Steven A. Eide, James L. Jones, and Thomas E. Wierman
Lockheed Martin Idaho Technologies Company
Idaho National Engineering and Environmental Laboratory
Idaho Falls, Idaho 83415

This paper summarizes the process and results of human health risk assessments of the U.S. Department of Energy (DOE) complex-wide programs for high-level waste, transuranic waste, low-level waste, mixed low-level waste, and spent nuclear fuel. The DOE baseline programs and alternatives for these five material types were characterized by disposition maps (material flow diagrams) and supporting information in the May 1997 report A Contractor Report to the Department of Energy on Environmental Baseline Programs and Integration Opportunities (Discussion Draft). Risk analyses were performed using the Simplified Risk Model (SRM), developed to support DOE Environmental Management Integration studies. The SRM risk analyses consistently and comprehensively cover the life cycle programs for the five material types, from initial storage through final disposition. Risk results are presented at several levels: DOE complex-wide, material type program, individual DOE sites, and DOE site activities. The detailed risk results are documented in the February 1998 report Human Health Risk Comparisons for Environmental Management Baseline Programs and Integration Opportunities (Discussion Draft).

In July 1996, the U.S. Department of Energy (DOE) Assistant Secretary for Environmental Management (EM) chartered a government contractor-led effort to develop a suite of technically defensible, integrated alternatives that meet the EM mission. The contractor teams were challenged to "think outside the box" for solutions that cross traditional site boundaries and enable the programs to get the job done at an earlier date and at a lower cost. The results of that effort are summarized in the May 1997 report A Contractor Report to the Department of Energy on Environmental Baseline Programs and Integration Opportunities (Discussion Draft). ¹ In that report are presented DOE complex-wide baseline programs and alternatives for six areas: high-level waste (HLW), transuranic waste (TRUW), low-level waste (LLW), mixed low-level waste (MLLW), spent nuclear fuel (SNF), and waste generated from environmental restoration activities (ER).

This paper summarizes a related effort, the development of simplified human health risk models for the baseline programs and alternatives. Details of the effort and risk results are documented in the February 1998 report Human Health Risk Comparisons for Environmental Management Baseline Programs and Integration Opportunities (Discussion Draft). ² The simplified risk models were developed to:

• provide a consistent and comprehensive, quantitative human health risk picture for all of the important activities involved in the baseline programs and alternatives
• evaluate the alternative program risks relative to the baseline program risks
• provide a simplified risk model that can be efficiently used to answer stakeholder questions about risk.

The technical basis for the SRM is documented in the September 1996 report A Simplified Method for Quantitative Assessment of the Relative Health and Safety Risk of Environmental Management Activities. ³ Use of the SRM is explained in the October 1996 report User's Guide for the Simplified Risk Model (SRM). ⁴ The SRM models in an approximate, quantitative manner the human health risk from radionuclide and chemical exposure from both accidents and normal, incident-free operation. Exposure pathways include airborne (inhalation) and groundwater (ingestion).

Risk is defined in DOE Order 5480.23, Nuclear Safety Analysis Reports as "... the quantitative or qualitative expression of possible loss that considers both the probability that a hazard will cause harm and the consequences of that event." ⁵ Types of risk that may be addressed in risk analyses include human health and safety, environmental impact, and programmatic impact. However, most risk assessments are focused on human health and safety. As shown in Table I, risk analyses can range from simple qualitative estimates obtained within hours or days to detailed, quantitative studies requiring years to complete. Risk analyses can be relative or absolute, with the absolute analyses resulting in human health risk estimates such as person-rem or cancer fatalities.

The SRM was developed to fulfill a need for a simple, absolute, and quantitative risk assessment methodology that could be used to generate risk estimates within weeks, as indicated by the shaded area in Table I. A quantitative risk assessment method was desired to maximize consistency in risk estimations covering a wide variety of material types, facilities and activities, and accident and normal exposure risks. Also, an absolute risk methodology was desired in order to compare results with more detailed risk assessments. Such comparisons are necessary to establish the validity of the SRM. Finally, because the EM integration studies cover a variety of options for handling material types across the entire DOE complex, a simplistic risk assessment methodology was needed to provide efficient and timely support.

The SRM analyzes human health and safety risk from normal, incident-free operations and accidents involving radionuclides and/or chemicals. Environmental impact and programmatic impact risk are not covered. Also, within the area of human health and safety, risk from standard industrial accidents (injuries and deaths) is not presently covered. Enhancements to the SRM are planned to cover standard industrial accidents and environmental impact.

SRM pathways for radionuclides and/or chemicals to affect human health and safety include airborne (inhalation) and groundwater (ingestion), as well as direct radiation from radionuclides. DOE site work force and surrounding population information is contained within the SRM, as well as representative DOE site airborne and groundwater dispersion characteristics.

As presently developed, the SRM is applicable to DOE waste management activities involving HLW, TRUW, LLW, MLLW, and SNF. Additional development would be required to comprehensively model ER or decontamination and decommissioning (D&D) activities.

The basic building block of the SRM is the risk matrix, shown in Table II. This matrix is filled out for each of the basic EM activities, such as storage, treatment, transportation, and disposal. The sample risk matrix shown in Table II is for a five-year storage of liquid HLW at the Idaho National Engineering and Environmental Laboratory (INEEL).

The basic equation for accident risk from both radionuclides and chemicals, upon which the matrix was developed, is the following:

where:

As indicated in the risk matrix in Table II, for accidents, the air and groundwater pathways are modeled separately. Also, radionuclides are divided into two categories: actinides and nonactinides. Chemicals are modeled by selecting several dominant chemicals and modeling each of those separately. (No chemicals have been included in the sample risk matrix in Table II.)

Normal, incident-free operational risk is estimated for two cases, as shown in Table II: airborne exposure to offgas or fugitive emissions (not modeled in the sample risk matrix), and direct radiation. The model for offgas of fugitive emissions is a variation of that used for accidents, with the assumption that a fraction of the curie or chemical inventory will escape into the atmosphere with the offgas. The model for direct radiation is based on the number of facility workers working directly with the hazardous materials and estimates of annual exposures for such work.

The risk results for each state, as indicated in Table II, are estimated separately for up to 35 distinct cases (combinations of actinide, nonactinide or chemical; pathway; normal or accident; and receptor). For example, the facility worker risk resulting from accidents releasing actinides into the atmosphere is 5.1E-1 person-rem. (The risk matrix in Table II has been calibrated to predict person-rem as the risk measure.) Similarly, the public risk is 8.4E-1 person-rem. The total risk for this state (last column in risk matrix) is 2.8 person-rem, comprised of 1.9 person-rem from accidents (actinide contribution), 1.6E-1 person-rem from accidents (nonactinide contribution), and 7.5E-1 person-rem from direct radiation exposure of workers resulting from periodic surveillances. Breaking down the risk another way (bottom row of risk matrix), 1.3 person-rem is incurred by facility workers, 5.5E-1 person-rem by DOE site workers, and 9.2E-1 person-rem by the public. This level of detail in modeling is necessary in order to cover all potentially important contributors to risk and to allow for risk results to be broken down into various types of categories (e.g., worker risk versus public risk or accident risk versus normal, incident-free operations risk).

The SRM has pre-defined look-up tables to assist the risk analyst in determining values for all of the elements in the risk matrix except for E_1a, the curie and/or chemical inventory, and E₆, the time duration (for storage and disposal states). However, the risk analyst must collect site-specific information for each state, such as the characteristics, size, and location (DOE site) of the facility or process and the form of the material being processed or stored. Then the analyst can select the appropriate values from the look-up tables.

A typical SRM analysis consists of seven steps:

1. information collection
2. development of a system flow diagram (disposition map)
3. breakdown of the system flow diagram into more basic SRM states (e.g., storage of specific materials, various treatment processes, onsite and offsite transportation, loading and unloading, and disposal)
4. characterization of curie and/or chemical flows through the system
5. development of SRM risk matrices for states
6. combining of individual risk matrix results into system risk results
7. comparison of SRM risk estimates with more detailed analyses (if available).

For the SRM analyses of HLW, TRUW, LLW, MLLW, and SNF, part of step 1 and all of step 2 had already been performed as part of the EM integration study. ¹

The SRM analysis process outlined above can take days to months to complete, depending upon the complexity of the system to be modeled, the difficulty in obtaining the necessary information (especially curie and/or chemical flow information), and the number of sites and states to be modeled. The SRM risk analyses covering DOE complex-wide programs for HLW, TRUW, LLW, MLLW, and SNF, each required approximately six weeks to complete.

DOE HLW is presently stored at four sites: Hanford, INEEL, Savannah River Site (SRS), and West Valley Demonstration Plant (WVDP). DOE plans to stabilize this waste, mostly liquid and sludge stored in underground tanks, and ultimately dispose of the stabilized HLW in a repository such as Yucca Mountain. The baseline program and alternative for HLW for the INEEL are outlined in the disposition maps presented in Figure 1. The baseline disposition map for the INEEL shows the process by which the stored HLW is retrieved, stabilized, and sent for disposal. The alternative involves additional separations activities, transport of HLW to other sites for processing or interim storage, schedule changes, and other activities.

The SRM HLW model was developed by subdividing the disposition map boxes into activities (risk states) such as storage, retrieval, loading and unloading, on-site and off-site transport, various types of waste treatment steps, and disposal. The subdivision of the disposition maps into activities is shown in Figure 1, where the numbers in parentheses refer to specific HLW activities. Figure 2 contains a description of each activity.

The SRM requires knowledge of the radionuclide curie and chemical flow through the disposition maps. The curie information was generally obtained from the report Integrated Data Base Report - 1995: U. S. Spent Nuclear Fuel and Radioactive Waste Inventories, Projections, and Characteristics. ⁶ Additional information required for the SRM, such as the waste form (liquid, sludge, grout, glass, etc.), characteristics of storage and processing facilities, types of on-site and off-site transport, and schedules were obtained from the site subject matter experts. In cases where such information was not available, assumptions were made.

SRM human health risk results (person-rem) for each of the HLW activities at the INEEL (shown in Figure 1) are presented in Figure 2. Shown in the figure are risk results for both the baseline program and the alternative. The activities modeled include all of the important steps in the baseline program (or alternative), from initial storage through final disposal. For the INEEL, the major activities with respect to human health risk involve various HLW treatment processes such as dissolution and separation, grouting, vitrification, and drying and packaging. For activities in Figure 2 that show risks from both the baseline and the alternative, risk differences are mainly the result of different fractions of the total HLW being covered by the activity. For example, the loading for off-site transportation (activity 11 in Figure 2) indicates almost double the risk for the baseline compared with the alternative. This results from the alternative plan to dispose of part of the waste on-site, rather than ship it off-site for disposal. Other differences shown in Figure 2 are typically the result of changes in treatment processes or treatment locations (e.g., vitrification being performed at Hanford rather than the INEEL in the alternative program).

SRM human health risk results (person-rem) for HLW, TRUW, LLW, MLLW, and SNF are summarized at the DOE complex level in Figure 3. As shown in the figure, TRUW has the highest risk (1.3E+4 person-rem), and SNF has the lowest risk (1.5E+3 person-rem). The total DOE complex risk is approximately 3.1E+4 person-rem for both the baseline programs and the alternatives. Therefore, the alternatives outlined in the May 1997 report A Contractor Report to the Department of Energy on Environmental Management Baseline Programs and Integration Opportunities (Discussion Draft) are considered to be risk neutral at the DOE complex-wide level with respect to human health risk from radionuclides and chemicals, compared with the baseline programs. However, at specific sites the risks may differ between baseline and alternative programs.

Changes in alternative programs that could result in significant risk differences between the baseline programs and the alternatives include the following: large changes in the total number of off-site shipment miles, using off-site rail shipping rather than truck shipping, and large changes in treatment volumes at sites with large off-site populations. However, the overall program must usually be analyzed to determine if potentially larger risks for some activities are offset by lower risks for other activities.

For transportation risk, the SRM results are most sensitive to the type of shipment (rail versus truck), the exposure (millirem per hour at one meter) from the shipping container, and the total number of shipping miles. For most other types of activities, the SRM results are sensitive to the waste form assumed (e.g., sludge versus glass), the total number of actinide curies present, the site population characteristics, and the assumed number of workers receiving direct radiation from normal operations.

Figure 4 presents the DOE complex-wide risk results by site for the baseline program. (The results for the alternative program are similar.) Included are 12 DOE sites, one "site" covering all off-site transportation, three disposal sites not located at the 12 DOE sites (TRUW, HLW/SNF, and LLW/MLLW), and a commercial treatment site. The highest risk for both the baseline and alternative programs is off-site transportation, with TRUW contributing approximately 75 % to the total off-site transportation risk. Among the DOE sites, SRS, Rocky Flats Environmental Technology Site (RFETS), Hanford, and the INEEL have the largest risks. Each site has a different composition of material type risks, with SRS dominated by HLW and RFETS dominated by TRUW. Disposal risk is a minor contributor to the overall results.

The SRM risk results for each of the five material types covered in this paper were compared with results from other, more detailed risk models. Comparisons were made with results from the DOE complex-wide May 1997 report Final Waste Management Programmatic Environmental Impact Statement ⁷, various other environmental impact statements, detailed risk assessments, and actual exposure information for various types of activities. In general, the comparisons indicate that the SRM risk predictions are reasonably accurate. In particular, the off-site transportation predictions generally agree to within 30 % with more detailed analyses, when comparable assumptions are made concerning the total number of trips, the trip lengths, amount of material carried by each truck or rail shipment, and the exposure rate at one meter from the shipping container(s).

SRM complex-wide models have been developed for HLW, TRUW, LLW, MLLW, and SNF to support Environmental Management Integration (EMI) activities. The SRM models provide consistent, comprehensive, and quantitative human health risk estimates for complex-wide programs for these five material types. Results indicate that at the DOE complex-wide level, the EMI alternative program human health risks are comparable to those for the baselines. As structured, the SRM models contain information on accident versus normal, incident-free operation risks, public versus worker/site risks, actinide versus nonactinide risks, and many other risk breakdowns. The models and model results can be used to efficiently answer a wide variety of potential questions that stakeholders may have. For example, DOE sites can use the SRM results to compare proposed alternative program risks with their baseline program risks. Because the SRM risk model includes five material types, trade-offs (e.g., higher TRUW risk but lower LLW risk) can be evaluated at the site level to determine how the overall site risk changes.

Several follow-on activities for the SRM are being considered. Ongoing work in the area of EMI involves the development of more detailed, up-to-date disposition maps covering the DOE complex-wide programs for HLW, TRUW, LLW, MLLW, and SNF. Future plans for the SRM may include the analysis of these more refined DOE complex-wide baseline programs. The risk results could then be used to determine the risk reduction per year as the waste management (and SNF) programs progress. Future SRM work may also include the development of companion modules to estimate standard industrial risks (injuries and deaths) and environmental impacts. These other types of risks could then be presented in formats similar to the human health risks from radionuclides and chemicals discussed in this paper. The SRM could be expanded to include other types of material types such as highly-enriched uranium or plutonium and other types of facilities such as experimental reactors found at DOE sites. Analyzing risks from these other material types and facilities in a manner consistent with the material types already considered would result in a more comprehensive risk picture for DOE sites and the entire DOE complex. Finally, the SRM might also be appropriate for scoping risk analyses associated with future waste management EAs and EISs. In such cases, the SRM could help risk analysts focus their more refined studies in the areas of highest risk.

1. U.S. DOE Complex-Wide EM Integration Team, "A Contractor Report to the Department of Energy on Environmental Management Baseline Programs and Integration Opportunities (Discussion Draft)," INEL-EXT-97-00493, Idaho National Engineering and Environmental Laboratory (May 1997).
2. S. A. Eide, J. L. Jones, and T. E. Wierman, "Human Health Risk Comparisons for Environmental Management Baseline Programs and Integration Opportunities (Discussion Draft)," INEEL-97-01253, Idaho National Engineering and Environmental Laboratory (February 1998).
3. S. A. Eide et al., "A Simplified Method for Quantitative Assessment of the Relative Health and Safety Risk of Environmental Management Activities," INEL-95/0645, Idaho National Engineering and Environmental Laboratory (September 1996).
4. R. G. Peatross and S. A. Eide, "User's Guide for the Simplified Risk Model (SRM)," INEL-96/0397, Idaho National Engineering and Environmental Laboratory (October 1996).
5. DOE Order 5480.23, "Nuclear Safety Analysis Reports," U.S. Department of Energy (April 10, 1992).
6. "Integrated Data Base Report -- 1995: U.S. Spent Nuclear Fuel and Radioactive Waste Inventories, Projections, and Characteristics," DOE/RW-0006, Rev.12, Oak Ridge National Laboratory (December 1996).
7. "Final Waste Management Programmatic Environmental Impact Statement," DOE/EIS-0200-F, U.S. Department of Energy (May 1997).

BACK