The NRC Committee on the WIPP has followed the WIPP project from its infancy in 1978. There have been some changes in committee membership, but there has been sufficient continuity that the committee has remained aware of the significant scientific and technological issues, as well as the associated policy and budgetary constraints on the WIPP project. This report focuses on the committee's understanding of the specific scientific and technological issues that form the core of DOE's compliance application to the U.S. Environmental Protection Agency and their influence on the perceived ability of WIPP to isolate TRU waste from the biosphere.
During its deliberations, the committee also has arrived at several general conclusions with respect to the overall integrity of WIPP, the difficult problem of human intrusion, and the role of Performance Assessment. Because of their importance to this report and the overall integrity of WIPP, the committee provides a general discussion of these issues in this chapter.
Integrity Of The Waste Isolation System
A previous National Research Council report provides a useful introduction to the topic of geologic isolation at WIPP. The report states that the objective of a geologic waste isolation system is
to protect humans now and in the future by isolating the waste from the environment effectively enough and for a period of time long enough that the amount of radioactive material ever reaching the biosphere will present no unacceptable hazard. To achieve this objective, a hierarchy of mechanisms exists to reduce the release of radionuclides to the biosphere and thus have the waste isolation system meet the criterion of overall performance. One or more of the following release control mechanisms must be sufficient to meet the panel's criterion:
- delay of ingress of water
- slow dissolution of radionuclides
- slow release from the waste package
- long ground-water travel time
- delay due to sorption in the geologic medium
The system must guard against an unacceptable release of the radioactive material into ground-water and the transport of this contaminated water to the biosphere—a principal pathway by which some portion of the buried radioactive material may eventually reach humans. (NRC, 1983, pp. 3-4)
Although the 1983 report gives primary emphasis to geologic isolation for high-level, spent reactor fuel or reprocessed waste (hence the high ranking given above to ''slow release from the waste package"), the list of important criteria is essentially the same for a transuranic (TRU) waste repository. The total radioactivity per unit volume of TRU waste planned for WIPP is roughly on the order of one-thousandth of the activity for an equal volume of spent fuel waste, but many of the same long-lived radioactive isotopes are present, and so geologic isolation is necessary.
In the committee's opinion, location of the WIPP repository within the 600-m-thick Salado Formation, some 400 m below the overlying, permeable water-saturated sediments is an exceptionally robust way in which to delay ingress of water to the repository or, equivalently, egress of radionuclides from the repository to the accessible biosphere. Natural salt
creep will result in the encapsulation of waste within what becomes essentially intact salt a few hundred years at most after emplacement. Sealing of the shafts with crushed and compacted salt also should ensure that, by natural salt creep processes, these potential conduits to the biosphere are sealed effectively. Thus, the waste should be sealed, permanently, within the Salado Formation, which has been stable for over 200 million years.
The committee is confident in asserting that, provided the repository is sealed effectively and remains undisturbed by human activity after final closure, WIPP will isolate TRU waste for long enough (i.e., many tens of thousands of years) to ensure that the radioactivity in the waste will have decayed to very low levels.
From the point of view of waste isolation, salt does have one drawback—it is frequently found in association with mineral resources, especially oil and gas. Drilling for these resources could result in one or more holes penetrating the repository and would provide a pathway to bring the waste into direct contact with ground-water and the accessible biosphere.
An intrusive drilling event could remove the isolation from ground-water provided by the 200-m-thick envelope of salt. Because dispersion and dilution are not significant factors at WIPP, the two remaining mechanisms to control radionuclide releases noted in the NRC (1983) report would be (1) slow dissolution of radionuclides, and (2) delay due to sorption in the geologic medium. Although WIPP does not include a "waste package," the backfill and seals around each waste room and panel serve a similar function. In fact, the salt "package" does not have the limited lifetime of a high-level waste canister. For WIPP, assurance of protection from excessive release of radionuclides due to drilling intrusion relies on (1) engineering efforts to improve repository performance (see Chapter 4); (2) improved understanding of the actinide source term (see Chapter 5); and (3) understanding of non-Salado Hydrology (see Chapter 6)—the same three mechanisms that are identified in the NRC (1983) report.
Role Of Performance Assessment
What contribution can be expected from each of these component factors, and how much is needed, to ensure effective long-term isolation of the waste?
Total System Performance Assessment, or PA, is the method that has been used at WIPP to identify and examine all credible pathways by which radionuclides could be released to the biosphere from the repository and to estimate the contribution of each component of the pathway. Although both the undisturbed and disturbed (by drilling) cases have been examined, the DOE PA team has placed appropriate emphasis on the more complex and critical disturbed case.
The latest "complete" performance assessment available to the committee was published in 1992. The 1992 PA identified the same three key factors noted above, and its results suggest that WIPP could comply with the EPA release standards if certain levels of performance were attained from each component.
- The assumed permeabilities governing the flow of brine through the underground excavations were much too high. As a result, predicted flow through the repository due to an E1E2 intrusion (see Chapter 2, Figure 2.5) greatly overestimated brine flow, and hence radionuclide release.
- In the 1992 PA, no credit was given for the ability of the excavations, especially when back-filled with crushed salt, to close within the order of a few hundred years, or for the potential of room and/or panel engineered seals and various other available engineering techniques to reduce both the consequences and the probability of serious drilling intrusions. For example, if the values for repository permeability used in the assessment are reduced progressively towards the value of intact salt (10-23 m2) over 10,000 years, rather than maintained at a constant 10-15 m2, the total predicted brine flow and radionuclide release would be much lower than indicated in the 1992 PA.
- There was insufficient experimental justification for the assumptions made with respect to actinide
- solubility, colloidal transport, and ground-water flow/radionuclide transport in the Rustler Formation overlying the Salado Formation.
- Experimental studies to improve the basis for estimating actinide solubility and Rustler Formation hydrology and transport have now started, but these will require several years of additional work in order to arrive at reliable results. At this stage—with the results of the 1996 PA not yet available, and important experimental studies still in progress—the committee does not know how the results will affect predicted releases and compliance. It does appear to the committee that engineered design features to reduce the adverse consequences of an intrusion are available, but not all of these features may be cost effective or needed if low actinide solubilities and/or sufficiently long travel times and radionuclide retardation can be demonstrated by the PA. Thus, to inform decisions on implementing costly engineering design features, it is important that these experimental studies be continued to completion.
- The 1992 PA was directed at demonstrating compliance with the radionuclide release limits stipulated in 40 CFR 191. No analyses were made of the health risks associated with these releases. At WIPP, radionuclides would be transported in brine that is considered to be, to a large extent, unpotable even to cattle. Consequently, the health risk will be very much smaller than if the water were potable.
In considering the radionuclide exposure hazard of WIPP to humans (see also Box 1.2), the committee recognizes the unpotable nature of much of the water in the Rustler above the Salado. The committee believes that accounting for the unpotable nature of the water could reduce the health risk by one or more orders of magnitude compared to the same release into a freshwater aquifer that is similar in size and has similar regional flow characteristics.
Taking into account all of the above considerations, the committee is confident in its judgment that DOE should be able to demonstrate that radionuclide releases at WIPP will be within the limits allowed by EPA, for both the undisturbed and disturbed cases, even with the severe criteria defined in 40 CFR 194. The associated health risks are likely to be well below the levels allowed under international standards. The results of the 1996 PA are awaited with interest as a test of this judgment.
The heavy emphasis placed in the 1992 PA on the analysis of compliance with 40 CFR 191 for the "disturbed case"—that is, assessing the consequences of human intrusion—is appropriate in the committee's judgment because of the added complexity of the radionuclide pathways introduced by intrusion. Even so, it is important to recognize the remarkable ability of the salt at WIPP to isolate waste under undisturbed conditions.
The Nature And Frequency Of Human Intrusion
Although the consequences of some form of human intrusion should be assessed, it is also evident to the committee that there is no scientific justification for estimating the precise nature or frequency of such intrusions over the next 10,000 years. For compliance purposes at WIPP, EPA has decided that the frequency of drilling for oil and gas exploration and production during the past century will continue unchanged for the next 10,000 years, even though it is acknowledged that the oil and gas reserves near WIPP will be practically exhausted within the near future (i.e., in less than 50 to 100 years).
The EPA assumes that today's drilling activities are representative of the type of drilling activities that will occur in the future. Although the resources drilled for today may not resemble those drilled for in the future, the EPA
believes it is reasonable for the average rate to be projected over the next 10,000 years, based on the assumption that while oil and natural gas may be depleted, other resources (which are not currently economical to recover, or whose uses are not yet evident) may become more valuable. This assumption leads [the EPA] to the conclusions that it is reasonable to project oil and gas drilling rates (based on the historical record) over the regulatory time frame; and second, that since these rates are surrogates for other potential resources, it is inappropriate to include consequences of activities or secondary
recovery techniques specific to oil drilling. (EPA, 1996, pp. 12-7, 12-8)
In effect, the EPA assumes that resources equivalent to more than 100 oil and gas fields, comparable to those now being exploited, will be explored for and developed at the WIPP site by borehole extraction methods, using the drilling density of the past century over the 10,000-year regulatory time frame. (Note: the mining of potash, using techniques in current use near WIPP, would not involve borehole extraction.)
The EPA prescription in 40 CFR 194 with respect to human intrusion illustrates both the difficulty and the subjective judgment involved in arriving at a regulatory standard for WIPP. Prescribing a future frequency of intrusive drilling based on past and present-day technologies is inherently speculative, highly subjective, and untestable by the scientific method (NRC, 1995). It is disconcerting to the committee that a repository such as WIPP, which appears to be an excellent choice based on geological considerations, could be judged "non-compliant" solely on the basis of a criterion which has a poor scientific basis.
The committee considers the approach proposed in the NRC (1995) report, Technical Bases for Yucca Mountain Standards (TYMS), to be a more appropriate way in which to assess the significance of human intrusion. The TYMS report recommends that the consequences of a single or combined drilling intrusion (such as E1E2—see Figure 2.5 of this report) be examined qualitatively, rather than quantitatively as in 40 CFR 191, in order to assess the severity of the consequences and to indicate what repository design measures may be appropriate to reduce the severity.
The incremental risk assumed by accepting human intrusion could also be compared with the long-term risks associated with leaving TRU waste on the surface.
The WIPP Compliance Certification Application is the first test of a unique legal framework, which necessarily will not be based on prior experience. Inevitably, such a legal framework will contain controversial and weak elements. DOE and EPA will need to work closely, in the public interest, to ensure that such elements of the complex legal framework are resolved expeditiously, through negotiation based upon scientific understanding.
WIPP also is a pioneering effort in the assessment of geological site suitability and design procedures for a waste repository. It is the first repository in the nation for which an application to begin permanent, deep, geological disposal is being submitted for a decision.
The scientific and technical questions posed, particularly to earth scientists, required unprecedented levels of quantitative assessment of the response of a rock mass over tens of thousands to hundreds of thousands of years. A great deal of original research was necessary to demonstrate the applicability of laboratory data to field-scale conditions and to develop and assess the reliability of computer models to reproduce these complex systems. Teams of specialists were assembled, and scientific programs defined. Total System Performance Assessment was introduced and refined. Inevitably perhaps, the scientific emphases, budget allocations, and the most critical information needs identified by PA were not always in concert. These difficulties notwithstanding, a great deal has been accomplished.
In retrospect, one valuable lesson from WIPP is the importance of establishing site-specific PA early in the repository site characterization and design process, both to identify the critical research needs and to establish the overall goals of the project. Scientific programs, priorities, and budget allocations must be identified and periodically re-examined, all in the context of PA.
Significant uncertainties are implicit in extrapolating from one or two decades of data, at most, to tens or hundreds of thousands of years—but a well coordinated integrated program, using PA effectively, should be able to produce sufficiently credible answers to establish, with "reasonable expectation" and within a finite time and budget, whether or not a proposed repository site and design are appropriate for effective long-term isolation of radioactive waste.