Environmental Impact Statement and Environmental Impact Report for Continued Operation of Lawrence

D.2 ACCIDENTS WITH POTENTIAL RELEASE OF RADIOACTIVE MATERIAL

LLNL and SNL, Livermore use radioactive materials in a wide variety of operations including scientific and weapons research and development, diagnostic research, research on the properties of materials, in isotope separation, and as calibration and irradiation sources. Radioactive materials are collected as waste products in forms varying from contaminated laboratory equipment and metal filings to contaminated trash and liquids. There is, therefore, a potential for releases of radioactive materials due to human error, failure or malfunctioning of equipment, accidents during the treatment, handling, or transportation of radioactive wastes, and severe natural events like earthquakes.

This section analyzes postulated accidents that could result in radioactive material releases. It describes how buildings and operations were selected for analysis, discusses the computer code that was used in the analysis as well as assumptions about weather conditions and atmospheric dispersion, presents the bounding scenarios, and estimates the potential health effects.

D.2.1 The Selection of Buildings and Operations for Accident Scenarios

The accidents involving radioactive material were divided into two groups: those that may occur in a specific building or operation and those that may occur during transportation. An accident involving multiple buildings was also postulated resulting from an earthquake (see section D.5). Table D.2-1 shows the process used for selecting buildings for radiological scenarios.

A total of 653 buildings were initially considered for accident scenarios involving radioactive material. Administrative buildings without hazardous materials were excluded. For the remainder, a facility hazard classification list for LLNL and a similar building list for SNL, Livermore (Fisher, 1989; SNL, Livermore, 1991a) were used to identify those with a significant potential for release (DOE Order 5481.1B). This hazard classification is as follows:

Low hazard: minor onsite and negligible offsite impacts on people or the environment.
Moderate hazard: considerable potential onsite impacts on people or the environment, but at most only minor offsite impacts.
High hazard: potential for onsite or offsite impacts on large numbers of people or for major impacts on the environment.

All LLNL Livermore site and SNL, Livermore buildings ranked as low hazard and those without radioactive materials were initially eliminated from consideration. This left 21 buildings with a high or moderate radiation hazard classification at the LLNL Livermore site and SNL, Livermore.

All LLNL and SNL, Livermore buildings that appear on the radionuclide inventory lists (Fisher, 1990; SNL, Livermore, 1991c) were reviewed to determine if the quantities, physical forms, and dispersion potentials corroborated the hazard classifications. This inventory review verified that low-hazard buildings contained only small quantities of radioactive materials that are in solid, not easily dispersible form such as sealed sources. It also identified the following buildings as not having a radiation hazard designation on the facility hazard classification lists: Buildings 194, 227, 412, and 910, all of which contain solid sealed sources with very small (millicurie) quantities, and 13 buildings at LLNL Site 300 that contain small sealed sources of tritium, cobalt, or curium. Buildings that contain only solid material or only small quantities of material were eliminated from consideration. This initial elimination and verification left 15 of the original 21 for further consideration.

For all the buildings, including some previously rejected, specific operations with the potential for radioactive material releases, potential initiating events, administrative limits on the permissible quantity of radioactive material; and material form, quantity, method of confinement and storage were reviewed (refer to Appendix A for building descriptions). One scenario was added based on this review (Building 625). If available, accident scenarios developed for safety analysis reports were reviewed, the buildings were visited, and discussions with personnel were held to frame the range of postulated scenarios for each building. Source terms were developed, starting with the largest radioactive-material administrative limit for an operation. Appropriate release fractions and reduction or removal factors were applied to calculate quantities of dispersible material. This source term was then compared to those for other scenarios for the same building to determine a scenario that would result in the highest radiation dose at the site boundary.

The process was completed after Laboratory and DOE personnel further reviewed the selected buildings to ensure all radioactive materials and operations using radioactive materials were considered. This review did not eliminate postulated scenarios; rather, the review provided verification of the selection process and ended with the addition of one scenario to bound possible accident releases from two additional buildings (Buildings 298 and 391). Table D.2-2 summarizes the final screening process starting with 15 buildings, adding Buildings 625 and 298 and dropping nine buildings from consideration.

Eight buildings were thus identified for nine scenarios involving radioactive material. These were Buildings 251, 331, and 332 at the LLNL Livermore site, which contain uranium, tritium, and transuranics including plutonium; Building 968 at SNL, Livermore, which has a large inventory of tritium; the Building 490 complex and Building 612 at the LLNL Livermore site; Building 298 which has a (proposed) small tritium laboratory; and Building 625, which is likely to be damaged by an earthquake. It should be noted that there are plans to reduce the tritium inventories for both Buildings 331 and 968 to accommodate programmatic changes. Currently Building 331, the Hydrogen Research Facility, has an administrative limit for tritium of 300 g and an inventory of less than 20 g. Under the proposed action, the administrative limit would be reduced from 300 g to 5 g and the inventory reduced accordingly. A portion of the tritium operations in Building 331 may be moved to Building 298, the Fusion Target Fabrication Facility, or to Building 391, the Nova-Upgrade/National Ignition Facility. In this event Building 331 and 298 or Buildings 331 and 391 would have a combined administrative limit of 10 g with no more than 5 g in any one building. Building 298 was selected for the accident scenario as described in section D.2.8.10.

Table D.2-1 Process for Selecting Buildings for Radiological Scenarios

Initial number of buildings: 653

1. Initial Elimination step:

Eliminate administrative buildings without hazardous material of any type.
Eliminate buildings with no radiation hazard and no radioactive materials.
Eliminate all low-hazard buildings.

Number of buildings remaining for consideration: 21

Eliminate all buildings with solid sealed-source radioactive materials.

Number of buildings remaining for consideration: 15

2. Verification step:

Using radioactive-material inventories (Fisher, 1990; SNL, Livermore, 1991c), verify elimination of low-hazard buildings with solid sealed-source radioactive materials and small quantities of radioactivity (mostly in the millicurie range).
Using the inventories, verify elimination of moderate-hazard buildings that contained no radioactive materials.

No buildings added or removed.

3. Selection for scenario development:

Consider radioactive-material type, quantity, and physical form.
Consider material confinement, use, and storage.

Added 1 building (Building 625) and eliminated 9 others.

Consider additional buildings identified by Laboratory and DOE personnel.

Two buildings added (Buildings 298 and 391).

Eliminate Building 391—Releases bounded by Building 298

Number of buildings selected for accident scenarios: 8 (Table D.2-2)

Source: Fisher, 1989; SNL, Livermore, 1991a.

Table D.2-2 LLNL and SNL, Livermore Buildings Considered for Radiological Accident Scenarios

Building	Name	Type of Research	Accident Scenario Modeled or Reason Not Modeled
233	Materials Management	Classified storage	Not modeled: small inventory of solid radionuclides
251	Heavy Element Facility	Heavy-element nuclear chemistry	Modeled: americium release
255	Calibration Laboratory	Radiation-instrument calibration	Not modeled: well-contained inventory of sealed sources
292	Rotating Target Neutron Source (RTNS) Facility	Physics	Not modeled: RTNS shut down, no radioactive material present other than contamination
321	Materials Fabrication Shop	Mechanical fabrication	Not modeled: small inventory of transient solid material
331	Hydrogen Research Facility	Tritium research	Modeled: tritium release
332	Plutonium Facility	Plutonium research	Modeled: inadvertent nuclear criticality Modeled: plutonium release
334	Hardened Engineering Test Building (HETB)	Intrinsic radiation and environmental testing	Not modeled: sealed sources, Building 332 plutonium-release scenario bounding
412	Environmental	Environmental science	Not modeled: small inventory of sealed sources
419	Size Reduction and Solidification Facility	None	Not modeled: not in use, small inventory of solid waste
490 Complex	Demonstration Facility	Laser isotope separation	Modeled: uranium fire in Building 493; other buildings in complex bounded by other scenarios.
513	Treatment and Container Storage	None	Not modeled: uranium turnings stored in liquid bounded by Building 612 scenario
514	Liquid Waste Disposal	None	Not modeled: potential for release of radioactive-filter cake waste bounded by Building 612 scenario
612	Waste Treatment and Storage	None	Modeled: transuranic-waste release
625	Container Storage	None	Modeled: transuranic-waste release initiated by earthquake
968	Tritium Research Laboratory (SNL, Livermore)	Tritium research	Modeled: tritium release
298	Fusion Target Fabrication Facility	Target development/ fabrication, Tritium research	Modeled: tritium release (proposed action)
391	NOVA-Upgrade/National Ignition Facility	Laser fusion irradiation	Not Modeled: tritium release is bounded by release for Building 298

D.2.2 Selection of Scenarios for Transportation Accidents

Two low-specific-activity (LSA) shipments of radioactive waste were modeled using multiple criteria including the maxium values allowed for a type A package (type A package limits are detailed in 49 C.F.R. Pt. 435). The first accident modeled was based on the waste stream from LLNL, which would have a mix of nuclides including some TRU. The second was based on the waste stream from SNL, Livermore, which would contain tritium oxide only. These scenarios are suitable for these activities specifically associated with the LLNL Livermore site, LLNL Site 300, and SNL, Livermore in terms of shipments of radioactive materials (nonwaste) and radioactive waste.

For the first LSA scenario, the maximum truck load (19,200 kg) and the total individual packages allowed per shipment (24) from a Hazardous Waste Management (HWM) procedure (LLNL, 1990b) were used to develop the source term. The details of this scenario are described in section D.2.8.9.1.

For the second LSA scenario, the maxium curie content allowed without highway route control (30,000 Ci of tritium oxide) was used to develop the source term. The details of this scenario are also described in section D.2.8.9.1.

A review of the prior studies done elsewhere on radioactive material (nonwaste) shipments is summarized in section D.2.8.9.2, below. From this it was concluded that the likelihood of a release from these transportation accidents would be low. Radioactive material shipments were not considered further.

The Waste Isolation Pilot Plant (WIPP) SEIS considered a bounding accident scenario for TRU offsite transport and is summarized in section D.2.8.9.3, below. As described in D.2.8.9, the TRU portion of the LLNL LSA source term was developed specifically for LLNL Hazardous Waste Management shipments; therefore, the WIPP dose conclusions cited below are not directly comparable to the LSA results later developed in section D.2.8.9. However, the LSA and WIPP results may be compared by the calculated chance that a member of the general public incurs a health effect. These calculated values are given in the transportation summary table in section D.2.9.

The potential radiological impact of the shipment of classified nuclear materials is assumed to be bounded by the DOE analysis of the proposed Special Isotope Separation Project (SIS), summarized below in section D.2.8.9.4. DOE concluded that the transportation risks would be less than 19 person-rem. The health risk associated with the shipment of classified waste is bounded by the LSA accident analysis shown below.

D.2.3 The Computer Code Used to Estimate Radiation Doses

The computer code chosen for computation of doses was GENII, Edition B, Version 1.473 (Napier et al., 1988), Environmental Radiation Dosimetry Suite. It was developed for DOE Field Office, Richland and is in the public domain. It incorporates the internal dosimetry models recommended by the International Commission on Radiological Protection. GENII is actually a coupled system of seven different programs that constitute the dosimetry system (Generation II) used at DOE Hanford reservation.

GENII estimates potential radiation doses to individuals or populations from routine or accidental releases of radionuclides to air or water and from residual contamination. GENII can account for direct exposures from water, soil, and air as well as exposures from inhalation and ingestion. The code can calculate annual doses, committed doses, and accumulated doses.

Before being selected, GENII was compared with two other codes used in accident analyses (Table D.2-3). The factors considered in GENII's selection were the code's ability to calculate radiation dose from a variety of exposure pathways, interactive menu-driven programs which help in input preparation, the code's history of use for a previous DOE environmental impact statement, and independence from codes developed at LLNL Livermore.

Table D.2-3 Accident Analysis Radiological Code Comparison

NAME OF CODE	GENII	MACCS	ARAC (MATHEW/ADPIC)
Developed by	Pacific Northwest Laboratory	Sandia National Laboratory	Lawrence Livermore National Laboratory
Funded by	Department of Energy (DOE)	Nuclear Regulatory Commission (NRC)	DOE, NRC, Department of Defense
Computers	IBM-PC	VAX	VAX
Past use	Endorsed by DOE for use in New Production Reactor Environmental Impact Statement (NPR-EIS) and Hanford Environmental Dosimetry Upgrade Project.	Endorsed by NRC for use in radiological analysis for NUREG-1150 reference plants: Zion, Surry, Sequoyah, Peach Bottom and Grand Gulf. NPR-EIS for severe accident analysis.	Used in preparing LLNL Safety Analysis, and real time analysis of exposures of accidental releases using actual meteorological conditions.
Application	Facility siting, EIS, and Safety Analysis Reports	Level 3 Probabilistic Risk Assessment and EIS	Emergency Planning, Safety Analysis Reports
Modeling Capability
Air Dispersion Model	Gaussian distribution	Gaussian distribution	Particle in cell
Terrain effect	No	No	Yes
Dosimetric model
Early exposure	Yes	Yes	Yes
Chronic exposure	Yes	Yes	Yes
Health Effect Model
Early injury or fatality	No	Yes	No
Latent cancer	No	Yes	No
Exposure Pathways
Air	Yes	Yes	Yes
Ground	Yes	Yes	Yes
Inhalation	Yes	Yes	Yes
Ingestion	Yes	Yes	Yes
Surface water	Yes	No	No

D.2.4 Meteorological Parameters

To model the atmospheric transport of released radioactive materials from the LLNL Livermore and SNL, Livermore sites, site-specific meteorological data (section 4.4) were reviewed to determine the meteorological conditions prevailing at the LLNL Livermore and SNL, Livermore sites. To make the analysis conservative, with results that represent the upper bound on consequences, stable meteorological conditions that give rise to minimal dispersion were assumed: the wind speed was set at 1 m per second; atmospheric stability was assumed to be Class F, or very stable; and the wind direction was constrained to the direction with the highest consequences for the general population. A wind direction of east to west was chosen for the population consequences, even though it is not the prevailing wind direction at the LLNL Livermore and SNL, Livermore sites. An eastern wind occurs about 4 percent of the time in a year, and the 1 m/sec wind speed occurs approximately 38 percent of the time in a year (section 4.4.3).

D.2.5 Dispersion Parameters

The radiation dose by exposure to an air concentration that has been transported to a specific point of interest from an airborne release depends on the type and quantities of radionuclides in the local environment. The atmospheric dilution depends on the altitude at which the material is released and on the environmental conditions traversed between the release point and the receptor. Atmospheric dispersion models can accurately describe the transport and turbulent diffusion of effluents released to the atmosphere for the selected meteorological conditions. The depletion of the plume as a result of deposition onto surfaces, washout by precipitation, and radioactive decay of the radionuclides can be routinely treated.

The selection of a dispersion model for a specific application is based on the intended use, the characteristics of the source, and available meteorological data. Simplified atmospheric models are appropriate for estimating potential doses because the releases are hypothetical, the receptors are not in the immediate vicinity of the release, and the evaluation is based on assumed meteorological conditions.

The GENII code uses the atmospheric-dispersion model of Pasquill, as modified by Gifford, to estimate the dispersion of the radioactive-material plume at various distances downwind from the point of release. This model is based on a Gaussian distribution of concentration in the lateral (y) and vertical (z) directions. The lateral (sy) and vertical (sz) dispersion parameters are themselves a function of the downwind distance (x).

The resulting doses from hypothetical accident scenarios modeled by GENII may differ from doses calculated by the ARAC models for similar scenarios analyzed in LLNL facility Safety Analysis Reports. It should be noted, however, that no actual comparative model calculations have been made. In general, for the same assumptions about radioactive material releases and meteorological conditions, GENII will probably calculate a higher dose that may be explained by the following model differences.

ARAC uses the ADPIC (Atmospheric Diffusion Particle-In-Cell) model, which is a numerical, three-dimensional particle dispersion model capable of simulating the time- and space-dependent distribution of air pollutants. ADPIC utilizes a spatial grid consisting of three-dimensional rectangular cells and performs grid averaging to calculate the average concentration of pollutants within grid cells. GENII utilizes straight-line Gaussian diffusion in its dispersion model.

Close to the release point, ADPIC's vertical diffusion coefficients are similar to those used in a standard Gaussian plume model. ADPIC's close-in horizontal diffusion coefficients, however, include the meandering effect of a plume for the F stability class usually assumed in accident scenarios. Plume meandering under very stable conditions is frequently observed and when it occurs it results in increased horizontal diffusion and reduced plume concentrations. Plume meander and the cell averaging employed in ADPIC may be able to account for a factor of 2 or 3 difference in the final dose calculated, compared to a standard straight-line Gaussian plume model with no plume meander.

Both APDIC and GENII account for particulate deposition from the plume to the soil surface; however, ARAC physically depletes the plume's concentration in proportion to the material lost to deposition, but GENII does not. GENII employs the plume deposition for predicting entry of the radionuclides into the ingestion and water pathways, but, in its present software configuration, does not numerically decrease the plume atmospheric concentration as a function of the downwind distance, x. The code user, however, may direct GENII to utilize a user-supplied dilution factor, C/Q, which does account for plume depletion. Initially, a correction for plume depletion was taken from the depletion fraction curve in Meteorology and Atomic Energy 1968 (Slade, 1968), where the deposition velocity for particulate nuclides is assumed to be 0.1 cm/sec. Subsequently, a mathematical procedure was implemented into a software module for ensuring numerical consistency and repeatability. The computer code was rigorously verified and validated against the data in the published literature (Slade, 1968). All plume depletion calculations performed externally to GENII utilized GENII's algorithms for the horizontal (sy) and vertical (sz) dispersion parameters.

In the Building 332 inadvertent criticality scenario, GENII accounted for the radioactive decay of released fission product gases with short half-lives compared to the plume transit time. In the Building 332 inadvertent criticality scenario, no correction was made to account for plume deposition. Regulatory guidance recommends against accounting for any type of plume depletion (NRC, 1979). However, GENII did account for the radioactive decay of the plume nuclides that occurs during plume transport.

Scenarios that postulated a fire assumed an elevated release with the atmospheric dispersion factor modeled as a line source to the height of the fire plume. For all other accident scenarios, the releases of radioactive material were assumed to be at ground level. The ground-level release assumption is conservative because the slower dispersion compared to elevated releases results in higher ground-level concentrations and, therefore, higher estimates of radiation exposures.

D.2.6 Dose Estimates

Because this analysis addresses accidents, doses are calculated for discrete releases of specific quantities of radioactive material. The GENII code is used to calculate the resulting radiation doses from the time-dependent spatial distribution of concentrations estimated for environmental media such as air and soil, or in foods (Napier et al., 1988). Once a source term is defined for a particular scenario and the atmospheric dispersion parameters are specified, the code calculates the environmental transfer, uptake, and human exposure. Individual doses were calculated at specific distances onsite, at the site boundaries, and away from the site for four exposure pathways: external exposure from the airborne plume, external exposure from radioactivity deposited on the ground, internal exposure from inhalation, and internal exposure from ingestion of terrestrial food and animal products at the western site boundary and beyond. Ingestion doses on site were not calculated because it can be reasonably expected that protective actions involving confiscating contaminated foodstuffs would follow an accident involving the release of radioactive materials. This action would preclude any individual onsite from ingesting more than a single meal of contaminated food products. The ingestion dose from a single meal was judged to be insignificant compared to doses estimated by GENII assuming 10 percent annual intake of contaminated foodstuffs.

Generic data from an NRC regulatory guide (Nuclear Regulatory Commission, 1977b) or the GENII code default parameters were used for food production and consumption rates. It was assumed that food for the population within the contaminated, primarily urban western sector was produced in that sector; however, consumption rates for contaminated terrestrial food and animal products were assumed to be 10 percent of the generic consumption rates. Since emergency response and appropriate protective actions would be recommended to minimize public exposure to contaminated foods, and since western Alameda County food production is much less than the assumed 10 percent consumption rate, the 10 percent assumption conservatively accounts for commercially and noncommercially grown products. This means that 10 percent of the yearly consumption of terrestrial food and animal products by the entire population in the western sector would be contaminated. Additional assumptions and factors used to calculate radiological doses are as follows:

Fraction of time spent in plume was 1.0 (used for inhalation intake and external plume exposure).
For onsite doses, the maximally exposed individual parameters were used; soil exposure time was 2 hours, which assumed evacuation from the site 2 hours after accident initiation; no ingestion was assumed; and the calculated doses were 50-year committed doses from plume inhalation.
For the nearest site boundary dose, the maximally exposed individual parameters were used, soil exposure time was 0.7 year (Nuclear Regulatory Commission, 1977b), no ingestion was assumed, and a 70-year committed dose was calculated from plume inhalation.
For the western boundary dose (sector with highest population), the maximally exposed individual parameters were used, soil exposure time was 0.7 year (Nuclear Regulatory Commission, 1977b), 10 percent ingestion was assumed, and a 70-year committed dose was calculated from 1 year of ingestion intake and plume inhalation.
For population exposures beyond the site boundary, the average individual parameters and options in GENII were used, soil exposure time was 0.5 year (Nuclear Regulatory Commission, 1977b), 10 percent ingestion was assumed, and a 70-year committed dose was calculated from 1 year of ingestion intake and plume inhalation.

The absorbed radiation dose from any type of ionizing radiation is the energy deposited per unit mass in a given material. Historically, this has been quantified by using the unit "rad," defined as 100 ergs per gram.

To quantify biological effects for a given energy and radiation type, the "dose equivalent" is used. The unit of dose equivalent is the rem. The rem is the product of the absorbed dose and the quality factor, Q, which characterizes the type of radiation. Beta and gamma radiations typically have a quality factor of 1 while alpha radiation has a quality factor of 20. Neutrons have a quality factor ranging from 2 to 11 depending on the energy of the neutron (DOE, 1990c). While the concept of applying a quality factor is valid for estimating internal effects, external effects of alpha and beta radiation are limited. Alpha radiation does not penetrate the dermal layer; beta radiation does not penetrate clothing. For both alpha and beta internal effects can be eliminated by breathing purified air, by avoiding ingestion of contaminated foodstuffs and water, and by protecting abrasions and cuts from contact with contaminated materials. However, the dose calculations assume no protective action is accomplished with respect to ingestion and inhalation.

GENII calculates a first-year dose resulting from 1 year of external and internal exposure and a committed dose resulting from 1 year of external and internal exposure plus the internal dose resulting from 1 year of intake. The committed dose is calculated for 50 years for radiation workers. For the general population, the committed dose is calculated for 70 years, which approximates a human lifetime.

As recommended by the International Commission on Radiological Protection (International Commission on Radiological Protection, 1977, 1979–1982), GENII uses weighting factors for various body organs to calculate an "effective dose equivalent" (EDE). Doses are calculated for the lungs, stomach, small intestine, upper large intestine, lower large intestine, bone surface, red bone marrow, testes, ovaries, muscle, thyroid, bladder, kidneys, and liver. The effective dose equivalent is the summation of the dose equivalents to specific organs weighted by the relative risk to that organ compared to an equivalent whole-body dose. For accident scenarios, the doses calculated include effective dose equivalents from external exposure to the plume and ground surface contamination for 1 year, and a "committed effective dose equivalent" (CEDE) resulting from the internal exposure pathways of plume inhalation and food ingestion for 1 year of intake. For convenience, the sum of the effective dose equivalent from external pathways and the committed effective dose equivalent from internal pathways is called the committed effective dose equivalent in this EIS/EIR. Population doses are calculated by using this 70-year CEDE at specific distances to the west of the LLNL Livermore site and SNL, Livermore. The western sector was chosen because it contains the largest number of people. The dose is then multiplied by the number of people in that particular sector. Sectors radiate in pie-shaped wedges from a central point at the LLNL Livermore site, and SNL, Livermore.

The protective action guides that have been established by the Environmental Protection Agency and the Food and Drug Administration for human exposures to radioactive material are discussed in section D.2.7. The health effects that can be expected from the estimated doses are discussed in section D.2.9.

D.2.7 Active and Passive Mitigations

Below are discussed mitigations to exposure and therefore mitigations to dose that would effect the postulated results of the accident scenarios. In general, no mitigation was assumed for emergency response while less than nominal mitigation was assumed for passive filtration when ventilation and building containment were shown by analysis to survive the postulated accident initiator.

D.2.7.1 Emergency Response and Protective Actions

The radiation doses estimated here for the various accident scenarios are the doses that would be received by the population if no protective actions were taken. However, both LLNL and SNL, Livermore have detailed plans for responding to accidents of the type described here, and the response activities would be closely coordinated with those of Alameda County (see Appendix J). Both LLNL and SNL, Livermore personnel are trained and drilled in the protective actions to be taken if a release of radioactive or otherwise toxic material occurs.

For the offsite population, the need for any protective action would be based on the predicted radiation doses. The emergency response would be based on the guidance provided in the protective action guides developed by the U.S. Environmental Protection Agency (1990). Generally, no protective action is required when the projected doses are less than 1 rem to the whole body or less than 5 rem to the thyroid, but radiation levels should be monitored, and an advisory to seek shelter may be issued. For whole-body doses of 1 to less than 5 rem and thyroid doses of 5 to less than 25 rem, the public should be warned to seek shelter, access to the contaminated area should be controlled, and evacuation should be considered unless constraints make it impractical. For whole-body and thyroid doses higher than 5 and 25 rem, respectively, evacuation would be mandatory, access to the contaminated area would be controlled and radiation levels would be monitored. As shown in Table D.2-4, higher doses are allowed for emergency workers.

The underlying principle for the protective action guides is that under emergency conditions all reasonable measures should be taken to minimize the radiation exposure of the general public and emergency workers. In the absence of significant constraints, protective actions may be implemented when projected doses are lower than the ranges given in the protective action guides.

The Food and Drug Administration (FDA) also recommends protective actions to protect the public health from food contamination resulting from radiation incidents (FDA, 1982). Preventive protective action guides (PAG) are projected dose commitments of 1.5 rem to the thyroid, or 0.5 rem to the whole body, bone marrow, or any other organ for which actions should be taken to prevent or reduce the radioactive contamination of human food or animal feeds. Emergency protective action guides are projected dose commitments of 15 rem to the thyroid or 5 rem to the whole body, bone marrow, or any other organ. At this level, responsible officials should isolate food containing radioactivity to prevent its introduction into commerce and determine whether condemnation or another disposition is appropriate.

If emergency response and protective actions were to be considered effective, then the offsite predicted doses reported for tritium releases would be reduced by a factor of 4 and inadvertent criticality doses would be reduced by a factor of 2.

Table D.2-4 Recommended Protective Actions to Reduce Whole Body and Thyroid

Dose from Exposure to a Gaseous Plume

Projected Dose (rem)	Recommended Actions^a	Comments
To the Population
Whole Body<1 Thyroid<5	No planned protective actions^b; the State may issue an advisory to seek shelter and await further instructions. Monitor environmental radiation levels.	Previously recommended protective action may be reconsidered or terminated.
Whole body1 to <5 Thyroid5 to <25	Seek shelter as a minimum. Consider evacuation. Evacuate unless constraints make it impractical. Monitor environmental radiation levels. Control access.	If constraints exist, special consideration should be given for evacuation of children and pregnant women.
Whole body5 and above Thyroid25 and above	Conduct evacuation. Monitor environmental radiation levels and adjust area for mandatory evacuation based on these levels. Control access.	Seeking shelter would be an alternative if evacuation were not immediately possible.
To Emergency Team Workers
Whole body25 Thyroid125	Control exposure of emergency team members to these levels except for lifesaving missions. (Appropriate controls for emergency workers include time limitations, respirators, and stable iodine.)	Although respirators and stable iodine should be used where effective to control dose to emergency team workers, thyroid dose may not be a limiting factor for lifesaving missions.
Whole body 75	Control exposure of emergency team members performing lifesaving mission to this level. (Control of time of exposure will be most effective.)

^aThese actions are recommended for planning purposes. Protective action decisions at the time of the incident must take existing conditions into consideration.
^bAt the time of the incident, officials may implement low-impact protective actions in keeping with the principle of maintaining radiation exposures as low as reasonably achievable.
Source: EPA, 1990.

D.2.7.2 High Efficiency Particulate Filtration

In all areas where plutonium can be handled and can exist in a dispersible form, high efficiency particulate air (HEPA) filters provide a final barrier against the inadvertent release of radioactive aerosols into the outside environment. However, these filters would not trap volatile fission products such as the noble gases and iodine; such gases will be released into the outside environment.

HEPA filter efficiencies are 99.99 percent or greater with the minimum efficiency of 99.97 percent for .3 micron particles, the size most easily passed by the filter. To maximize containment of plutonium particles and provide redundancy, two HEPA filters in series are used. A "housekeeping" filter is placed ahead of these two filters to mitigate dust loading and any potential for chemical assault on the HEPA filters. For purposes of calculating the release fraction of plutonium that would be released to the atmosphere for the postulated accidents in Building 332, degraded filter efficiencies of 99.9 percent for the first filter and 99.8 percent for the second filter as provided in A Guide to Radiological Accident Considerations for Siting and Design of DOE Non-Reactor Nuclear Facilities (Elder, 1986) are used. None of the accident scenarios approach or exceed the design parameters assumed in the referenced guide. LLNL experience also supports the use of the referenced efficiencies. Actual data from an analysis of HEPA filter replacement records covering the last 10 years in Building 332, show that none of the filters used to prevent a potential for release of plutonium to the atmosphere have degraded to the overall efficiencies assumed for the accident scenarios.

D.2.8 Description of Accident Scenarios

Following the screening process that reduced the initial buildings considered from 653 to nine buildings, evaluation of potential bounding accident scenarios including such relative elements as meteorological parameters, dispersion parameters, dose estimates, and emergency response and protective actions, a total of ten bounding accident scenarios involving nuclear materials were developed. Those scenarios are described below.

D.2.8.1 Inadvertent Nuclear Criticality in Building 332, Plutonium Facility, LLNL Livermore Site

The bounding radiological accident scenario for individual LLNL and SNL, Livermore facilities is an inadvertent nuclear criticality yielding 10¹⁸ fissions within Building 332 Plutonium Facility. In a review of 41 recorded inadvertent criticalities that have occurred nationally, 10 have estimated yields of 10¹⁸ or greater fissions (Stratton and Smith, 1989). Of these 10, 3 occurred in aqueous processing plants, 2 occurred in heavy-water-natural-uranium systems, 4 occurred in water moderated reactors or reactor prototypes, and 1 occurred in an aircraft engine prototype reactor. All involved a combination of physical parameters and/or processes that are found in reactor or nuclear materials processing facilities but are not present in the research and plutonium metals fabrication activities in Building 332. No criticality accidents have occurred in DOE plutonium metals fabrication facilities. Previous safety analyses have analyzed the consequences of an inadvertent criticality yielding 10¹⁸ fissions without postulating a method of initiation. The estimated frequency of occurrence is less than 1×10^-6/year and, in fact, such an event may not be possible under the operational conditions and procedures existing in Building 332. However, despite the extremely low probability of occurrence, the consequences of this accident have been analyzed with the initiator left undefined.

This scenario is based on a review of the Safety Analysis Report for Building 332 (LLNL, 1990c), the Facility Safety Procedure (LLNL, 1989c), discussions with personnel, and guidance for evaluating the potential consequences of inadvertent criticalities in nonreactor facilities (Elder et al., 1986; Nuclear Regulatory Commission, 1979). An inadvertent criticality would produce the highest first-year effective dose equivalent at the site boundary and is therefore the bounding single-building scenario for radioactive materials.

Development of Scenario and Assumptions About the Radioactive Source Term

The Plutonium Facility allows operations utilizing plutonium, uranium, and other fissile materials individually and in combinations when all conditions specified in the Safety Analysis Report, the Facility Safety Procedure, and other applicable controlling documents are satisfied. Operations involving uranium require the preparation of an Operational Safety Procedure, and no fissile material operations are allowed that can produce a calculated fission yield greater than 10¹⁸ fissions. This limitation applies to all operations and to all fissile materials (Garcia, 1991).

An inadvertent nuclear criticality would result in an uncontrolled release of fission energy. It could be caused by the inadvertent assembly of a critical mass of plutonium or by the introduction of a moderating material such as water, to a subcritical mass of plutonium. The inadvertent criticality was postulated to occur in a glovebox through the addition of water to a dispersible quantity of plutonium in an appropriate geometric configuration. The criticality was postulated to yield 10¹⁸ fissions, with the nuclear reaction terminating as the water evaporated. The fission products produced by the reaction could breach the glovebox and be released into the room.

To prevent such accidents, a criticality-control system has been developed for Building 332 based on the double-contingency principle. This requires two independent and unlikely failures or errors to occur before an accident is possible. Since the postulated scenario requires multiple breakdowns in nuclear criticality safety controls, no specific accident initiating event was postulated for this accident. It was instead just assumed that an inadvertent criticality occurs, with a fission yield of 10¹⁸ based on Building 332 operations involving dispersible plutonium mixtures. This fission yield represents a maximum credible yield for a Building 332 criticality accident (Garcia, 1991).

The radioactivity that would be released by a criticality accident can be estimated from Table 1 of NRC Regulatory Guide 3.35 (Nuclear Regulatory Commission, 1979), which lists the important radionuclides released in a criticality accident with a fission yield of 10¹9. Since the fission yield for the accident postulated here is 10 times lower (i.e., 10¹⁸), the total curie content given in the cited table was reduced by a factor of 10 for all nuclides and further reduced by a factor of 4 for the iodines to allow for the depletion discussed in the regulatory guide.

This scenario assumes that the building ventilation would remain operable. No mechanical failures in the high efficiency particulate air (HEPA) filters or electrical failures in building systems would occur coincident with the criticality due to the fission energy being dissipated locally in the glovebox and the filters being located in the building loft. The air in the room is exhausted through a high efficiency particulate air filter installed in the room, but no credit in the analysis was taken for this filter. Then the air is exhausted through two-stage HEPA filters in another part of the building to the environment. These filters were assumed to have efficiencies of 99.9 and 99.8 percent (Elder, 1986). The assumption about filter efficiency is conservative because the nominal efficiency for these filters is 99.97 percent. The filters would not trap volatile fission products of the criticality (noble gases and iodine); these gases would be released to the environment. The HEPA filters would, however, collect most of the plutonium and fission-product particulates released by the accident. (For a 10¹⁸ fission yield, 0.1 mg of plutonium dioxide would be released into the room before filtration.) After both the reduction and removal factors are applied, the source term for a 10¹⁸ fission yield would be as shown in Table D.2-5. The dispersible plutonium released into the room is essentially contained by the high efficiency particulate air filters and does not significantly contribute to the offsite radiation dose delivered by this accident.

Computer Modeling and Results

The GENII code was run with the activities given in Table D.2-5 and the atmospheric-dilution factors (?C/Q) calculated for given radial distances from the Building 332 release point for a ground-level release. A ground-level release was assumed because the Building 332 stacks are not more than two and one-half times the height of the building (NRC, 1979). For this scenario, the GENII atmospheric model was used to account for the radioactive decay of released fission product gases with short half-lives compared to plume transit time. Depletion of the plume by dry deposition was not assumed.

The results are given in Table D.2-6. The first three entries in Table D.2-6 give the results for onsite exposure 0.1 km from the release point, the nearest LLNL Livermore site boundary (0.4 km to the south), and the western boundary (0.9 km from the release point). The other entries represent increasing radial distances from the point of release. Table D.2-6 gives the committed effective dose equivalents for the inhalation and ingestion internal exposure pathways. The table also includes effective dose equivalents (EDE) for the external exposure pathways of immersion in the passing radioactive plume and exposure to ground surface contamination.

The calculated effective dose equivalent of 1.9 rem from the external pathways of air immersion and ground surface contamination (Table D.2-6) at the nearest site boundary (0.4 km to the south) fall within the whole-body dose range (1 to 5 rem) at which some protective action is recommended by the EPA (see section D.2.7). The thyroid dose of 3.9×10^-2 rem, from radioiodines, is below the 5 rem limit at which protective action is recommended by the EPA and below the 1.5 rem limit at which protective action is recommended by the FDA (EPA, 1990; FDA, 1982). The ingestion pathway dose at the western boundary is well below the preventive protective action guide of 0.5 rem for whole body and organs other than the thyroid at which protective action is recommended by the FDA (see section D.2.7).

The radiation doses received by personnel in the immediate vicinity of the accident during the first minute of the criticality were also calculated using equations given in Regulatory Guide 3.35 (Nuclear Regulatory Commission, 1979) for calculating prompt (i.e., without measurable delay) gamma and prompt neutron doses. The prompt gamma dose is calculated from the following equation:

Dg = 2.1×10^-20 Nd-2e-3.4d

whereDg is the gamma dose in rem, N is the number of fissions, and d is the distance from the source in kilometers.

The prompt neutron dose is calculated from

Dn = 7×10^-20 Nd-2 e-5.2d

whereDn is the neutron dose in rem and the other terms are as defined above.

For a fission yield of 10¹⁸ without shielding or personnel evacuation, the prompt gamma and prompt neutron doses at 10 m are calculated to be

Dg = (2.1×10×-20)(1×10¹⁸)(0.01)-2e-3.4(0.01) = 205 rem

Dn = (7×10^-20)(1×10¹⁸)(0.01)-2e-5.2(0.01) =665 rem

The combined prompt external dose for unshielded personnel at a distance of 10 m would be 870 rem, which is likely to be lethal. The doses received by other people in the room and in the building would depend on distance and any shielding that would be provided by equipment, furniture, or features of the building. Immediate evacuation of personnel would reduce the prompt dose received by those not in the immediate vicinity of the inadvertent criticality. An assessment of the potential fatalities is discussed in section D.2.9. The combined prompt external dose at the nearest and the western site boundaries would be 0.088 and 0.002 rem, respectively, without shielding.

If a criticality did occur, it would be considered an emergency situation at the highest level (level 4, disaster), requiring response by the Crisis Manager and the appropriate safety and emergency-response organizations (see Appendix J) in addition to high-level management.

The total projected doses to the population in the western sector from LLNL Livermore site and SNL, Livermore at preselected distances from the release point are shown in Table D.2-7. The western sector was identified as the sector with the greatest population from the 1990 Census (Appendix C).

The estimated doses for the offsite population are below the levels at which protective action is recommended. Furthermore, since the analysis uses conservative assumptions to maximize the dose, the actual doses would be lower. The health effects predicted for this bounding release event are discussed in section D.2.9.

Preventative Measures

The prevention of nuclear criticality is one of the principal safety considerations at LLNL. In addition to engineered safety features such as workstation separation and the use of safe geometric configurations, administrative controls and safety procedures are strictly enforced. These controls and procedures include limits on the quantities of fissile material that is present at any workstation, verification that a critical mass is not likely to occur, and glovebox cleanup procedures. The Criticality Safety Group of the Health and Safety Division of the Hazards Control Department advises Laboratory personnel on fissile materials that may present a criticality concern. Mitigating features such as criticality alarms, emergency-response procedures and drills, emergency exits, the building ventilation system, and the shielding provided by building construction would all help reduce doses to personnel outside the room and building.

Table D.2-5 Source Term for an Inadvertent Criticality

Nuclide	Total Activity (Ci)
Krypton-83m	11.0
Krypton-85m	7.1
Krypton-85	0.000081
Krypton-87	43.0
Krypton-88	23.0
Krypton-89	1,300
Xenon-131m	0.01
Xenon-133m	0.22
Xenon-133	2.7
Xenon-135m	330
Xenon-135	41.0
Xenon-137	4,900
Xenon-138	1,100
Iodine-131	0.275
Iodine-132	30.0
Iodine-133	4.0
Iodine-134	107.5
Iodine-135	11.25

Table D.2-6 Calculated Individual Doses for an Inadvertent Criticality Accident in Building 332

Distance from Building 332 (km)	CEDE (rem)		EDE (rem)
Distance from Building 332 (km)	Inhalation	Ingestion	Air Immersion	Ground Surface
0.1 (Onsite)	7.8×10^-1	N/C	30*	1.7×10^-4
0.4 (Nearest boundary)	8.2×10^-2	N/C	1.9	4.8×10^-2
0.9 (Western boundary)	2.2×10^-2	2.8×10^-2	3.2×10^-1	1.2×10^-2
1.6	9.5×10^-3	2.8×10^-3	1.1×10^-1	3.3×10^-3
2.4	4.8×10^-3	1.5×10^-3	4.3×10^-2	1.6×10^-3
4.0	2.0×10^-3	7.4×10^-4	1.4×10^-2	7.0×10^-4
5.6	1.1×10^-3	4.7×10^-4	5.5×10^-3	3.7×10^-4
7.2	6.9×10^-4	3.4×10^-4	2.7×10^-3	2.4×10^-4
12.0	2.9×10^-4	1.8×10^-4	5.6×10^-4	9.2×10^-5
24.0	9.8×10^-5	7.6×10^-5	7.9×10^-5	2.5×10^-5
40.0	4.6×10^-5	4.5×10^-5	2.1×10^-5	1.0×10^-5
56.0	2.9×10^-5	3.2×10^-5	8.7×10^-6	5.4×10^-6
72.0	2.0×10^-5	2.3×10^-5	4.3×10^-6	3.4×10^-6
80.0	1.7×10^-5	2.1×10^-5	3.1×10^-6	2.7×10^-6

CEDE = Committed Effective Dose Equivalent.
EDE = Effective Dose Equivalent.
N/C = Not calculated. Protective actions would limit ingestion onsite. See section D.2.6.
* Does not include prompt dose of 6 rem from criticality event.

Table D.2-7 Collective Population Dose for an Inadvertent Criticality in Building 332

Distance (km)	Population Western Sector*	Individual CEDE (rem)	Collective Population Dose (person-rem)
2.4	2,676	5.1×10^-2	136.5
4.0	3,982	1.7×10^-2	67.7
5.6	5,786	7.4×10^-3	42.8
7.2	10,068	4.0×10^-3	40.3
12	26,776	1.1×10^-3	29.5
24	81,772	2.8×10^-4	22.9
40	305,746	1.2×10^-4	36.7
56	436,096	7.5×10^-5	32.7
72	544,684	5.1×10^-5	27.8
Total: 1,417,586		Total: 436.9

CEDE = Committed Effective Dose Equivalent.
* Source: Educational Data Systems, 1991.

D.2.8.2 Tritium Release from Building 968, Tritium Research Laboratory, SNL, Livermore

An accidental release of tritium gas was identified as a reasonably foreseeable scenario for the Tritium Research Laboratory at SNL, Livermore. Potential for releases was identified for the radioactive-waste accumulation area and the uranium hydride handling system. Releases were also postulated for an accident involving a primary container and an earthquake.

These potential releases were identified from a review of the Tritium Research Laboratory Safety Analysis Report (SNL, Livermore, 1981), the report of the Technical Review Committee on the Tritium Research Laboratory Incident of August 18, 1987 (Bauer, 1987), information provided in Appendix I of this EIS/EIR, and a building walkdown. The release from an earthquake was estimated to provide the highest doses and was therefore analyzed. Other potential releases were eliminated from further consideration for reasons summarized below.

The quantities of radioactive waste that accumulate in the Tritium Research Laboratory (Building 968) are much lower than those present at the waste facilities of the LLNL Livermore site (Appendix B). Therefore, since a waste release from the LLNL Livermore site would be bounding, this scenario was eliminated.

The purpose of the uranium hydride storage beds is to store tritium in the uranium. Heat is required to release the tritium followed by failure of the containment system and a gas-purification system for an accident to occur. Moreover, the amount of tritium released would be bounded by the release of tritium gas from the failure of a primary container.

Human error would be the most probable initiator of a release of tritium gas from a primary container, but such a release would require removal of the primary container from its secondary container or glovebox. Procedural controls prevent the opening of a primary container outside a glovebox. The administrative limit per container is 50 g of tritium, but operational experience indicates that 25 g is seldom at risk; therefore, any postulated scenario is expected to be more severe than an actual release. If tritium gas were released from a primary container it would be discharged through the building ventilation system, resulting in an elevated release and consequently greater dispersion than that of a ground-level release. Since the earthquake scenario described below postulates a ground-level release of the same quantity of tritium, a release from a primary container was not analyzed further.

Development of Scenario and Assumptions About the Radioactive Source Term

During an earthquake with a ground acceleration of 0.8g (i.e., an earthquake so severe it is likely to occur only once in 5000 years), the primary container inside a glovebox is assumed to fail. It is also assumed that the emergency power system fails because the control panel necessary to start the emergency diesel generator is not resistant to an earthquake of this magnitude. Emergency power failure results in ventilation system failure. Debris is assumed to strike the glovebox, breaching its integrity and allowing a release of tritium gas to the environment through earthquake-generated cracks in the building's structure. Fire is not postulated as the tritium gas is being released due to the lack of an ignition source and the quick dilution of the gas into the room volume and subsequent release to the environment. Fire in the building from sources other than tritium is not postulated because of the installation of seismic shutoff valves throughout the natural gas pipeline system and the limited amounts of combustible and flammable material in the building.

Administrative controls limit the maximum amount of tritium allowed at the SNL, Livermore site to 50 g (SNL, Livermore, 1991d). The accident scenario assumes that the entire building inventory of tritium is out of the Tritium Research Laboratory vault and involved in experiments at the time of the earthquake.

The present inventory in the Tritium Research Laboratory, however, is about 26 g (SNL, Livermore, 1992). This amount is undergoing a reduction in order to prepare for the Tritium Research Laboratory's decontamination and decommissioning (D & D). After FY 1993, the tritium at risk would be limited to residual contamination. Secondly, a large fraction of the current inventory would typically be stored in the Tritium Research Laboratory vault, which is designed to withstand an 0.8g earthquake. Therefore, the amount of tritium that is at risk of release in any accident in the Tritium Research Laboratory is usually much less than 50 g.

For the purposes of this EIS/EIR, however, it was assumed that the administrative limit of 50 g of gaseous tritium was at risk. Additionally, it was assumed that 1 percent would oxidize to tritium oxide or tritiated water after escaping from the building (NCRP, 1979). This results in 0.50 g, or 4825 curies, of tritium oxide vapor as the source term. Tritiated water behaves like ordinary water in the body and can be absorbed through the skin. Tritiated water diffuses rapidly with body water and is at least 25,000 times as hazardous as tritium gas (SNL, Livermore, 1990c).

Appendix C reported that about 55 percent of the chronic release of tritium is in the form of tritium oxide vapor. This may be attributed to the tritium oxide vapor released from systems in which the tritium is already in the vapor form. Tritium used by experimenters is all in a dry gaseous form (SNL, Livermore, 1991b).

Computer Modeling and Results

A ground-level release of this quantity of tritium, with no calculated plume depletion by decay or dry deposition, would result in the doses shown in Table D.2-8 when external exposure from the passing plume, internal exposure from inhalation, and internal exposure from food ingestion are considered. The estimated collective doses received by the population in the western sector from SNL, Livermore are given in Table D.2-9.

The calculated CEDE of 0.37 rem at the nearest site boundary (400 m to the east) is significantly lower than the whole-body dose range (1 to 5 rem) at which the EPA recommends protective action (EPA, 1990) for accident releases. This 0.37 rem dose is essentially all internal exposure from inhalation of tritium vapor. There is no dose contribution from ground surface contamination, and negligible contribution from air immersion, the external exposure pathways. At the western boundary (850 m from Building 968), the dose of 0.92 rem is primarily from ingestion and greater than the whole body preventive PAG (0.5 rem) at which the FDA recommends protective action to prevent or reduce the radioactive contamination of human food or animal feeds. Health effects are discussed in section D.2.9.

The specific mechanism by which elemental tritium becomes oxidized in the environment is known to be first, the deposition of gaseous elemental tritium (HT or T2) into the ground where soil microorganisms catalyze/metabolize it into tritiated water (HTO or T2O), then the subsequent emission of tritiated vapor (resuspension) into the air. Direct oxidation of elemental tritium in the atmosphere is a very slow process and has been shown by experiments to be negligible compared to the soil conversion process (Fusion Technology, 1988). However, the magnitude of the resuspension coefficient (which is used to predict the conversion of HT into HTO) and its dependence upon soil parameters is still being investigated by the scientific community (Fusion Technology, 1988).

Although the tritium source term in these accident scenarios assumes that 1 percent of the elemental tritium ultimately becomes oxidized to HTO upon release, there is enough quantitative experimental data on the resuspension coefficient when coupled with the application of Gaussian Plume Models (Gulden and Raeder, 1988; Taschner et al., 1991) to assure that this 1 percent conversion is a conservative assumption. Future scientific investigation is expected to lead to a better understanding of the details of tritium interaction with the soil and is expected to lead to a further reduction of this 1 percent estimate.

Once the tritiated vapor is resuspended, its transport downwind can be well described by validated models for atmospheric dispersion of plumes (Slade, 1968; Hanna et al., 1982).

Preventative Measures

In this postulated accident scenario, most of the features and controls that would prevent or mitigate a tritium release from the Tritium Research Laboratory were defeated. They include physical features like systems for preventing overtemperatures and overpressures and gloveboxes for secondary containment; procedures such as periodic checks for leaks in the primary containment systems; and safe work permits for all maintenance operations. Systems and administrative controls that would mitigate exposure to onsite personnel from a tritium release include monitoring and alarm systems and emergency response (building evacuation) upon multiple tritium alarms. Finally, while the entire building inventory of tritium is assumed at risk, normal practice is to store tritium in the vault when not specifically required. Although a 50 g release is foreseeable, it is not likely based on demonstrated non-administrative constraints when conducting concurrent operations and experiments.

D.2.8.3 Plutonium Release from Building 332, Plutonium Facility, LLNL Livermore Site

Research and development activities in the Plutonium Facility, Building 332, are varied and involve a number of different processes. Potential accidents caused by natural phenomena, human error, and equipment failure were analyzed in the Safety Analysis Report (LLNL, 1990c). Of these, an accident was postulated to establish the greatest consequences from a release of plutonium into a laboratory in Building 332. This release event assumed a high concentration of aerosolized plutonium oxide in the largest Building 332 laboratory saturating the air to the maximum reported equilibrium concentration (Elder et al., 1986). This approach provides an upper limit for plutonium concentrations released to the room. The room ventilation system then exhausts the plutonium-laden air to the environment through a bank of dual-stage high efficiency particulate air filters.

Because of the wide range of operations and processing environments, particular initiating events were not identified for each plutonium activity. The postulated accident does not have a specific initiating event either; it was assumed that the primary confinement barriers to the release of plutonium (a container or a glovebox) were breached. Two mitigating safety features, the building confinement system and the ventilation exhaust system, which are common to all building operations, were assumed to remain operational after the release of plutonium into the laboratory. Thus, though no credit is taken for the glovebox filter or the room filter, the release path to the environment is through two high efficiency particulate air filters in series.

Several other potential accidents were analyzed in the Draft Safety Analysis Report (LLNL, 1990c). They included accidents initiated by a design-basis earthquake, a design-basis wind or design-basis tornado, a design-basis fire, spills of radioactive material, a hydrogen explosion, an inadvertent nuclear criticality, and equipment failures. None of these accidents, however, would result in a source term as large as that for the hypothetical release analyzed here: 60 g of plutonium in an airborne respirable aerosol form. This maximum initial source term scenario incorporates the assumption that an accidental initiator of unknown origin saturates the volume of the largest laboratory (Room 10¹⁰) to a nominal bounding value for respirable plutonium oxide saturation in air. Event-specific source terms were derived for each accident in the Draft Safety Analysis Report from the physical state and quantity of plutonium at risk, and each plutonium release was found to be of lesser consequence than the scenario postulated here.

The seismic analyses documented in Appendix I found that no significant damage to the structure is expected for the postulated seismic event. Further, those analyses determined that failures of equipment (such as gloveboxes, overhead piping, or ducts) in ways that would cause release of radioactive materials were also unlikely to occur. Some localized failures of the fire protection system's threaded piping could be postulated, resulting primarily in water leakage. However, these localized failures do not represent a significant potential for multiple laboratory room glovebox ruptures causing multiple releases in excess of the quantities assumed to be released in this scenario. If an isolated failure of some unspecified component did, in fact, occur during a seismic event in such a way as to impact and break open a glovebox, the administrative limit (20 kg) for a single laboratory would be at risk of release. Therefore, the maximum quantity of material at risk in a seismic event is limited to 20 kg for an individual unspecified component failure. Because the seismic analysis did not identify any significant potential failures, multiple failures releasing more than 60 g of plutonium in an airborne respirable aerosol form are not reasonably foreseeable.

The Plutonium Facility is designed to withstand the winds generated by the design-basis wind or design-basis tornado and the associated wind-driven missiles. In addition, no release would be expected because all plutonium-handling operations are terminated during a severe wind situation.

A fire in a Building 332 laboratory with a maximum fuel loading would be limited to a single laboratory, even without other mitigation, because of the fire-resistive construction of the interior and exterior walls. The quantity of plutonium released from a solid form or ignited from a more dispersible form (e.g., cutting chips, fines, or powder) would be very small. It is not reasonably foreseeable to have a fire involving the entire material at risk in a laboratory or a fire which could significantly degrade the building's containment.

A spill of radioactive material, which could result from dropping an improperly closed can containing plutonium oxide powder during removal from a glovebox, would result in the release of less than 1 g of material. All other spill scenarios would not exceed this powder spill.

A hydrogen explosion is unlikely, as shown by a detailed fault-tree analysis (LLNL, 1990b); a combination of four or more independent equipment failures and human errors would be needed for such an explosion. The explosion scenario was not further evaluated as it is not reasonably foreseeable.

The inadvertent plutonium criticality described in section D.2.8.1 releases noble gases and iodines that deliver offsite doses. However, experimental data show that for plutonium releases in the laboratory, the 60 g scenario described here represents an upper limit. The building leakpath containment would remain functional as no pressure or temperature transients could be postulated. Therefore the HEPA filters were also assumed to remain functional thereby mitigating the release of plutonium. Because the offsite doses delivered by the plutonium criticality event are greater than those for the 60 g source term, the criticality event is bounding for the Plutonium Facility; however, the 60 g source term bounds plutonium source terms within the building.

Random equipment failures can potentially contribute to the release of plutonium, but analysis has shown that they would not release plutonium in and of themselves. For example, the failure of a glovebox exhaust fan and ventilation fire dampers or a leak in the exhaust-ventilation ductwork would not result in a release because potential release paths to the environment are automatically isolated and the exhaust passes through high efficiency particulate air filters. If a glovebox exhaust fails to close, by design both fans would trip on interlock. Glovebox air would then be stagnated and contained by the room's controlled ventilation. If the ventilation fire dampers fail, the dampers in the supply system fail closed, and room exhaust dampers fail open, allowing room air to continue to exhaust through the high efficiency filters. Similarly, leaks in the exhaust ventilation ductwork would not pose a problem because all high efficiency particulate air filters are upstream of the exhaust fans. If a leak is upstream of a high efficiency particulate air filter, air is drawn into the leak toward the filter. If a leak is downstream of the fans and air filters, clean air would be exhausted.

Development of Scenario and Assumptions About the Radioactive Source Term

It was assumed that the air in the largest laboratory in Building 332, which has a volume of about 600 m3, is filled with a plutonium concentration of 100 mg/m3 of airborne particle size less than 10 mm (Elder et al., 1986), resulting in an initial source term of 60 g.

Two high efficiency particulate air filters in series, with assumed efficiencies of 99.9 percent and 99.8 percent (Elder et al., 1986), would trap much of the plutonium released into the room, and the quantity of plutonium released to the environment would be 0.00012 g. (The assumption about filter efficiency is conservative because the normal efficiency for these filters is a minimum of 99.97 percent.) Fuel-grade plutonium was assumed in developing the source term because it delivers a higher potential radiation dose than weapons-grade plutonium. Assuming that the 0.00012 g released has the composition and the specific activities shown in Table D.2-10, a source term by isotope was calculated.

Computer Modeling and Results

Table D.2-11 summarizes the calculated doses from this postulated plutonium release fromground level accounting for plume depletion by dry deposition. The first three entries in the table correspond to an onsite dose 100 m from the release point, the dose at the nearest LLNL Livermore site boundary (0.4 km to the south), and the dose at the western site boundary (0.9 km to the west). The other entries are for various radial distances from the point of release.

The committed effective dose equivalent of 6.6×10^-3 rem at the nearest site boundary (Table D.2-11) is essentially all from inhalation and much lower than the whole body dose range (1 to 5 rem) at which the EPA recommends protective action (EPA, 1990) for accident releases. There is no significant contribution to the dose from the external exposure pathways of air immersion and ground surface contamination. The ingestion pathway dose at the western boundary is well below the preventive protection action guide of 0.5 rem for the whole body and organs other than the thyroid at which protective action (FDA, 1982) is recommended by the FDA. The estimated collective dose received by the population in the western sector of the LLNL Livermore site is shown in Table D.2-12. Health effects are discussed in section D.2.9.

Preventative Measures

Because of engineered safety features, such as the use of primary and secondary containment, administrative limits on the quantities of plutonium, various safety procedures, and other controls, the postulated release is not likely. Building confinement, ventilation, and emergency power are all seismically qualified and would survive the design basis earthquake. High efficiency particulate air filters are fully protected from fire. Thus, these safety features were considered to remain operational during the accident, with only degraded performance by the filters. If a release did occur, its consequences would be mitigated by emergency response actions designed to protect both workers and the public.

Table D.2-8 Calculated Individual Doses for the Tritium Release from Building 968

Distance from Building 968 (km)	CEDE (rem)		EDE (rem)
Distance from Building 968 (km)	Inhalation	Ingestion	Air Immersion	Ground Surface
0.1 (Onsite)	4.2	N/C	3.4×10^-9	0
0.4 (Nearest boundary)	3.7×10^-1	N/C	3.1×10^-10	0
0.85 (Western boundary)	1.0×10^-1	8.2×10^-1	8.7×10^-11	0
1.6	4.1×10^-2	1.4×10^-1	3.3×10^-11	0
2.4	2.2×10^-2	7.4×10^-2	1.8×10^-11	0
4.0	1.1×10^-2	3.6×10^-2	8.7×10^-12	0
5.6	6.8×10^-3	2.3×10^-2	5.6×10^-12	0
7.2	4.9×10^-3	1.6×10^-2	4.0×10^-12	0
12.0	2.6×10^-3	8.6×10^-3	2.2×10^-12	0
24.0	1.1×10^-3	3.9×10^-3	9.4×10^-13	0
40.0	6.6×10^-4	2.3×10^-3	5.4×10^-13	0
56.0	4.8×10^-4	1.6×10^-3	3.9×10^-13	0
72.0	3.6×10^-4	1.2×10^-3	3.0×10^-13	0
80.0	3.2×10^-4	1.1×10^-3	2.7×10^-13	0

CEDE = Committed Effective Dose Equivalent.
EDE = Effective Dose Equivalent.
N/C = Not calculated. Protective actions would limit ingestion onsite. See section D.2.6.

Table D.2-9 Collective Population Dose for the Tritium Release from Building 968

Distance (km)	Population Western Sector*	Individual CEDE (rem)	Collective Population Dose (person-rem)
2.4	2,676	9.6×10^-2	256.9
4.0	3,982	4.7×10^-2	187.2
5.6	5,786	3.0×10^-2	173.6
7.2	10,068	2.2×10^-2	221.5
12	26,776	1.1×10^-2	294.5
24	81,772	5.0×10^-3	408.9
40	305,746	2.9×10^-3	886.7
56	436,096	2.1×10^-3	915.8
72	544,684	1.6×10^-3	871.5
Total: 1,417,586		Total: 4,216.6

CEDE = Committed Effective Dose Equivalent.
* Source: Educational Data Systems, 1991.

Table D.2-10 Source Term for the Fuel Grade Plutonium Release from Building 332

Radionuclide	Concentration (wt%)	Specific Activity (Ci/gm)	Weight (gm)	Activity (Ci)
Pu-238	0.10	1.7×10¹	1.2×10^-7	2.04×10^-6
Pu-239	78.0	6.2×10^-2	9.36×10^-5	5.8×10^-6
Pu-240	18.0	2.3×10^-1	2.16×10^-5	4.97×10^-6
Pu-241	1.6	1.0×10²	1.92×10^-6	1.92×10^-4
Pu-242	0.49	3.9×10^-3	5.88×10^-7	2.29×10^-9
Am-241	1.9	3.4	2.28×10^-6	7.75×10^-6
Total:			1.20×10^-4	2.13×10^-4

Table D.2-11 Calculated Individual Doses for the Plutonium Release from Building 332

Distance from Building 332 (km)	CEDE (rem)		EDE (rem)
Distance from Building 332 (km)	Inhalation	Ingestion	Air Immersion	Ground Surface
0.1 (Onsite)	7.8×10^-2	N/C	4.9×10^-10	5.5×10^-12
0.4 (Nearest boundary)	6.6×10^-3	N/C	4.2×10^-11	1.4×10^-9
0.9 (Western boundary)	1.6×10^-3	6.3×10^-5	9.9×10^-12	3.4×10^-10
1.6	6.5×10^-4	8.8×10^-6	4.2×10^-12	9.9×10^-11
2.4	3.4×10^-4	4.6×10^-6	2.1×10^-12	5.2×10^-11
4.0	1.6×10^-4	2.1×10^-6	9.9×10^-13	2.4×10^-11
5.6	9.5×10^-5	1.3×10^-6	6.1×10^-13	1.5×10^-11
7.2	6.6×10^-5	9.1×10^-7	4.2×10^-13	1.0×10^-11
12.0	3.2×10^-5	4.4×10^-7	2.0×10^-13	5.0×10^-12
24.0	1.3×10^-5	1.7×10^-7	7.8×10^-14	1.9×10^-12
40.0	6.5×10^-6	8.8×10^-8	4.2×10^-14	9.9×10^-13
56.0	4.3×10^-6	5.9×10^-8	2.8×10^-14	6.7×10^-13
72.0	3.1×10^-6	4.2×10^-8	2.0×10^-14	4.8×10^-13
80.0	2.7×10^-6	3.6×10^-8	1.7×10^-14	4.1×10^-13

CEDE = Committed Effective Dose Equivalent.
EDE = Effective Dose Equivalent.
N/C = Not calculated. Protective actions would limit ingestion onsite. See section D.2.6.

Table D.2-12 Collective Population Dose for the Plutonium Release from Building 332

Distance (km)	Population Western Sector*	Individual CEDE (rem)	Collective Population Dose (person-rem)
2.4	2,676	3.4×10^-4	0.9
4.0	3,982	1.6×10^-4	0.6
5.6	5,786	9.6×10^-5	0.6
7.2	10,068	6.7×10^-5	0.7
12	26,776	3.2×10^-5	0.9
24	81,772	1.3×10^-5	1.1
40	305,746	6.6×10^-6	2.0
56	436,096	4.4×10^-6	1.9
72	544,684	3.1×10^-6	1.7
Total: 1,417,586		Total: 10.4

CEDE = Committed Effective Dose Equivalent.
* Source: Educational Data Systems, 1991.

Table D.2-13 Calculated Individual Doses for the Tritium Release from Building 331

Distance from Building 331 (km)	CEDE (rem)		EDE (rem)
Distance from Building 331 (km)	Inhalation	Ingestion	Air Immersion	Ground Surface
0.1 (Onsite)	2.8×10^-1	N/C	2.5×10^-10	0
0.4 (Nearest boundary)	2.6×10^-2	N/C	2.1×10^-11	0
0.9 (Western boundary)	6.7×10^-3	5.3×10^-2	5.4×10^-12	0
1.6	2.8×10^-3	9.6×10^-3	2.3×10^-12	0
2.4	1.5×10^-3	5.1×10^-3	1.2×10^-12	0
4.0	7.5×10^-4	2.5×10^-3	6.1×10^-13	0
5.6	4.7×10^-4	1.6×10^-3	3.9×10^-13	0
7.2	3.3×10^-4	1.2×10^-3	2.8×10^-13	0
12.0	1.8×10^-4	6.1×10^-4	1.5×10^-13	0
24.0	7.9×10^-5	2.6×10^-4	6.5×10^-14	0
40.0	4.6×10^-5	1.5×10^-4	3.9×10^-14	0
56.0	3.3×10^-5	1.1×10^-4	2.8×10^-14	0
72.0	2.5×10^-5	8.4×10^-5	2.1×10^-14	0
80.0	2.3×10^-5	7.7×10^-5	1.9×10^-14	0

CEDE = Committed Effective Dose Equivalent.
EDE = Effective Dose Equivalent.
N/C = Not calculated. Protective actions would limit ingestion onsite. See section D.2.6.

D.2.8.4 Tritium Release from Building 331, Hydrogen Research Facility, LLNL Livermore Site

One of the projects in the proposed action is establishing a Hydrogen Research Facility in Building 331. The release of tritium during a large, beyond design basis earthquake (peak ground acceleration of 0.8g) during tritium operations was selected as the bounding scenario for Building 331. The Draft EIS/EIR evaluated an accident scenario involving 3.5 g of tritium at risk. This information was presented in the Draft EIS/EIR section 5.6 and the Executive Summary, but the supporting calculations in this appendix were for a 2.0 g at risk scenario. The information revised here is to be consistent with section 5.6 and the Executive Summary.

During a 0.8 g earthquake, the Hydrogen Research Facility could lose all electrical power because its electrical and emergency power systems have not been constructed to current seismic design criteria (see Appendix I). Furthermore, the gloveboxes are not seismically qualified and could also be damaged during an earthquake. As a result, the ventilation system would be inoperable and undiluted tritium could be released at ground level.

Two other accident scenarios considered but not developed for Building 331 were a fire/explosion in the Hydrogen Research Facility, and procedural error when removing a container from a glovebox. The fire scenario was eliminated because the following conditions would have to exist simultaneously for a fire to cause a release of tritium: (1) the presence of a large amount of fuel, (2) the failure of people to respond, (3) the failure of a fire alarm to sound or the LLNL Fire Department to respond, and (4) a failure of the fire protection system. That all of these conditions would occur at once is extremely unlikely. Additionally, design features such as inert argon environments in gloveboxes, glovebox oxygen monitors, and tritium containments, along with administrative controls such as flammable gas cylinder controls, hazardous gas handling and process operations procedures all ensure that a tritium mixture will not reach the lower explosive limit of 4 percent hydrogen concentration in air by volume. Therefore, a hydrogen fire/explosion was not considered for analysis.

The procedural-error scenario involved an individual removing a container of tritium gas from a secondary containment unit and opening it. Such procedure violations have occurred at other facilities. Two factors limit this accident scenario: (1) more than one procedural violation is required, and (2) the impact on and off the site would be very low, as normal operations call for only small fractions of the tritium-gas inventory to be in each container. The operation of the ventilation system would also reduce the impact of this release. The system is designed to dilute the gas and release it 30 m above the ground, thus allowing the gas to disperse before reaching the ground; therefore, the procedural-error scenario was eliminated.

Development of Scenarios and Assumptions About the Radioactive Source Term

Five grams of elemental tritium was the total proposed tritium inventory for the Hydrogen Research Facility. At any given time, a portion of this inventory would be stored in uranium hydride beds and traps, while the tritium gas would be stored in containers with strict limits on quantity (1 to 3.5 g). For this scenario, the release of 3.5 g of tritium gas was evaluated; however, the Building 331 tritium inventory may be reduced to accommodate programmatic changes. Currently Building 331, the Hydrogen Research Facility, has an administrative limit for tritium of 300 g and an inventory of less than 20 g. Under the proposed action, the administrative limit would be reduced from 300 g to 5 g and the inventory reduced accordingly. A portion of the tritium operations in Building 331 may be moved to Building 298, the Fusion Target Fabrication Facility, or to Building 391, the Nova- Upgrade/National Ignition Facility. In this event, the three buildings would have a combined administrative limit of 10 g with no more than 5 g in any one building. For these facilities, the administrative limit would therefore be reduced from 300 g in one facility (Building 331) to 10 g total in three facilities (Buildings 298, 331, and 391).

Administrative restrictions now limit operations to procedures that affect one primary container at a time.

It was assumed that an earthquake with a peak ground acceleration of 0.8g occurred while a laboratory technician was opening or transferring the contents of a primary container holding 3.5 g of tritium gas. All electrical power including emergency power is lost, shutting down the ventilation system. The glovebox is breached, allowing 3.5 g of tritium gas to enter the room; because the roof in the room was assumed to be damaged by the earthquake, tritium gas is released into the environment. It was assumed that all of this tritium gas (3.5 g or 33,775 Ci) is released and that 1 percent, or 338 Ci, was oxidized into tritiated water (NCRP, 1979). The ventilation system is assumed to be inoperable, causing the tritium to be released at ground level instead of through the 30-m stack.

Appendix C reported that about 55 percent of the chronic release of tritium is in the form of tritium oxide vapor. This may be attributed to the tritium oxide vapor released from vapor recovery systems in which the tritium is already in the vapor form. Tritium from the primary containers is in a dry gaseous form (Toy, 1991).

Computer Modeling and Results

Doses from external exposure to the passing plume, and internal exposures from inhalation and food ingestion, were calculated without assuming plume depletion for a ground-level release and presented in Table D.2-13. Exposure to an atmosphere containing tritium results in intake of that material by absorption through intact skin and by inhalation. To account for skin absorption, GENII assumes that the combined total rate of tritium intake in air is 150 percent of the inhalation intake rate alone. The onsite dose, set at 100 m from the release point, was estimated at 0.28 rem, which is a small fraction of the current annual limit for occupational exposures of 5 rem/yr (DOE, 1990c). The calculated CEDE of 0.026 rem at the nearest site boundary (400 m to the south) is significantly lower than the whole body dose range (1 to 5 rem) at which the EPA recommends protective action (EPA, 1990) for accident releases. This 0.026 rem dose is essentially all internal exposure from inhalation of tritium vapor. There is no dose contribution from ground surface contamination and negligible contribution from air immersion, the external exposure pathways. At the western boundary (900 m from Building 331), the dose of 0.060 rem is primarily from ingestion and less than the whole body preventive PAG (0.5 rem) at which the FDA recommends protective action to prevent or reduce the radioactive contamination of human food or animal feeds (FDA, 1982).

The tritium source term calculations did not quantify the specific contribution from resuspension originating from the deposition of gaseous tritium onto the ground, subsequent oxidation in the soil followed by emission of tritiated vapor. For further discussion see section D.2.8.2.

The estimated collective doses received by the population in the western sector from the LLNL Livermore site are shown in Table D.2-14. Health effects are discussed in section D.2.9.

Preventative Measures

Engineered safety features (e.g., multiple containment barriers, systems for recovering glovebox effluents, and inerting of gloveboxes), administrative limits, and safety procedures will reduce the probability of a tritium release from the laboratory.

D.2.8.5 Americium Release from Building 251, Heavy Element Facility, LLNL Livermore Site

A release of 5 curies of americium-241 after an earthquake was selected as the bounding accident scenario for Building 251, the Heavy Element Facility. The following accidents were considered but not analyzed further: a material-handling accident in the facility's hardened area, a material-handling accident in the unhardened area, a material-handling accident in a glovebox, a chemically induced explosion or fire in a glovebox, and an inadvertent nuclear criticality. An inadvertent criticality was eliminated because the quantities of fissile material currently allowed in the Heavy Element Facility preclude such an event. Future operations involving quantities of fissile material that have the potential for an inadvertent criticality would have to be analyzed before being allowed into the facility. Of the other accidents considered and eliminated, the most probable are spills inside gloveboxes or spills during the transport of material. Both accidents would be mitigated by the double containment of material. Scenarios initiated by an explosion or fire from an inadvertent mixing of chemicals or gases during an experiment are less likely. They would be mitigated by procedures that restrict the use of explosives, chemicals, or gases; by the 5-Ci maximum limit on the quantity of alpha-emitting radionuclides allowed in the unhardened area; and by the use of double high efficiency particulate air filters in the ventilation system. A spill followed by a fire outside a glovebox in the unhardened area would have an offsite dose consequence similar to the earthquake scenario only if the ventilation system also failed. These human-error-induced scenarios would be bounded by the release of radioactive material initiated by a severe earthquake.

The Heavy Element Facility is divided into two distinct areas. The area in the northwest portion (Rooms 1027, 1035, 1047, 1048, and 1052) called "the hardened area," is designed to withstand a 0.8g earthquake. The remainder of the facility, referred to as "the unhardened area," may not withstand a 0.5g earthquake. The facility handbook (LLNL, 1990a) states widely differing limits for actinide alpha-emitting radionuclides in these areas: 5000 curies for the hardened area, and 5 curies of alpha and neutron emitters and fissile material for the unhardened area. The actinide materials are packaged and stored in approved stainless-steel containers. These containers provide secondary or tertiary containment and are stored in specially designed, locked concrete pits equipped with alarms. The storage containers are also used when 5 or more curies of transuranic material is transported in the facility.

In the unhardened area, a 0.8g earthquake could create, at most, a 5-Ci spill of actinide material in a glovebox. If the unhardened area is damaged during such an earthquake, and if falling debris breaches the glovebox, the 5 Ci would be exposed directly to the environment. In this scenario the ventilation and air-recirculation system are rendered inoperable (either by loss of power or by damage), and since an unimpeded release path is open to the outside, a ground-level release would occur.

Development of Scenario and Assumptions About the Radioactive Source Term

The number of nuclides used in Building 251 varies and is never limited to one isotope. Of the actinide materials used in Building 251, americium-241 was selected as the radionuclide with the highest potential for offsite impact. The maximum amount of americium-241 allowed to be out of storage and in a glovebox in the unhardened area is 5 Ci. The actual amount of radioactive material at risk varies but is seldom above 3 Ci, and is distributed among many gloveboxes. The chemical form of the americium-241 was assumed to be metal oxide; in this form it may be possible to have all of the material staged within a single glovebox in preparation for chemical dissolution for further processing. This arrangement would require all other at-risk materials to be removed from the unhardened area to comply with the 5 Ci unhardened area limit. Experiments involving uranium oxide (Mishima and Schwendiman, 1973) indicate that, at most, 0.015 percent of this oxide powder could be dispersed in air during an earthquake. In this scenario, the ventilation system would be rendered inoperable, and the shell of the building would fail, allowing most, if not all, of the americium-241 aerosol to be released into the environment. The scenario was modeled as a ground release, and the source term was calculated to be 0.00075 Ci of americium-241.

Computer Modeling and Results

Table D.2-15 summarizes the doses calculated for this ground-level release with plume depletion from external exposure to the passing plume, external exposure from ground deposition, and internal exposures from inhalation and food ingestion. The first entry in the table shows the onsite doses at 0.1 km distance; the 0.6 km distance is the nearest LLNL Livermore site boundary and the closest to the nearest population center to the west. The other table entries are for radial distances out to 80 km from the LLNL Livermore site.

The committed effective dose equivalent of 0.14 rem at the nearest site boundary is essentially all from inhalation and is much less than the whole body dose range (1 to 5 rem) at which the EPA recommends protective action for accidents (EPA, 1990). There is no significant dose contribution from the external exposure pathways of air immersion and ground surface contamination. The ingestion pathway dose at the western boundary is well below the preventive PAG of 0.5 rem (FDA, 1982) for whole body and organs other than the thyroid at which protective action is recommended by the FDA. The total projected collective population dose received in the sector to the west of the LLNL Livermore site is shown in Table D.2-16. Health effects are discussed in section D.2.9.

Preventative Measures

Five curies of transuranic material is the maximum amount allowed in a glovebox in the unhardened section of Building 251. The norm is usually other materials in lesser quantities.

The severe earthquake postulated for this event is expected to occur only once in 5000 years (i.e., in any given 5000-year period, one would expect to record only one earthquake producing an 0.8g acceleration.) The unhardened building shell and some, if not all, gloveboxes may survive unscathed and undamaged (see Appendix I). Other mitigating factors include administrative controls, the use of primary and secondary containment for radioactive materials, and the use of high efficiency particulate air filters.

Table D.2-14 Collective Population Dose for the Tritium Release from Building 331

Distance (km)	Population Western Sector	Individual CEDE (rem)	Collective Population Dose (person-rem)
2.4	2,676	6.5×10^-3	17.4
4.0	3,982	3.2×10^-3	12.7
5.6	5,786	2.1×10^-3	12.2
7.2	10,068	1.5×10^-3	15.1
12.0	26,776	7.9×10^-4	21.2
24.0	81,772	3.5×10^-4	28.6
40.0	305,746	1.9×10^-4	58.1
56.0	436,096	1.5×10^-4	65.4
72.0	554,684	1.1×10^-4	59.9
Total	1,417,586	Total	290.6

CEDE = Committed Effective Dose Equivalent
*Source: Educational Data Systems, 1991.

Table D.2-15 Calculated Individual Doses for the Americium-241 Release from Building 251

Distance from Building 251 (km)	CEDE (rem)		EDE (rem)
Distance from Building 251 (km)	Inhalation	Ingestion	Air Immersion	Ground Surface
0.1 (Onsite)	3.1	N/C	4.7×10^-8	5.1×10^-10
0.6 (Nearest and western boundary)	1.3×10^-1	1.1×10^-2	2.0×10^-9	6.7×10^-8
1.6	2.6×10^-2	7.0×10^-4	3.9×10^-10	9.4×10^-9
2.4	1.4×10^-2	3.6×10^-4	2.0×10^-10	4.8×10^-9
4.0	6.6×10^-3	1.7×10^-4	9.7×10^-11	2.3×10^-9
5.6	3.9×10^-3	1.0×10^-4	5.9×10^-11	1.4×10^-9
7.2	2.7×10^-3	7.1×10^-5	4.1×10^-11	9.6×10^-10
12.0	1.3×10^-3	3.5×10^-5	2.0×10^-11	4.7×10^-10
24.0	4.8×10^-4	1.3×10^-5	7.6×10^-12	1.8×10^-10
40.0	2.6×10^-4	7.0×10^-6	3.9×10^-12	9.4×10^-11
56.0	1.8×10^-4	4.6×10^-6	2.6×10^-12	6.3×10^-11
72.0	1.3×10^-4	3.3×10^-6	1.9×10^-12	4.5×10^-11
80.0	1.1×10^-4	2.9×10^-6	1.7×10^-12	3.9×10^-11

CEDE = Committed Effective Dose Equivalent.
EDE = Effective Dose Equivalent.
N/C = Not calculated. Protective actions would limit ingestion onsite.
See section D.2.6.

Table D.2-16 Collective Population Dose for the Americium-241 Release from Building 251

Distance (km)	Population Western Sector*	Individual CEDE (rem)	Collective Population Dose (person-rem)
2.4	2,676	1.4×10^-2	37.5
4.0	3,982	6.8×10^-3	27.1
5.6	5,786	4.0×10^-3	23.1
7.2	10,068	2.8×10^-3	28.2
12	26,776	1.4×10^-3	37.5
24	81,772	5.0×10^-4	40.9
40	305,746	2.7×10^-4	82.6
56	436,096	1.8×10^-4	78.5
72	544,684	1.3×10^-4	70.8
Total: 1,417,586		Total: 426.1

CEDE = Committed Effective Dose Equivalent.
* Source: Educational Data Systems, 1991.

D.2.8.6 Transuranic-Waste Release Outside Building 612, Waste Treatment and Storage, LLNL Livermore Site

Waste-management operations were examined to identify potential release scenarios for the Building 612 complex. This analysis determined the accident scenario with the highest potential for offsite impacts to be a release of transuranic waste, which is more radioactive than the low-level waste handled in the facility.

Manual handling of 55-gal drums of sealed transuranic waste is a routine operation in the Building 612 Complex. Drums of transuranic waste are received in the high bay of Building 612, where they are logged, weighed, and labeled. These drums are transported to Building 625 for storage. It was postulated that during transportation one drum is punctured by a forklift, and the entire contents spill onto the ground just outside Building 612, where they are ignited by a spark. The transuranic material, consisting mainly of contaminated laboratory waste, burns completely, releasing a fraction of the container's maximum radioactivity content (6 alpha curies of americium-241) directly into the environment. Since this incident was postulated to occur outside Building 612, no credit was taken for mitigating features, with the exception that cleanup and decontamination commences immediately after the fire is extinguished, thereby limiting the total release fraction to 0.05 percent (Mishima and Schwendiman, 1970).

Only one drum of transuranic waste is justified for this scenario because a direct hit by the forklift is required to puncture the drum. The assumption that the entire contents of the drum would spill, ignite, and burn completely is conservative because much of the waste material is typically solid and would not be dispersed completely from the drum. An ignition source of unspecified origin was postulated to provide a mechanism for dispersing the waste in the atmosphere, thus maximizing the offsite consequences.

Both Buildings 612 and 625 have high bays with fire-protection sprinklers. Since there is little or no combustible material in these areas, a fire intense enough to breach more than one drum is not likely.

Development of Scenario and Assumptions About the Radioactive Source Team

The quantity of transuranic waste at risk was postulated to be 6 Ci of alpha radiation versus 3 Ci of plutonium-239 equivalent material (LLNL, 1988a), which is the current maximum limit for a 55-gal drum. When the contents of the drum spill and burn, 0.05 percent (Mishima and Schwendiman, 1970) of the initial activity is released. Thus, 0.003 Ci of americium-241 equivalent material is dispersed directly into the environment in a 10-m elevated release modeled as a line source. Plume depletion by dry deposition was included in the dose calculations.

Computer Modeling and Results

The radiation doses calculated for this scenario from external exposure to the passing plume, external exposure from ground deposition, and internal exposures from inhalation and food ingestion are given in Table D.2-17. The 0.1 km distance is within the site, the 0.2 km distance is the nearest LLNL Livermore site boundary to the south, and the 1.6 km distance is the western boundary, closest to the nearest population center.

The committed effective dose equivalent of 2.0 rem at the nearest site boundary is from inhalation and within the whole body dose range (1 to 5 rem) at which some protective action is recommended by the EPA for accidents (EPA, 1990). There is no significant dose contribution from the external exposure pathways of air immersion and ground surface contamination. The ingestion pathway dose at the western boundary is well below the preventive PAG of 0.5 rem (FDA, 1982) for whole body and organs other than the thyroid at which protective action is recommended by the FDA.

The projected collective population doses for people living in the western sector are shown in Table D.2-18. Health effects are discussed in section D.2.9.

Preventative Measures

Since this accident scenario was postulated to occur outside the building, no credit was taken for mitigating features with the exception that cleanup and decontamination commences immediately after the fire is extinguished. The source term was based on the projected new limit of 6 Ci of alpha radiation versus the current limit of 3 Ci of plutonium-239 equivalent material.

Table D.2-17 Calculated Individual Doses for the Americium-241 Release Near Building 612

Distance from Building 612 (km)	CEDE (rem)		EDE (rem)
Distance from Building 612 (km)	Inhalation	Ingestion	Air Immersion	Ground Surface
0.1 (Onsite)	3.7	N/C	5.7×10^-8	6.2×10^-10
0.2 (Nearest boundary)	2.0	N/C	3.0×10^-8	1.0×10^-6
1.6 (Western boundary)	1.1×10^-1	9.0×10^-3	1.7×10^-9	5.6×10^-8
2.4	6.2×10^-2	1.6×10^-3	9.0×10^-10	2.2×10^-8
4.0	2.9×10^-2	7.9×10^-4	4.5×10^-10	1.0×10^-8
5.6	1.8×10^-2	4.9×10^-4	2.8×10^-10	6.6×10^-9
7.2	1.3×10^-2	3.3×10^-4	1.9×10^-10	4.5×10^-9
12.0	6.2×10^-3	1.6×10^-4	9.0×10^-11	2.2×10^-9
24.0	2.2×10^-3	5.8×10^-5	3.3×10^-11	7.7×10^-10
40.0	9.7×10^-4	2.6×10^-5	1.5×10^-11	3.6×10^-10
56.0	5.7×10^-4	1.6×10^-5	9.0×10^-12	2.1×10^-10
72.0	4.0×10^-4	1.1×10^-5	6.1×10^-12	1.5×10^-10
80.0	3.4×10^-4	9.1×10^-6	5.2×10^-12	1.3×10^-10

CEDE = Committed Effective Dose Equivalent.
EDE = Effective Dose Equivalent.
N/C = Not calculated. Protective actions would limit ingestion onsite.
See section D.2.6.

Table D.2-18 Collective Population Dose for the Americium-241 Release Near Building 612

Distance (km)	Population Western Sector*	Individual CEDE (rem)	Collective Population Dose (person-rem)
2.4	2,676	6.4×10^-2	171.3
4.0	3,982	3.0×10^-2	119.5
5.6	5,786	1.8×10^-2	104.1
7.2	10,068	1.3×10^-2	130.9
12	26,776	6.4×10^-3	171.4
24	81,772	2.3×10^-3	188.1
40	305,746	1.0×10^-3	305.7
56	436,096	5.9×10^-4	257.3
72	544,684	4.1×10^-4	223.3
Total: 1,417,586		Total: 1,671.6

CEDE = Committed Effective Dose Equivalent.
* Source: Educational Data Systems, 1991.

D.2.8.7 Natural Uranium Release from Building 493, Separation Support Facility, LLNL Livermore Site

A fire involving 5000 kg of natural or depleted uranium was selected as the bounding accident for Building 493, the Separation Support Facility, a part of the Building 490 Complex. Building 493 is a warehouse that stores the uranium used in the laser isotope-separation process. This building is surrounded by a standard chain-link security fence; it may contain up to 80,000 kg of uranium at one time. The uranium is in metal form with massive shapes and is stored in separate, sealed metal containers in a caged area. Nominal storage per container is less than 500 kg. Flammable materials are not stored in Building 493; however, combustible materials are present. Fire protection is provided by sprinklers above the work areas. A fire-induced release from this building requires an outside source of energy, such as a tanker truck, that breaches the containers of uranium and provides fuel for a fire of approximately 30-minutes duration. (A fire involving 7000 gal of ethanol with a burning rate of 0.015 kg/s-m2 covering 12,435 sq ft of floor area would result in a 25-minute fire.) This fire would oxidize and disperse a fraction of the uranium into the atmosphere. The amount of uranium oxidized and available for dispersal is limited by the outside energy source rather than the quantity of uranium in storage.

Other events that were considered and eliminated were a material-handling accident in Building 493; a fire involving 100 kg of uranium being transported between buildings, and a fire involving 600 kg of uranium in Building 490. These accidents all involve natural or depleted uranium. A small quantity of enriched uranium is created by the separation process in Building 490; however, these quantities are not allowed to accumulate to levels that would create a criticality concern (LLNL, 1991d). Any accidents involving these small quantities of enriched uranium are bounded by accidents in other facilities at the LLNL Livermore site.

In the event of a material-handling accident in Building 493, only personnel inside the building would be affected. Even if a large quantity of natural uranium were spilled, the uranium would remain in the building, posing no threat to offsite persons or other onsite personnel. As in the spill of any other heavy metal, there is no source of energy for dispersing the material beyond the building.

The uranium used in the separation process is carried between Buildings 490, 491, and 493 in transporters, typically in quantities of approximately 100 kg. Transporting these materials creates a potential for a vehicle accident; however, the transporters are built strongly and are not likely to be breached during a traffic accident. The quantity involved is much smaller than the quantities used in other Building 490 Complex activities, and therefore this scenario was not developed.

The maximum uranium fire in Building 490 would require that the maximum number of isotope-separation modules be in operation and breaching of both of the primary and secondary containments. The at-risk inventory is less than 600 kg of uranium at temperatures high enough to promote rapid oxidation if air enters the vacuum modules. If all at-risk uranium oxidizes, the maximum release fraction from the module tanks (primary containment) is 0.1 percent (LLNL, 1987). Taking no credit for building secondary confinement or high efficiency particulate air—filtered ventilation used for ALARA purposes during normal operations, this is also the maximum release fraction that can be dispersed to the environment. The consequences of this event are bounded by the major fire event in Building 493.

Development of Scenario and Assumptions About the Radioactive Source Term

Bounded by the maximum credible external energy source that could breach no more than 10 sealed uranium metal storage containers in Building 493 and oxidize the massive pieces, 5000 kg was determined to be a quantity that could be oxidized during a 30-minute fire. From experiments with uranium oxide powders in gasoline fires, the release fraction was conservatively estimated at 0.1 percent (LLNL, 1987), for a total release of 5 kg of uranium, modeled as a line source in an elevated release of 30 m. Although much of the uranium stored in Building 493 is depleted of uranium-235, the source term (Table D.2-19) was computed using the average natural uranium isotopic mix. This is conservative since natural uranium has a higher specific activity than depleted uranium.

Computer Modeling and Results

Table D.2-20 summarizes the doses calculated for this elevated release (including plume depletion by dry deposition) at various distances from external exposure to the passing plume, external exposure from ground deposition, and internal exposures from inhalation and food ingestion. The first entry shows the onsite doses at 0.1 km distance; the 0.3 km distance is the nearest LLNL Livermore site boundary to the north, and the 1.3 km distance is the site boundary to the west.

Radiological Impact

The committed effective dose equivalent of 0.15 rem at the nearest site boundary is from inhalation and below the whole body dose range (1 to 5 rem) at which some protective action is recommended by the EPA for accidents (EPA, 1990). There is no significant dose contribution from the external exposure pathways from air immersion and ground surface contamination. The ingestion pathway dose at the western boundary is well below the preventive PAG of 0.5 rem (FDA, 1982) for whole body and organs other than the thyroid at which protective action is recommended by the FDA.

The collective doses received by the population in the western sector from the LLNL Livermore site are estimated in Table D.2-21. Health effects are discussed in section D.2.9.

Toxicological Impact

Since a fire is postulated, the chemical form of uranium released is assumed to be UO2 and U3O8, both insoluble forms of uranium. If inhaled, insoluble uranium is cleared from the pulmonary tissue to the ciliated airways and is eliminated by the gastrointestinal tract. Thirty-day feeding experiments in rats with compounds of insoluble uranium were found to be nontoxic. In contrast, soluble uranium compounds, especially if ingested chronically, can pose a threat to the kidney and causes lesions in the renal tissues. The renal threshold concentration of uranium that results in significant damage is a matter of controversy, and estimates range from 3 to <1 mg per gram of renal tissue. However, it is premature to attempt a risk estimate for developing renal damage from an acute inhalation and chronic ingestion of insoluble uranium compounds as there may be no risk at all. Therefore, the toxicological impact of this postulated release was dismissed as insignificant when compared to the radiological impact (BEIR IV, 1988).

The total impact of this release on the public and the environment would be negligible offsite and of a minor consequence at LLNL.

Preventative Measures

The scenario assumed that the uranium fire burned for 30 minutes; however, the fire-protection sprinklers would reduce the severity of the fire, and the fire department would respond within 3 minutes to begin fire control procedures.

Table D.2-19 Building 493 Source Term

Isotope	Halflife (years)*	Atomic Weight*	Specific Activity	Natural Abundance (percent)*	Curies
U-234	2.46×105	234.0409	6.2×10^-3	0.0055	1.71×10^-3
U-235	7.04×108	235.0439	2.2×10^-6	0.720	7.78×10^-5
U-238	4.47×109	238.0508	3.4×10^-7	99.2745	1.67×10^-3

* Source: General Electric Company, 1989.

Table D.2-20 Calculated Individual Doses for the Uranium Release from Building 493

Distance from Building 493 (km)	CEDE (rem)		EDE (rem)
Distance from Building 493 (km)	Inhalation	Ingestion	Air Immersion	Ground Surface
0.1 (Onsite)	4.1×10^-1	N/C	4.7×10^-9	6.5×10^-10
0.3 (Nearest boundary)	1.5×10^-1	N/C	1.7×10^-9	7.7×10^-7
1.3 (Western boundary)	3.8×10^-2	1.0×10^-4	4.3×10^-10	1.8×10^-7
1.6	2.8×10^-2	3.1×10^-5	3.3×10^-10	1.0×10^-7
2.4	1.8×10^-2	1.9×10^-5	2.1×10^-10	6.2×10^-8
4.0	9.2×10^-3	9.9×10^-6	1.1×10^-10	3.2×10^-8
5.6	5.9×10^-3	6.3×10^-6	6.7×10^-11	2.1×10^-8
7.2	4.4×10^-3	4.5×10^-6	4.9×10^-11	1.5×10^-8
12.0	2.1×10^-3	2.2×10^-6	2.4×10^-11	7.2×10^-9
24.0	7.9×10^-4	8.5×10^-7	9.3×10^-12	2.8×10^-9
40.0	3.6×10^-4	3.8×10^-7	4.1×10^-12	1.2×10^-9
56.0	2.1×10^-4	2.2×10^-7	2.4×10^-12	7.2×10^-10
72.0	1.4×10^-4	1.5×10^-7	1.6×10^-12	4.8×10^-10
80.0	1.2×10^-4	1.3×10^-7	1.3×10^-12	4.1×10^-10

CEDE = Committed Effective Dose Equivalent.
EDE = Effective Dose Equivalent.
N/C = Not calculated. Protective actions would limit ingestion onsite.
See section D.2.6.

Table D.2-21 Collective Population Dose for the Uranium Release from Building 493

Distance (km)	Population Western Sector*	Individual CEDE (rem)	Collective Population Dose (person-rem)
2.4	2,676	1.8×10^-2	48.2
4.0	3,982	9.2×10^-3	36.6
5.6	5,786	5.9×10^-3	34.1
7.2	10,068	4.4×10^-3	44.3
12	26,776	2.1×10^-3	56.2
24	81,772	8.0×10^-4	65.4
40	305,746	3.6×10^-4	110.1
56	436,096	2.1×10^-4	91.6
72	544,684	1.4×10^-4	76.3
Total: 1,417,586		Total: 562.8

CEDE = Committed Effective Dose Equivalent.
* Source: Educational Data Systems, 1991.

D.2.8.8 Transuranic-Waste Release from Building 625, Container Storage, LLNL Livermore Site

The effect of a severe earthquake on waste management operations was examined to identify potential consequences in the Building 612 Complex and specifically for Building 625. The objective was to postulate a credible event that could disperse a quantity of plutonium-equivalent material to the environment as a result of an earthquake with a peak ground acceleration of 0.8g (see Appendix I).

The following observations were made during a walkdown of the Building 612 complex:

The existence of an overhead rail crane in Building 625.
A cement berm divides the building into two halves used for separately storing polychlorinated biphenyls and radioactive materials.
Type A 55-gal drums containing transuranic waste were stacked two high in eight rows within the half of the building used for radioactive-material storage. These rows are perpendicular to the overhead crane's axis.

Since this scenario is initiated by a 0.8g earthquake, it follows that (as detailed in Appendix I) the building shell is likely to suffer significant structural damage. Simplified structural analysis demonstrates that while Building 625 is unlikely to collapse completely, falling debris (such as the overhead crane) has sufficient energy to rupture eight Type A containers. With the current storage scheme, eight drums would receive a direct impact if the crane fell flatly. Failure of the building shell would allow respirable particulates ejected from the crushed drums to be released into the environment.

The consequences of this scenario would be increased by fire. Building 625 has a high bay with fire-protection sprinklers. If, as postulated, the building shell fails, the derangement of piping will prevent the sprinklers from working. A fire that could not be rapidly extinguished would raise the release fraction and create an additional mechanism for the dispersal of the transuranic waste. Building 625, however, does not contain any flammable materials, and a fire is not likely to result from the earthquake.

Development of Scenario and Assumptions About the Radioactive Source Term

Various waste-handling operations in the Building 612 complex involve materials with low levels of radioactivity. The highest potential offsite doses would result from a release of transuranic waste, specifically americium-241. This material was chosen for this scenario because of limits proposed by the Hazardous Waste Department for the Building 612 Complex and its potential for the highest offsite doses. The transuranic radionuclides handled in Building 625 are listed in Table D.2-22 (LLNL, 1989f).

A dose calculation was performed to compare the offsite impacts of these nuclides. Based on these results, americium-241 was chosen as the nuclide that would cause the worst offsite impact.

In summary, during this scenario, a 0.8g earthquake ruptures the Building 625 shell; the overhead crane falls, crushing 8 drums, each loaded to 6 Ci of americium-241. The sudden compression and decompression of the drums results in a fractional release of 0.015 percent (DOE, 1990a). No credit was taken for the building's containment; therefore, 100 percent of the particulate plume was assumed to be released at ground level. While the building shell would suffer damage, no complete failure of the shell could be postulated, and therefore, no additional wind-driven contribution to the release fraction could be justified. Only a small fraction of americium-241 (less than 1 percent) contributes to the curie loading in Building 625 and in the 612 Complex (LLNL, 1989f); however, no credit was taken for a nominal inventory.

The source term was calculated to be

6 Ci×8 drums×0.015% = 0.0072 Ci

This source term is based on a proposed limit of 6 Ci of alpha radiation per drum, not the currently allowed limit of 3 Ci of plutonium-239–equivalent material per drum. The plume was depleted assuming dry deposition for the dose calculations.

Computer Modeling and Results

Table D.2-23 presents the doses calculated for this postulated scenario. The 70-year committed effective dose equivalent of 30 rem onsite is significant, with 1.4 rem being delivered in the first year for any personnel exposed to the airborne radioactive plume for its entire passage. The 70-year committed effective dose equivalent of 4.2 rem at the nearest site boundary of 0.3 km is within the guidelines at which the EPA recommends evacuation, access control to the contaminated area, and monitoring of radiation levels (see section D.2.7). This dose is essentially all from inhalation, and there is no significant dose contribution from air immersion and ground surface contamination, the external exposure pathways. The ingestion pathway dose at the western boundary is well below the preventive PAG of 0.5 rem (FDA, 1982) for whole body and organs other than the thyroid at which protective action is recommended by the FDA. The projected doses received by the population in the western sector from the LLNL Livermore site are as shown in Table D.2-24. Health effects of this release are discussed in section D.2.9.

Preventative Measures

It is quite unlikely that all the drums would contain the maximum amount of americium-241 and would be stored in precisely the same rows and columns of drums under a fully loaded crane during a seismic event expected to occur once in 5000 years. These factors contribute to the low probability of occurrence. Additionally, each drum has a rigid liner and the contents are wrapped in individual polyethylene bags, which provide additional containment. This accident is mitigated by administratively parking the overhead crane in a position away from loaded transuranic waste drums, and by seismic stops, which are installed on the overhead crane.

Table D.2-22 Transuranic Nuclides of Interest in Building 625

Radionuclide	Half-life (years)	Atomic mass unit	Specific Activity (Ci/g)	Initial Source Term (Ci)	Final Source Term (Ci)	Dose^a (rem)	Dose^b (rem)
Pu-238	8.77×101	238.0496	1.71×101	48	7.2×10^-3	3.1	2.1×10^-1
Pu-239	2.41×104	239.0522	6.23×10^-2	48	7.2×10^-3	3.3	2.2×10^-1
Pu-240	6.57×103	240.0538	2.27×10^-1	48	7.2×10^-3	3.3	2.2×10^-1
Pu-241	1.44×101	241.0568	1.03×102	48	7.2×10^-3	5.4×10^-2	3.6×10^-3
Pu-242	3.76×105	242.0587	3.93×10^-3	48	7.2×10^-3	3.0	2.0×10^-1
Am-241	4.33×102	241.0568	3.43×100	48	7.2×10^-3	4.8	3.4×10^-1
Cm-244	1.81×101	244.0627	8.09×101	48	7.2×10^-3	2.6	1.9×10^-1

^a 70-year committed effective dose equivalent at the eastern site boundary (0.3 km).
^b 70-year committed effective dose equivalent at the western site boundary (1.6 km).

D.2.8.9 Analysis and Review of Transportation Accidents Involving Radioactive Materials and Waste

Table D.2-23 Calculated Individual Doses for the Americium-241 Release from Building 625

Distance from Building 625 (km)	CEDE (rem)		EDE (rem)
Distance from Building 625 (km)	Inhalation	Ingestion	Air Immersion	Ground Surface
0.1 (Onsite)	30	N/C	4.6×10^-7	4.9×10^-9
0.3 (Nearest boundary)	4.2	N/C	6.3×10^-8	2.1×10^-6
1.6 (Western boundary)	2.5×10^-1	2.0×10^-2	3.8×10^-9	1.3×10^-7
2.4	1.3×10^-1	3.5×10^-3	1.9×10^-9	4.7×10^-8
4.0	6.2×10^-2	1.6×10^-3	9.7×10^-10	2.2×10^-8
5.6	3.7×10^-2	9.9×10^-4	5.6×10^-10	1.4×10^-8
7.2	2.6×10^-2	6.8×10^-4	3.9×10^-10	9.3×10^-9
12.0	1.3×10^-2	3.3×10^-4	1.9×10^-10	4.5×10^-9
24.0	4.8×10^-3	1.3×10^-4	7.6×10^-11	1.8×10^-9
40.0	2.5×10^-3	6.7×10^-5	3.8×10^-11	9.1×10^-10
56.0	1.7×10^-3	4.5×10^-5	2.6×10^-11	6.0×10^-10
72.0	1.2×10^-3	3.2×10^-5	1.8×10^-11	4.3×10^-10
80.0	1.1×10^-3	2.7×10^-5	1.6×10^-11	3.8×10^-10

CEDE = Committed Effective Dose Equivalent.
EDE = Effective Dose Equivalent.
N/C = Not calculated. Protective actions would limit ingestion onsite.
See section D.2.6.

Table D.2-24 Collective Population Dose for the Americium-241 Release from Building 625

Distance (km)	Population Western Sector*	Individual CEDE (rem)	Collective Population Dose (person-rem)
2.4	2,676	1.3×10^-1	347.9
4.0	3,982	6.4×10^-2	254.8
5.6	5,786	3.8×10^-2	219.9
7.2	10,068	2.7×10^-2	271.8
12	26,776	1.3×10^-2	348.1
24	81,772	4.9×10^-3	400.7
40	305,746	2.6×10^-3	794.9
56	436,096	1.7×10^-3	741.4
72	544,684	1.2×10^-3	653.6
Total: 1,417,586		Total: 4,033

CEDE = Committed Effective Dose Equivalent.
* Source: Educational Data Systems, 1991.

D.2.8.9.1 Low Specific Activity Radioactive Waste Release During Shipment

Several factors were taken into account to define a transportation accident scenario including the radionuclides present in the waste, the quantities transported in Type A and Type B packagings, the frequency with which each nuclide has historically been handled by Hazardous Waste Management, the half-lives of the radionuclides, and potential doses. These factors were analyzed to postulate two potentially bounding accident scenarios involving a Low Specific Activity (LSA) shipment of waste. See the Glossary for the definition of LSA.

The first accident modeled was based on the waste stream from LLNL which would have a mix of nuclides including some TRU. The second was based on the waste stream from SNL, Livermore which would contain tritium oxide only.

D.2.8.9.1.1 Low Specific Activity Radioactive Waste Release During Shipment (First Scenario)

Development of Scenario and Assumptions About the Radioactive Source Term

Data from LLNL operations (LLNL, 1989d) and values tabulated in 49 C.F.R. section 173.435 for Type A packagings were used to estimate the maximum amount of each radionuclide shipped in Type A packaging. Radionuclides with half-lives less than 5 days were eliminated because their activity would decrease significantly by decay during onsite storage.

The nuclides identified in waste requisitions as having half-lives longer than 5 days were analyzed with the GENII code to compare potential doses. The source term used was the smaller of (1) the A2 value listed in 49 C.F.R. (maximum amount allowed in a Type A package), or (2) the maximum amount handled in 1989 and 1990 (LLNL, 1989d). GENII calculations identified the radionuclides that would deliver the highest doses at 100 m. Combining the top five from this list with the top five from a list of the nuclides most frequently handled by Hazardous Waste Management produced the radionuclides selected for analysis in this accident scenario. The waste containing these radionuclides was assumed to be shipped by truck.

The procedure followed for Low Specific Activity shipments allows a maximum of 24 packages on a truck. These 24 packages were assigned hypothetical contents that include the radionuclides selected for analysis. The truck shipment was thus assumed to contain seven packages of plutonium-239, six packages of uranium-238, three packages of tritium oxide, two packages of phosphorus-32, one package of uranium-233, one package of uranium-235, one package of thorium-232, one package of americium-241, and two packages containing other radionuclides with negligible radioactivity.

The accident scenario begins with a severe accident (category 6 or higher) in which all containers are breached. A category 6 accident requires a crushing force of more than 4.40×105 newtons with no fire, ranging down to a crush force of more than 2.2×104 newtons with a fire of 1027·C for more than 1.5 hours (Nuclear Regulatory Commission, 1977a). The highest accident category is 8. Release fractions were taken from the supplemental EIS for the Waste Isolation Pilot Plant (DOE, 1990a) which analyzed failures of Type A containers. Table D.2-25 summarizes the source terms for this accident. Except for tritium, the total respirable release fraction for each radionuclide in Table D.2-25 was calculated to be 0.01; for tritium oxide this fraction was assumed to be 1.

Computer Modeling and Results

An acute 10-meter elevated release was modeled, and individual doses were calculated at specified distances from external exposure to the passing plume, external exposure from ground deposition, and internal exposures from inhalation and food ingestion. The plume was depleted for nuclides released in a particulate form. No specific location was postulated, and food ingestion was assumed for all distances from the point of release. The calculated individual doses from this scenario with Low Specific Activity waste are summarized in Table D.2-26. The calculated population doses are summarized in Table D.2-26A.

The inhalation pathway dominates the dose delivered to an individual in the path of the radioactive plume, with the ingestion pathway also contributing significantly. There is negligible dose contribution from the external exposure pathways. The doses calculated are based on very conservative assumptions about food ingestion. For example, it was assumed that an individual at 0.1 km would eat 10 percent of 1 year's intake of vegetables, fruits, grains, meat, poultry, eggs, and milk contaminated with radioactivity at a level found at 0.1 km from the release point. The CEDE of 4.7 rem at 0.1 km is less than the emergency PAG of 5 rem to the whole body, bone marrow, or any other organ at which the FDA requires isolating food to prevent its introduction into commerce. Committed effective dose equivalents decrease with distance from the release point and between 0.5 km and 1.0 km the CEDE drops below the preventive PAG of 0.5 rem at which the FDA recommends protective actions to prevent or reduce the radioactive contamination of human food or animal feeds.

Preventative Measures

This accident scenario used conservative assumptions about the radioactivity content of Type A packages. The average radioactivity content of typical shipments is usually much lower (by a factor of 100 or more) than the values used in this scenario. This scenario assumed that tritium is shipped in oxide form; however, tritium can also be shipped in its hydride form. If tritium hydride is shipped, the total release fraction will drop from 1.0 to 0.0002, lessening the severity of this scenario. In practice, all applicable regulations of the U.S. Department of Transportation are met for Low Specific Activity shipments, and together with inspections of packages, this reduces the probability of accidents severe enough to breach the containers.

Table D.2-25 Source Term for Low Specific Activity Waste Transportation Accident

Radio- nuclide	Number of Containers	Concentration per Container (Ci)	Total Radioactivity (Ci)	TRRF^a	Source Term (Ci)
Pu-239	7	2.0×10^-3	1.4×10^-2	0.01	1.4×10^-4
U-238	6	5.6×10^-2	3.4×10^-1	0.01	3.4×10^-3
H-3^b	3	1000	3000	1	3000
Am-241	1	8.0×10^-3	8.0×10^-3	0.01	8.0×10^-5
Th-232	1	6.0×10^-3	6.0×10^-3	0.01	6.0×10^-5
U-235	1	2.0×10^-1	2.0×10^-1	0.01	2.0×10^-3
U-233	1	1.0×10^-1	1.0×10^-1	0.01	1.0×10^-3
P-32	2	20	40	0.01	4.0×10^-1
Other	2	No consequence	N/A	N/A	N/A

^a Total respirable release fraction.
^b H-3 is in the form of HTO.
N/A = Not applicable.

Table D.2-26 Calculated Individual Doses for Low Specific Activity Transportation Accident Waste (First Scenario)

Distance from Accident (km)	CEDE (rem)		EDE (rem)
Distance from Accident (km)	Inhalation	Ingestion	Air Immersion	Ground Surface
0.1	3.1	1.6	2.9×10^-6	3.0×10^-5
0.5	5.2×10^-1	2.1×10^-1	5.0×10^-7	5.4×10^-6
1.0	1.7×10^-1	4.8×10^-2	1.7×10^-7	1.9×10^-6
1.6	8.1×10^-2	1.5×10^-2	8.5×10^-8	8.9×10^-7
2.4	4.4×10^-2	5.1×10^-3	4.8×10^-8	5.0×10^-7
4.0	2.0×10^-2	1.2×10^-3	2.1×10^-8	2.3×10^-7
5.6	1.2×10^-2	4.6×10^-4	1.3×10^-8	1.4×10^-7
7.2	8.5×10^-3	2.1×10^-4	4.5×10^-9	1.0×10^-7
12.0	4.0×10^-3	4.7×10^-5	4.3×10^-9	4.7×10^-8
24.0	1.4×10^-3	1.1×10^-5	1.6×10^-9	1.7×10^-8
40.0	6.5×10^-4	4.8×10^-6	2.1×10^-10	7.5×10^-9
56.0	3.6×10^-4	2.8×10^-6	4.0×10^-10	4.3×10^-9
72.0	2.3×10^-4	1.8×10^-6	2.5×10^-10	2.7×10^-9
80.0	1.9×10^-4	1.5×10^-6	2.1×10^-10	2.2×10^-9

CEDE = Committed Effective Dose Equivalent.
EDE = Effective Dose Equivalent

Table D.2-26A Collective Population Dose for Low Specific Activity Transportation Accident Waste (First Scenario)

Midpoint (km)	R+	R-	Sector Population 104/sq mi×5·/360·×p×(R+2-R-2)	Individual CEDE (rem)	Collective Population Dose (person-rem)
0.1	0.3	0.0	15	4.7	71
0.5	0.8	0.3	80	0.73	58
1.0	1.3	0.8	190	0.22	42
1.6	2.0	1.3	389	9.7×10^-2	38
2.4	3.2	2.0	1,051	5.0×10^-2	53
4.0	4.8	3.2	2,156	2.1×10^-2	45
5.6	6.4	4.8	3,019	1.2×10^-2	36
7.2	9.6	6.4	8,626	8.7×10^-3	75
12.0	18.0	9.6	39,058	4.1×10^-3	160
24.0	32.0	18.0	117,928	1.5×10^-3	177
40.0	48.0	32.0	215,640	6.5×10^-4	140
56.0	64.0	48.0	301,896	3.7×10^-4	112
72.0	80.0	64.0	388,152	2.3×10^-4	89
Total: 1,078,201				Total: 1,096

CEDE = Committed Effective Dose Equivalent

D.2.8.9.1.2 Low Specific Activity Radioactive Waste Release During Shipment (Second Scenario)

Development of Scenario and Assumptions About the Radioactive Source Term

Scenario development was similar to that for the first LSA accident scenario.

After reviewing the list of nuclides, percentages, and amounts, tritium oxide was selected to be packaged on the truck for the accident scenario as a special case analysis.

The implied limit in 49 C.F.R. 173 was used to bound the source term, which states that 30,000 Ci of LSA may be transported in a single shipment without imposing highway route control. Other DOE guidance recommends minimizing the total number of waste shipments (DOE, 1988b) with the implication to minimize overall highway collision risk. SNL, Livermore Hazardous Waste Management attempts to minimizes the number of waste shipments without exceeding the highway route control limit. Therefore, the source term prior to release was set at 30,000 Ci HTO.

Historically, the maximum number of Type A containers that SNL, Livermore packaged was 25, or 25,000 Ci of HTO (the A2 limit for HTO is 1000 Ci per container).

The accident scenario begins with a severe accident (category 6 or higher) in which all containers are breached. All waste was assumed to be shipped in Type A containers. A fire of at least 30 minutes duration was assumed. The total respirable release fraction for tritium oxide was set at 100 percent based on the assumption that a fire of at least 30 minutes duration occurs. The tritium oxide would therefore be released as steam.

Computer Modeling and Results

The computer modeling was similar to that for the first LSA accident scenario except that the only plume depletion velocity used was 1.8 cm/second. This velocity applies to tritium oxide, which was the minimum (conservative) value found in the literature. As before, individual doses were calculated at specified distances from external exposure to the passing plume, external exposure from ground deposition, and internal exposures from inhalation and food ingestion. Computed individual doses are summarized in Table D.2-26B.

The ingestion pathway dominates the dose delivered to the individual in the path of the radioactive plume, with the inhalation pathway also contributing significantly. There is negligible dose contribution from the external exposure pathways. The doses were calculated based on the same conservative assumptions concerning food ingestion described above for the first transportation accident scenario.

The calculated CEDE of 21 rem at 0.1 km, which includes both internal and external dose, exceeds the 5 rem EPA Protective Action Guides (PAG) level to a distance of almost 0.5 km. Above the 5 rem level, the EPA PAGs would require evacuation (EPA, 1990). The ingestion component of CEDE is 15.8 rem at 0.1 km and is greater than the emergency FDA PAG level of 5 rem to the whole body. Above this level, the FDA requires isolating food to prevent its introduction into commerce (FDA, 1982). Emergency protective actions would be required out to a distance of between 0.1 km and 0.5 km to limit the ingestion of potentially contaminated food stuffs.

The population dose reported in Table D.2-26C assumes the same urban population density (10,000 persons per square mile) as was assumed in reporting the population dose for the first LSA accident scenario (Table D.2-26A). Should this second accident occur in an urban area, the land use would not be agricultural; therefore, little or no food would be out in the open subject to contamination from the plume passage. As a consequence, the individual CEDE at 0.1 km would be composed of only the inhalation component of 4.7 rem instead of the total CEDE of 21 rem. In this case, because the EPA's PAG limit of 5 rem (EPA, 1990) is not exceeded, evacuation would not be necessary.

Preventative Measures

This accident scenario used conservative assumptions about the radioactivity content of Type A packages. The average radioactivity content of typical shipments is usually much lower (by a factor of 100 or more) than the values used in this scenario. In practice, all applicable regulations of the U.S. Department of Transportation are met for Low Specific Activity shipments, and together with inspections of packages, this reduces the probability of accidents severe enough to breach the containers.

Table D.2-26B Calculated Individual Doses for Low Specific Activity Transportation Accident Waste (Second Scenario)

Distance from Accident (km)	CEDE (rem)		EDE (rem)
Distance from Accident (km)	Inhalation	Ingestion	Air Immersion	Ground Surface
0.1	4.7	15.8	3.8×10^-9	0
0.5	6.1×10^-1	2.02	5.0×10^-10	0
1.0	1.4×10^-1	4.69×10^-1	1.1×10^-10	0
1.6	4.2×10^-2	1.42×10^-1	3.5×10^-11	0
2.4	1.4×10^-2	4.79×10^-2	1.2×10^-11	0
4.0	3.2×10^-3	1.09×10^-2	2.7×10^-12	0
5.6	1.1×10^-3	3.64×10^-3	8.7×10^-13	0
7.2	4.3×10^-4	1.47×10^-3	3.5×10^-13	0
12.0	5.1×10^-5	1.68×10^-4	4.1×10^-14	0
24.0	1.0×10^-6	3.43×10^-6	8.7×10^-16	0
40.0	1.9×10^-8	6.35×10^-8	1.5×10^-17	0
56.0	6.4×10^-10	2.17×10^-9	5.3×10^-19	0
72.0	3.1×10^-11	1.04×10^-10	2.5×10^-20	0
80.0	7.5×10^-12	2.52×10^-11	6.1×10^-21	0

CEDE = Committed Effective Dose Equivalent.
EDE = Effective Dose Equivalent.

D.2.8.9.2 Review of Prior Radioactive Material (Nonwaste) Shipment Studies

Radioactive material shipments into and out of the LLNL Livermore site, LLNL Site 300, and SNL, Livermore are strictly regulated by the Department of Transportation (DOT), the Nuclear Regulatory Commission (NRC), and DOE. The safety of radioactive material shipments in the United States has been analyzed in a generic environmental impact statement (Nuclear Regulatory Commission, 1977a). This analysis shows that the packaging required by regulation for the shipment of radioactive materials makes releases during an accident highly unlikely. A study conducted by Sandia Laboratories (McClure and Tyron-Hopko, 1986) describes the low probability of a release.

The LLNL Livermore site uses both type A and B packaging as defined in DOT regulations (49 C.F.R. section 173.403) to ship radioactive materials. Type A packages have limits on the quantity of radioactive material and are designed to retain the integrity of containment and shielding under normal transportation conditions, as demonstrated by water spray, free drop, compression, and penetration tests (49 C.F.R. section 173.465 or section 173.466). Type B packages are designed to retain their integrity during normal transportation and under severe accident conditions (10 C.F.R. Part 71.73). The LLNL Livermore site ships very small quantities of radioactive material using type A packaging. The effects of breaching a type A package holding limited amounts of radioactivity would be small and would not be a bounding scenario. Therefore, radioactive material shipments by air or truck were not further considered.

Table D.2-26C Collective Population Dose for Low Specific Activity Transportation Accident Waste (Second Scenario)

Midpoint (km)	R+	R-	Sector Population 10⁴/sq mi×5·/360·×p×(R+2-R-2)	Individual CEDE (rem)	Collective Population Dose (person-rem)
0.1	0.3	0.0	15	21	311
0.5	0.8	0.3	80	2.6	209
1.0	1.3	0.8	190	6.1×10^-1	116
1.6	2.0	1.3	389	1.8×10^-1	71
2.4	3.2	2.0	1,051	6.2×10^-2	65
4.0	4.8	3.2	2,156	1.4×10^-2	30
5.6	6.4	4.8	3,019	4.7×10^-3	14
7.2	9.6	6.4	8,626	1.9×10^-3	16
12.0	18.0	9.6	39,058	2.2×10^-4	9
24.0	32.0	18.0	117,928	4.4×10^-6	1
40.0	48.0	32.0	215,640	8.3×10^-8	0
56.0	64.0	48.0	301,896	2.8×10^-9	0
72.0	80.0	64.0	388,152	1.4×10^-10	0
Total: 1,078,201				Total: 843

CEDE = Committed Effective Dose Equivalent

D.2.8.9.3 Review of Waste Isolation Pilot Plant (WIPP) TRU Study

Shipments of radioactive waste consist of transuranic waste and low–specific-activity waste. Appendix D of the Final Supplement Environmental Impact Statement for the Waste Isolation Pilot Plant (the SEIS; DOE, 1990a) contains a detailed analysis of transuranic waste shipments by truck and rail. It contains a bounding case accident analysis for shipments from several laboratories, including the Los Alamos National Laboratory (LANL) in New Mexico. LLNL is not shipping mixed waste offsite at the present time, although these shipments will probably resume at some time in the near future, and at that time the shipments will resemble those from LANL. The SEIS calculated the impacts using contact-handled transuranic waste compositions that were scaled up to the maximum curie content allowed either by the WIPP Waste Acceptance Criteria or the limits established for the TRUPACT-II (the NRC-approved shipping container). The results of that scenario are given below.

The bounding accident scenario for a transportation accident offsite involving transuranic waste was analyzed in the Final Supplemental Environmental Impact Statement for the Waste Isolation Pilot Plant (DOE, 1990a). This analysis assumes a shipment of three fully loaded TRUPACT-II containers, the number being limited by highway weight limits. Each TRUPACT-II contains 14 drums and each has a curie content set at the maximum thermal or fissile gram limits set by the WIPP Waste Acceptance Criteria or by the limits established for the TRUPACT-II, approximately 3600 plutonium equivalent curies in all.

All TRUPACT-II containers are assumed to be breached in the accident and then engulfed in fire for 2 hours. About 0.02 percent of the hazardous particulate materials are postulated to be released in respirable form (less than 10 mm Aerodynamic Equivalent Diameter in size). The accident is assumed to occur during a period of very stable atmospheric meteorological conditions, which would hold down the contaminant plume and limit its dispersion, thus maximizing radiation doses and hazardous chemical concentrations. The accident occurs in the urban portion of a large metropolitan area with a mean population density of 10,000 persons per square mile.

The RADTRAN and AIRDOS computer codes were used to predict the consequences of the transuranic waste scenario. Health effects resulting from the radiation exposure could be from external exposure as the contaminated plume passed by and by internal exposure (inhalation, resuspension, and ingestion) to the dispersed contaminated material. Since it was assumed that the accident occurred in a densely populated urban zone, ingestion associated with contamination of agricultural products was not applicable.

Assuming a resuspension half-life of 60 days, because the material would either be cleaned up or washed away, the RADTRAN calculation predicts a collective dose of 5.58×10⁴ person-rem (committed effective dose equivalent) and the AIRDOS calculation predicts a collective dose of 3.97×10⁴ person-rem.

D.2.8.9.4 Review of Special Isotope Separation Project(1) (SIS) SST Study

Both LLNL and SNL, Livermore receive and ship classified nuclear materials (e.g., plutonium, tritium, and weapon components) in Type B packaging, using safe secure transports (SSTs), a special highway transport system providing enhanced security and accident protection. Much of the information regarding these shipments is classified, including the actual types and amounts of material shipped, the number of shipments made, the date of shipments, the actual operating procedures of the SST, and the routes followed.

Shipments of these types of radioactive materials using SSTs are not unique to LLNL and SNL, Livermore; they occur between other facilities in DOE weapons complex. Shipments of classified nuclear materials from LLNL and SNL nonetheless must adhere to the regulatory limits established by the U.S. Department of Energy and usually meet the regulations of the U.S. Department of Transportation for shipping hazardous materials (49 C.F.R. 400–177). In 1988, DOE assessed the potential impacts of SST shipments between several of its weapons facilities in an environmental impact statement for the proposed Special Isotope Separation (SIS) project (DOE, 1988b).

Using the RADTRAN transportation risk analysis model (see discussion of RADTRAN model in Appendix D), the SIS EIS assessed normal and potential transportation accident impacts from shipping weapon-grade plutonium in SSTs from three possible locations for the SIS facility to the Rocky Flats Plant near Denver, CO. The three possible SIS sites were the Hanford site near Hanford, WA; the Idaho National Engineering Laboratory near Idaho Falls, ID; and the Savannah River Plant near Aiken, SC. Shipments of feed materials, the radioactive materials from which plutonium would be made, were also analyzed for routes from Hanford Site to the three proposed locations for the SIS (including Hanford Site to Hanford Site shipments).

The RADTRAN model analyzes transportation risks in two contexts: normal transportation, where exposure to workers and the public is limited to radioactivity emanating from the transportation package; and potentially severe accident conditions, where radioactive material may be released to the environment. Because much of the specific information needed for input into the RADTRAN model to analyze the SIS shipments is classified (e.g., the exact nature of the materials shipped, the possible release fractions in case of an accident, the actual routes to be used, and the number of shipments expected to be made), conservative estimates were used for the RADTRAN analysis of SIS shipments. In addition, where extraordinary safety procedures were present (e.g., SSTs do not travel in poor weather conditions), the analysis took no credit for reducing risk impacts.

The shipment of nuclear materials from the Savannah River Plant near Aiken, SC, to the Rocky Flats Plant (2672 km) plus shipment of feed materials from the Hanford Site to the Savannah River Plant (4539 km) represent the longest potential routes for SST shipments for the SIS analysis. Because of the distances involved, the analysis of the risks for these shipments would bound the potential shipments from LLNL and SNL, Livermore to other DOE weapons facilities. In addition, the SIS transportation risk analysis assessed annual radiological dose to the populations from shipments of SIS material (which includes not just plutonium, but feed material, potentially by-product, transuranic waste, and onsite low–level waste).

For normal transport, the transportation risk was calculated to be less than 19 person-rem, resulting in 5.3×10^-3 latent cancer fatality and 2.4×10^-3 genetic disorder (using the method of reporting risk in the SIS environmental impact statement). The annual radiological risk of health effects in the event of a transport accident was calculated to be 2.9×10^-4 latent cancer fatality and 1.3×10^-4 genetic disorder. Most of the radiological risk is attributed to incident-free (i.e., normal transport); potential accidents contribute little to the total radiological risks. Nonradiological risks resulting from mechanical injuries from traffic accidents are about 10 times higher than radiological risks (DOE, 1988b).

D.2.8.10 Tritium Release from Building 298, Fusion Target Fabrication Facility, or from Building 391, NOVA Upgrade/National Ignition Facility

One of the projects in the proposed action is establishing Building 298 as an inertial confinement fusion (ICF) target/tritium research facility with the ability to fill targets for laser fusion experiments. In Building 298, a 5 g maximum at-risk tritium research facility and target fill system is proposed. Filled targets would be transported to Building 391 for each experiment, one at a time. This scenario was originally characterized in the Draft EIS/EIR by a bounding release of 4 g of elemental tritium based on design configurations projected for the NOVA Upgrade/National Ignition Facility. This amount has been revised to 5 g at risk based on a subsequent analysis of the tritium handling process used in target fabrication.

During operation of the NOVA Upgrade/National Ignition Facility, tritium will be transported to Building 391 and ultimately will accumulate in the exhaust collection system. The most likely scenario calls for a peak inventory of approximately 60 mg tritium; however, the maximum allowable inventory in the building would be 5 g of tritium. This inventory of 5 g will be shared between the two buildings, 298 and 391. As the inventory in Building 391 increases, the inventory in Building 298 will decrease correspondingly, keeping the combined inventory at or below 5 g. Because Building 298 is closest to the site boundary, a release of the 5 g inventory from this building will bound the consequences for both buildings.

Development of Scenarios and Assumptions About the Radioactive Source Term

Scenario development was similar to that of Building 331 (see section D.2.8.4). No seismic analysis, however, was conducted on either Building 298 or 391. It was therefore assumed that these buildings fail during an earthquake with a peak ground acceleration of 0.8 g.

The source term is 1 percent of 5 g of tritium (483 Ci) and the release is postulated to occur in Building 298. As in Building 331, the ventilation systems are assumed to be inoperable (no elevated release), causing the tritium to be released at ground level. No plume depletion by dry deposition is assumed in the atmospheric dispersion model.

Computer Modeling and Results

Doses from external exposure to the passing plume, and internal exposures from inhalation and food ingestion, were calculated without assuming plume depletion for a ground-level release and are presented in Table D.2-27. Exposure to an atmosphere containing tritium results in intake of that material by absorption through intact skin and by inhalation. To account for skin absorption, GENII assumes that the combined total rate of tritium intake in air is 150 percent of the inhalation intake rate alone. The onsite dose, set at 100 m from the release point, was estimated at 0.40 rem (for Building 298), which is a small fraction of the current annual limit for occupational exposures of 5 rem/yr (DOE, 1990c). The calculated CEDE of 0.20 rem (for Building 298) at the nearest site boundary (150 m to the north) is significantly lower than the whole body dose range (1 to 5 rem) at which the EPA recommends protective action (EPA, 1990) for accident releases. This 0.20-rem dose is essentially all internal exposure from inhalation of tritium vapor. There is no dose contribution from ground surface contamination and negligible contribution from air immersion, the external exposure pathways. At the western boundary (750 m from Building 298), the dose of 0.11 rem is primarily from ingestion and less than the whole body preventive PAG (0.5 rem) at which the FDA recommends protective action to prevent or reduce the radioactive contamination of human food or animal feeds (FDA, 1982).

The tritium source term calculations did not quantify the specific contribution from resuspension originating from the deposition of gaseous tritium onto the ground, and subsequent oxidation in the soil followed by emission of tritiated vapor. For further discussion see section D.2.8.2.

The estimated collective doses received by the population in the western sector from the LLNL Livermore site are shown in Table D.2-28. Health effects are discussed in section D.2.9.

Preventative Measures

Table D.2-27 Calculated Individual Doses for the Tritium Release from Building 298

Distance from Building 298 (km)	CEDE (rem)		EDE (rem)
Distance from Building 298 (km)	Inhalation	Ingestion	Air Immersion	Ground Surface
0.1 (Onsite)	4.0×10^-1	N/C	3.5×10^-10	0
0.15 (Nearest boundary)	2.0×10^-1	N/C	1.7×10^-10	0
0.75 (Western boundary)	1.3×10^-2	1.0×10^-1	1.1×10^-11	0
1.6	4.0×10^-3	1.4×10^-2	3.3×10^-12	0
2.4	2.1×10^-3	7.3×10^-3	1.8×10^-12	0
4.0	1.1×10^-3	3.5×10^-3	8.8×10^-13	0
5.6	6.8×10^-4	2.3×10^-3	5.5×10^-13	0
7.2	4.8×10^-4	1.7×10^-3	4.0×10^-13	0
12.0	2.5×10^-4	8.8×10^-4	2.2×10^-13	0
24.0	1.1×10^-4	3.8×10^-4	9.3×10^-13	0
40.0	6.5×10^-5	2.2×10^-4	5.5×10^-14	0
56.0	4.8×10^-5	1.6×10^-4	4.0×10^-14	0
72.0	3.5×10^-5	1.2×10^-4	3.0×10^-14	0
80.0	3.3×10^-5	1.1×10^-4	2.8×10^-14	0

CEDE = Committed Effective Dose Equivalent.
EDE = Effective Dose Equivalent.
N/C = Not calculated. Protective actions would limit ingestion onsite.
See section D.2.6.

Table D.2-28 Collective Population Dose for the Tritium Release from Building 298

Distance (km)	Population Western Sector*	Individual CEDE (rem)	Collective Population Dose (person-rem)
2.4	2,676	9.4×10^-3	25.2
4.0	3,982	4.6×10^-3	18.3
5.6	5,786	3.0×10^-3	17.4
7.2	10,068	2.1×10^-3	21.4
12	26,776	1.1×10^-3	29.5
24	81,772	4.9×10^-4	40.1
40	305,746	2.9×10^-4	88.7
56	436,096	2.1×10^-4	91.6
72	544,684	1.6×10^-4	87.1
Total: 1,417,586		Total: 419.3

CEDE = Committed Effective Dose Equivalent.
* Source: Educational Data Systems, 1991.

D.2.9 Estimated Health Effects and Risk

The radiation doses calculated for each of the accident scenarios described in the preceding section have been analyzed to estimate the associated risk to the general population and to persons working at the LLNL Livermore site and SNL, Livermore. The method used to estimate health risks is described in detail in Appendix C and is based on methods and factors developed by the International Commission on Radiological Protection, and other authorities. These factors are summarized in Table D.2-29.

As explained in Appendix C, the effects of human exposure to ionizing radiation depend on a number of factors, including the dose received, the type of exposure (external or internal), the type of radiation delivering the dose (alpha, beta, or gamma), and the duration of exposure. The damage caused by ionizing radiation is traceable to the chemical alteration of molecules in various tissues of the body. As a result, in living organisms, the absorption of equal amounts of energy per unit mass does not result in the same biological effect. The alteration results from the ionization or excitation caused by the passage of a high-energy particle (the radiation). The severity of these changes is directly related to the local rate of energy deposition along the track of the particle inside the body. Heavy charged particles like alpha particles release energy over shorter distances than do photons or electrons and therefore tend to produce more biological damage. However, to reach vulnerable body tissues, alpha particles must be inhaled or ingested because they cannot penetrate the skin.

High doses of ionizing radiation (i.e., 100 rem and more) usually cause immediate, or "acute," effects, such as nausea and vomiting; if the doses reach as high as several hundred rem, death can be expected within a few weeks unless appropriate medical treatment is administered. For chronic exposures to low levels of radiation, the most significant risk is that of cancer. Genetic effects (i.e., effects that are not evident in the exposed person but are inherited by his or her children) can be induced if the exposure occurs during the reproductive period, and birth defects can result from exposure of the fetus during gestation.

The collective radiation doses estimated for the exposed population and the associated health risks are summarized in Table D.2-30. The exposed population was assumed to be the 1,417,586 persons residing to the west of the site boundary. As already explained, the western sector was selected for the analysis because it contains the largest number of people; the choice of this sector is conservative because an eastern wind occurs only 4 percent of the time during a given year. Table D.2-31 summarizes the onsite and site boundary doses and associated health risks to the maximally exposed individual. The health risks are stated in terms of fatal cancers and the total health detriment, which is defined as the sum of the total number of fatal cancers, the total number of other cancers, and the total number of severe genetic effects.

The collective radiation doses and associated health risks estimated for transportation of radioactive materials and waste are summarized in Table D.2-32. The exposed population was assumed to be 1,100,000 persons residing downwind of the postulate accidents.

The data in the tables show that for the general population the largest doses and associated health risks would be incurred from two of the postulated accident scenarios: a tritium release from Building 968 and a release of americium-241 from Building 625 (the multiple-building scenario initiated by a severe earthquake). For the general population, the tritium release would result in 2 cases of fatal cancer and a total health detriment of 3 in a population of 1,417,586. For comparison, background radiation is expected to cause 10,000 cases of fatal cancer and 20,000 total health detriments. The health effects of the accident scenarios are much smaller than those expected from background radiation.

For persons present onsite, the largest doses and associated health risks would be incurred from an inadvertent criticality in Building 332. For the persons exposed onsite, Table D.2-31 shows the risks predicted for exposure at 100 m from the accident, at the nearest site boundary, and at the western boundary. In the case of a 100-m exposure to an individual from the inadvertent criticality, the probability of fatal cancer is 0.005 to 0.02, and the probability of a total health detriment is 0.009 to 0.02. Additional analysis was conducted to quantitatively characterize the impact of this postulated accident and is discussed below.

Assessment of Potential Fatalities and Elevated Health Effect Risks for the Bounding Radiological Accident

This assessment was performed for the Inadvertent Criticality discussed in section D.2.8.1 to estimate the maximum number of health effects associated with the postulated event and hence bound the number of severe health effects from any other postulated radiological accident at LLNL or SNL, Livermore. The assumptions for dispersion modeling were the same as those used for the lesser radiological accident scenarios with the exception of wind direction.

Three wind directions were chosen for analysis. The first direction was toward the nearest site boundary, the second in the direction that maximized both onsite and offsite effects, and the third in the direction that would have the potential for the greatest number of severe health effects.

For this assessment one receptor point was modeled at the plume centerline at 775 m corresponding to the maximum distance from Building 332 where the plume would expose personnel to levels which would exceed a radiation dose equivalent of 0.5 rem. The plume width at 775 m was assumed to be 60 m (»2sy). It was conservatively assumed that the resulting doses at these receptors would be the same in any direction without regard to protection or dispersion from obstacles such as buildings and trees. Additionally, the entire population in the path of the plume was assumed to be outdoors and exposed to the plume for the entire duration of plume passage.

Although the dose from the plume decreases with increasing lateral distance from the plume centerline, the concentration and duration were conservatively assumed to be uniform laterally at the centerline values. The resulting plume and prompt dose outlines were drawn to scale on an overlay and then rotated on a map of the combined LLNL Livermore site and SNL, Livermore to determine the directions identified above and the affected population. The outline was then applied to the exposed population to estimate the number of people in the areas of concern (Figure D-2).

To account for the effects of the prompt critical dose, two receptor arcs at 40 m (prompt dose ³ 50 rem corresponding to elevated health effect risks) and at 250 m (prompt dose ³ 0.5 rem) were also modeled. Shielding provided by Building 332's concrete walls was not considered. This would be expected to attenuate the prompt dose.

Finally, an estimate of potential fatalities in Building 332 was computed using the average expected population in the building (19), an estimate of personnel in the Radioactive Materials Area (RMA) based on square footage ratio ((20,800)/30,647×19 » 13), and an estimate of the "LD50" zone or area exposed to ³ 450 rem (Turner, 1986). In this estimate, the attenuation of the concrete walls was considered. About 2100 square feet of lab would be exposed to ³ LD50 which predicts that one person would be exposed to a potentially lethal dose. However, since procedures require two personnel to perform most operations (this requirement mitigates the likelihood of personnel error causing this accident), up to four people would be exposed to a radiation dose that exceeds the LD50. The consequences are summarized in Table D.2-33.

Table D.2-29 Risk Estimators for Health Effects from Exposure to Ionizing Radiation

Effect	Risk Factor * (probability per rem per 70 years)
Fatal Cancer	5.0×10^-4
Fatal, nonfatal, and severe genetic effects	7.3×10^-4

* Source: ICRP, 1991.

Table D.2-30 Calculated Collective Radiation Doses to Members of the Public and the Associated Health Risks from Postulated Accident Scenarios

Accident Scenario	Collective 70-yr Dose (person-rem)^a	Range of Estimated Health Effects
Accident Scenario	Collective 70-yr Dose (person-rem)^a	Fatal Cancer	Total Detriment^b
Inadvertent criticality in Building 332	440	0.2	0.3
Tritium release from Building 968	4220	2	3
Plutonium release Building 332	10	0.005	0.007
Tritium release from Building 331	290	0.1	0.2
Americium-241 release from Building 251	430	0.2	0.3
Americium-241 spill and fire near Building 612	1670	0.8	1
Uranium fire at Building 493	560	0.3	0.4
Americium-241 release from Building 625	4030	2	3
Tritium release from Building 298	420	0.2	0.3
Background^c	29,800,000	10,000	20,000

^a Distance out to 80 km in western sector.
^b Defined as the sum of the total number of fatal cancers, nonfatal cancers, and severe genetic effects.
^cAnnual background dose of 300 mR/year.

Table D.2-31 Calculated Onsite and Site Boundary Radiation Doses and the Associated Health Risks from the Postulated Accident Scenarios

Accident Scenario	Individual 70-yr CEDE (rem)	Range of Estimated Probability for a Health Effect for an Individual
Accident Scenario	Individual 70-yr CEDE (rem)	Chance of Incurring a Fatal Cancer in 70 Years	Total Chance of Incurring Health Effects in 70 Years*
Inadvertent criticality in Building 332 - at 100 meters - at nearest site boundary - at western boundary	37 2.0 0.38	2×10^-2 1×10^-3 2×10^-4	2×10^-2 1×10^-3 3×10^-4
Tritium release from Building 968 - at 100 meters - at nearest site boundary - at western boundary	4.2 0.37 0.92	2×10^-3 2×10^-4 5×10^-4	3×10^-3 3×10^-4 7×10^-4
Plutonium release from Building 332 - at 100 meters - at nearest site boundary - at western boundary	0.078 0.0066 0.0016	4×10^-5 3×10^-6 8×10^-7	6×10^-5 5×10^-6 1×10^-6
Tritium release from Building 331 - at 100 meters - at nearest site boundary - at western boundary	0.28 0.026 0.060	1×10^-4 1×10^-5 3×10^-5	2×10^-4 2×10^-5 4×10^-5
Americium-241 release from Building 251 - at 100 meters - at nearest site boundary - at western boundary	3.1 0.14 0.14	2×10^-3 7×10^-5 7×10^-5	2×10^-3 1×10^-4 1×10^-4
Americium-241 spill and fire near Building 612 - at 100 meters - at nearest site boundary - at western boundary	3.7 2.0 0.12	2×10^-3 1×10^-3 6×10^-5	3×10^-3 1×10^-3 9×10^-5
Uranium fire at Building 493 - at 100 meters - at nearest site boundary - at western boundary	0.41 0.15 0.039	2×10^-4 8×10^-5 2×10^-5	3×10^-4 1×10^-4 3×10^-5
Americium-241 release from Building 625 - at 100 meters - at nearest site boundary - at western boundary	30 4.2 .27	2×10^-2 2×10^-3 1×10^-4	2×10^-2 3×10^-3 2×10^-4
Tritium release from Building 298 - at 100 meters - at nearest site boundary - at western site boundary	0.40 0.20 0.11	2×10^-4 1×10^-4 6×10^-5	3×10^-4 2×10^-4 8×10^-5

Background (70 years)	21	1×10^-2	2×10^-2

* Defined as the total chance of incurring fatal cancers, nonfatal cancers, and severe genetic effects.
CEDE = Committed Effective Dose Equivalent.

Table D.2-32 Consequences of Radioactive Transportation Accident Scenarios

Transportation Scenario	Description	Collective Population* Dose (Person-Rem)	Calculated Chance that a Member of the General Population May Contract a Cancerous Health Effect
LSA	Analysis of transportation accidents involving low specific activity radioactive waste shipments (Scenario 1). Type A packaging.	10⁹6	7.3×10^-7
LSA	Analysis of transportation accidents involving low specific activity radioactive waste shipments (Scenario 2). Type A packaging.	843	5.6×10^-7
TRU	TRUPACT analysis for WIPP (DOE, 1990a). Type B packaging.	5.58×10⁴	3.8×10^-5
SST	DOE analysis of safe secure transport of classified nuclear material (DOE, 1988)	19	1.3×10^-8

* 1.1×10⁶ people. Population density = 10⁴/sq mile.

NEWSLETTER

Join the GlobalSecurity.org mailing list

Weapons of Mass Destruction (WMD)