B.8.0 MAJOR ASSUMPTIONS
To develop engineering data required to perform impact analyses for each of the alternatives discussed in the EIS, assumptions were made regarding the technologies that have been configured to create a remediation alternative. These assumptions were based either on the best information available, applications of a similar technology, or engineering judgement. By definition when an assumption is made there is some level of uncertainty associated with it that can be expressed as a range for the assumed value that reasonably could be expected. This section identifies the major assumptions used for the alternatives and the uncertainties associated with the cost estimates . Uncertainties associated with the engineering data are discussed in Volume Five, Appendix K.
B.8.1 IN SITU ALTERNATIVES
It was assumed that there would be no leaks from the SSTs or DSTs during the administrative control period for the No Action, Long-Term Management, or In Situ Fill and Cap alternatives. This assumption is based on ongoing SST interim stabilization to remove pumpable liquids, the ability to detect and recover leaks from the space between the inner and outer liners of the DSTs, and ongoing monitoring activities within the tank farms. The SSTs and DSTs were assumed to maintain their structural integrity throughout the administrative control period under the No Action and Long-Term Management alternatives.
The In Situ Vitrification, In Situ Fill and Cap, and the in situ portion of the Ex Situ/In Situ Combination alternatives were assumed to require additional characterization data to evaluate the acceptability of in-place disposal and address RCRA land disposal requirement considerations. This requirement would be in addition to the current characterization requirements for the ex situ alternatives. These additional characterization efforts could involve extensive laboratory analysis of additional tank samples and may require modifications to the tanks to install additional risers for sampling access.
In Situ Vitrification
The in situ vitrification system was assumed to be capable of vitrifying each of the tanks to the required depth resulting in a consistent waste form. It was also assumed that the variation in waste composition and inventory from tank to tank would not impact the ability to produce an acceptable waste form.
In Situ Fill and Cap
The concentrated liquid waste contained in the DSTs was assumed to be acceptable for gravel filling. Under the In Situ Fill and Cap alternative, the DST liquids would be concentrated using the 242-A Evaporator to remove as much water from the waste as possible but would still contain substantial volumes of liquid. It has been estimated that concentration by the 242-A Evaporator would reduce the current liquid volumes contained in the tanks by approximately one-third (WHC 1995f).
B.8.2 EX SITU ALTERNATIVES
Waste Retrieval Efficiency
The waste retrieval function described for the ex situ alternatives was assumed to remove 99 percent of the waste volume contained in each tank during waste retrieval. Under this assumption, 1 percent of the tank volume would be left in-tank as a residual. It was further assumed that the 1 percent waste volume represented 1 percent of the waste inventory on a chemical and radiological basis.
The amount and type of waste that would remain in the tanks after retrieval is uncertain. The Tri-Party Agreement (Ecology et al. 1994) set a goal for the SSTs that no more than 1 percent of the tank inventory would remain as a residual following waste retrieval activities. The engineering data for the waste retrieval and transfer function common to all ex situ alternatives was developed using 99 percent retrieval as a goal. However, achieving this level of tank waste retrieval may require extraordinary effort and cost, and it may not be practicable to achieve 99 percent retrieval from all tanks.
Releases During Retrieval
Retrieval of SST waste under each of the ex situ alternatives was assumed to result in the release of 15,000 L (4,000 gal) from each SST to the soils surrounding the tank during retrieval operations. It was also assumed that the contaminant concentrations in the liquids released were at maximum predicted concentrations using the congruent dissolution model. See Volume Four, Section F.2.2.3 for a discussion on the congruent dissolution model. No leakage was assumed to occur from the DSTs during retrieval operations because DSTs have provisions for leak containment and collection . This assumption is based on having 67 known or suspected SSTs that have leaked in the past (Hanlon 1995). Most of the SSTs were built in the 1940's and now are about 50 years old. The leakage volume estimate was based on current information from the waste retrieval program and on the assumption that the average leakage from an SST would be one order of magnitude lower than the maximum release estimated for tank 241-C-106 during sluicing operations. The maximum leak estimated from tank 241-C-106 during sluicing operations was 150,000 L (40,000 gal). Th e leak estimate for tank 241-C-106 assumes that the leak occurs early in the sluicing operation, leak detection devices and controls fail, sluicing operations proceed without these leak detection devices, the leak(s) occur at the bottom of the tank, and the remaining sludge does not plug any leaks (DOE 1995d).
The assumption that each of the 149 SSTs leaks 15,000 L (4,000 gal) during retrieval is conservative and provides an upper bound of 2,260,000 L (596,000 gal) on the calculated impacts from tank leakage during retrieval. Total leakage from all SSTs during retrieval operations would be expected to be lower than the bounding values used because of the following assumptions.
- Seventy-five percent of the tanks that are known or suspected leakers are assumed to have leaked at the air-water interface on the sidewall of the tank and would remain above the liquid level during sluicing (51 tanks).
- Twenty-five percent of the tanks that are known or suspected leakers are assumed to have leaked at or near the tank bottom and would be retrieved using a robotic arm based system (16 tanks). The robotic arm based system would not use the large volumes of liquids required for sluicing operations.
- Leak detection systems would be used during waste retrieval operations, and indications of tank leakage during retrieval would result in actions taken to minimize leakage. These actions could include switching to robotic arm based systems or limiting the amount of sluicing liquid in the tank.
- Administrative controls would be used to monitor liquid inventories.
- There is a tendency for solids in the sludge to plug any leaks.
- The free liquid in the tanks during sluicing could be pumped out in a short time using the transfer pumps.
The most probable occurrence of a leak during sluicing would involve the sluicers opening a plugged leak in the tank wall. The waste leakage during sluicing would be any free-standing liquid above the level of the leak point and the sluicing stream as it impacts the tank wall. Based on historical leak rates of other SSTs, the actual leaked volume is expected to be on the order of a few thousand liters (a few thousand gallons) (DOE 1995d). DOE currently is working with Ecology to define the operating envelope for allowable leakage during retrieval. Final design of the waste retrieval systems would include measures to detect control leakage.
Tank Residuals
The residual contaminants left in the tanks would either be insoluble and hardened on the tank walls and bottom or be of a size that could not be broken up and removed from the tanks. In either case, the residual would have low solubility because the retrieval technologies proposed would use substantial quantities of liquid in an attempt to dissolve or suspend the waste during retrieval. Because of the uncertainties regarding the amount and type of residual waste that would remain in the tanks, a conservative assumption was made to bound the impact of the residual waste. For purposes of the analysis, it was assumed that 99 percent recovery would be achieved for ex situ alternatives, and the residual waste left in the tanks would contain 1 percent of all the original tank inventory, including the water-soluble contaminants.
The assumption that the 1 percent tank residuals following retrieval represent 1 percent of the original tank inventory is conservative because it assumes that soluble and insoluble constituents would remain as residuals in the same proportions as the original tank inventory. The effect of retrieving less than 99 percent of the waste volumes from the tanks during retrieval would be an increase in the amount of waste left in the tanks and corresponding increases in groundwater contaminant concentrations and post-remediation risk. The in situ and combination alternatives leave substantially more waste onsite for disposal and provide an upper bound on the impacts associated with the amount and type of waste that is disposed of onsite.
A nominal case retrieval release and residual tank inventory was developed to assess the impacts that would result from nominal, as compared to bounding, assumptions for tank releases during retrieval and the residual waste left in the tanks following retrieval. Additional information on the nominal case is provided in Section B.3.0.
The nominal retrieval release inventory was developed by assuming that the waste would be diluted by one-third by adding water during waste retrieval. Possible dilution ratios that would be used during waste retrieval range from 3:1 to 10:1. Thus, the dilution factor of one-third assumed for the nominal case is a conservative assumption and is substantially lower than the dilution factions that would be obtained using 3:1 or 10:1 dilution ratio. These dilution ratios represent the amount of liquid required to mobilize the waste solids and would be made up of existing tank liquids and water additions. The nominal case retrieval release volume was assumed to be 15,000 L (4,000 gal) from each SST and the contaminant concentrations were assumed to be two-thirds of the bounding case. The average volume of waste released from each SST during retrieval was not reduced for the nominal case because insufficient information is available to support a lower average release volume. The volume of waste released during retrieval would depend on the ability to detect a leak and take corrective action.
The nominal tank residual inventory was developed by modifying the bounding tank residual inventory to reduce the mobile constituents of concern based on solubility. The mobile constituents of concern were evaluated because of their contribution to post-remediation risk. The isotopes C-14, Tc-99, and I-129 were reduced for the nominal case tank residual inventory to 10 percent of the bounding tank residual inventory. This is based on the assumption that 90 percent of the residual inventory of these isotopes would be soluble in the retrieval liquids and would be retrieved from the tanks for ex situ treatment. Typical sludge wash factors representing the solubility in water for each of these isotopes are as high as 99 percent. The nominal case residual was limited to 90 percent to account for conditions where the scale and hardened sludges were not exposed to the sluicing liquid during retrieval. Table B.8.2.1 shows the nominal and bounding residual inventories for select mobile constituents.
Table B.8.2.1 Tank Residual Inventory in Curies
Assumptions Affecting HLW Volume
The major factors that affect the volume of HLW produced by any of the ex situ alternatives include waste inventory, waste loading (glass specifications), blending, and the efficiency of the separations processes.
The waste inventory that has been used for all alternatives is provided in Volume Two, Appendix A along with a discussion on data accuracy and uncertainty.
Waste loading is the mass fraction of the nonvolatile waste oxides in the vitrified waste. The waste oxide loading would be controlled by the amount of glass formers that are added during the vitrification process. The higher the waste loading, the more waste that would be contained in the vitrified glass and the lower the waste volume.
Blending is the mixing of the waste from different tanks during retrieval to obtain an average waste feed stream for treatment. Because there are 177 tanks that contain waste and the waste composition varies from tank to tank, it would be difficult to achieve a completely uniform blending of the waste during retrieval.
Separating the waste into HLW and LAW streams for treatment would involve various processes to physically or chemically separate specific constituents in the waste stream. The separations efficiency would be a measure of how well these processes work and would define the amount of each constituent that would be processed in the HLW and LAW treatment facilities.
The assumptions used for each of the factors described previously and their combined affect on the overall volume of HLW and LAW are discussed in the following sections.
Waste Loading
The waste loading for all ex situ treatment alternatives except for Ex Situ No Separations was assumed to be 20 weight percent waste oxides for the HLW and 15 weight percent sodium oxide for the LAW. The waste loading for the Ex Situ No Separations alternative was assumed to be 20 weight percent sodium oxide.
Waste loading was assumed to be 20 weight percent waste oxides (this includes all waste constituents that would be converted to oxides in the vitrified waste form, excluding the sodium and silica contained in the tank waste) for HLW glass for each alternative that would involve separating the HLW and LAW. Because the No Separations alternative would not separate the HLW and LAW, all of the sodium in the waste inventory would be converted into the HLW glass and the methodology described for the other alternatives would not be valid. The 20 weight percent sodium oxide loading for the No Separations alternative would result in a glass that would be equivalent to established glass compositions defined in the Waste Acceptance System Requirements Document (DOE 1995s). The Waste Acceptance System Requirements Document does not set specific limits for the different constituents that make up waste loading, but instead requires that for acceptance a waste form must be equal to or better than the reference glass.
The waste loading would affect the volume of waste that would be produced from a given amount of waste. This volume, along with the operating schedule and the assumed operating efficiency, would determine the size of the processing facilities and the operating resource requirements required to support the process. A decrease in waste loading would then translate into a larger volume of vitrified waste, larger treatment facilities or longer operating schedules, increased resource requirements, and higher disposal cost.
Waste loading may typically range from 20 to 40 weight percent waste oxides with 30 to 35 weight percent loading used as a target value. The Defense Waste Processing Facility glass has a design basis waste loading of 25 weight percent and a maximum waste loading of 38 percent (DOE 1995s).
The waste loading for all alternatives that would produce LAW was assumed to be 15 weight percent sodium oxide. The volume of LAW produced affects the size and number of LAW disposal vaults that would be built onsite.
Waste Blending
Each of the ex situ alternatives that use vitrification as an immobilization technology have assumed a waste blending factor of 1.2 for the HLW to account for variations in the composition of the waste during retrieval operations. Variations in the waste feed composition would not affect the calcined product that would be produced by the Ex Situ No Separations (calcination) alternative. Uniform blending would require simultaneous retrieval from specific groups of tanks to deliver a uniform average feed stream to the treatment facilities. The blending factor would be multiplied by the volume of HLW produced under uniform blending conditions to calculate the volume of HLW expected due to variations in the waste feed. Because variations in the waste feed composition would not be expected to affect the LAW vitrification process, a blending factor of 1.0 was assumed for LAW. One of the major sources of uncertainty associated with developing a retrieval sequence that would achieve a uniform blending was the lack of accepted tank-by-tank inventory data. The HLW blending factor for Hanford tank waste was the recommendation of an independent technical review team (Taylor-Lang 1996) .
Separations Efficiencies
The volume of vitrified HLW produced would be a function of the waste loading and the mass of waste to be vitrified. Reducing the HLW volume through separations processes would therefore require separating the nonradiological constituents from the HLW constituents during the pretreatment process. The lower bound on the number of canisters that could be produced would be controlled by the heat-generating limit of 1,500 W per canister (DOE 1995q). This heat-generating limit would provide a lower bound on the number of 1.2- m3 ( 41 -ft3) canisters of 177 for the tank waste and 298 for the tank waste combined with the Cs and Sr capsules (WHC 1995e). The following flowsheet assumptions would affect the volume of HLW produced:
Intermediate Separations, Phased Implementation, and ex situ portion of the Ex Situ/In Situ Combination 1 and 2 alternatives :
- The enhanced sludge washing process would solubilize 85 percent of the aluminum, 75 percent of the Cr, and 70 percent of the phosphate into the liquid phase, and following solid-liquid separations these would be included in the LAW feed;
- Solid-liquid separation would assume gravity settling in the tanks followed by decanting of the liquid. The solids settling process was assumed to achieve 50 weight percent solids.
Extensive Separations:
- Solid liquid separations would use centrifuges capable of achieving 0.1 percent solid in clarified liquids.
- Acid-side dissolution of the solid phase species would assume between 50 and 90 percent dissolution in a two-step dissolution process. This would include recycling 95 percent of the undissolved solids from the second acid dissolution step back to the caustic leaching step to begin another dissolution cycle. The remaining 5 percent of the undissolved solids would be sent to the HLW process. There is uncertainty in the optimistic acid-side dissolution assumptions that are critical to the volume of HLW produced by the extensive separations process.
The volume of HLW produced would directly impact the number of HLW packages requiring disposal at the potential geologic repository, which in turn would affect the cost associated with disposal. The number of HLW packages produced would also determine the number of offsite shipments required to transport the immobilized HLW to the potential geologic repository. The waste loading would also determine the concentration of radiological contaminants in the waste form. There would be a relationship between the waste loading, number of shipments (probability of an accident), and the concentration of contaminants in the waste form (consequence of an accident). As the waste loading increased, the probability of an accident would go down because there would be fewer trips required to transport the waste, but the consequences of an accident would go up because there would be a higher concentration of contaminants in the waste form (see Appendix E, Section E.15.0 for a discussion of accident uncertainties).
Canister Size and Type
Two sizes of HLW canisters were assumed for the ex situ alternatives. All of the ex situ alternatives except the No Separations alternative assumed a canister size of 0.6-m inside diameter by 4.57-m long (2-ft inside diameter by 15 ft long) with a net volume capacity of 1.17 m3 (41 ft3). The Ex Situ No Separations alternative (both vitrification and calcination) assumed a canister size of 1.7-m diameter by 4.6 m long (5.5-ft diameter by 15 ft long) with a net volume capacity of 10 m3 (360 ft3).
It is recognized that these sizes are larger than the 0.62-m3 (22-ft3) standard size canister that is identified for canistered HLW in the current waste acceptance requirements at the potential geologic repository (DOE 1995q). However, the DOE Office of Civilian Radioactive Waste Management has recently acknowledged the technical acceptability of a longer canister (e.g., 0.6-m diameter by 4.6 m long [2-ft diameter by 15 ft long]) for Hanford HLW (Milner 1996). The larger 10-m3 (360-ft3) canister assumed for the No Separations alternative was not evaluated for acceptance by the Office of Civilian Radioactive Waste Management. The large canister would occupy the same space as a standard waste package. The standard waste package would consist of four 1.2-m3 (41-ft3) canisters within a large disposal container. The design of the waste package and canister sizing has not been finalized.
B.8.3 COST UNCERTAINTY
Cost uncertainty for the various tank waste treatment alternatives has been evaluated using Decision Science Corporation's Range Estimating Program for personal computers. The Range Estimating Program has been applied to thousands of diverse problems by thousands of users. The Range Estimating Program inputs allow the user to specify a simple range rather than require selection of a probability density function. The Range Estimating Program outputs identify, quantify, and rank the risks.
The upper level of the cost range for new technologies was estimated such that there was a high certainty that its capital or operating cost would not be exceeded. This upper level (as a percent of the estimated cost) varied up to a high of plus 200 percent based upon the degree of uncertainty and complexity of the technology. The use of this high-range level addressed the concerns expressed in the System Requirements Review, Hanford Tank Waste Remediation System Final Report issued April 1995, which indicated that actual costs of new technology facilities of the type under consideration herein can often exceed estimated costs by a factor of two or more (DOE 1995s).
The information presented in Table B.8.3.1 identifies a range for the total estimated cost of each alternative. This range represents the calculated variation in estimated cost that could occur for any of the alternatives. This range is a function of input parameters such as the level of design development, uncertainties associated with implementability, and assumptions made for the relative uncertainty of different cost components. The total estimated cost range is statistically based and was obtained through a Monte Carlo simulation. The input parameters are based on the alternatives described in the EIS; however, major changes to the waste inventory, conceptual designs, or major assumptions would change the estimated cost range.
Table B.8.3.1 Comparison of Tank Waste Alternatives Cost Uncertainty
Input to the Range Estimating Program was based on best available information, conceptual cost estimates, and engineering judgement (Jacobs 1996).
|
NEWSLETTER
|
| Join the GlobalSecurity.org mailing list |
|
|
|
