An End State Methodology for Identifying Technology Needs for Environmental Management, with an Example from the Hanford Site Tanks (1999)

A conceptual approach to deriving a technology development program is based on end state criteria. This chapter discusses end states for wastes, the justification for using end states as a basis for defining a technology development program, and a conceptual approach to defining a technology development program based on end state considerations.

The establishment of fully defined program objectives and of clear program priorities for technology development investments have become increasingly important to the Office of Environmental Management (EM) for meeting schedules, cost constraints, and other requirements. A systematic planning process should provide the framework for identification of technology development needs. The use of a systematic process is imperative for efficient and effective completion of EM's mission to manage the environmental problems at its sites. The end state approach recommended by this committee is such a systematic and disciplined process.

The committee notes that the end state based approach is similar in principal to the widely used systems engineering process to define technology development needs.¹ Top-level requirements defined as a part of the systems engineering process include end state specifications, and the flowdown therefrom provides specific requirements for each process step. The functional flowsheets defined as a part of the end state approach are the same as the architectures that are part of the systems engineering process. Both approaches call for explicit consideration of alternatives. The primary distinction between these two approaches is that the portions of the systems engineering process related to definition of a technology development program have been elaborated in the end state based approach to fulfill the purposes of this report.

In an earlier report, the National Research Council (1996a) concluded that end state specification of the products resulting from remediation activities are an appropriate and necessary basis for planning and conducting a waste-related technology development program. In this context, an end state can be expressed as the desired composition, configuration, performance, and location of a particular waste product at the completion of remediation activities, frequently wastes emplaced in a disposal facility. If the phased-decision approach

previously recommended by the National Research Council (1996a) were to be used, the end state may be that associated with the end of one of the phases.

The process of using end state specifications to derive technology development requirements is shown in Figure 3. The committee proposes a seven-step approach for identifying technology development needs for remediation of the waste in tanks and the tank sites:

In the final analysis, the approach for identifying the technology development needs noted in this study was driven by end states that are believed to be reasonable in terms of regulatory, budgetary, and technical acceptability. Critical to managing the end state based approach is the level of detail associated with the definition of scenarios and functional flowsheets. Consideration of functional operations rather than specific processes is required. The framework of scenarios in which technologies are discussed in this report is defined at a relatively high level to avoid having the study consumed in the details of flowsheet process engineering and chemistry.

The development of the end state approach begins by characterizing the initial state of the subject waste to provide the data for subsequent processes and evaluations. The initial state represents the existing condition of the waste being managed and is often thought to be only the compositional and physical characteristics of the stored waste. In systems as complex as those at Department of Energy (DOE) sites, many other aspects, such as the contaminated environment surrounding the tanks, must be included in the definition of the initial state. These are discussed in a following section on characterization of the initial state.

Based on knowledge of the initial state, a few plausible approaches are developed that will result in the transition of the waste from its initial state to an end state. These approaches, called scenarios, are qualitative descriptions of the transition path of waste from its initial state to an end state. Examples of scenario descriptions are 'exhume all waste to yield a site suitable for unrestricted use' or 'stabilize waste in place with long-term institutional controls.' Scenario development would focus on a reference scenario that is based on the best current judgment of the most desired processes for remediation. In general, scenario definition should also include a few plausible alternative scenarios that incorporate reasonable changes in circumstances.

Then, end state specifications are developed for each scenario. Selection of the scenario and its associated end state(s) is based on such considerations as risk reduction, cost minimization, environmental regulations, and stakeholder values. A complete specification of an end state will include not only compositional and physical requirements that meet product

acceptance or other criteria but also the location, legal, societal, and institutional requirements for the products. The list of products for which end state specifications are to be developed should be shortened by eliminating those for which technology development is unlikely to be needed.

Given the initial state and end state specifications, a relatively simple functional flowsheet is then developed for each scenario to achieve its associated end states. These flowsheets identify the steps needed to accomplish a waste treatment scenario, expressed in terms of the function to be performed instead of the specific processes to be employed. These steps define the actions and the intermediate products required to translate the initial state of the waste into one that meets its end state specifications. When the list of functions and associated requirements is defined for the selected range of plausible scenarios, the duplicate functions are consolidated. This yields a unique set of functions and their associated performance requirements that provides the basis for deriving the technology development requirements.

The performance requirements inferred from the resulting end state specifications are then allocated to each function as functional requirements, which define how well the function must perform. Then, for each function a technology assessment is performed to compare technology requirements for each function to the state of existing technology capabilities. If existing technology or science is inadequate, a gap is identified that should be pursued as part of a technology development program.

The following section discusses the justification for using an end state based approach. Each step is discussed in more detail from the vantage point of the technology developer.

The primary benefits of an end state based approach as a basis for a technology development program are (1) the technology development needs are specifically tied to a plausible set of end states for the initial wastes by an explicit decision logic, (2) the technology development program is designed to support multiple plausible sets of end states until a final decision on the preferred end state or end states is made, and (3) the technology development program that uses this process properly has the integrity to withstand scrutiny from the research community at large, Congress, stakeholders, and various DOE review committees.

The explicit connection of the technology development program to the desired end states of the initial wastes is intended to impose discipline and efficiency on the technology development program. The need for each technology development project may be derived from a specification of the end state to be achieved and a technology assessment to determine whether additional development is required. A proposed technology development project that cannot lead to achieving a plausible end state should not be funded unless it addresses other technical needs related to implementation of new technology. If technology to achieve the end state already exists, then justification of additional technology development would require that such development lead to increased benefits, such as reduction of implementation cost or risk, that would compensate for the projected cost of the development. In the absence of an end state based approach, there is the risk that some research projects would address inconsequential needs.

Because technology development typically requires years to produce deployable results, whereas knowledge concerning remediation problems and decisions on how to best manage them are changing much more frequently, a technology development program must proceed in the face of considerable uncertainty. As technology development nears completion, its results, when combined with an analysis of relevant externally imposed constraints, will provide decision

makers with reliable information to make informed decisions on technology implementation. If technology development supporting only a reference scenario is pursued, changes in externally imposed constraints (such as resource limitations or changes in allowable risks) may inhibit or prevent implementation of the reference approach due to inadequate technology. Without pursuing technology development for the alternative scenarios, the information needed to select the best course of action will not be available and, if the reference approach is even partially deficient, there will be costly delays stemming from the time required to develop additional technology.

The issue of how to allocate technology development resources among reference and alternative scenarios is a policy decision that should be explicitly addressed by DOE. Investing a significant fraction of technology development resources in functions supporting alternative scenarios and their associated end states is a useful form of technology portfolio management.

Not all remediation problems require technology development or consideration of alternative scenarios. If remediation can be completed in the near term with an acceptable² demonstrated technology, then technology development is not required. If remediation is to be completed in the near term, but technology is inadequate, it is likely that only a reference scenario and its associated set of end states need to be addressed. However, the remediation problems in which EM is investing most of its technology development resources involve complex, long-term projects [e.g., high-level waste (HLW) tank remediation, subsurface contamination, facility decontamination and decommissioning] where changes in such external factors as budget, regulations, and stakeholder values are likely to occur. In these cases, an end state based approach that includes reference and alternative scenarios should be used.

The consideration of reference and alternative scenarios and their associated end states as described here is not intended to address the issue of whether redundant technology development should be supported to meet the end state specifications of a specific scenario. That is, the scope of alternatives does not address whether two or more different technology development projects should be pursued to meet a specific functional requirement. Such redundancy is justifiable when the need is critical or the probability for success of a single technology is judged to be low.

Another benefit of using the end state based approach to define an appropriate technology development program is its clarity (i.e., it can be readily understood). When properly documented, there is a clear path from the problem to the solution through specification of the initial problem, definition of a reference scenario and alternatives to accommodate uncertainty, identification of functional approaches to move from the initial problem to the solution, assessment of the adequacy of existing technology, and support for technology development only in those areas where technology is inadequate. The existence of this traceable path provides clear linkage of the proposed technology development projects to the ultimate desired end state of the waste, which, after appropriate independent reviews, should provide adequate justification to decision makers to support the technology development program. Achieving this linkage requires documentation of the various steps taken to implement the end state based approach. Detailed documentation should be provided to those directly involved in the process (i.e., problem owners, technology providers, reviewers). The committee notes that this documentation tends to be voluminous and frequently incomprehensible to decision makers, who need summary formats that focus on the relationship of technology development projects to bridging the gap between the initial and end states.

The first step in defining a technology development program based on end state considerations is to characterize the initial state of the waste (e.g., contaminated liquid, solid wastes, buried wastes) to the extent necessary to evaluate process performance needs and compliance with regulations. Common waste characterization parameters are location, age, volume, and chemical and physical characteristics, including their variability and uncertainty. Characterization parameters that are less commonly considered should include the regulatory status of the target waste, the nature of nearby contaminated sites, and proposed future land uses. Also important are compliance agreements and applicable regulations.

Typically, the technology development organization is not in a position to characterize the waste or associated waste management approaches. If the initial state cannot be satisfactorily defined, a definitive and timely plan for characterizing the initial state should be established and implemented by the problem owner. In the interim, the technology development organization must rely on provisional information, including that which it develops internally. In many cases this information is likely to be uncertain or incomplete, and defining an end state based technology development program may require the technology development organization to use speculative calculations or assumptions subject to future validation. In all cases the data, analyses, and assumptions used as a basis for defining technology development requirements should be clearly documented at the outset.

Using the Hanford tanks as an example, characterization of the initial state could include the amount, chemical and physical properties and radioactive content of waste in the tanks, condition of the tanks, characteristics of contamination in the immediately surrounding soil and ground water, classification of the waste (i.e., mostly HLW with constituents regulated under the Resource, Conservation, and Recovery Act of 1976, as amended, or RCRA), features of existing compliance agreements, and changes that may take place during subsequent processing steps. Uncertainties in the preceding should be explicitly addressed, but only the characterization information necessary to support initial decision making, regulatory compliance, safety and environmental risk assessments, and the design of functional flowsheets should be obtained.

A scenario defines the qualitative transition path of waste from an initial state to a specified end state. For example, one Hanford tank scenario might be cleanup of the tank farms to levels acceptable for unrestricted future use (i.e., a 'greenfield'), and retrieved tank contents processed to yield HLW sent to off-site disposal and low-activity waste (LAW) for on-site disposal. Another Hanford tank scenario might be stabilization of tanks and their contents to the point that the site is acceptable only for certain types of restricted industrial use with continuing institutional control (i.e., a 'brownfield').

Definition of plausible scenarios for products should be based on consideration of the specific situation, including any legal precedents, applicable regulations and compliance agreements, stakeholder values, and foreseeable budgetary constraints, as well as the technical state of the art. Any of these factors may ultimately constrain the viability of the scenarios. For example, limitations on any future budget may render certain scenarios essentially unachievable. Similarly, stakeholder concerns may favor a particular scenario as being the preferred remediation path. While scenarios should be plausible, it is important to recognize that the range

of alternative scenarios need not be limited solely to those that are now considered to be optimal or that meet current compliance agreements and regulations. Instead, the alternatives should represent outcomes that have a reasonable chance of being considered for accomplishment in the future, given the potential for changes in both technology and public values between the present time and the time when final remediation decisions will be made.

Admittedly, determining which scenarios have a reasonable chance of being considered for deployment is a subjective decision, since it is essentially a prediction of a highly uncertain future state of affairs in an institutionally driven process. It is nonetheless possible to identify outcomes with differing degrees of plausibility. For example, while a 'greenfield' solution (i.e., removal of all wastes and tanks, and decontamination of all soil) may be financially and technically impractical at many sites, it is relatively easy to imagine situations in which stakeholders and regulators may eventually require a higher degree of cleanup than specified by current regulations and agreements. Similarly, while it may be societally unacceptable simply to walk away from unremediated tanks, it is plausible that situations will occur in which taxpayer concerns and Congressional action result in cleanup budgets that are significantly lower than current projections. This could lead to the need to remediate lower-risk tanks in place. Since final disposal sites have not yet been established for many product wastes, factors such as acceptable waste forms, the volume of wastes that can be accommodated, and the cost of disposal could change from current plans, again leading to possible alternative scenarios. Finally, it is likely that a case-by-case combination of remediation scenarios may be beneficial in many cases. The end state based approach provides the basis for this option.

The committee notes that the number of scenarios carried forward into subsequent steps must be limited to remain tractable. Although each case will vary, a minimum preferred outcome is for one scenario and associated end state set to represent the current baseline approach (the approach currently preferred and planned by the problem owner); a second scenario to represent a plausible scenario based on a highly risk-averse, resource-available environment; and a third scenario to represent a risk-tolerant, resource-constrained environment.

Other issues arise from privatization of cleanup activities. At present, privatization is the planned approach for some cleanup activities, which would suggest that the private sector (rather than DOE) is responsible for any needed technology development in support of these activities. As noted in ''Harnessing the Market: The Opportunities and Challenges of Privatization" (U.S. Department of Energy, 1997a), several factors may cause privatization to be less than fully successful. These include DOE's lack of experience in overseeing privatized activities and the absence of firm end state specifications, as well as some contractors' lack of experience with the kinds of complex, multi-regulator environments that characterize many DOE sites. It may be reasonable in some situations to postulate failure or limited use of privatization as one scenario to be evaluated, resulting in a need for DOE to develop technologies as a backup in case suitable technologies are not forthcoming from the private sector.

The breadth of scenarios beyond the current baseline that need to be considered may depend in part on the level of the organization performing the evaluation. In particular, individual sites are subject to strong and immediate operational and stakeholder pressures, and hence site managers may not be able to do much more than plan for their current baseline activities. By contrast, central organizations, such as those in DOE headquarters, are at least to some degree insulated from these pressures and potentially able to take a longer and broader view. Thus, a function of a central environmental management organization may be to specify a broader range of potential future scenarios and ensure that technologies are developed to meet the needs of those scenarios that are considered plausible, as well as avoiding duplication of technology development efforts.

The committee recognizes that stakeholder concerns and compliance agreements are important factors that could limit DOE's ability to implement an end state based approach to defining an appropriate technology development program. Some stakeholders do apparently recognize that readjustments to the baseline may become necessary if a particular approach proves to be infeasible, for whatever reason (technical, programmatic, or political). At present, many public stakeholders apparently desire that DOE follow the existing baseline approach. They view the investment of significant resources in technology development for alternative scenarios as a diversion from that effort. This is unfortunate in the committee's view, because more explicit consideration of alternatives as proposed herein could make the overall DOE program more robust with respect to unexpected developments, as well as provide a more transparent rationale for adoption of a particular approach from among the candidate approaches.

One set of end state specifications is developed for each scenario. The end state specification set defines an objective of remediation for a scenario and should be quantified to the extent possible. To determine end state specifications for use in establishing an appropriate technology development program, the specifications can be grouped into those related to human health and ecological risk (e.g., safety, hazard, land and resource use) and those related to cost (e.g., process reliability, performance, efficiency). These groups are not always separable, as in the case where the cost of remediation must be reduced to affordable levels to allow it to occur while still reducing risk to acceptable limits; moreover, any of the end state specifications may be dictated by technical, institutional, or societal factors.

Risk-related end state specifications are normally fixed limits that must be shown to be met, with due consideration for uncertainties, for a remediation activity to be acceptable under regulations, but where further risk reductions are not required except in the case of some occupational or public health and safety situations (e.g., as low as reasonably achievable, or ALARA, requirements). Risk-related specifications are translated into usable end state specifications for products using a modified risk assessment. Normally, the risk assessment process begins with a specific initial state for a hazardous material, postulates mechanisms by which the material can be released, and then estimates the likelihood of releases and the resulting doses to humans. To define the scenario and the end state set of specifications, the allowable impact is specified (the risk specification) and this, in combination with postulated release mechanisms, is used to determine the acceptable pre-release state of the hazardous material, which is the end state specification of the product. The end state specification frequently consists of information on the allowable inventory or concentration of hazardous species in a product, the allowable release rate of the species under various conditions, and the packaging of the product.

Conceptually, translating a risk-related end state specification into a useful end state specification for waste products from a scenario is straightforward. In reality, there are numerous factors that can yield different results or results with large uncertainties. Some of the most common factors are use of different measures of risk (e.g., different affected populations, different locations, population vs. individual impacts), inadequate knowledge of ground water movement, and imperfect knowledge of the interaction of hazardous species with water, soil, and air. To achieve the best results, all of these need to be consistently and accurately specified in an appropriate way (National Research Council, 1983, 1994a, 1996a, and 1996b) in the context of the problem. This should include explicit characterization of uncertainties that are expressed quantitatively to the maximum extent possible.

In contrast to regulated risk, cost-related specifications have no fixed limits, and there is presumably always the desire to invest in technology development to reduce the cost of operations more than the amount of the investment. In many cases risk limits can be met and the primary driving force for technology development is to reduce the cost of achieving the same risk-related end state specification. Using a process similar to a risk assessment, the costs of producing and dispositioning the products are estimated, the largest cost contributors identified, and the most important factors that might be altered to reduce the cost are determined.

To continue the Hanford tank example, end state specifications could contain a limit on dose to an individual from LAW in an on-site, near-surface disposal facility, and a requirement to reduce the number of very costly HLW glass logs. This, when coupled with a modified risk assessment, could result in an end state specification to lower the maximum allowable release rates of important radionuclides in the LAW (e.g., technetium-99, cesium-137). A cost analysis could result in an analysis to reduce the volume of the HLW.

As in the case of specifying the initial state of the waste, if the end states of the products are not completely specified, a plan leading to the timely resolution of the open items should be prepared and executed. The clearly preferred option is for end state specifications to come from the DOE site problem owners. If these are not forthcoming, the next preferred option is for them to be provided by relevant EM headquarters' remediation programs. Finally, if they are not forthcoming from either of these sources, DOE technology development organizations should develop and clearly state enabling assumptions. The committee expects that it will frequently be the case that end states and related technology development requirements will require provisional specifications from technology development organizations in the face of major uncertainties in costs, benefits, public acceptability, and many other factors. Regardless of who develops the scenarios and end state specifications, the uncertainties are likely to be so large that only a modest analytical effort, coupled with a substantial amount of judgment, will be useful—that is, it will be necessary to avoid the trap of over-analysis in the absence of reliable data.

Given a set of end state specifications for the products from each scenario, the next step is to create for each set a conceptual flowsheet composed of functions that will take the waste from its initial state into products that meet the end state specifications. This is called a functional flowsheet. It is functional because it does not state the means (i.e., specific process) by which the goal is accomplished. Instead, this flowsheet states only the functions that must be performed. For example, one function related to the previously mentioned brownfield scenario could be to stabilize buried waste. The flowsheet is stated in this form so that any technology that might perform the required function (e.g., in situ vitrification, grouting) is not precluded from consideration during technology evaluation. The functional flowsheets developed for each scenario could include activities that meet the general requirements of the scenario's end state specifications for each product.

The committee notes that, in many cases, knowledge of available and projected technology can define or limit the functions that can be considered. As a consequence, the end state based approach in general, and development of functional flowsheets in particular, often requires iteration to achieve closure.

In many cases different scenarios and associated functional flowsheets will have a number of functions in common. These must be consolidated into a unique list of functions, listing each function only once to avoid possible subsequent duplication. This needs to be performed with recognition that functions having the same general description may have different functional requirements. For example, flowsheets for two scenarios could both contain the function tank stabilization. However, in one flowsheet the stabilization function could need to meet more stringent end state specifications than in another because waste may not be postulated to be retrieved in the first. Thus, the second flowsheet could need to have only the function tank stabilization, while the first flowsheet could be modified to have this same function followed by an enhanced tank stabilization function to elicit better stabilization technologies.

The consolidated functions in each flowsheet must be accompanied by the performance requirements (called functional requirements) that are derived from and will achieve the specifications of the associated end state set. For example, the function exhume buried waste would be accompanied by specification of the maximum allowable residual contamination levels. This, in effect, defines what is waste to be exhumed and what is the residual and soil contamination that can remain.

The process of allocating functional requirements to meet end state specifications requires some attention. An end state specification on the allowable release rate of radionuclides from LAW could result in the requirement that the radionuclide removal function remove a certain fraction of each radionuclide from the LAW stream. Alternatively, it could be equally possible to achieve the same release rate with an improved waste form or a combination of removal and waste form, thereby meeting the specification. While noting that the allocation of the end state specifications to yield functional requirements could be necessary in many cases, the committee suggests that the requirements always be accompanied by the ultimate end state specification, with some explanation of their relationship.

Continuing the Hanford tank example, two of the consolidated functions could be removal of radiocesium from the supernatant liquid and solidification of the supernatant liquid to yield immobilized LAW. The end state specification for the LAW is postulated to limit the release rate of radiocesium from the immobilized LAW. The requirement for meeting this specification could be allocated entirely to the cesium removal function as a very high removal fraction, to the solidification function in the form of a very low release rate limit for cesium, or intermediate combinations of these.

Given the consolidated list of functions and associated requirements for the scenarios, the next step in the end state based methodology is to ascertain whether acceptable technology already exists to perform each of the functions identified above and meet the functional requirements. If so, no further technology development is required. If not, the function becomes a candidate for technology development. The process of determining the adequacy of available

technology to perform a specific function is called technology assessment, and should be performed and explicitly documented for each function.

A typical approach to technology assessment would be to consult with experts in a particular field to survey similar activities and experience. For example, in the brownfield scenario, experts in site stabilization would be consulted and a survey of burial site stabilization projects would be conducted. The objective is to provide a database to assess the applicability of the technologies to each function. Inquiries concerning the technology status must be undertaken very carefully, because situation-specific differences can render useless a technology proven to work in an ostensibly similar situation.

The technology status related to a particular function, when compared to the requirements of that function, provides the basis to decide whether technology development is required through answers to a few key questions:

When demonstrated technology is not available, several developing technologies that might eventually be adapted to perform the required function will exist. In this case, it is often necessary to select among the candidate developing technologies because limited technology development resources are available. This decision is typically based on subjective evaluation of such considerations as projected costs, the potential for successful completion of technology development, and the likely performance of the process. Approaches for selecting, prioritizing, and decision making for technology development projects to meet EM needs are currently the subject of a National Research Council study in support of DOE's Office of Science and Technology (OST) program.³

Completion of the entire process described above results in a portfolio of technology development activities that, if successfully completed, would allow the scenarios, or combinations thereof, to be implemented and their associated end state specifications to be achieved. That is, upon completion of the technology development, decision makers would have the information necessary to confidently select the best scenario or scenarios to achieve the preferred end state(s), and the information necessary to proceed to implementation would be available. The key elements of this process are (1) the functional requirements that the new technology must meet, because the requirements were derived from consideration of alternative end states of the waste that specify this performance, and (2) the selection of the best candidate technology for meeting these requirements based on formal technology assessment and selection process.

There is the need for active feedback in the end state based approach. As technology development proceeds, the potential for successfully completing development and the performance of the new technology will become increasingly clear. As it does, the desirability of continuing technology development should be periodically reviewed, probably on an annual basis. If new information indicates that a technology development project will have higher costs or lower performance than previously estimated, it should again be subjected to the decision making sequence described earlier in this chapter to determine whether it should continue.

The establishment of fully defined program objectives and the setting of clear program priorities for technology development investments have become increasingly important to meeting schedules, cost constraints, and other requirements. A systematic technical planning process is needed to provide the framework for identification of technology development needs. The end state based approach is such a systematic and disciplined process.

It is possible to specify a generic approach to determine waste-related technology development needs based primarily on consideration of the end state to be achieved for the waste. This approach consists of (1) characterizing the initial state or condition of the wastes or waste site to be remediated, (2) identifying reference and alternative scenarios to accomplish a general remediation objective, (3) specifying the product forms and requirements of the desired end states of such scenarios, (4) defining the functional flowsheets required to transform the initial waste or waste site into the desired end states, (5) combining essentially identical functions in the flowsheets into a unique set of functions, (6) allocation of end state specifications to each processing function as functional requirements, and (7) comparatively assessing the respective development or deployment status of the technology required for each function to yield technology development needs. The end state based approach incorporates key elements of the widely used systems engineering process. A difference between the traditional systems engineering approach and the end state based approach is that the portions of the systems engineering process related to definition of a technology development program, such as the use of alternative scenarios, are specifically focused for explication in the end state approach.

Technology development typically requires years to produce deployable results, whereas regulations and stakeholder values concerning remediation problems and decisions on how to best manage them change more frequently. This uncertainty requires that an appropriate technology development program be based on a range of plausible scenarios. The committee

believes that the end state based approach should encompass (1) a reference scenario that is essentially the current site baseline or most likely approach, plus (2) at least one plausible alternative scenario to efficiently accommodate uncertainty. An end state based approach that considers a range of plausible scenarios will identify technology gaps associated with the site baseline flow sheet and technology development needs associated with credible alternative functional flowsheets. Furthermore, it will identify uncertainties and potential improvements to the existing site baseline process technology that often lead to additional development needs. Allocation of technology development resources among a reference approach and alternatives is a policy decision requiring explicit attention by EM, with the aid of cost and risk assessment results.

The committee notes that not all remediation problems require technology development or consideration of alternative sets of end states. In particular, short-term problems with well-defined solutions should not require the end state based approach. Consideration of reference and alternative end states does not address the issue of whether multiple technology development projects should be supported to meet a specific processing functional requirement. That is, the scope of alternatives does not address whether two or more different technology development projects should be pursued to meet a specific technology development need. Development of multiple processes is justifiable when the need is critical to implementing the scenario and the probability of success of a single technology development project is judged to be low.

End state specifications can be grouped into those related to human health risk (e.g., safety, hazard) and those related to cost (e.g., process reliability, performance, efficiency) for use in establishing an appropriate technology development program. These specifications must provide sufficient information to allow technology development to achieve processes that will produce acceptable waste products.

Performance assessment (whether described as risk assessment, decision analysis, or scenario analysis), cost assessment, and trade-off studies are necessary to translate risk and cost specifications into usable end state specifications. However, the risk- and cost-related information necessary to establish end state specifications is often not available. For this reason, plans leading to the timely resolution of the open items must be prepared and executed, and in the interim enabling assumptions may be developed by problem owners or, if necessary, technology providers. It will always be the case that end states and related functional requirements will have to be identified in the face of uncertainties in costs, benefits, public acceptability, and many other relevant factors. The uncertainties are likely to be so large that only a modest analytical effort, coupled with substantial expert judgment, will be useful to avoid over-analysis in the absence of reliable data.

Tank waste remediation activities yield primary and secondary products resulting from waste processing and site remediation. It is necessary to identify both of these groups of products as a basis for developing end state specifications. In the case of Hanford tank remediation, products might include HLW, LAW, the tanks and tank farms themselves, secondary solid wastes (radioactive, mixed, or chemically hazardous), processed gaseous and liquid effluents, sanitary wastes, uncontaminated construction wastes, materials suitable for recycling, and the byproducts of facility decontamination and decommissioning. For many of the products the end state specifications and technology for achieving them are well-known and commercially available, and their explicit inclusion in the functional flowsheet would serve no useful purpose

related to technology development. Thus, in the case of technology development for the current baseline approach to Hanford tanks remediation, the list of products might be reduced to HLW, LAW, and the tanks and tank farms.

The primary benefits of the end state based approach to determining a technology development program are that (1) requirements are explicitly tied to a plausible set of end states for the wastes, (2) the program is designed to efficiently support multiple sets of plausible end states for the wastes until a final decision on the preferred end state is made, and (3) the program evolved using this process has the integrity to withstand external scrutiny. This results from the presence of a clear path from the problem to the solution through specifying the initial problem, defining a reference goal and alternatives to accommodate uncertainty, identifying functional approaches to transition from the initial problem to the solution, assessing the adequacy of existing technology, and supporting technology development only in those areas where technology is inadequate.