Micah D. Lowenthal and William E. Kastenberg
Center for Nuclear and Toxic Waste Management
University of California
Berkeley, California

Most of the current classes for radioactive waste in the United States are based on sources and jurisdictions rather than on the hazards and risks of disposal. A risk-based approach may be more desirable. Some have suggested alternative classification schemes that incorporate heat-generation rates, exposure rates, and/or half-lives. But these systems are based solely on the characteristics of the waste. We argue here that risks assessed outside of a context are of limited value. In particular, the risks posed by disposal of waste cannot be properly determined without examination of the behavior of the waste in a specific disposal environment. Radiation protection regulations and environmental laws governing disposal of radioactive waste demand site-specific assessments resulting in waste-acceptance criteria. We argue that within the DOE complex these waste-acceptance criteria (WAC) constitute a form of de facto waste classification for disposal. We suggest that WAC based on site-specific performance assessments could replace existing classes for disposal while other hazard-based classes could be retained for handling, storage, and transportation. This approach could lead to more appropriate distribution of waste in disposal environments.

The current classification system for radioactive wastes in the United States is an evolved descendent of the first waste classes. As pointed out by Kocher [1], early definitions of radioactive-waste classes in the United States were based on the physical form (liquid, solid), the heat generation rate, and the specific activity of the waste. These were matters of operational concern-handling and management of reprocessing wastes-not concern for the impacts of disposal on public health [1]. As the problems of casual or improper management of other radioactive wastes presented themselves, or as management plans changed, more classes were incrementally introduced [2].

The basis for newer classes was consistent with the first distinctions between high-level waste and "other than high-level" waste [3]. Some of the classes, namely high-level waste (HLW), spent nuclear fuel (SNF), and uranium and thorium mill tailings (UMT), are source-defined. That is, they are based entirely on the generating process or source of the waste, rather than on generic features of the waste itself. Transuranic waste (TRUW) is defined by exclusion of the source-defined classes and by elemental and concentration restrictions, whereas low-level waste (LLW) is defined solely by exclusion of the other waste classes and by jurisdiction. Waste managers use classes relevant to handling, namely "contact-handled material" and "remote-handled material."

Observers and critics point out that the categories have irrelevant consistency and relevant inconsistencies [1,2,4]. For example, low-level wastes share a feature of dubious importance, namely that they are not HLW, SNF, TRUW, or UMT. The wastes can differ in their surface dose rates, half-lives, radiotoxicity and disposal hazards by several orders of magnitude. These critics call for a waste classification scheme that reflects the relevant features of the wastes and the actual risks posed by disposal.

But what do we mean by relevant features and actual risks?

The purpose of a classification is to simplify management of the multiple elements of a diverse system (adapted from [6]). Classification of radioactive waste can be helpful at every stage between generation and disposal of the material. But a single system for classifying wastes that integrates all of the desired purposes has proven elusive. Berg and Brennecke list some factors for classification [5] (reproduced in Table I). This table illustrates that different activities require different considerations. This context dependence suggests that we might want different classification systems for different steps. (Some of the features tend to correllate, e.g. heat generation and surface dose rate, but others may not.)

The United States is not alone in constructing waste classification systems for waste management activities. A superficial examination would suggest that different nations have chosen very different features of radioactive waste as the basis for classification3/4 half-life in France [7], heat-generation rate in Germany3/4 but deeper exploration reveals many commonalities in the actual regulations and practices for disposal in different countries. The correllated features mentioned above underly some of the commonalities, and others are the result of safety assessments discussed later in this paper.

Other papers have proposed or explored alternative systems of classification [8,9,10]. Reference [10] is of particular interest because it suggests that for current waste streams, the characteristics used to divide classes do not change the classification if the divisions are chosen properly. But while these systems incorporate some of the relevant features discussed above, they still do not reflect actual risks of disposal. The most sophisticated waste classification currently utilized is the subclassification used by the Nuclear Regulatory Commission (NRC) for LLW.

The NRC limits for Class C LLW3/4 the most hazardous radioactive waste found suitable by the NRC for shallow-land burial3/4 are the products of scenario-based hazard calculations. The NRC established dose limits for exposure of the public from LLW-disposal facilities. The limits pertain to a set of scenarios including intruder scenarios and radionuclide-migration scenarios. To establish simple, uniform criteria for determination of waste class, the NRC worked backwards from the dose limits using the intruder scenarios to find activity-concentration limits for some of the most hazardous long-lived fission and activation products (for a description of these see Ref. [11]). The intruder scenarios establish volumetric concentration limits. These calculations were performed on a site-generic basis.

The NRC concluded that, unlike the intruder scenarios, the migration scenarios depended on the total inventory of a radionuclide (rather than the concentration) disposed at the site and would therefore require site-specific analysis. The company managing a LLW facility does performance assessments to establish the volume of waste that can be accepted given projections of the character of the waste stream. In rare cases, the total-inventory limits must be so restrictive that they impact the specific-activity limits (²²⁶Ra at Ward Valley). But generally, the facilities simply place greater waste-form requirements on waste containing some problem radionuclides, such as ³H, and limit the total disposable inventory of others. Any NRC-licensed LLW with concentrations of radionuclides that exceed the limits in 10 CFR 61.55, that is waste designated as greater-than-class-C (GTCC) waste, is the responsibility of the DOE [12] and must be disposed in a facility licensed by NRC.

The NRC thus recognized a context-dependence for the risks associated with disposal. Likewise, the Department of Energy (DOE), in response to environmental laws, has conducted performance assessments that establish the waste-acceptance criteria (WAC) for each of their disposal sites (see e.g., Reference [13]). As with the NRC methodology, DOE used scenarios and worked backward from dose limits to establish concentration and package limits for each radionuclide expected to be important. Unlike the NRC limits, the DOE WAC are site-specific. As mentioned above, other nations ostensibly have different approaches to waste classification, but Germany and France both perform assessments similar to those of the DOE to establish package or concentration limits at their waste-disposal facilities (see e.g. [14,15,16,17,18]).

The WAC can (practically, rather than legally) be applied to any waste, regardless of its origin, provided that the composition is known. The waste form might be factored in to the concentration limits as well, as they are in the German repositories (see [18]). The WAC should represent the best available estimates of the risks associated with disposal of waste at a particular site. The WAC, then, are a logical basis for a risk-based classification system for disposal of radioactive waste. And because waste generators must show that their waste meets the WAC in order to dispose of the waste at a particular site, the WAC already constitute a form of de facto waste classification.

To some, the particularism of the WAC may seem to run counter to an ideal goal of classification, which is to simplify management of wastes without knowledge of the final destination of the waste. The United States does not currently have disposal facilities that will accept all of its waste. How should we deal with those wastes? From a risk perspective we should do two things: use classes applicable for storage and handling until appropriate waste-disposal facilities are available, and construct facilities appropriate for disposal of the wastes.

It is important to examine the practical implications such changes have for waste handlers; we should consider the risks and other costs associated with assay of radioactive wastes. But, as mentioned above, waste generators already have a burden of assay due to the WAC. The concentrations of radionuclides in the waste can be determined by radioassay or by process calculations and verification, but the waste cannot exceed limits. Further, we would have to retain the classes associated with handling of waste (contact and remote) and those for transportation, because those are already hazard-based classes associated with particular contexts (site specificity is much less relevant for these contexts).

A logical step in pursuing this approach is to investigate the value of information provided by greater characterization of wastes. If costs and exposures were of no concern, one might institute a characterization protocol for all wastes that highlights or eliminates factors (such as criticality) from consideration and then examines highlighted factors in detail. There is a point past which further characterization is unproductive. Prudent choices can be made based on analysis of the propagation of uncertainty from waste characterization through assessed health risk.

Major legal, institutional, and political barriers resist the changes suggested in this paper. But the benefits of adopting this approach are in appropriate matching of wastes and disposal facilities. Due to differences in geochemistry a radionuclide of great concern in one location may be of modest concern in another. Thus, appropriate distribution of wastes among disposal facilities can result in greater assurance of public health as well as reduced costs.

We gratefully acknowledge the assistance and input of Jesse Yow, Larry Ramspott, Tom Isaacs, Jor-Shan Choi, and Jessica Booher for their thoughtful comments and suggestions and for providing information. This work was funded under the Nuclear Materials Stewardship Initiative at Lawrence Livermore National Laboratory and the Campus Laboratory Collaborations Program of the University Office of the President and was carried out at the Center for Nuclear and Toxic Waste Management at the University of California at Berkeley.

1. D.C. KOCHER, "Classification and Disposal of Radioactive Wastes3/4 History and Legal and Regulatory Requirements," Radiation Protection Management, 1990, 7(4), 58-78.
2. M.D. LOWENTHAL, "Radioactive Waste Classification in the United States: History and Current Predicaments" UCRL-CR-128127, Lawrence Livermore National Laboratory, (1997).
3. R.E. BERLIN and C.C. STANTON, Radioactive Waste Management, John Wiley & Sons, New York (1989): p28
4. MAKHIJANI and S. SALESKA, High-Level Dollars Low-Level Sense, IEER Press, Takoma Park, MD, (1992).
5. H.P. BERG and P. BRENNECKE, "Basic considerations on radioactive waste classification regarding the different waste management steps," Kerntechnik, 1993, 58(5).
6. International Atomic Energy Agency, "Classification of Radioactive Waste: A Safety Guide," Safety Series No. 111-G-1.1, Vienna, (1994).
7. Agence Nationale pour la Gestion des Déchets Radioactifs; "Centre de l'Aube" Fontenay-aux-Roses, France, (1993).
8. D.C. KOCHER and A.G. CROFF, "A Proposed Classification System for High-Level and Other Radioactive Wastes," ORNL/TM-10289, Oak Ridge National Laboratory, (1987).
9. M. ZUCCHETTI and P. ROCCO, "Classification of ITER Waste: the Recycling Option for ITER Materials," Technical Note No. I.94.166, Joint Research Centre, Italy. (1994).
10. C.F. SMITH and J.J. COHEN, "Development of a Comprehensive Waste Classification System," in Waste Management '89, Proceedings of the Symposium on Waste Management, Vol. 2, ed. R.G. Post. University of Arizona, Tucson, Arizona February 26-March 2, 1989
11. Office of Nuclear Materials Safety and Safeguards, "Draft Environmental Impact Statement on 10 CFR Part 61 "Licensing Requirements for Land Disposal of Radioactive Waste"," US Nuclear Regulatory Commission, NUREG-0782, vol.2-4, Washington, DC September 1981
12. "Responsibilities for disposal of low-level radioactive waste" 42 U.S.C. § 2021c (b) (1) (D)
13. Martin Marietta Energy Systems, Inc., EG&G Idaho, Inc., Westinghouse Savannah River Company, "Radiological Performance Assessment for the E-Area Vaults Disposal Facility (U)," Westinghouse Savannah River Company WSRC-RP-94-218, (1994).
14. PITIOT, "The French Integrated Waste Management System for LLW" Low Level Radioactive Waste Technical Seminar, April 23-25, 1996, Troyes, France; Institute of Nuclear Materials Management.
15. P.W. BRENNECKE and E.H. WARNECKE, "Preliminary Waste Acceptance Requirements3/4 Konrad Repository Project" Nuclear Energy 30(3), 1991: pp 173-181
16. W. NOACK, K. KUGEL and P. BRENNECKE, "The Morsleben Repository3/4 Waste Acceptance Requirements And Radioactive Wastes To Be Disposed Of" Proceedings of the Fifth International Conference on Radioactive Waste Management and Environmental Remediation, ICEM '95, Sept. 3-7, 1995 Berlin, Germany; ASME; NYC (1995): v2, pp 1285-1288
17. K. KUGEL, W. NOACK, C. BARD, H. GILLER, B.-R. MARTENS, and P. BRENNECKE, "Anforderungen an enzulagernde radioaktive Abfälle und Mab nahmen zur Produktkontrolle radioaktiver Abfälle, Endlager für radioaktive Abfälle Morsleben (ERAM), Teil 1: Endlagerungsbedingungen, Stand: September 1993" Bundesamt für Strahlenschutz, Salzgitter, September 1993; Bfs-ET-14/92-REV-2
18. P. BRENNECKE, "Anforderungen an endzulagernde radioaktive Abfälle (Endlagerungs-bedingungen, Stand: Dezember 1995)3/4 Schachtanlage Konrad" Bundesamt für Strahlenschutz, Salzgitter, Dezember 1995; ET-IB-79