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 71
Alternative High-Level Waste Treatments at the Idaho National Engineering and Environmental Laboratory 8 Tank and Bin Closure The 11 tanks that contain the sodium-bearing waste (SBW) liquids and the seven bin sets that contain the solid high-level waste (HLW) calcine will eventually be closed regardless of the final disposition of these wastes. This chapter evaluates the closure options for these two storage systems. Because liquids in storage are not considered a long-term solution for disposition, the tanks will be "emptied" before closure. Storage of the calcine solids in the bin sets have potentially a longer term stability than tank storage of liquids; therefore, the bin sets may be considered as either "emptied" or full when closed. The quotation marks around emptied emphasize that the bins and tanks will contain some residual waste due to limitations of the present drawdown systems and the difficulty of achieving complete decontamination. These storage systems are presently monitored continuously and actively controlled, and it is assumed that such monitoring and control will continue for several decades while final disposition is being carried out. The starting points for the site's evaluation, as contained in a bin set closure study (Dahlmeir et al., 1998) and a tank closure study (Spaulding et al., 1998), are the following: The 11 tanks will be emptied to the level of capability of the existing steam jet and airlift systems leaving a residual 4- to 12-inch layer of highly acidic heel (probably a mixture of both solids and liquids) at the bottom interspersed in the cooling pipes. Deposits on the walls behind and on the cooling pipes will be present also (Spaulding et al., 1998). The bin sets will have the calcine removed to a capability of the presently planned retrieval method (see Chapter 2). Transport piping and the existing 8-inch risers used in the retrieval task are the hardware available to remove calcine prior to closure. Additional secondary containment will be in place to enable new openings to the bin sets to be made to carry out further decontamination activities as needed (Dahlmeir et al., 1998). The regulations impacting on the technical issues are those currently in place, although changes are likely and new regulations may be promulgated before closure is initiated (Spaulding et al., 1998; Dahlmeir et al., 1998). The interaction between the technical issues of practicality and regulatory issues of what is sufficient will always be dynamic. Opportunities for trade-offs exist, and the Department of Energy (DOE) and its contractors should identify, define, and evaluate these
OCR for page 72
Alternative High-Level Waste Treatments at the Idaho National Engineering and Environmental Laboratory trade-offs in order to permit an understanding of the possible benefits and risks among several options being considered for closure. Closure of the two storage systems is evaluated separately in the following discussion; however, there may be common technical issues for both systems. There are three technical actions involved in the closure process: characterization—knowing what remains, decontamination—knowing what added effort at retrieval is needed, and stabilization—how to immobilize the residual waste. The closure of both of the two storage systems (i.e., tanks and bins) is evaluated for each of these three actions. TANK SYSTEMS: CHARACTERIZATION A heel composed of both solution and sludge deposits is anticipated in the bottom of the tanks (Spaulding et al., 1998). The present characterization data on solutions in the tank are available for process control and engineering design purposes, but these data are probably not a reliable predictor of the composition of the bottom heel, because there may be an enrichment of hazardous elements and radioactive isotopes in the sludge portion (Spaulding et al., 1998).1 To determine the level of decontamination and the degree of immobilization for closure, characterization of the heel is essential. The tank sludge has accumulated through progressive processing campaigns; therefore, old analyses are probably not relevant to current conditions. For this reason, old analyses (>5 years old) of the composition of the SBW sludge were mainly discarded (Garcia, 1997). The remaining analyses are sparse and vary considerably from sample to sample. The characterization database does not describe how the samples of sludge are obtained. Sample size appears to be a problem issue. Larger sample sizes involve more risk to personnel and make handling more difficult than smaller sample sizes. In this vein, microanalytical methods to assay just the hazardous and radioactive constituents needed for closure could be considered. Variations between samples are expected to be large compared to the instrumental accuracy; therefore, reliance on a sufficient number of small size samples (as opposed to the large sample sizes) may improve the overall characterization effort. Sampling of sludges at other DOE sites is a challenge also, and the Idaho National Engineering and Environmental Laboratory (INEEL) should continue to capture any benefits from these other investigations. The tank closure study (Spaulding et al., 1998) identifies a task to characterize the heel (presumably including sludge) in all but one of the six options for tank closure (Spaulding et al., 1998). Two of the options characterize each tank over a 1-month period, 6 months prior to initiation of heel removal. Heel removal of all tanks is spread over about a decade starting in year 2007, and removal is done one tank at a time. Three options start characterization of heels in year 2000 and continue until 2016. Some options are premised on a risk-based analysis before qualification and acceptance by regulators. A risk assessment requires characterization to identify the source terms of the hazardous and radioactive constituents. The current characterization database is insufficient to define the risk unless large margins of conservatism of the key constituents can be justified without a heel analysis. Ideally, all eleven 1 Calculations of radionuclide inventories for tank closure use data from Garcia (1997). There is no traceability from the original sources.
OCR for page 73
Alternative High-Level Waste Treatments at the Idaho National Engineering and Environmental Laboratory tanks should be sampled and characterized before selecting a closure option. If a tank heel can be sampled and characterized in 1 month, then the sampling and characterization schedule for all eleven tanks could extend over a period of 2 to 3 years or more, depending on the ease of deployment of sampling equipment and the analytical laboratory capacity. Characterization of the tanks has the goal of providing data for the following two goals: process control of liquids that are removed by the existing steam jet and airlift for subsequent treatment (i.e., parameters needed for the design and testing of chemical process steps used in the downstream processes to treat the liquid waste), and risk assessment evaluations to leave or retrieve the remaining heel. These two goals may allow for different protocols to be developed in sampling and in the use of analytical resources. For example, process control analyses may need large sample sizes of solution to analyze for major constituents that define downstream adjustments of process parameters or chemical additions, and large liquid samples might then be collected for this purpose. However, the risk assessment may only need smaller samples sufficient in size to identify the existence of hazardous and radioactive species in the sludge and their concentration relative to the solution. For the solution portion of the heel, this need may be met by analysis of a sample collected for process control purposes. In this approach, the sludge (i.e., the solid part) of the heel would be the primary focus for near term sampling development. Important answers provided by sludge sampling would include the following: the content of Resource Conservation and Recovery Act (RCRA) materials such as heavy metals (i.e., Hg) and chlorinated hydrocarbons that were used at the INEEL site and that, if present in the tanks, would likely settle to the bottom; the total organic carbon content, which could be tested by a procedure such as the carbon dioxide released on ignition of sludge; the curie content of alpha entitling isotopes and their spatial distribution in the heel; the curie content of beta/gamma emitting isotopes; and the mass of sludge per area of tank floor surface. Sampling and analysis of sludge to obtain this information would probably settle the issue of establishing a "representative sample" to satisfy regulatory and stakeholder needs. Sludge sampling has been done in the past, so it is possible and practical. In collecting sludge data for performance assessments and other uses, the degree of analytical rigor that is required should be assessed to determine whether innovative thinking would permit simpler procedures to be used. Full advantage should be taken of any regulatory flexibility. TANK SYSTEMS: DECONTAMINATION Decontamination of a tank involves removing as much as possible of the liquid and sludge portions of the heel on and below the piping array. Six options for decontamination of the heel have been studied. One involves complete ("clean") removal of the heel and dismantlement of the tank and vault. The remaining options involve decontamination to an extent sufficient for a regulated closure based on a risk assessment. The tank walls and cooling pipes are washed down to remove suspected sludge deposits. The remaining heel that is not
OCR for page 74
Alternative High-Level Waste Treatments at the Idaho National Engineering and Environmental Laboratory removable by steam jets is reduced from about 12 inches off the tank floor to about 1 inch using a disposable peristaltic pump. It is not evident how much energy2 is needed to suspend the sludge for removal. The preferred approach of Spaulding et al. (1998) appears to be a grout sweep-out system that directs the residual heel to an existing steam jet or airlift system. The process involves using a grout slurry that lifts and displaces the solution, and presumably the sludge as well, so that it raises the liquid to the level of a steam jet. The grout slurry is assumed to capture the sludge. A maneuverable remote transfer arm will be placed inside the tank to deliver the grout to the entire floor area. The design must consider the cooling pipe array and the brackets holding the pipes that may resist flow of the slurry. The sweep-out system may be evaluated in a mock-up facility before implementation (Spaulding et al., 1998). The Savannah River Site has grouted a tank in this fashion but in a different environment and under different regulatory rules (Spaulding et al., 1998). Success of this process depends on the reliability of mock-up testing. Unsuccessful testing will raise concerns and extend the development costs. If the grout solidifies prematurely in any of the transfer systems, recovery measures will have to be pretested as part of the development. Most tanks have two to three injection ports, except for one tank, which has only one port (Spaulding et al., 1998). The conditions for grouting the sludge in the tank are uncertain at present. Preliminary testing has shown that the pH of the heel is an important criterion for making a strong grout (Spaulding et al., 1998). To adjust the pH to between 0.5 to 2.0, the heel will be diluted by a series of washdowns. During this dilution, additional precipitation is expected to add to the sludge. There are concerns that sludge deposited under highly acidic (>2 M) conditions and sludge deposited under lower acidic (near 0.5 to 1.0 M) conditions may have different mixing qualities with the grout slurry. In any case, the degree and quality of mixing is not measurable when performed in situ. With no assurance of a well-mixed and strong slurry, the release rate of hazardous and radioactive constituents by the grout will be in question, as will the ability to model this release rate conservatively. Valid and reliable sludge simulant experiments may be required, including large-scale mock-up testing to certify the reliability of the concept. The criteria for the acceptable degree of decontamination of a tank should be established before implementation of a technical approach. If the residual sludge can be reduced sufficiently, the certification of the grout/sludge mixing may be unnecessary for a given level of risk. TANK SYSTEMS: STABILIZATION The degree of stabilization (e.g., via addition of grout) needed for the hazardous chemical elements and radioactive isotopes remaining in the tank is determined by the mobility of these species in the environment after the loss of physical containment (e.g., a breach in the tank wall by corrosion). After closure there is a period in which one or more barriers is the key means for assuring continued containment of the residual waste. During this period some radioactive elements decay, and those that remain to contribute to the time-dependent activity are elements that can be stabilized to varying degrees. For hazardous elements such as mercury or heavy metals, the hazard remains forever; therefore, their long-term 2 An example of one of these energy-adding techniques is to bubble an inert gas into the tanks while the liquid is drawn off, in order to agitate and suspend solids in the liquid to expedite their removal.
OCR for page 75
Alternative High-Level Waste Treatments at the Idaho National Engineering and Environmental Laboratory concentrations in the surrounding environment, as determined by their rate of release from stabilized residues and their mobility, should be within accepted regulatory standards. The regulations governing tank closure do not completely specify the technical means for compliance. Although regulatory procedures often do not use risk assessment calculations as a means for establishing a safe condition, risk assessments provide a rationale for trade-offs among various technical considerations. Two such considerations are the degree of decontamination of the tank interior that is to be achieved and the requisite quality of mixing between injected grout and residual sludge. These two issues are related—for instance, if the grout does not mix with the sludge attached to the steel tank surface (the primary environmental barrier), then no credit can be derived from the grout, and the decontamination specification would need to be correspondingly more stringent. Some options for tank closure after stabilization of the grouted heel include continued void filling with grout made from uncontaminated materials or from materials containing one or more substances that are hazardous or radioactive. If the tank void is filled with these substances, the degree of stabilization of the original heel may be irrelevant. Comments are given below about stabilization of specific constituents that remain in the tank after closure. Water A primary requirement for closure is that the tanks contain no free liquid. If the tank is grouted, the grout usually bonds the water as a hydrated species, and the water stays bonded under containment at ambient temperatures. However, the water may become mobile if there is a thermodynamic driving force such as temperature gradients in the tanks, which could, over long periods, allow free water to collect in localized areas within the tank. Thus, thermal gradient testing may be as important as thermal cycling. Halides Halides (chloride and fluoride) are potential accelerators of intergranular stress corrosion of stainless steel. Aluminum nitrate is used to complex the halides. Thus, materials used to fill the tank voids must be monitored for high salt content. Mercury Until there is an understanding of where the mercury (used in spent fuel reprocessing) is distributed in the waste streams, it will be important to monitor the mercury concentration in grout, and the effect of this mercury contribution on the performance of the system upon closure. Radioisotopes The standards for the U.S. Nuclear Regulatory Commission (USNRC) waste Classes A and C define the disposition of allowed radioactive waste in a shallow land disposal facility. The isotopes of concern are 90Sr and 137Cs in the early period when containment is the principal method to control dispersal. After a few hundred years, these isotopes will decay and
OCR for page 76
Alternative High-Level Waste Treatments at the Idaho National Engineering and Environmental Laboratory the overall activity decreases substantially. Thereafter, the principal isotopes contributing to the activity are alpha emitters such as 239Pu and 240Pu and a beta (β) emitter, 99Tc. It is important to know if these isotopes are concentrated in the sludge part of the heel in order to show that stabilization meets the performance goals derived from applicable concentration and release limits. After loss of containment, the 99Tc leach rate may be the key parameter in a risk assessment. BIN SET SYSTEMS: CHARACTERIZATION The amount of calcine in the bins after decades of storage has been estimated from calculations of calcine added to the bins from a series of processing campaigns (Dahlmeir et al., 1998). If mechanisms such as caking have occurred to make the calcine stationary, operations such as characterization and retrieval might have to rely on an in-situ comminution process. In-situ sampling of this calcine is difficult and may not be useful unless the bin contents have been homogenized3 sufficiently before calcine removal, which seems impractical and unlikely. Access mechanisms into the bin systems are very limited so that obtaining representative samples of the residuals would probably require extensive alteration of the monitoring and control systems. Consequently, characterization of the calcine composition prior to retrieval would be a challenge. As a result of this challenge, the proposed closure program for the bin sets (Dahlmeir et al., 1998) does not include a significant characterization task for unretrieved calcine. Risk assessments of the options for closure will rely on characterization of retrieved calcine and on conservative bounding estimates of the residual calcine left after retrieval. The proposed strategy of Dahlmeir et al. (1998) is to apply a fall-back sequence to closure using best-technology efforts that start with an attempt to achieve "clean" closure first. If this is not attainable, then Resource Conservation and Recovery Act (RCRA) closure and, finally, closure to "landfill standards" criteria would be applied. BIN SET SYSTEMS: DECONTAMINATION Decontamination of the bins depends on the success of the airlift transport system for removal of the calcine. A planning assumption is that 5 percent of the calcine will remain in the bins after completion of the retrieval program (Dahlmeir et al., 1998). Cylindrical bins and annular bins may require different approaches to air lifting of the residue. There are also internal supports in the bins that may trap calcine. As described in the characterization section above, the bin sets have limited access points, and the monitoring and control system would probably require alteration before any insitu decontamination operations could occur. However, alterations of hardware to properly satisfy eventual closure would best be done before the start of calcine removal operations, and actions taken during calcine retrieval will impact the success of bin decontamination. Therefore, the retrieval and bin closure operations are related, and the two engineering projects—the Calcine Retrieval and Transport Project (CRTP) and the Bin Set Closure Project (BSCP)—responsible for these operations have a potentially complicated interface. 3 The previous chapter considered "sufficient" homogenization of already retrieved material to ensure product consistency in physical or chemical treatment processes. This chapter also considers waste homogeneity, but in relation to sampling stored waste prior to retrieval. In particular, in situ characterization is of interest for the residuals likely to be unretrieved and therefore to remain in the tanks or bins during their closure.
OCR for page 77
Alternative High-Level Waste Treatments at the Idaho National Engineering and Environmental Laboratory As an example, CRTP actions can influence the ability to decontaminate the bins (Spaulding et al., 1998). The rush to get CRTP operations under way to retrieve calcine from the bins, to feed downstream HLW and transuranic (TRU) separations, may put pressure on funding and management to ignore the closure task. This situation splits the system development activity and narrows potential innovations to a subsystem perspective. The system function is to move the calcine, which has three subfunctions of retrieve, transport, and decontaminate. Since airlift mechanisms for both CRTP and BSCP are only in the conceptual stage, a closure cycle system outlook is important. Because the lift mechanisms during retrieval and decontamination have the same principle of operation, the proof-of-principle and mock-up testing can be developed in concert, rather than under separate project requirements. A similar synergism is to apply the energy dissipation of the decontamination mechanism (e.g., a CO2 decon system) to the retrieval operations, where this energy may be useful to solve the problem of calcine inhomogeneity in the bin during the retrieval phase and to break up potential caking. Clean closure of bins requires nitric acid treatment. Undissolved calcine is expected to remain after treatment so that more aggressive treatments may be necessary, but their effectiveness may be questionable. The calcine has a much higher specific activity than the SBW in the tanks and the bins were not designed for transport of liquids. As a result of these physical conditions and high radiation fields, the intensive decontamination requirements associated with clean closure of the bins are considered by Dahlmeir et al., (1998) to be impractical. BIN SET SYSTEMS: STABILIZATION As stated above, the clean closure option for the bins has been deemed impractical (Dahlmeir et al., 1998). Therefore, the approach for defining closure specifications is based on a risk assessment of the residual calcine left in the bin. The assumption in Dahlmeir et al. (1998) is that any hazardous constituents will be delisted (see Chapters 9 and 10 for further discussion); hence, the risk is determined by the stabilization of residual radioisotopes. As with the tanks in the tank farm, there is an incentive to fill the bins with low-level radioactive grout that contains radionuclide concentrations at USNRC Class A or C levels. The option for closure becomes dependent on the acceptance criteria, qualification of source data, and algorithms used in the assessment. This strategy requires a credible characterization of the remaining calcine, which is not now planned. It is possible to provide a conservative bounding of the key characterization data to show that the risk is acceptable in lieu of accurate measurements. Selection of a particular bin closure option appears possible using this risk-based approach. To qualify for filling with clean grout or grout containing radioisotopes at Class A or C standards, an upper limit on the amount of residual calcine remaining in the bin sets must be known. Further, the residual calcine must be well mixed with the grout to receive credit for its stabilization. These issues are similar to those for tank farm closure. The curing of the grout in the bin environment will have to be prequalified for leachability, coherence, and strength, since it cannot be tested in situ. Assurance that no free liquid remains after a 7-day cure would also be needed. The full set of trade-offs between exposure to personnel and risk to the public have yet to be assessed. This situation apparently requires a negotiation with the Idaho state regulators (Dahlmeir et al., 1998).
OCR for page 78
Alternative High-Level Waste Treatments at the Idaho National Engineering and Environmental Laboratory USE OF A RISK ANALYSIS TO DEVELOP CLOSURE SPECIFICATIONS AND TO MODEL RELEVANT RISKS A probabilistic risk assessment was done (Slaughterbeck et al., 1995) to evaluate the health and environmental risk from current INEEL operations and future waste treatment options. The risk assessment was a high-level analysis and did not focus on specific designs for future facilities. It evaluated consequences of events such as fires, earthquakes, explosions, and an aircraft accident. Risks to the public and to on-site personnel 100 meters or more from the origin of the accident were evaluated. All the options considered in Slaughterbeck et al. (1995) included tank and bin closure. The risk from tanks and bins after closure was assumed to be nonexistent. Within the limits of the Slaughterbeck et al. (1995) analysis of accident scenarios, all waste treatment options for INEEL HLW, including the "no action" option, have similar risks. This is due in part because of the limitations of that analysis, which did not model in detail the operations internal to any plant facility, did not intercompare various treatment options, and did not consider risk to workers in close proximity to such operations. The committee found that not enough information about the various technical alternatives is available to detail the risks associated with each. In general, however, the more complicated the process steps that are conducted, the greater the technical risk. A risk analysis of different scope, to model long-term environmental effects and processing risks for workers, could and should be done to develop closure specifications for the tanks and bin sets. One potential outcome of such a site risk analysis is that all treatment options may have similar public risk, but that risk to personnel will be lowest for the closure option with the least manpower activity involved. If the calcine is retrieved and separated into fractions that are stored and/or disposed of at the site but not in the bins or tanks, then one would expect little change in public risk. If the closure assumption is that as much as 5 percent of the calcine may be unrecovered by retrieval and separation for offsite disposition, then one must ask how much increase in public and personnel risk occurs if 100 percent of the calcine is left. In other words, is the current bin storage of calcine acceptable on a relative risk basis, and can treatment options be justified in risk terms? One way to probe these issues is to ask the following questions in the event that 100 percent (versus 5 percent) of the calcine is left. Is the health and safety risk to the public increased 20 times? If there is a loss of containment (i.e., a leak), does the dispersal rate increase by a factor 20? Is a factor of 20 within the margin of conservatism of data and algorithms used in the assessment? Many risk analyses work with margins of a few decades; hence, a factor of 20 is a relatively modest one. If 100 percent of the calcine is left in the bins, rather than 5 percent, monitoring and controlling operations should not result in 20 times greater risk from radiation exposure because the calcine is self-shielding. The calcine has been safely contained for decades and will be monitored and controlled for at least several more decades. If there is a gradual degradation of containment (e.g., a rapid rate of corrosion of the steel vessel, which has not been observed yet), forecasts of life expectancy should be possible on at least a several-decade timeframe. The structural integrity of the bin sets, improved and updated as needed, should protect against an earthquake during the monitoring and controlled period. In the meantime, the radioactive sources are decaying and risk of radiation exposure is decreasing. The
OCR for page 79
Alternative High-Level Waste Treatments at the Idaho National Engineering and Environmental Laboratory conclusion is that whatever the risk level is now to workers and the public, it is decreasing and can be managed adequately for the foreseeable future. Similar risk-based arguments can also be used to derive tank closure specifications and strategy. Cursory analysis of the content of the calcine indicates that the principal contributors to risk (e.g., heavy metals and actinides in the heels) are the same as for the SBW heels in the tanks. The integrity of the containment vessel, and monitoring for releases, would be important components of the strategy to control risk in the next several decades. After the integrity of containment can no longer be assumed, releases to the environment would be controlled by the waste form (from which radioactive and hazardous chemical constituents would leach into the surrounding environment) left in the tank. This would be in combination with other factors such as environmental mobility and rainfall that are less amenable to direct control. The bin sets in their present form have pressure relief valves to release gas if it accumulates in the bins. Characterization of samples of calcine show that gas evolution in sealed containers occurs when the calcine is heated above 200 °C due to decomposition of the nitrate (Garcia, 1997). Over time, the temperatures from internal heat sources will diminish so that heating would have to occur from an external source. Closure would require containment of gas as well as solid when the facility goes to a passively monitored state from an actively monitored and controlled state. The risk analysis would need to include a potential pressure event. The degree of severity depends on the free space in the bin and the amount of calcine retained.
OCR for page 80
Alternative High-Level Waste Treatments at the Idaho National Engineering and Environmental Laboratory This page in the original is blank.
Representative terms from entire chapter: