied geologic, topographic, and climatological differences among and within sites and that the design of the tanks and their waste contents vary widely, so unique solutions may be required for each tank or group of tanks. However, given the uncertainties and the challenges ahead, the committee judges that it would be desirable to have general guidelines on technologies for tank waste remediation that apply to all tanks. There is no dispute that all snowflakes or fingerprints are unique, but that does not mean that there are no similarities or common features among them. The same is true of the solutions for tank waste retrieval. Tank waste retrieval is not like an assembly line in which everything is the same except the color. The distinctive design of the tank, the mixtures of the wastes in the tank, and the local geology, hydrogeology, and precipitation can influence the best methods for the removal of wastes to the maximum extent practical and how much needs to be removed. The tanks also have many similarities: They are all steel, they are all located below ground, and most are above the water table. They have similar limitations in the size of the tools that can be inserted into the tank, and the techniques for removing the most difficult wastes from the tanks are similar. Therefore, many things can be learned from each waste removal operation, but the details of what removal operations to use in a particular tank cannot necessarily be determined in advance because conditions in the tank may be very different from earlier experience. However, it may be possible on the basis of earlier experience to identify some techniques that are more or less likely to work in that particular tank’s environment. FINDINGS AND RECOMMENDATIONS Despite this being the background chapter, the first findings and recommendations are given here because they establish the framework for the report. They also explain why this report discusses processes (e.g., waste retrieval, waste processing, tank closure, monitoring, performance assessment, decision making) in a “global” way, while each site is treated individually at a more detailed level. Finding II-1: There is great diversity in the natural and manmade conditions within and among sites. Recommendation II-1: Each tank or group of tanks with similar problems should be addressed using an approach specifically tailored to best address the particular situation by taking into account site- and tank-specific conditions, previous experience, and advancements in technology. Finding II-2: There has been limited operational experience acquired in tank remediation and closure so far at the Hanford Site, Idaho National Laboratory, and Savannah River Site. Recommendation II-2: Given the early stage of the tank waste remediation program and the challenges ahead, the committee judges that it would be desirable to have general guidelines on applicable technologies for tank waste remediation that apply to all sites. 20 As mentioned earlier, the Hanford Site has 177 tanks; the Idaho National Laboratory has 11 tanks and 7 calcine vaults; and the Savannah River Site has 51 tanks, two of which are already closed.