X
Illustrative Example of the Recommended Decision-Making Process

The technical complexities of the tanks and their interactions with the geosphere are included where possible in the Department of Energy’s (DOE’s) performance assessments and in the models they use. Therefore, the performance assessment and sensitivity analyses that are performed using different scenarios can indicate the impact of various decisions on projected risks. Indeed, DOE’s draft Section 3116 waste determinations for the closure of Tanks 18 and 19 (see DOE-SRS, 2005a, Figure 7-14, p. 137) and for the Idaho National Laboratory tank farms show how these two sites used performance assessments to make tank closure decisions. However, as discussed in Chapter VIII, the trade-offs between risks and benefits were not always transparent.

EXAMPLE OF HOW TO STRUCTURE DECISIONS ABOUT WHETHER A TANK IS SUFFICIENTLY CLEAN TO GROUT AND CLOSE

The following is an example of a risk-informed, transparent decision-making framework for tank closure. This example illustrates how a single, consistent approach to thinking about both waste removal to the maximum extent practical and tank closure can lead to very different choices from site to site and even from tank to tank within a single site. It also shows that the choice of the point and time of compliance impacts the decision to close a tank or to carry out additional waste removal.

However, the following are important caveats about this example. The example uses grossly simplified (indeed hypothetical) assumptions about waste retrieval methods, scenarios, and groundwater flow patterns to focus on an illustration of the decision-making framework and to avoid distracting the reader with the assumptions used or the numbers obtained.

Also, the example does not account for the increase in risks to workers associated with decisions to clean tanks further. The trade-off between increasing health risks to workers living today in favor of decreasing these risks to residents and intruders living 10,000 years from now is a major consideration in tank closure decision making that could not be dealt with in this example.

The example is intentionally simple so that readers can gain an appreciation of the value of a consistently structured approach to thinking about tank closure decisions and gain better insight into the merits of some of the committee’s recommendations regarding tank closure decisions. Application to a real tank would require far more technical detail and consideration of more decision-relevant factors, but it would not have to differ in any fundamental way from the framework presented in this chapter.

The Illustrative Tank and Its Hypothetical Risk Profile

Risks associated with future release of radionuclides in the tank heels vary over time and space, which is called the “risk profile” to distinguish it from the narrower concept of risks at a single “point of compliance,” as required for establishing whether performance objectives have been met.1 Figure X-1 summarizes the illustrative example’s risk profile for grouting and closing a tank that already has some soil contamination around it, which will start to be cleaned up as soon as the tank has been closed. To economize on data presented in what is intended as a simple illustrative example, the committee summarizes this profile by showing how the dose level would change over time at three specific locations around the tank (identified in Figure X-1 as shaded circles) since any plume of radionuclides released from the tank after closure would move through the surrounding vadose zone

1

The term “risk” is used here to describe the doses at a particular point in time and space, under the standard assumptions of exposure of individuals at that location. This is not the usual concept of risk, which would account for the probability of such exposure occurring and also for the health effects of such exposure. In this example, the committee also assumes for simplicity of exposition that the performance objective is that the risk may not exceed 1.0 at the point of compliance.



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 112
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report X Illustrative Example of the Recommended Decision-Making Process The technical complexities of the tanks and their interactions with the geosphere are included where possible in the Department of Energy’s (DOE’s) performance assessments and in the models they use. Therefore, the performance assessment and sensitivity analyses that are performed using different scenarios can indicate the impact of various decisions on projected risks. Indeed, DOE’s draft Section 3116 waste determinations for the closure of Tanks 18 and 19 (see DOE-SRS, 2005a, Figure 7-14, p. 137) and for the Idaho National Laboratory tank farms show how these two sites used performance assessments to make tank closure decisions. However, as discussed in Chapter VIII, the trade-offs between risks and benefits were not always transparent. EXAMPLE OF HOW TO STRUCTURE DECISIONS ABOUT WHETHER A TANK IS SUFFICIENTLY CLEAN TO GROUT AND CLOSE The following is an example of a risk-informed, transparent decision-making framework for tank closure. This example illustrates how a single, consistent approach to thinking about both waste removal to the maximum extent practical and tank closure can lead to very different choices from site to site and even from tank to tank within a single site. It also shows that the choice of the point and time of compliance impacts the decision to close a tank or to carry out additional waste removal. However, the following are important caveats about this example. The example uses grossly simplified (indeed hypothetical) assumptions about waste retrieval methods, scenarios, and groundwater flow patterns to focus on an illustration of the decision-making framework and to avoid distracting the reader with the assumptions used or the numbers obtained. Also, the example does not account for the increase in risks to workers associated with decisions to clean tanks further. The trade-off between increasing health risks to workers living today in favor of decreasing these risks to residents and intruders living 10,000 years from now is a major consideration in tank closure decision making that could not be dealt with in this example. The example is intentionally simple so that readers can gain an appreciation of the value of a consistently structured approach to thinking about tank closure decisions and gain better insight into the merits of some of the committee’s recommendations regarding tank closure decisions. Application to a real tank would require far more technical detail and consideration of more decision-relevant factors, but it would not have to differ in any fundamental way from the framework presented in this chapter. The Illustrative Tank and Its Hypothetical Risk Profile Risks associated with future release of radionuclides in the tank heels vary over time and space, which is called the “risk profile” to distinguish it from the narrower concept of risks at a single “point of compliance,” as required for establishing whether performance objectives have been met.1 Figure X-1 summarizes the illustrative example’s risk profile for grouting and closing a tank that already has some soil contamination around it, which will start to be cleaned up as soon as the tank has been closed. To economize on data presented in what is intended as a simple illustrative example, the committee summarizes this profile by showing how the dose level would change over time at three specific locations around the tank (identified in Figure X-1 as shaded circles) since any plume of radionuclides released from the tank after closure would move through the surrounding vadose zone 1 The term “risk” is used here to describe the doses at a particular point in time and space, under the standard assumptions of exposure of individuals at that location. This is not the usual concept of risk, which would account for the probability of such exposure occurring and also for the health effects of such exposure. In this example, the committee also assumes for simplicity of exposition that the performance objective is that the risk may not exceed 1.0 at the point of compliance.

OCR for page 112
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report FIGURE X-1 Schematic of a tank site and locations where risk profile is summarized. NOTE: P.O. = performance objective. and groundwater. The three graphs at the bottom show the assumed projected risk levels at each of the three locations. (The gray concentric lines encircling the tank represent loci of approximately equal risk levels as estimated at each of the three illustrative locations.) Location 3 is intended in this example to be the designated “point of compliance.” The associated figure for Location 3 shows that radionuclides released from the closed tank would not reach that location until almost 400 years from now. The simplified example assumes that no other plume of contamination will reach the point of compliance before then. At its maximum (about 500 years from now), the risk from that contamination would not exceed 1.0, consistent with this example’s starting assumption, which is that if the tank in question could be closed with its residual heels as-is, it would meet the legally required performance objectives. Risks are not below the level of the performance objectives at all locations, however. Closer to the tank, the risks are larger and occur earlier in time. Location 1 reflects the temporal pattern of risks in the immediate vicinity of the tank, and Location 2 is a point about half-way between Location 1 and the point of compliance, Location 3. The hypothetical and grossly simplified spatial-temporal risk pattern in Figure X-1 should not be used to draw conclusions about the risks at the three receptor locations at any specific site. The example is hinged on a simplified situation where transport through the vadose zone flow is assumed to be approximately vertical and flow in the saturated zone is assumed to be approximately horizontal. Although these assumptions are not inconsistent with groundwater flow at the Savannah River Site, they are not representative of the conditions at Idaho National Laboratory or the Hanford Site, where horizontal transport in the vadose zone can mean that the highest dose from exposure to groundwater does not occur at the point closest to the tanks.2 Figure X-1 illustrates the committee’s statement that an entire profile of risks is relevant to decisions and not just the legally required meeting of performance objectives at a point of compliance. It also illustrates the committee’s statements 2 At the Idaho National Laboratory, the point of calculation was moved farther away from the facility boundary because estimated radionuclide concentrations are higher several hundred meters away from this boundary.

OCR for page 112
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report about the significance of the policy decision regarding what location will serve as the point of compliance in estimating whether a closure plan can meet performance objectives. If Location 2 were designated the point of compliance, the tank could not legally be closed under Section 3116 without further removal of residual radionuclides. The committee is not suggesting where the point of compliance should be placed because this is a policy decision made by the state and DOE. However, the committee believes that the full spatial and temporal variations of risks are not explicitly indicated in the performance assessment for tank closure. Figure X-1 also shows some hypothetical risks associated with the present soil contamination near this tank. It is reflected as high current risks at Location 1. If this soil contamination is remediated once the tank is closed, it never appears as far from the tank as Locations 2 or 3. Although this source of contamination is not related to the tank heels, it is not necessarily included in a performance assessment for tank closure. However, decisions that might affect the timing of tank closure might be affected by the interaction of this timing with the ability to rapidly reduce near and present contamination.3 For this reason, existing soil contamination in the vicinity of the tanks is included in this decision example. Alternatives to Immediate Tank Closure with As-Is Residual Heel Quantities In this example, once the bulk of the waste has been removed and the heel reduced to the limit of a given technology, the options are either to grout and close the tank or to proceed with further heel removal with alternative technologies. Two types of alternative technologies are chosen in this example based on considerations in the waste retrieval chapter (Chapter III): Further waste removal using an established technology, such as chemical cleaning with oxalic acid; or Further waste removal using advanced technologies that are still in the development stages but might become available within the next 10 years and have the ability to substantially reduce the heel in the tank before it is actually closed (see Chapter IX for examples). Both options involve trade-offs. Additional cost is one consideration, but this example relies solely on the offsetting risks that each alternative may create while offering a potential to further reduce the risk from heels that remain in a tank. The assumption is that the tank is still connected to the rest of the tank farm so additional removal does not involve reconnecting it or removing abandoned equipment from inside a tank that was prepared for final closure. The example arbitrarily introduces one risk from using oxalic acid:4 If the tank to be washed has a crack that is not an active leak site because it is plugged by the solid sludge matrix, the plug might be cleared during the washing process and some of the mobilized radionuclides may be released through the crack before the wash liquids can be pumped out of the tank. The example also takes into account two risks of waiting for an emerging new waste removal technology:5 The new technology may never work as expected, making the wait fruitless. The sludges in the ungrouted tank may become mobilized in the intervening years. The sludges are in a solid form at present and therefore are unlikely to “leak” from the tank under current conditions. However, if water (e.g., from rain or flood) were to enter the tank despite current precautions to prevent its ingress, mingle with the residuals, mobilize some of the radionuclides, and then exit again before evaporating (e.g., through a crack in the tank), some of the residuals could enter the environment before they were stabilized. Another risk scenario might be a catastrophic tank failure, due to subsidence or possibly an earthquake. Therefore, once the bulk of the waste has been removed and the heel reduced to the limit of a given technology, DOE faces three basic choices: Grout the tank immediately, with the residual heels at their current levels. Delay closure for about two years to perform an oxalic acid wash to reduce the heels to lower levels. Delay closure for about 10 years in the hopes that a specific emerging cleanup technology will allow heels to be reduced to lower levels during that interval. Each of these three options produces a different profile of “risks” over space and time around the tank site. Figure X-2 shows a decision tree that summarizes the trade-offs in this decision. To keep the example simple, the tree includes only the risk of sludge leaking from the tank during the wash process for oxalic acid washing. Under Option A, there is a 3 For example, staff at Idaho National Laboratory indicated that they feel urgency to close their emptied high-level waste tanks because they cannot start to clean up very high dose rate soil contamination from earlier transfer pipe leaks until the tanks are closed. 4 Other potential or perceived drawbacks of using oxalic acid, such as the criticality concerns and downstream problems described in Chapter III, are not considered in this example. 5 Once a specific technology has been identified as the one worth waiting for, a more specific set of risks can be added to the list that follows and included in the risk analysis that supports this decision.

OCR for page 112
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report FIGURE X-2 Illustrative tank decision tree structure. Decisions are made by comparing risk profiles EV(A), EV(B), EV(C), which combine risks now in soils (near-term) with risks possible if tanks leak (long term). NOTE: EV = expected value. particular risk profile that is consistent with the results of the performance assessment that finds the performance objectives will be met at the point of compliance at a time when exposure would be at a maximum. This was illustrated in Figure X-1 for a particular tank. Under Option B, the risk profile (including risk at the point of compliance) will be changed in one of two ways. If there is no leakage as a result of the acid washing, the main effect will be that risks at each point in time and space will be reduced from those under Option A by the amount by which the washing step reduces the overall radionuclide content that will be grouted in the tank. With some probability p, however, there would be a release from the tank at the time of the washing, and this will increase potential exposures at each location much earlier than in the case of immediate grouting. These two risk profiles can be combined probabilistically to obtain a single expected risk profile for Option B by weighting each profile by its probability and then adding them together. Option C in this illustrative analysis can be characterized by four different profiles. The four outcomes reflect the combinations of the tank leaking or not leaking before it can be grouted, and whether the technology becomes available or not. In the event that the tank does not leak before grouting at year 10, but the technology never becomes available during that time, the risk profile is identical to that of Option A. However, it now has a probability less than one, equal to (1 − q) × r. This profile is then probabilistically combined with the three other profiles to obtain a single expected risk profile for Option C. To the extent that there is a high risk of leakage in the 10 years of waiting, the expected risk profile for Option C will have higher risks than Option A in the early decades. Additionally, the greater the probability that the technology will never become available to reduce the quantity of heels in the tank, the less is the expected benefit from Option C in terms of reducing risks relative to those estimated for Option A in later years when risk would be at its maximum at the point of compliance. For a strictly risk-based comparison, these three options create some trade-offs between higher risks in the near term, and potentially lower risks in the long term, which is the time that is the focus of the performance objectives. These trade-offs can be assessed only by considering the full temporal and spatial profile of risks, because any comparison based merely on the long-term risk at the point of compliance would always suggest that an alternative to Option A would be “better” solely on a risk basis. Of course a more complete risk-benefit analysis would consider cost, worker exposures, and other important factors for decision making. There is another risk-based consideration that adds further richness to the comparison of the alternatives. At many of the tank sites, there is soil contamination already present around the tank. In some cases, the existing contamination implies very high potential doses now, if any individuals were to be exposed to it. Cleanup of these soils cannot proceed safely until the tanks in the vicinity have been stabi-

OCR for page 112
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report TABLE X-1 Assumptions for Options B and C in Figure X-2 Variable in Figure X-2 Full Interpretation of Variable Numerical Assumption for Example Variables Related to Oxalic Acid Wash (Option B) P Probability that the oxalic acid wash will cause some release of mobilized radionuclides due to existing crack in tank underneath sludge cake .01 Y Percentage of residual sludge now in tank that will be removed as a result of the oxalic acid washing 75% A Delay before tank will be closed due to oxalic acid wash 2 years Variables Related to Waiting for Emerging Technology (Option C) Q Probability that radionuclides will be released from ungrouted and unstabilized tank during 10-year wait for new technology .05 R Probability that the emerging technology in question fails to become a working method of cleaning within the 10-year wait .50 Z Percentage of residual sludge now in tank that will be removed if the new technology does become workable 90% B Delay before tank will be closed due to waiting for the new technology 10 years lized and closed. Thus, any delay in tank closure to limit projected risks in the long term can cause delays in the reduction of large and known risks in the present. This tradeoff has also been incorporated into the illustrative example that follows. Illustration of Impact of Alternative Tank Options on Risk Profiles To illustrate how the decision tree in Figure X-2 can actually be used to provide a risk-based evaluation of the three alternatives, specific numerical assumptions have to be introduced in addition to those that generated the initial risk profile shown in Figure X-1. Table X-1 summarizes the key assumptions needed to assess the risk profiles for Options B and C, given a starting point that is the risk profile under Option A (i.e., close the tank with no further reduction of the residual radionuclides presently in it). The numerical assumptions shown in Table X-1 are intended for a single specific tank, with its specific forms of residuals and specific new technology needs. As will be shown later, the numerical assumptions will vary from tank to tank, as will the Option A initial risk profile. These tank-specific factors affect the relative desirability of each of the alternative options. Table X-1 lists pessimistic assumptions about the efficacy of the oxalic acid wash, namely that the process would remove only about 75 percent of the remaining radioactive material. The probability that washing would create a leak is assumed to be fairly small, .01 (1 percent).6 Table X-1 also gives a fairly pessimistic assumption about the likelihood that the new technology will ever function, with only a 50-50 chance that it will become a viable option. It would, however, allow 90 percent heel removal. Figure X-3 shows the expected risk profiles for Options A, B, and C when using the tank example in Figure X-1 and the assumptions in Table X-1 about the alternative options. On the basis of a pure risk-risk comparison, it would appear that the oxalic acid wash (Option B) would provide the best overall outcome, even after considering that this would delay cleanup of the substantial existing soil contamination. Since the performance objectives are met at Location 3 (the designated point of compliance) even under Option A, a comparison of risks based solely on the ability to meet performance objectives does not demonstrate a convincing case to stakeholders for the extra residual waste reduction with oxalic acid washing unless the risk-benefits trade-offs required by ALARA (as low as reasonably achievable) are laid out explicitly. However, Figure X-3 shows how much overall benefit to the risk profile may be obtained from Option B. The performance objectives could be met at a wider range of locations and even are almost met right near the tank itself (i.e., at Location 1). This would not be true under the expected outcomes for Options A or C for this tank situation. Additionally, the two-year delay in reducing the current contamination affecting Location 1 has relatively little impact on the overall timing and level of risks. The near-term risks at Location 1 would have to be deemed a very significant hazard for actual individual exposures—and thus given very high weights relative to the longer-term, more hypothetical risks for the delay in cleaning up present site contamination—to affect the trade-offs in favor of waiting for an oxalic acid wash to occur.7 6 This assumes that the quantity of residuals and their constituent elements are such that there is no concern with criticality in this tank; otherwise the oxalic acid wash would not be an option for the tank. 7 Alternatively, one would have to believe that the two-year delay in initiating remediation of the existing soil contamination would create risks that contamination would never be cleaned up at all and would cause a plume of risk to appear at Locations 2 and/or 3 in the next several decades. Given that Options B and C imply relatively brief delays in cleanup relative to the life span of these DOE site cleanup programs, the committee considers this too unlikely a concern to be incorporated into the risk profiles of locations distant from the tank.

OCR for page 112
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report FIGURE X-3 Expected risk profiles for illustrative tank under Options A, B, and C. NOTE: P.O. = performance objective. This illustrative example of a decision has been based strictly on risk considerations, although the concept of risk has been extended beyond the single point-in-space-and-time analysis that is currently the modus operandi for tank waste determinations. It has also been extended to incorporate risks from contamination other than that associated solely with future releases from the heels left in the tank. Just these extensions have added insights to the case for considering alternatives that might further reduce the heels. It is important to remind ourselves that there are other important considerations in choosing among the options, including costs of alternative options, worker risks, and so forth. These would also be important to incorporate into the decision, and doing so could lead to a different choice than has been suggested might make sense on a purely risk-based comparison. However, it is the committee’s view that this comparison of options would be greatly enhanced for decision makers and the public if it starts from a rigorous and structured decision tree such as has been highlighted here. Different Decisions Will Be Warranted for Different Tanks The example in the preceding section has shown how a structured analysis of risks in time and space can be used to assess alternatives to the immediate closure of a tank that may contain some amount of residual heels. This was a single hypothetical tank situation, and there is nothing in the way of a conclusion from it that can be applied to all tanks generally. Site-specific and tank-specific conditions can substantially alter the comparison among options, even within the risk-only framework illustrated thus far. This section, demonstrates how tank-specific conditions can alter the appearance of the best option by extending the initial example to a wider range of tank situations. Table X-2 shows a set of assumptions for three different tanks. All are assumed to have the same general geological conditions (e.g., all exist in the same tank farm), so we do not alter the assumptions for these tanks are not altered regarding the timing and spatial patterns of risk once radionuclides are released from each tank. However, in the real world, these conditions also will vary, further adding to the ways in which decisions may differ from tank to tank. Tank X in Table X-2 is the same as the illustrative tank already used in the preceding section. Tank Y differs from Tank X in two ways. First, Tank Y has far less contamination of the surrounding soils, although it has a comparable degree of risk with the heel remaining in the tank. Second, the materials in the Tank Y heel are not expected to be as easily removed via oxalic acid washing, because some of those residuals are in zeolite form. Tank Z has very little residual but is in a location that has intense existing surrounding soil contamination. Also, Tank Z’s small remaining heel will not likely benefit from oxalic acid wash (because, for example, the materials may have never been in sludge form). Additionally, it seems unlikely any new mechanical or chemical cleaning method will be able to further reduce the small amounts that remain in the tank, simply because so little physical material remains to be removed. Of the three illustrative tank examples, Tank Z might be viewed as relatively more like the situation with Idaho National Laboratory’s liquid high-level waste or sodium-bearing waste tanks. Figure X-4 shows the relative risk profiles for all three options at each of the three hypothetical tanks. It can be seen that the best choice, based solely on comparisons of risk profiles, is likely to vary. Whereas oxalic acid wash (Option B) appears to be a preferred choice for the original example Tank X, waiting for the new technology becomes preferred for Tank Y, and immediate closure seems to be a reasonable choice for Tank Z. In the case of Tank Z, the long-term risks are very low to start with and hardly vary for Options B or C. At the same time, there is substantial near-term risk at Location 1, and the delay in reducing this near-term risk under Options B and C presents the only visible way in which risks differ from option to option.8 8 There are real differences in the risk profiles for Options A, B, and C at Tank Z, but they require that a smaller scale be used on the graph. However, all of the graphs in Figure X-4 purposely were drawn with the same vertical scales, to allow the differences among the three tanks to be more apparent.

OCR for page 112
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report TABLE X-2 Assumptions for Options B and C in Three Different Tanks Variable in Figure X-4 Full Interpretation of Variable Tank X Tank Y Tank Z Conditions of Tank Heels and Surrounding Soil Contamination       L Risk levels at Location 1 due to existing soil contamination (1.0 = performance objective risk level) 5 0.5 5 M Maximum risk levels that will occur at Location 1 if existing heels are released to environment (1.0 = performance objective risk level) 5 5 1 Variables Related to Oxalic Acid Wash (Option B)       P Probability that the oxalic acid wash will cause some release of mobilized radionuclides due to existing crack in tank underneath sludge cake .01 .01 .01 Y Percentage of residual sludge now in tank that will be removed as a result of the oxalic acid washing 75% 30% 10% a Delay before tank will be closed due to oxalic acid wash 2 years 2 years 2 years Variables Related to Waiting for Emerging Technology (Option C)       q Probability that radionuclides will be released from ungrouted and unstabilized tank during 10-year wait for new technology .05 .05 .05 r Probability that the emerging technology in question fails to become a working method of cleaning within the 10-year wait .50 .50 .90 Z Percentage of residual sludge now in tank that will be removed if the new technology does become workable 90% 90% 10% b Delay before tank will be closed due to waiting for the new technology 10 years 10 years 10 years FIGURE X-4 Risk profiles for Options A, B, and C for three hypothetical tanks facing different situations. NOTE: P.O. = performance objective.

OCR for page 112
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report These comparative risk figures also bring to mind some issues related to the need for reliance on long-term institutional controls and the ways in which tank management decisions can lessen that concern. The situation for Tank Z is one in which risks at many locations on the site from the long-term potential release of radioactivity in tank heels are small. Long-term institutional controls are therefore of lesser concern than strong action in the near term to reduce present risks. Tanks X and Y, on the other hand, impose substantial risks throughout the site over the very long term, even though performance objectives are “met.” Without additional tank cleanout, there may be unacceptable levels of risk at some locations unless institutional controls are assumed to be effective for hundreds of years. In the case of Tank Y, near-term risks are small relative to the long-term risk from the tank heels, so it may make sense to delay tank closure long enough to obtain a reasonably large reduction in the area that would have long-term contamination above levels deemed acceptable at assumed points of compliance. Tank X faces a more complex trade-off between near-term needs for remediation and long-term risks that could be quite high in the absence of institutional controls. In this example, however, the slight delay to perform an oxalic wash (combined with the fact that the wash could be very effective for this particular tank) may be worth the substantial reduction in need for long-term institutional control to ensure that future exposures do not exceed limits. Again, the additional decision-relevant considerations of cost, worker exposure, et cetera, might further modify these choices and should not be ignored in a full analysis for actual tank decisions. Additionally, decisions to weight near-term or long-term risks more heavily, or to weight the risks at one location more heavily than another, could affect the choices. Nevertheless, the point remains that different choices may be reasonable for different tanks and that the decision to close a tank immediately may not be the best option even when a tank is projected to be able to meet its performance objectives. (Note that all three of these hypothetical tanks would meet the performance objectives with Option A—immediate closure—if the point of compliance is Location 3.) Table X-3 provides an example of an approach that DOE could adopt to decide when to stop waste retrieval in a specific tank. The table requires that DOE quantify the various inputs to the risk-benefit analysis described in the table, such as public risk, worker risk, costs, compliance with Federal Facility Agreements, effect on nuclear power receptivity, and how these risks vary with time (e.g., in 30, 300, and 1,000 years). The table takes into account whether additional time has been spent to achieve additional waste reduction (second column). The third column entries would be the assumed rate of monetary inflation, and the fourth column represents the projected change in level of public trust in DOE. The fifth column lists times at which the inputs are measured. Each parameter is attributed a weighing factor according to DOE’s TABLE X-3 Determining How Much Waste to Leave in a Tank

OCR for page 112
Tank Waste Retrieval, Processing, and On-Site Disposal at Three Department of Energy Sites: Final Report priorities. In the next-to-last column there is a number that captures all of these risks and benefits for each amount of waste proposed to be left in the tanks. The table also takes into account the potential presence of other site risks with radioactivity already committed to the ground. In this scenario, if the residual left in the tanks is one or more orders of magnitude smaller than the radioactivity already committed to the site, pursuing additional waste retrieval may not be a good use of resources and workers. Summary of the Illustrative Example In summary, the radioactive materials that pose the risks in tank wastes have widely varying levels of activity, half-lives, and physical properties. This situation results in risks that are extremely long term and highly variable over both space and time. Because risks can change over space and time, a risk-informed approach would help in considering the long-term consequences and risks of tank waste management and disposition decisions. Each of the three sites should undertake a separate examination of how the tank risks vary over space and time (see Chapter VIII, Recommendation VIII-1). An example discussed in this chapter shows how A structured approach to decision making about tank closure helps clarify the trade-offs, making it a more transparent process; Each decision has implications for an entire risk profile and adds to the information provided by the performance assessment; The interaction between tank closure decisions and the need for long-term institutional controls can be better informed by a structured assessment of risk profiles; and “Maximum extent practical” could be communicated more transparently to regulators and stakeholders in the context of changes in risk profiles than in the context of the limit of a waste retrieval technology. As explained in Chapter VIII, the structured assessment of risk profiles is still only a part of the information needed for risk-informed decision making. Costs and worker risks are not found in the illustrative example but a real-world analysis would include these considerations as well. The structuring activity helps identify the additional concerns that are relevant for a full risk-informed approach and can promote the kind of discussion about choices that is a hallmark of a participatory decision process. Different end points for waste removal and final tank closure can emerge for different types of tank, sites, and waste types.