Comments on Proposed Radiation Protection Standards for Yucca Mountain, Nevada (1999)

Advisers to the Nation on Science, Engineering, and Medicine

National Academy of Sciences

National Academy of Engineering

Institute of Medicine

National Research Council

November 26, 1999

Dear Administrator Browner:

The Board on Radioactive Waste Management (BRWM) has prepared thisNational Research Council report ¹in response to the Environmental Protection Agency’s (EPA’s) request for comments on Environmental Radiation Protection Standards for Yucca Mountain,Nevada; Proposed Rule ²and requests that it be entered into public docket A-95-12. Thisreport reflects a consensus of the BRWM and has been reviewed inaccordance with the procedures of the National Research Council.

The Energy Policy Act of 1992 directed EPA to arrange for an analysisby the National Academy of Sciences of the scientific bases for radiationprotection standards to be applied at Yucca Mountain. The Act alsodirected EPA to develop radiation protection standards that wereconsistent with the National Academy of Sciences’ recommendations:

... the Administrator shall, based upon and consistent with the findingsand recommendations of the National Academy of Sciences, promulgate,by rule, public health and safety standards for protection of the public fromreleases from radioactive materials stored or disposed of in the repository atthe Yucca Mountain site. Such standards shall prescribe the maximum annualeffective dose equivalent to individual members of the public from releasesto the accessible environment from radioactive materials stored or disposedof in the repository. (P.L. 102-486, Title VIII, Section 801. 42 U.S.C. Section 10141 note)

Findings and recommendations to EPA on the technical bases for YuccaMountain standards were provided in the National Research Council³ report entitled Technical Bases for Yucca Mountain Standards,⁴ hereafter referred to as the “TYMS report.” This report was authored by a committee of experts that was appointedby the Chairman of the National Research Council and that operatedwith oversight from the BRWM. The TYMS committee was disbanded afterits report was completed in 1995. However, the BRWM is a continuingNational Research Council committee⁵ and can, at its discretion and with prior approval from the NationalResearch Council’s Governing Board, prepare reports on significantand timely issues.

The Board’s objectives in preparing this self-initiated letter report are thefollowing: (1) to provide a comparison of the technical bases forYucca Mountain standards recommended by the TYMS report with thoseproposed by EPA in its draft 40 CFR Part 197 (the “proposed standards”); ⁶and (2) where differences are noted, to provide comments on how EPA’s draft standards could be changed to make them consistent with therecommendations in the TYMS report. This letter report describesthe major recommendations in the TYMS report that EPA followed inits proposed rule, and it also highlights several areas in whichthe recommendations were not followed.

For nine major issues, the TYMS report provided recommendations onthe technical bases for Yucca Mountain standards. A brief summaryof each of those issues, the technical bases recommended by the TYMSreport, and the standards proposed by EPA are given below. Severalof these issues are discussed in more detail later in this report.

1. THE FORM OF THE STANDARD. Although the Energy Policy Act stipulated that EPA should developa standard that prescribes dose equivalents, the TYMS report recommendedthat EPA develop a standard that sets a limit on the risk to individuals of adverse health effects from releases from the repository.EPA did not accept the TYMS recommendation. Instead, it proposesa dose-based standard.

2. WHO IS PROTECTED. The TYMS report recommended that EPA apply the standards to a critical group representative of those individuals in the population who, basedon cautious, but reasonable, assumptions, have the highest risk resultingfrom repository releases. EPA proposes standards based on a reasonably maximally exposed individual (RMEI). As discussed below, the proposed standards are broadly consistentwith the recommendations in the TYMS report in light of the way inwhich EPA has proposed to implement them.

3. LEVEL OF PROTECTION. The TYMS report noted that the level of protection was a policydecision to be established through the rulemaking process. EPA hasaccepted the

   3	The National Research Council is the operating arm of the NationalAcademies. Its reports do not National Research Council, 1995, TechnicalBases for Yucca Mountain Standards: Washington, Academy of Engineering,or the Institute of Medicine.

   4

   necessarily represent the views of the members of the National Academyof Sciences, the National D.C., National Academy Press. This reportcan be viewed online at the following URL: http://books.nap.edu/catalog/4943.html.This report is referenced in the text as “TYMS, p. xxx”, where xxx is the relevantpage number.

   5	A standing National Research Council entity on radioactive wastemanagement has been operating continuously since 1958 and is responsiblefor oversight of studies on radioactive waste management and disposal.The current BRWM membership roster is given in Attachment A.

   6	The Board used the TYMS report and the Federal Register notice asreference materials for this letter report. The Board did not examinethe additional documentation available from EPA.

rulemaking suggestion and is proposing to establish two levels of protection, one for an allpathways standard to protect the RMEI, and another standard to protect users of ground water and ground water as a resource. The proposed numeric value for the all-pathways RMEI standard falls within the range of values suggested in the TYMS report. The proposed groundwater standard is discussed under Protection of the General Public.

4. PROTECTION OF THE GENERAL PUBLIC. The TYMS report concluded that an individual-risk standard wouldprotect the health of the general public, provided that policy makersand the public were prepared to accept that very low radiation dosespose a negligibly small risk. EPA appears to agree with this conclusion.However, some parts of EPA’s proposed standards are inconsistentwith this conclusion, most notably with respect to estimation andconsideration of collective doses, design alternatives to reducecarbon-14 releases from the repository, and a standard for protectionof ground water as a resource.

5. HOW LONG. The TYMS report recommended that compliance assessment be conductedfor the time when the greatest risk occurs, within the limits imposedby the long-term stability of the geologic environment, but alsonoted that EPA might choose a different compliance time to be consistentwith its management of risks from long-lived hazardous non-radioactivematerials. EPA proposes a 10,000-year standard, but also requiresthe peak dose within the period of geologic stability to be calculated.The proposed standard is consistent with the recommendations in theTYMS report in some respects, but there are also differences thatmay have a major impact on the licensing process without real benefitto protection of the public, as discussed below.

6. HUMAN INTRUSION. The TYMS report recommended that the standards developed by EPAshould require active and passive institutional controls in the nearterm, should be based on an explicitly assumed intrusion scenarioin the long term, and should set limits for the human intrusion scenariothat are no more stringent than the undisturbed case. EPA proposesa standard that is broadly consistent with these recommendations,as discussed below. It is based on two alternative scenarios fora single human intrusion as a result of exploratory drilling forground water.

7. ALARA.⁷The TYMS report noted that there is no scientific basis for incorporatingthe ALARA principle into the standard. The proposed EPA standarddoes not explicitly require ALARA, but does suggest an ALARA-likeapproach to reduce carbon-14 releases from the repository (see Protection of the General Public).

8. EXPOSURE SCENARIOS. The TYMS report concluded that there is no scientific basis for predictingthe societal factors required to establish exposure scenarios and,therefore, the report recommended that such scenarios be establishedthrough the rulemaking process. EPA has used rulemaking to defineexposure scenarios based on the state of society, human biology,and knowledge that exists at the time of submission of the licenseapplication for the repository. The proposed standard is consistentwith the TYMS report’s recommendations.

7

ALARA, or “As Low As Reasonably Achievable,” is usually applied after threshold regulatory limits have been met.Under ALARA, all exposures shall be kept as low as reasonably achievable,social and economic factors being taken into account (e.g., InternationalCommission on Radiological Protection Report ICRP-60, 1991).

9. TECHNOLOGY-BASED STANDARDS. The TYMS report recommended against imposing repository subsystemperformance requirements in the Yucca Mountain standards. The proposedEPA standards do not contain subsystem requirements and, therefore,are consistent with the recommendations in the TYMS report.

• THEFORMOFTHESTANDARD. Both the TYMS report and the draft EPA standards draw a clear distinctionbetween the form of the individual-protection standard and the numeric value of that standard, whatever its form. We will follow that distinctionin this report and focus the discussion in this section on the formof the standard. The numeric value of the standard is discussed inthe section entitled Level of Protection elsewhere in this report.

The TYMS report recommended that EPA adopt a risk-based standard for protection of individuals for the following two reasons(TYMS, p. 64-65):

1. A risk-based standard would not have to be revised in subsequentrulemaking if advances in scientific knowledge reveal that the dose-responserelationship is different from that envisaged today. Such changes have occurred frequentlyin the past, and can be expected to occur in the future ....

2. Risks to human health from different sources, such as nuclear powerplants, waste repositories, or toxic chemicals, can be compared in reasonably understandable terms. Doses or releases have to be stated in radiationunits ... that are not easily understood by the general public and that canonly be compared conveniently with other sources of radiation or radioactivity.

In the Supplementary Information preamble (hereafter referred toas the “preamble”) to its proposed standards, EPA states that (EPA, p. 46984)“We have reviewed and evaluated the merits of a risk-based standardas recommended by NAS. However, we are proposing a dose-based standardfor the following reasons.” EPA then provides the following fourreasons for its decision to propose a dose-based standard for protectionof individuals:

First, both national and international radiation protection guidelinesdeveloped by bodies of non-governmental radiation experts ... generally have recommended that radiation standards be established in terms of dose. Also, nationaland international radiation standards ... are established almost solelyin terms of dose or concentration, not risk. Therefore, a risk-based standardwill not allow a convenient comparison with the numerous existing radiation guidelinesand standards that are stated in terms of dose.

Second, we have an established methodology for calculating dose ....The development of this methodology was a combined effort of many Federal agencies involved in radiation protection and has become Federalpolicy. ... By contrast, there is currently no [agreed upon guidance] for calculating risk from radiation exposure.

Third, we have based the proposed dose-based standard upon the riskof developing a fatal cancer as a result of that level of exposure basedupon a linear, non-threshold, dose-response relationship. We would establisha risk-based standard in the same manner. Thus, a risk-based standard, likea dose-

based standard, depends upon current knowledge and assumptions aboutthe chance of developing a fatal cancer from a particular exposure level.Dose and risk are closely related; one can be converted to the other simplyby using the appropriate factor. Therefore, both dose- and risk-based standardsare based upon scientific assumptions that could change and no matter how itis expressed, the standard is based upon risk.

Finally, [the Energy Policy Act] specifically calls for a dose-based standard.

The Board believes that EPA’s rationale for proposing a dose-based standard is flawed for thefollowing reasons. EPA’s statement in its reason #3 (above) that“a risk-based standard ... depends upon current knowledge and assumptionsabout the chance of developing a fatal cancer from a particular exposurelevel” is incorrect. A risk-based standard is not based on scientific assumptions. Instead, it is based on a public-policydetermination of acceptable risk levels to individuals or populations,usually expressed in the form of a probability—for example, a 10⁻⁶ (or one-in-a-million) lifetime probability of developing a fatalcancer.

As noted in the TYMS report, there is no scientific basis for determiningacceptable risk levels. Rather, the determination of acceptable risklevels is a policy decision established through the rulemaking processand is based on the risk that the public is willing to bear frompotential radiation releases from a repository. Once the level ofacceptable risk is established, then a numerical standard (e.g.,a dose standard) that relates that risk to individual or populationexposures to radiation can be obtained using data that are subjectto change with improvements in scientific knowledge. For example,acceptable risk then can be characterized in terms of dose limitsestablished from knowledge of the dose-response relationship andexposure pathways.

There have been several adjustments in the dose-response relationshipbetween radiation and fatal cancers over the past several decades,as EPA acknowledges in the preamble to its proposed standards (EPA,p. 46979). Several BEIR (Biological Effects of Ionizing Radiation)reports have addressed the changing knowledge of dose-response relationshipsand another BEIR committee has just begun the latest examination.⁸However, establishing an acceptable risk level does not require knowingthe dose-response relationship or exposure pathways. It is not ascientific issue but a public policy choice. Once established, thisrisk level could be changed,⁹but it would be, again, a public policy decision.

The EPA itself acknowledges the nature of the changing dose-responserelationship in its discussion of the proposed standards: EPA statesthat the current proposal for a 15 millirem limit for whole-bodydose is not a change from the previous 40 CFR Part 191.03(a) limitof 25 millirem because “the 25-mrem [millirem]/yr, whole-body-doselimit established in 1985 is essentially equivalent to the risk associatedwith today’s proposed limit of 150 µSv [microsieverts] (15 mrem) CEDE [committed effectivedose equivalent]/yr ...” (EPA, p. 46985). Thus, EPA writes that “the proposed dose level essentially maintainsa similar risk level ...” (EPA, p. 46983).

In explaining its decision to adopt a dose-based standard, EPA notesin reasons #1 and #2 that most national and international standardsare expressed in terms of dose and that it has an established methodologyfor calculating dose. But what is EPA’s basis for proposing whatit believes the numeric value of the dose standard should be? EPAanswers this question directly in its reason #3: “... we have basedthe proposed dose-based standard upon the risk of developing a fatalcancer as a result of that level of exposure based upon a linear,non-threshold, dose-response relationship. We would establish a risk-basedstandard in the same manner ....” In other words, EPA acknowledgesthat it is using risk to establish its dose-based standard.

EPA also states in reason #3 that dose and risk are closely relatedand can be converted by using the appropriate factor. Although thisis certainly true arithmetically, EPA’s statement obscures the relationshipbetween these two factors. For the purposes of public policy making,these two factors are not equivalent—acceptable risk is derived froma public policy decision, whereas dose is derived from risk throughthe appropriate conversion factors.

By proposing a standard using dose as a proxy for risk, EPA is precludingthe public from easily comparing the proposed individual-protectionstandards for Yucca Mountain with standards for regulating otherkinds of hazardous materials. Furthermore, the use of dose as a proxystandard also makes it difficult for the public to compare the proposedindividual-protection standard, which is now expressed by EPA inunits of microsieverts per year and millirems per year, directlywith EPA’s proposed ground-water standard, which is expressed byEPA in terms of both millirems per year and picocuries per liter.Additional comments on the ground-water standard are made elsewherein this letter.

The TYMS report’s recommendation of a standard based on a specified level of acceptablerisk that is determined through the rulemaking process representsa major departure from traditional EPA practice for setting radiationstandards, as EPA itself acknowledges in reason #2 above. EPA hasgenerally used dose-based radiation standards¹⁰and would no doubt find it difficult to adopt a risk-based approach.Indeed, asking the public about acceptable risk levels may be farmore difficult than following established precedents. Nevertheless,a risk-based form for the standard was recommended by the TYMS committeebecause it requires public involvement in a critical decision onacceptable levels of risk from a Yucca Mountain repository, and because,fundamentally, this is a policy, not a scientific, judgment.

EPA points out in reason #4 that it was directed by the Energy PolicyAct to develop a “dose-based standard.” In fact, as noted at the beginning of this report,Congress directed EPA to develop standards that “prescribe the maximumannual effective dose equivalent to individual members of the public....” In the Board’s view, this directive does not prohibit EPA from basing sucha standard on acceptable risk as determined through the rulemakingprocess. In other words, the TYMS report’s recommendation that theform of the individual-protection standard be based on risk does not preclude EPA from expressing the numeric value of the standard in units of risk and in derivative units of dose,so long as the risk value is clearly understood as the underlyingbasis for the proposed dose standard. This approach would achievethe aims of the

TYMS report’s recommendations and it would allow the EPA to meet its Congressionalmandate.

The Board strongly recommends that EPA adopt an explicit risk-basedindividual-protection standard for Yucca Mountain.

• WHOISPROTECTED. The standard proposed by EPA departs, in a strict sense, from theTYMS recommendation of using a “critical group” (CG) as the basis for the standard.EPA has chosen instead to write the standard for a “reasonably maximallyexposed individual” (RMEI). As explained below, however, the Boarddoes not view EPA’s choice of the RMEI approach over a CG approachas a significant departure from the recommendations in the TYMS report because ofthe way in which EPA has proposed to implement this standard. Inother words, the Board believes that the proposed EPA standard isbroadly consistent with the TYMS report’s recommendations.

EPA appears to recognize that its standard must be written in a waythat provides appropriate protection to the individuals who havethe highest potential for exposure to radioactivity released froma repository, while avoiding unrealistic and unnecessarily conservativeassumptions for individual exposure. EPA’s proposed standard setsforth a detailed definition for the implementation of the RMEI approach(EPA, p. 47014-47015). The Board believes that such detailed specificationis needed for both the RMEI concept and EPA’s reasonable expectation concept so that both can be appropriately implemented.

The CG concept is described in the TYMS report on pages 52-54. Akey idea is that the individuals in the group should be “homogeneousin risk” (TYMS, p. 53). The TYMS report defines the critical groupas being “representative of those individuals in the population who,based on cautious, but reasonable, assumptions, have the highestrisk resulting from repository releases. The group should be smallenough to be relatively homogeneous with respect to diet and otheraspects of behavior that affect risks” (TYMS, p. 53). The reportgoes on to say that, for individuals living thousands of years inthe future, “it will be necessary to define hypothetical personsby making assumptions about lifestyle, location, eating habits, andother factors” (TYMS, p. 54).

Whether the new EPA standard is written for a small homogeneous “critical group” of “a few to a few tens of persons” (TYMS, p. 53) or for a single “reasonably maximally exposed individual,” the key issue in the implementation of the standard is the specificationof the assumptions about location, behavior, and life style thatwill determine exposure. The choice of one individual or a homogeneoussmall group as the basis for the standard is less important thanthe specification of assumptions that will determine the potentialexposure of that individual or small group.

EPA has provided considerable specificity in the proposed standardsin its definition of RMEI. In so doing, the EPA has chosen to excludesome exposure situations, such as from a small quantity of contaminatedground water close to the repository in a location where no peoplelive at present and where future agricultural activity would appearunlikely.

In preparing the final rule, EPA may wish to be even more specificin setting forth the location, behavior, and lifestyle of the RMEIfor purposes of calculating risk in determining compliance with thestandard. Alternatively, EPA may choose to pass to the U.S. Nuclear

Regulatory Commission (USNRC) the challenge of establishing specificassumptions for the compliance calculations. These specifics needto be clearly established prior to the process for evaluating whethercompliance with the standard has been achieved in an applicationfor a repository license.

• LEVELOFPROTECTION. The TYMS report made no recommendation about a specific level (i.e.,a specific numeric value) for the individual-protection standardto be set by EPA:

The level of protection established by a standard is a statementof the level of risk that is acceptable to society. We [the TYMS Committee] acknowledge that determining what is acceptable is not ultimately a question of sciencebut of public policy. Whether posed as ‘How safe is safe enough’ or as ‘What is an acceptable level of risk?’, the question is not solvable by science alone. The rulemaking process ... is an appropriate method of addressing thequestion of an appropriate level of protection. (TYMS, p. 49)

Table 2-3 of the TYMS report (TYMS, p. 43-46) provides a summaryof the approaches that have been recommended or established by variousadvisory or regulatory bodies, both in the United States and abroad,for regulating high-level radioactive waste disposal. Table 2-4 ofthe TYMS report (TYMS, p. 50) lists protection levels in terms ofannual dose limits, annualized individual-risk limits, or both, thathave been established in the United States by either the USNRC orEPA for a variety of different activities, some similar to and othersquite different from the activity associated with Yucca Mountain.The values range from about 10⁻⁴ to 10⁻⁶ (i.e., one chance in ten thousand to one chance in one million) interms of annualized individual risk. The TYMS report recommends notonly that rulemaking be used for establishing levels of protection,but also suggests that a “reasonable starting point” for the rulemaking would be for EPA topropose a numeric value somewhere within the range presented in Table2-4 (TYMS, p. 49).

EPA is establishing the level of protection through the rulemakingprocess and is proposing a dose-based, all-pathways individual-protectionstandard of 150 microsieverts (15 millirems) CEDE/year, which, accordingto EPA, corresponds to an annualized risk level of 7 × 10⁻⁶ (EPA, p. 46985). EPA also proposes that the individual-protectionstandard should apply to the so-called “undisturbed” case, that is, a postulatedfuture scenario in which no disturbances such as human intrusionoccur. EPA has carefully spelled out the public-policy issues thatsupport its position and notes, for example, that the proposed levelof protection is consistent with other current standards. However,EPA provides only one example of such consistency, and it is notevident from the discussion if the standards are actually consistent.¹¹

Although the dose-based form of the standard is inconsistent with the recommendations in the TYMSreport, as noted previously, the risk-equivalent numeric value of the standard given by EPA falls within the range of values inTable 2-4 of the TYMS report. Thus, the magnitude of the proposed numeric value of the individual-protection standardis consistent with the recommendations in the TYMS report. However,EPA has failed to make clear the underlying risk basis for this proposedlevel of protection or to present this risk basis in a way that iseasily understood by the public.

In addition to an all-pathways individual-protection standard, EPAis proposing a separate ground-water standard that appears to bedesigned to protect both individuals living near the repository andthe general public living at some distance from the repository. Thisstandard is discussed in more detail under Protection of the General Public below.

• PROTECTIONOFTHEGENERALPUBLIC. Section 801(a)(2)(A) of the Energy Policy Act specifically calledfor an answer to the question “whether a health-based standard basedupon doses to individual members of the public from releases to theaccessible environment ...will provide a reasonable standard forprotection of the health and safety of the general public.”

The TYMS committee answered this question as follows:

We conclude that an individual-risk standard would protect publichealth, given the particular characteristics of the site, provided that policymakers and the public are prepared to accept that very low radiation doses posea negligibly small risk.

The individual-risk standard that we recommend is intended to protecta critical group. In this context, the general public includes both global populationsas well as local populations that lie outside the critical group. Globalpopulations might be affected because radionuclide releases from a repository can intheory be diffused throughout a very large and dispersed population. In thecase of Yucca Mountain, the likely pathway leading to widely dispersed radionuclidesis via the atmosphere beginning with release of carbon dioxide gas containingthe carbon-14 ... radioactive isotope which might escape from the waste canisters.

The risks of radiation produced by such wide dispersion are likelyto be several orders of magnitude below those of a local critical group. (TYMS, p. 7)

In other words, the TYMS report concluded that individual-protectionstandards would also be protective of the health of the general public.Therefore, the TYMS report concluded that there was no need to establisha separate set of standards to protect the health of the generalpublic.

EPA’s proposed standard is, on first examination, in agreement with theTYMS report’s recommendations:

We are proposing to find that the individual-protection standardis sufficient to protect public health based upon the unique characteristics of thearea around the Yucca Mountain site. ... we are not proposing to adopt a separatelimit on radiation releases for the purpose of protecting the general population,but we are recommending that collective dose be estimated and considered [underlining added for emphasis] .... (EPA, p. 46992)

Moreover, with respect to carbon-14 releases from the Yucca Mountainrepository, EPA also notes that the risk to the general public isvery low:

The risk over an individual’s lifetime from exposure to gaseous [carbon-14] released from the Yucca Mountain repository, as estimated by NAS,... is extremely low and well below the risk level that we generally regulate.(EPA, p. 46991)

Nevertheless, EPA goes on to write

While we are not proposing to adopt additional regulatory requirementsfor collective exposures of the general population ... we urge DOE toexamine design alternatives for the disposal system, for the purpose of reducingpotential risk to the general population .... (EPA, p. 46992)

EPA’s requirement that the collective dose be “estimated and considered” is inconsistent with both the TYMS report and with EPA’s own conclusion (in the preamble to its proposed rule) that additionalstandards are not needed to protect the general public. Moreover,EPA has provided no guidance to either the licensee or the regulatoron what they are to do with these collective dose estimates.

Further, EPA’s urging of DOE to seek design alternatives to reduce potential risksto the general public from carbon-14 releases, even when individual-doselimits have been satisfied, is also without a firm scientific basisand will add little, if any, additional protection to the generalpublic. Indeed, EPA’s urging of DOE could be interpreted as an attemptto apply ALARA, which was also rejected by the TYMS committee becauseit lacks a scientific basis.

The TYMS report noted that “Defining the level of incremental risk that is negligible is a policyjudgment” (TYMS, p. 60). In the Board’s view, EPA is making policy judgments with respect to collectivedose estimation and design alternatives to reduce carbon-14 releasesbut has not explained clearly its reasoning for these judgments.The Board recommends that EPA revise the preamble discussion to explainthe embedded policy judgments, and that EPA explicitly include forpublic comment the option to exclude the estimation and considerationof collective dose and design alternatives to reduce gaseous carbon-14releases from the standard for the purpose of protecting the generalpublic.

EPA has also proposed to include a separate standard for protectionof future human users of ground water:

... we believe that ground water protection for the Yucca Mountainsite should focus upon the protection of the ground water as a resource for futurehuman use. It is the general premise of this proposal that the individual-protectionstandard would adequately protect those few current residents closestto the repository. The proposed ground water standards are directed to protecting theaquifer as a resource for current users, and a potential resource for larger numbers of future users either near the repository or for communities farther away comprised of as many as several hundred thousand people [underlining added for emphasis]. (EPA, p.47002)

... we are proposing additional ground water protection standardsthat would establish specific limits to protect users of ground water and groundwater as a resource. (EPA, p. 46992)

... In its license application to [US]NRC, DOE must provide a reasonable expectation that, for 10,000 years of undisturbed performance afterdisposal, releases of radionuclides from radioactive material in the YuccaMountain disposal system will not cause the level of radioactivity in therepresentative volume of ground water at the point of compliance to exceed [specified limits] .... (EPA, p. 47015)

... we are proposing to protect the ground water to the same levelas the maximum contaminant levels (MCLs) for radionuclides which we have established under the authority of the Safe Drinking Water Act ....(EPA, p. 47000)

EPA states that its ground-water standard is intended to protectground water as a resource for future human users. In fact, the proposedground-water standard appears to be designed to protect both individualsliving near the repository (“larger numbers future users ... nearthe repository”) and the general public living at some distance fromthe repository (“communities farther away comprised of as many asseveral hundred thousand people”).

The TYMS report concluded that an individual-protection standardwould be sufficient to protect public health from a repository atYucca Mountain. Therefore, the TYMS committee did not consider itnecessary to comment on the need for a ground-water standard suchas that established in the Safe Drinking Water Act and used in 40CFR 191:

40 CFR 191 includes a provision to protect ground water from contaminationwith radioactive materials that is separate from the 40 CFR 191 individual-doselimits. These provisions have been added to 40 CFR 191 to bring it into conformitywith the Safe Drinking Water Act, and have the goal of protecting groundwater as a resource. We make no such recommendation, and have based our recommendations on those requirements necessary to limit risks toindividuals [underlining added for emphasis]. (TYMS p. 121)

EPA agrees that an individual-protection standard is appropriateto protect public health and says that it is not recommending a separatelimit on radiation releases for the purpose of protecting the generalpopulation. Yet EPA is inconsistently proposing a carryover of 40CFR 191 to establish separate ground-water protection standards forpurposes of protecting individuals and the general population. Suchseparate limits may greatly complicate the licensing process andhave but a negligible impact on protection of the public.

In the preamble, EPA implies that there is a scientific basis forinclusion of separate ground-water limits in the standards—for example,EPA provides a detailed analysis of approaches to calculating suchlimits on page 47001-47003. The Board respectfully disagrees anddoes not believe there is a basis in science for establishing suchlimits for the reasons described above. The Board recognizes EPAhas the authority under the Energy Policy Act to establish separateground-water limits as a matter of policy, but if it does so it shouldexplicitly state the policy decisions embedded in the proposed standardand ask the public to comment on those decisions.

If EPA wishes to establish such standards on the basis of science,it must make more cogent scientific arguments to justify the needfor this standard. EPA should explain, for example, why a separateground-water standard is required, given that EPA has establisheda

standard for protection of individuals and has concluded that itdoes not need to establish a separate standard for protection ofthe general public. Given that most if not all of the estimated exposureto radiation from the Yucca Mountain repository is estimated to bethrough consumption and use of contaminated ground water, the proposedground-water protection standard appears to duplicate provisionsof the individual-protection standard. However, because the ground-waterstandard does not necessarily apply at the same location as the individual-protectionstandard, its application to Yucca Mountain lacks overall consistencyand coherence.

Additionally, the proposed ground-water protection standard and individual-protectionstandard may be inconsistent scientifically. The individual-protectionstandard is derived from EPA’s latest dosimetry models, whereas theproposed ground-water protection standard ¹² is based on a 25-year-old regulation (i.e., the Safe Drinking WaterAct), which in turn is based on 40-year-old dosimetry. ¹³ The use of such outdated standards is contrary to the overarchingthemes of the TYMS report: namely, the importance of using the bestavailable science in the standards and of designing those standardsso that they can be easily modified as scientific knowledge changes.¹⁴

The use of mixed dose-based units in the proposed ground-water standard(see footnote 12) reinforces an important point made in previoussections of this report: these units give the public no basis onwhich to compare levels of protection afforded by these standardsor to compare these proposed standards with standards already inplace for regulation of other hazardous materials. The Board recommendsthat EPA present such comparisons, preferably in terms of risk, togive the public a firmer basis for understanding and commenting onthese proposals.

• HOWLONG. The TYMS report noted that “The time scale for long-term geologic processes at Yucca Mountainis on the order of approximately one million years” (TYMS, p. 72),and concluded that “there is no scientific basis for limiting thetime period of the individual-risk standard to 10,000 years or anyother value” (TYMS, p. 55). Additionally, the report recommendedthe application of a risk standard “at times when the peak potentialrisks might occur” (TYMS, p. 55-56). The report also noted that “Selection of a time scale ... must therefore takeinto account the scientific basis for the performance assessmentitself” and “also involves policy considerations” (TYMS, p. 30) “that we [the TYMS committee] have not addressed” (TYMS, p. 56).

The TYMS report emphasizes that peak releases, doses, and risks atYucca Mountain are calculated to arise well beyond 10,000 years,and it quotes an earlier National Research Council report ¹⁵ which states that a 10,000-year limit “makes compliance rather easy” (TYMS, p. 55). The TYMS report also noted that uncertainties inthe non-biosphere part of the analyses do

not necessarily increase rapidly beyond 10,000 years; some uncertaintiesthat can dominate at shorter times (e.g., canister failure rates)can become less important at long time scales.

In its proposed standards, EPA asks for comments from the publicon two approaches to the time-scale issue. The first is effectivelythat proposed by the TYMS report, and the second is application ofa quantitative dose limit up to 10,000 years, together with a requirementto examine “disposal system performance after 10,000 years ... tosee if dramatic changes ... could be anticipated” (EPA, p. 46993).No indication is given of how one judges whether a change is “dramatic.” EPA states its preference for the second approach.

EPA notes that the majority of those commenting to EPA on the TYMSreport (EPA, p. 46981) stated that quantitative analyses of the disposalsystem should be carried out for compliance purposes at the timeof peak risk, but EPA has nevertheless retained its earlier recommendationfor quantitative compliance assessment only up to 10,000 years. EPAhas given a series of policy and technical arguments for this choiceand has also included a requirement for consideration of the periodbeyond. The TYMS committee excluded policy considerations from itsdeliberations on this issue. The TYMS committee had considered allof the technical arguments offered by EPA and concluded that “thereis no scientific basis for limiting the time period ... to 10,000years” (TYMS, p. 55). Clearly, the 10,000-year limit is strictlya policy choice and should be acknowledged as such.

Given that the proposed EPA standard requires that the performanceof the disposal system be examined after 10,000 years if the peakdose is calculated to occur then, there may be little practical difference between the TYMS report’s recommendations and the proposed EPA standards. The major issueis that EPA provides no guidance on how analyses should be done forthe period of geologic stability beyond 10,000 years and gives noindication of how the results should be used in judging acceptability.To mandate that these results become “part of the public record” but to give no indicationof how they will be taken into account seems to postpone rather thansolve problems associated with licensing and provide no real benefitsto protection of the public.

In preparing its final rule, EPA may wish to be more specific inproviding guidance on how the analyses beyond 10,000 years shouldbe used in determining compliance with the standard. Alternatively,EPA may choose to explicitly pass to the USNRC the task of establishingsuch guidance.

• HUMANINTRUSION. The TYMS report described at some length how the EPA standard shouldaddress the issue of human intrusion. On the issue of passive andactive institutional controls, the TYMS report concluded that “...the repository developer should be required to provide a reasonablesystem of active and passive controls to reduce the risk of intrusionin the near term” (TYMS, p. 108), even though the TYMS report concludedthat it is not possible, using such controls or by any other knownmeans, to preclude such an intrusion, nor is there a way to developan objective numerical value for the likely future rate of inadvertentintrusions. On this set of institutional-control issues, the EPA’s proposed standard adheres closely to the philosophy espoused inthe TYMS report.

On the issue of using a risk-based approach to regulating human intrusion,the TYMS report states that “... we do not believe that it is scientificallyjustified to incorporate alternative scenarios of human intrusioninto a fully risk-based compliance assessment that requires

knowledge of the character and frequency of various intrusion scenarios” (TYMS, p. 108). The report continues

We do however conclude that it is possible to carry out calculationsof the consequences for particular types of intrusion events, for exampledrilling one or more boreholes into and through the repository. We also believe thatcalculations of this type might be informative in the sense that they can provideuseful insight into the degree to which the ability of a repository to protect publichealth would be degraded by intrusion.

For these reasons, to address the human intrusion issue on an adequatebasis, we recommend that ... EPA should specify in its standard a typicalintrusion scenario to be analyzed for its consequences on the performance ofthe repository. Such an analysis will provide useful quantitative informationthat can be meaningful in the licensing process .... Because the assumed intrusion scenario is arbitrary and the probability of its occurrence cannotbe assessed, the result of the analysis should not be integrated into an assessmentof repository performance based on risk, but rather should be considered separately.The purpose of this consequence analysis is to evaluate the resilienceof the repository to intrusion. (TYMS, p. 108-109)

After discussing an approach using a single stylized but well-definedscenario, the TYMS report concludes

Because the probability is inherently unknowable, we are led to theconclusion that the most useful purpose of this type of analysis is to identifythe incremental effects from the assumed scenario .... Our recommendation is thatEPA should require that the conditional risk as a result of the assumed intrusionscenario should be no greater than the risk levels that would be acceptablefor the undisturbed-repository case. (TYMS, p. 112-113)

The key performance issue is whether repository performance wouldbe substantially degraded as a consequence of an inadvertent intrusion.... (TYMS, p. 121)

The EPA’s proposal in draft 40 CFR 197 follows the TYMS reasoning in mostof its particulars, specifically insofar as it concludes that onlyinadvertent intrusion should be covered, specifies a single stylizedinadvertent-human-intrusion scenario, requires that the scenariobe analyzed using the same methods as used for the undisturbed case,and requires that the doses arising from inadvertent human intrusionmeet the same standard as for the non-intrusion case. Specifically,

... we [EPA] are proposing to adopt the NAS-suggested starting point for a human-intrusion scenario. That scenario is a single, stylized intrusionthrough the repository to the underlying aquifer based upon current drillingpractices. The proposed scenario presumes that the intrusion occurs because of exploratory drilling for water. (EPA, p. 46999)

We are proposing to use the same RME [reasonable maximum exposure] descriptors for this analysis and scenario as in the assessment forcompliance

with the individual-protection standard .... we propose to interpret the NAS-recommended test of ‘resilience’ to be a longer-term test as measured by exposures caused by releases which occur gradually through the borehole,not suddenly as with direct removal. (EPA, p. 46999)

We also are proposing that the test of the resilience of the repositorysystem be the dose incurred by the same RMEI as determined for the individual-protection standard. (EPA, p. 47000)

The EPA proposed two alternative forms for its proposed rule, inlight of the possibility that an intrusion into the repository usingtoday’s drilling technology might not breach an intact canister inrelatively early times, before substantial canister degradation willhave occurred. The EPA approach here (EPA, p. 47015) is not discussedin the TYMS report, but insofar as it attempts to be realistic aboutthe likelihood of canister breach using today’s typical exploratory-drillingtechnology, it seems to be fully consistent with the TYMS philosophy.In summary, the proposed EPA approach to human intrusion appearsto have followed the TYMS recommendations very closely.

• CONCLUDINGOBSERVATIONS. The Congress recognized the special role of the National Academiesin providing scientific advice to EPA for use in establishing radiationprotection standards for Yucca Mountain. The text of the conferencereport (CR page H-12056; reproduced in TYMS, p. 141-143) describesthis role:

The Conferees do not intend for the National Academy of Sciences,in making its recommendations, to establish specific standards for protection ofthe public but rather to provide expert scientific guidance on the issues involvedin establishing those standards.

The Board’s main objective in writing this letter report is to assist EPA inensuring that the Yucca Mountain standards reflect the sound scientificthinking embodied in the TYMS report. The Board hopes that EPA findsthe information contained in this letter report to be helpful inachieving this end.

As discussed in detail elsewhere in this letter, the Board foundthat many of the standards proposed by EPA are consistent, eitherin design or in implementation, with the recommendations in the TYMSreport. However, the Board has also noted several significant differences.These differences reflect what the Board believes to be flawed reasoning,the misapplication of science to issues that have a clear policybasis, or a failure to explain fully the policy decisions embeddedin the proposed standards. The Board identified problems with thefollowing four issues:

• Form of the Standard. The Board finds that EPA’s various rationales for proposing a dose-based standard are flawedand that they miss a key point of public policy and public communication.The Board strongly recommends that EPA adopt an explicit risk-based,individual-protection standard for Yucca Mountain and has suggestedan approach for doing so that is consistent with both the TYMS reportand the Congressional directive to prescribe a “maximum annual effectivedose equivalent” for protection of individuals.

• Level of Protection. The Board finds that the numeric value proposed by EPA for its individual-protectionstandard is within the range of starting values suggested in theTYMS report. However, EPA has failed to make clear the underlyingrisk basis for this proposed level of protection or to present thisrisk basis in a way that is easily understood by the public.

• Protection of the General Public. EPA’s requirement that collective dose be estimated and considered isinconsistent with both the TYMS report and with EPA’s conclusionthat additional standards are not needed to protect the general public.Further, EPA’s urging of DOE to seek design alternatives to reducegaseous carbon-14 releases, even when individual-dose limits havebeen satisfied, is also without a firm scientific basis and couldbe interpreted as an attempt to apply ALARA, which was rejected bythe TYMS committee because it lacks a scientific basis. The Boardrecommends that EPA revise the preamble discussion to explain thepolicy judgments embedded in these requirements, and that EPA explicitlyinclude for public comment the option to exclude from the standardthe estimation and consideration of collective dose and design alternativesto reduce gaseous carbon-14 releases for the purpose of protectingthe general public.

  The Board also finds that EPA’s proposal to include a separate ground-water standard lacks a soundscientific basis and will add little, if any, additional protectionto individuals or the general public from radiation releases fromthe repository. If EPA wishes to establish ground-water standardson the basis of science, it must make more cogent scientific argumentsto justify the need for this standard. Moreover, such a standardshould be based on best-available science, not on 40-year-old dosimetryas EPA has proposed. If EPA wishes to establish a separate ground-waterstandard on the basis of policy, it should make clear the policydecisions embedded in the proposed standard and ask the public tocomment on those decisions.

• How Long. The Board finds that there may be little practical difference between the proposed EPA standardand the TYMS report recommendation because of the way that EPA proposesto implement this standard. However, EPA gives no guidance on howanalyses should be done for the period of geologic stability beyond10,000 years and gives no indication of how the results will be consideredin judging acceptability. To mandate that these results become “part of the public record” but to give no indication of how they will be taken into accountseems to postpone rather than solve problems associated with licensingand provide no real benefit to protection of the public. EPA maywish to provide more specific guidance on compliance determinationin its final rule. Alternatively, EPA may choose to explicitly passto the USNRC the task of establishing such guidance.