Final Environmental Impact Statement for the Continued Operation of the Pantex Plant

APPENDIX F TRANSPORTATION RISK ANALYSIS

F.1 INTRODUCTION

This appendix contains additional material for support of the impact analyses in the transportation sections of the EIS. Further information regarding Pantex Plant related transportation activities, including nonradiological hazardous wastes and materials, is provided in the Pantex Plant Safety Information Document and the Ross Aviation Transportation Analysis Report (Pantex 1996a; Ross 1994).

F.2 PACKAGING

Pantex Plant utilizes many different containers for packaging of nuclear explosives, explosives and explosive components, and radioactive materials. These containers provide protection against accidental release of hazardous material. Each container is designed for a specific application, and some have firm guidelines for their construction. Specifications for containers used during offsite transportation of radioactive materials are contained in 10 CFR 71. The following are examples of packaging guidelines for intersite shipment of radioactive material.

F.2.1 Limited Quantity/Instruments and Articles

These materials are packaged in strong, tight packages that will not leak the radioactive contents during normal transportation.

F.2.2 Type A Packaging

Type A packages must be designed according to the requirements of 49 CFR Sections 173.24, 173.411, and 173.412. This package type must be adequate to prevent any loss of radioactive material or shielding while subjected to normal (i.e., incident-free) transport conditions. This packaging will be in appropriate inner containers to prevent leakage of the contents with the inner containers overpacked in one of the following containers:

Strong fiberboard box.
Tightly sealed can or drum.
Strong wooden box.
Approved Department of Transportation (DOT)-7A equivalent packaging.
Original containers received from manufacturer, if in good condition.

Materials shall be cushioned to prevent movement, and containers shall be securely closed.

F.2.3 Type B Packaging

Type B packages must meet all of the requirements for Type A packaging as well as withstand test conditions that simulate serious accident damage. The test results must show that only a limited loss of shielding and essentially no loss of material occurs. The test conditions are determined by the Nuclear Regulatory Commission (NRC) in regulation 10 CFR 71.73 and include the following:

A 9-meter (30-foot) free drop onto an unyielding surface.
A dynamic crush test consisting of dropping
A puncture test consisting of a free drop greater than 1 meter (3.3 feet) onto a 15-centimeter (6-inch) diameter steel pin.
A 30-minute thermal exposure at 800 °C (1,475 °F).
Immersion in water for 8 hours (for fissile materials packaging only).

In addition to requirements for accident survivability, the DOT places limits on the external radiological hazards posed by radioactive materials shipments. For both Type A and Type B radioactive materials packaging, the dose rate at any point on the outer surface of the package may not exceed:

200 millirem per hour on the surface of the package.
10 millirem per hour at 1 meter (3.3 feet).

If the packages are transported in an "exclusive use" closed transport vehicle, with the exception of aircraft, then the following maximum dose rates are allowed:

1,000 millirem per hour on the accessible package surface.
200 millirem per hour at the external surface of the transport vehicle.
10 millirem per hour at 2 meters (6.6 feet) from the external surface of the transport vehicle.
2 millirem per hour in any area of the vehicle that a person occupies.

The Pantex Plant Safety Information Document provides a comprehensive listing of containers used at Pantex Plant (Pantex 1996a).

F.3 PIT CONTAINERS

Pit containers are used as a protective barrier for interim pit storage at Pantex Plant and as Type B packages for offsite shipments. The following is a discussion of the pit containers currently in use or planned for future use at Pantex Plant.

F.3.1 AL-R8 Container

These containers were developed by DOW Chemical in the late 1960's and are shown in Figure F.3.1-1. The AL-R8 container was certified as a Type B package in 1974, and was used mainly for movement of pits between Rocky Flats and Pantex Plant. In 1988, a Revised Safety Analysis Report for Packaging was issued stating that the AL-R8 container did not meet all the Federal transportation regulations. In 1991, the use of the AL-R8 for intersite shipments was discontinued. The AL-R8 is now the primary container used for pit staging at Pantex Plant. The containers have a uniform, nominal outside diameter of 51 centimeters (20 inches). All AL-R8 containers are constructed of 18-gauge carbon steel. Within an AL-R8 container, a pit is secured on a metal frame and is surrounded by Celotex (a high-density cane-fiber pressboard) insulation. The AL-R8 does not have an inner containment vessel.

Figure F.3.1-1. CrossSectional View of an AL-RE Container

F.3.2 FL-Type Container

The FL-Type container is currently the only certified container used for pit transport and shown in Figure F.3.2-1. This Type B package has a 16-gauge stainless steel outer containment drum surrounding a 12-gauge stainless steel inner containment drum. Celotex insulation is present between the inner and outer containment drums. The inner containment drum measures 35.1 centimeters (13.8 inches) in diameter and 97 centimeters (38 inches) in height. The outer containment drum measures 57.2 centimeters (22.5 inches) in diameter and 127 centimeters (50 inches) in height. Both the internal and external containment drums are constructed of stainless steel. The inner containment vessel is sealed with dual concentric silicone O-rings. DOE has only 300 of these containers. They are currently used for transporting pits from Pantex Plant to Los Alamos National Laboratory (LANL).

Figure F.3.2-1. Cross-Sectional View of an FL-Type Container.

F.3.3 AT-400A Container

DOE commissioned Sandia National Laboratories, Lawrence Livermore National Laboratory, and LANL to design and build the AT-400A container. This container is expected to replace the AL-R8 for pit staging at Pantex Plant, and be the primary container for offsite shipments. The AT-400A is constructed from 304L stainless steel with a high density insulator foam liner and a welded inner containment vessel (see Figure F.3.3-1). This pit storage container is a container within a container. The pit is contained within an inner containment vessel that slightly resembles a small propane tank used for camping. The wall of the inner containment vessel is 0.635 centimeters (0.250 inches) thick, with a 34.29- centimeter (13.5-inch) inner diameter. A sampling tube exists at the top that is used to periodically sample the gas inside. The environment within the inner containment vessel is controlled and guaranteed via an inert gas backfill. The inner containment vessel sits between two foam filled inserts inside the outer container. The outer container dimensions are approximately 51 centimeters (20 inches) in diameter by 71 centimeters (28 inches) high. This containment vessel is totally fabricated from 304L stainless steel. The AT-400A is scheduled for use at Pantex Plant in late 1996. Sandia National Laboratories has tested prototypes of the AT-400A and a DOT Type B certification is pending. Initial performance testing of the AT-400A package indicates that the container can withstand a considerably more severe accident environment than the FL-type container. The inner containment vessel of the AT-400A is completely welded, which provides superior behavior in fire environments. The AT-400A container also performs significantly better than the FL container in impact, puncture, and crush environments.

Figure 3.3-1. Cross-Section of an AT-400A Container

F.4 TRANSPORTATION VEHICLES AND SAFETY AND SUPPORT EQUIPMENT

Vehicles used for onsite transportation include hardened trailers, flatbed trailers, Safe Secure Tractor Trailers (SSTs), forklifts, tow motors, Stage Right trailers, tractors, electric carts, and transportation carts. This section discusses the Stage Right equipment that is in use at Pantex Plant. Further information on transportation vehicles and equipment is provided in the Safety Information Document and the Transportation Evaluation Report for Ross Aviation, Inc. (Pantex 1996a; Ross 1994).

F.4.1 Stage Right Forklift

The Stage Right forklift is a technologically advanced electric forklift, specifically designed for the Stage Right project for pit storage. This forklift is larger than the other electric forklifts at Pantex Plant and is fully enclosed. The Stage Right forklift rides on a special rail system that keeps it in a straight line throughout its operation. This rail system contains a sensor system that detects the position of the forklift on the rails throughout its operation. The Stage Right forklift can perform its loading operations only when it is correctly positioned between these light sensors. The operator has a display in front of him to notify him when the forklift is in position.

The Stage Right forklift has one boom arm for lifting the Stage Right pallets. This boom arm has a unique safety feature to prevent an inadvertent puncture of a pit container. On the extreme end of the boom there exists a compression sensor that, when compressed, completely stops the action of the Stage Right forklift. The pressure of a person's outstretched arm is enough to compress the sensor and shut down the forklift. The boom has to be in exactly the correct position to fit into the Stage Right pallet boom slot. Once positioned, the Stage Right forklift can continue with its operation. The Stage Right forklift will be replaced by an automated guided vehicle when development at Pantex Plant is completed.

F.4.2 Stage Right Trailer

Pits are transferred between Zone 12 South and Zone 4 in either a double-axle, non-hardened pallet trailer or a hardened trailer. These pallet trailers are able to withstand a 2,722-kilogram (6,000-pound) load or more, and are towed by stakebed trucks. The trailer box is approximately 127 centimeters (50 inches) wide, 203 centimeters (80 inches) high, and 356 centimeters (140 inches) long. The frame of the trailer is the load-bearing structure so that the outer skin and rivets are not subject to any load other than their own weight. An emergency electric braking system is in place to stop the trailer in the event that the connection with the tow vehicle is broken. These Stage Right trailers are specifically built to hold Stage Right pallets (see Figure F.4.2-1).

Figure f.4.2-1. Stage Right Trailer

F.4.3 Stage Right Pallet

Two types of these pallets exist. The first can accommodate four AL-R8 containers, and the second can hold six AL-R8 containers. In this pallet's design, the containers sit both in it and on it depending on the pallet's positioning (see Figure F.4.3-1).

Figure F.4.3-1. Stage Right Pallet.

A new Stage Right pallet that will be able to accommodate both the AL-R8 and AT-400A containers is in the design phase at Pantex Plant. This new pallet will be constructed of stainless steel. The Stage Right pallet that can accommodate four containers is approximately 104 centimeters by 99 centimeters by 74 centimeters (41 inches by 39 inches by 29 inches), and the Stage Right pallet that can accommodate six containers is approximately 155 centimeters by 99 centimeters by 74 centimeters (61 inches by 39 inches by 29 inches).

F.5 INTERSITE TRANSPORTATION IMPACT ASSESSMENTS

The Analysis of Dispersal Risk Occurring in Transportation (ADROIT) code was developed specifically for the Defense Programs Transportation Risk Analysis program and is currently being used in the development of safety analysis for DOE transportation activities. The code was designed specifically to evaluate event trees and conduct consequence assessments for SST shipments and air transport of tritium reservoirs. The code can also be used to evaluate the effect on the assessed risk of modifications to equipment or procedures (positive measures), changes in shipment campaigns, improvements in understanding of cargo, etc.

The assessment covers transportation of nuclear explosives and pits by SST and tritium reservoirs by DC-9 aircraft for the period fiscal year (FY) 1997-2006. Annual risks are computed based on projected shipments associated with current 10-year weapon planning documents.

For nuclear explosives, the baseline 10-year shipments were obtained from current weapons activity schedules. For security reasons, detailed information on the specific weapon systems and routes is classified.

For tritium reservoirs, the baseline shipment schedule corresponds to those for nuclear explosives. The number of tritium reservoirs to be transported between Pantex Plant and the Savannah River Site (SRS) is assumed to be equal to the number of nuclear explosives transported to or from Pantex Plant. All tritium reservoirs will be transported in DC-9 aircraft. Up to 75 tritium reservoirs may be transported per flight. The projected baseline shipments of tritium reservoirs was obtained from current 10-year weapons activity schedules.

Assessments were conducted for pit transportation from Pantex Plant to four alternative storage sites: Nevada Test Site (NTS), SRS, Hanford Site, and Kirtland Air Force Base (KAFB). An assessment was also conducted for transportation of pits between Pantex Plant and LANL and return (for the purpose of quality assurance and testing). Pits are expected to be transported in a DOT-certified container, such as, the AT-400A container.

The estimated risks for nuclear explosive shipments, pit shipments, and tritium reservoir shipments are documented in the following sections. All calculations were conducted with the ADROIT code, Version 1.1, which was developed specifically for transportation of Defense Programs materials. A description of the methodology used in the ADROIT code appears in section F.6.

F.5.1 Nuclear Explosive Shipments

F.5.1.1 Incident-Free Risk

Tables F.5.1.1-1 and F.5.1.1-2 summarize incident-free risks. The health risk from incident-free exposure is obtained by multiplying the exposure by the dose-health effects factor, which is typically taken to be 0.0005 latent cancer fatalities (LCFs) per person-rem (NAP 1990). It should be noted that for greater accuracy the ADROIT code uses a distribution curve rather than a single point estimate of 0.0005.

F.5.1.2 Accident Fatality Risk-Nuclear Explosive Shipments

Table F.5.1.2-1 summarizes the accident fatality risk. The estimated risk for the baseline shipment schedule is highest in FYs 1997 and 2005, which are the years in which the shipment volume is greatest. The risk curves for accident fatality risk in FY 1997 are shown in Figure F.5.1.2-1.

F.5.1.3 Accident Dispersal Risk-Nuclear Explosives

Table F.5.1.3-1 summarizes the dispersal risk from weapons shipments. The estimated risk for the baseline is FY 1997, which is the year with the greatest shipment volume. The risk curves for accident fatality risk in FY 1997 are shown in Figure F.5.1.3-1. The probability of having more than one LCF arising from a dispersal event associated with nuclear explosive shipments to or from Pantex Plant is most likely between 2 x 10^-9and 5 x 10^-7 (the 5th and 95th percentiles) with a median estimate of approximately 4 x 10^-8. The shape of each individual risk curve reflects randomness in the outcome. The difference between the 5th percentile, median, and 95th percentile risk curves reflects the calculated uncertainty in the estimates. The baseline risk values given in Table F.5.1.3-1 are the average values of the areas under the complete set of risk curves for each year considered.

The dominant risk environments are highway accidents involving severe collisions and fires and accidents involving very long duration fires. There is no single, dominant scenario associated with credible (>10^{-7 yr-1}), high-consequence events. For the purposes of this discussion high-consequence is defined to be greater than 1 LCF. Instead there are a collection of scenarios that together constitute credible, high-consequence events. These scenarios have the following common characteristics:

The accident includes either a severe collision and fire (such as a collision with a heavy truck or fixed object that also involves a fuel fire) or a very long duration fuel fire (such as an accident with a fuel tanker or train that involves a fuel fire).
The accident results in violent reaction of the high explosive.

Similarly, there is no single dominant scenario associated with credible, moderate consequence event. For the purposes of this discussion moderate consequence is defined to be less than 1 LCF. The scenarios that dominate the moderate-consequence events have the following common characteristics:

The accident includes either a severe collision and fire (such as a collision with a heavy truck or fixed object that also involves a fuel fire) or a very long duration fuel fire (such as an accident with a fuel tanker or train that involves a fuel fire).
The accident results in a fire-driven dispersal and/or the accident occurs in a rural area.

F.5.2 Pit Shipments

F.5.2.1 Incident-Free Risk

Table F.5.2.1-1 provides unit risk factors for intersite pit shipments. The incident-free exposure health risk is obtained by multiplying the exposure by the dose-health effects factor, which is typically taken to be 5.0 x 10^-4LCFs per person-rem (NAP 1990).

Approximately 63 percent of the collective exposure (and health risk) is received by people on the roadway. About 33 percent is received by members of the public at rest stops. The balance of the collective exposure is received by people off the roadway. By contrast, the maximum individual dose (and risk) is received by an individual off the roadway. This is because all of the shipments are either to or from Pantex Plant, so an individual living in the vicinity of the plant near the roadway is assumed to be exposed to the intrinsic radiation from all the shipments, whereas the people sharing the roadway or at rest stops are not likely to include the same individuals for all (or even most) shipments.

F.5.2.2 Accident Fatality Risk

Table F.5.2.1-1 summarizes the accident fatality risk. The estimated risk is highest for the alternative involving interim storage at SRS. The risk curves for annual accident fatality risk for 2,000 pit shipments to SRS are shown in Figure F.5.2.2-1.

Figure F.5.2.2-1. Fatality Risk Curves for 2,000 Pit Shipments to the Savannah River Site

F.5.2.3 Accident Dispersal Risk

The accident dispersal risk for pit shipments is based on the performance characteristics of the FL-Type container because performance data on the AT-400A could not be incorporated into the risk assessment in time for this EIS. However, performance testing of the AT-400A indicates that it can withstand considerably more severe accident environments than the FL-type container. The containment vessel in the AT-400A container is completely welded, which provides superior behavior in fire environments (in terms of preventing a dispersal). The AT-400A container also performs significantly better than the FL-type container in impact, puncture, and crush environments. Consequently, the estimates provided here for accident dispersal risk (based on the FL-type container) are believed to significantly overstate the actual risk of shipping pits in the AT-400A container. In fact, the probability of releasing plutonium from a pit in an AT-400A container during SST transportation may be negligible.

Table F.5.2.3-1 summarizes the accident dispersal risk for pit shipments in the FL-type container. The estimated risk is highest for the alternative involving interim storage at the Hanford Site. The risk curves for annual accident dispersal risk associated with 2,000 pit shipments between Pantex Plant and Hanford Site are shown in Figure F.5.2.3-1. The probability of having more than 1 LCF arising from a dispersal event associated with pit shipments from Pantex Plant is most likely less than 7 x 10^-8(the 95th percentiles). The shape of each individual risk curve reflects randomness in the outcome whereas the difference between the median and 95th percentile risk curves reflects the calculated uncertainty in the estimates. The mean risk values given in Table F.5.2.3-1 are the average values of the areas under the complete set of risk curves for each year considered.

Figure F.5.2.3-1. Accident Dispersal Curves for 2,000 Pit Shipments to the Hanford Site.

The dominant risk environments are highway accidents involving severe collisions and fires and accidents involving very long duration fires. There are no credible (>10^-7
yr-1), high-consequence events. For the purposes of this discussion high consequence is defined to be greater than 1 LCF.

There is no single dominant scenario associated with credible, moderate-consequence event. For the purposes of this discussion moderate consequence is defined to be less than 1 LCF. The scenarios that dominate the moderate-consequence events have the following common characteristics:

The accident includes either a severe collision and fire (such as a collision with a heavy truck or fixed object that also involves a fuel fire) or a very long duration fuel fire (such as an accident with a fuel tanker or train that involves a fuel fire).
The accident results in a fire-driven dispersal.

F.5.3 Tritium Reservoir Shipments

Tritium reservoirs are shipped using DC-9 aircraft operated by a DOE contractor. Because members of the public are not in close proximity to the aircraft, intrinsic radiation and accident fatalities are much less of a hazard to members of the general public than they are in highway transportation. Therefore, the incident-free risk and the accident fatality risk are not calculated for tritium reservoir shipments.

F.5.3.1 Accident Dispersal Risk

Table F.5.3.1-1 summarizes the accident dispersal risk. The estimated risk for the baseline shipments is highest in FY 1997, the year that the shipment volume is greatest. The risk curves for accident dispersal risk in FY 1997 are shown in Figure F.5.3.1-1. The probability of having more than 1 LCF arising from a dispersal event associated with tritium reservoir shipments to and from Pantex Plant is most likely between 4 x 10^-12 and 2 x 10^-9 (the 5th and 95th percentiles) with a median estimate of about 4 x 10^-11. The shape of each individual risk curve reflects randomness in the outcome. The difference between the 5th percentile, median and 95th percentile risk curves reflects the calculated uncertainty in the estimates. The mean risk values given in Table F.5.3.1.-1 are the average values of the areas under the complete set of risk curves for each year considered. The dominant risk environment is accidents at or near an airport that involve severe crashes and/or fires. There are no credible (>10-7 ^yr-1), high-consequence events. For the purposes of this discussion high-consequence is defined to be greater than 1 LCF.

Figure F.5.3.1-1.Accident Dispersal Risk Curves for Fiscal Year 1997, Tritium Shipments

There is not a single dominant scenario associated with credible, moderate consequence event. For the purposes of this discussion moderate consequence is defined to be less than 1 LCF. The scenarios that dominate the moderate consequence events have the following common characteristics:

The accident includes either a severe crash and/or a fire.
The accident occurs either at or near the airport.

F.6 ADROIT METHODOLOGY

The ADROIT code was used to calculate the incident-free risk, the dispersal risk, and the accident risk from intersite shipments expected in the timeframe of this EIS.

F.6.1 Incident-Free Risk

The ADROIT code utilizes the RADTRAN-IV methodology to calculate incident-free exposures. The transportation of radioactive material results in some radiological exposure to the general public along the route. Included in the ADROIT computer code incident-free risk calculations for transport are models describing exposures to persons (e.g., residents) adjacent to the transport route (off-link exposures), exposures to persons (e.g., passengers on passing vehicles) sharing the transport route (on-link doses), and exposures to persons at stops (e.g., residents and truck crews not directly involved in the shipment).

For calculational purposes, each SST is modeled as a point radiation source located at the geometric center of the trailer. The source strength of the radiation is usually given in terms of the Transportation Index, which is a measure of the source strength 1 meter (3.3 feet) from the "package" surface (For weapon shipments, a dose rate of 3 millirem per hour at the surface of the trailer was used (PC 1995a). For pit shipments, a dose rate of 1 millirem per hour at 1 meter from the trailer surface was used).

As an SST carrying radioactive material traverses its route, people in the area adjacent to the route will receive low levels of radiation. The dose each person receives depends on his distance from the route and the speed of the source as it passes them. A combination of geometric dispersion and absorption of the radiation in the environment outside of the SST produces a rapid drop-off of radiation dose with distance from the source. Also, the faster the shipment passes the less exposure time there is for people along the route. Data from the 1990 Census was used to compute population density as a function of location along the roadway.

Vehicles sharing the roadway with the shipment have the potential for exposure to the radiation source. The basic approach for treating people on the roadway is very similar to that used for people off the roadway. An effective population density for the roadway is obtained and in all the calculations the population distribution is treated as a continuum. To keep the model both simple and conservative, the shielding provided by the vehicles transporting people on the roadway is neglected.

Typical SST shipment schedules include stops for fuel, meals, and rest/driver change. During those stops the public in the vicinity of the SST is exposed to a stationary source of radiation. The total rest stop time is taken to be 25 percent of the total driving time.

F.6.2 Accident Fatality Risk

The probability of fatalities due to direct effects of the accident environment (i.e., blunt trauma and/or burns to vehicle occupants, pedestrians and/or bystanders) is calculated using the ADROIT code based on the event tree shown in Figure F.6.2-1.

Figure F.6.2-1. Event Tree for Calculation of Fatality Risk

The probability of a fatal accident involving an SST is calculated, first, by estimating the annual probability, based on a given shipment campaign, of an SST accident severe enough to require that the vehicle cannot be driven. This value is obtained from the overall SST tow-away rate per mile for a given type of roadway and the distance traveled on specific types of roadways. The probability of a fatal accident and the number of fatalities given a tow-away accident is obtained from historical databases describing tractor-trailer accident severities (NAP 1990; SNL 1993).

F.6.3 Accident Dispersal Risk

Radioactive materials transported between sites include plutonium, uranium, and tritium. Other than relatively low levels of intrinsic radiation (which is considered in the incident-free analysis), plutonium and uranium do not pose a significant health hazard unless they are converted to an aerosol with respirable size particles. Three mechanisms by which aerosol may be generated and released are considered in the ADROIT code: violent reaction of high explosive(s) (HE), oxidation in a fire, and spalling and breakup of the surface oxide layer by mechanical forces.

There are three basic elements of the accidental dispersal risk assessment: Probabilities, Consequences, and Uncertainties. Probabilities of release by the three mechanisms that can produce respirable-sized aerosols and specific consequence scenarios are developed based on an event tree analysis. Consequences are evaluated for each end event in the tree through an assessment which integrates dispersal calculations, route characterization, population data, and dose-health effects models to provide an estimate of excess LCFs. Uncertainties are evaluated by incorporating Latin hybercube sampling into the calculations for probabilities and consequences.

F.6.3.1 Event Tree

The basis for the probability analysis in the ADROIT code is an event tree analysis. An event tree is a logic diagram that describes accident sequences that can lead to specific consequence scenarios. The event tree is composed of questions that define the types and severities of transportation accidents that occur, the resulting damage to the transporter and cargo, release mechanisms, accident locations, and the meteorological conditions. The event tree developed for the ADROIT code to analyze transportation of radioactive material in SSTs consists of 17 questions. These questions are depicted in Figures F.6.3.1-1 through F.6.3.1-4. A similar event tree is used for air transportation but is not discussed further.

Figure F.6.3.1-1 ADROIT Dispersal Event Tree Questions 1-5.

Figure F.6.3.1-2 ADROIT Dispersal Event Tree Questions 6-9.

Figure F.6.3.1-3 ADROIT Dispersal Event Tree Questions 10-14.

Figure F.6.3.1-4 ADROIT Dispersal Event Tree Questions 15-17.

The initiating events for the tree are traffic accidents in one of four operating environments. The operating environments are based on road type and population area. Although the structure of the tree is the same for all four initiating events, the quantification of some of the branches depends on the operating environment. The initiating events are quantified in terms of an annual probability of occurrence. All other branches of the tree are quantified in terms of conditional probabilities.

The remainder of this discussion is regarding the initiating events and the branches of the event tree. Details of the response of the packaging system (the packaging system includes the SST, container, and cargo) are limited because of the sensitive nature of the information involved in the evaluation of package response.

F.6.3.2 Initiating Event Probabilities

Highway transportation accidents are evaluated in four operating environments based on road type and population (SNL 1994a). Road types are divided into limited access roads and non-limited access roads. The distinction in road type is made because accident rates and the frequency of different accident types vary significantly with road type. Population areas are divided into urban and rural areas. The distinction in population area is retained primarily to capture the difference in the size of the population exposed given a dispersal event.

The initiating event probabilities are calculated as the product of the following: the overall tow-away accident rate per mile for SST transport; the fraction of tow-away accidents that have severities comparable to fatal accidents; influence factors that relate the overall accident rate to the operating environment of interest; and the mileage in the operating environment of interest.

The operating history with SST transport is sufficiently long to define an overall tow-away accident rate. The mean estimate for the rate of tow-away accidents involving an SST is 0.066 per million miles (SNL 1994a). However, the number of accidents experienced with SST transports is not sufficient to quantify the accident rate in the operating environments of interest or the types and severities of accidents. Thus, general commerce data for heavy truck transportation is used as a surrogate for SST data to quantify the relative accident rates in different operating environments and the types and severities of accidents.

The most comprehensive database available on heavy truck accidents is the Trucks Involved in Fatal Accidents (TIFA) database (UM 1991; UM 1993), which is maintained by the University of Michigan Transportation Research Institute. This database contains information on fatal accidents. Fatal accidents provide a good match with accidents of sufficient severity to threaten cargoes carried within SSTs.

Because of the lack of details for SST accident rates in operating environments of interest and the types and severities of accidents it was necessary to introduce a factor which could provide a bridge between the tow-away accident rate and the severities and operating environments of interest. The bridge is provided by modifying the tow-away accident rate with an estimate of the fraction of tow-away accidents that are considered to have severities comparable to fatal accidents. The fraction is estimated to be between 0.042 and 0.51 (SNL 1994b). In general, the tow-away rate includes accidents of lesser severity than those associated with fatal accidents.

Because the probability of occurrence of accidents of varying severities is different for different road types and the consequences of accidents differ because of differing population densities, highway transportation accidents were evaluated in four operating environments based on road type and population area. Table F.6.3.2-1 presents the four influence factors (SNL 1994a).

The mileage in each operating environment is derived from shipment projections and route characterization. The TIGER/Line files (Census 1991; Census 1991a) were used to develop route segmentation data files, which include information on geographic location, operating environment, and applicable meteorological stations as a function of route mile marker (cumulative distance from the route origin) at closely-spaced points.

F.6.3.3 Accident Environments

Radioactive materials are shipped in packaging systems that mitigate accident environments and help prevent releases to the environment. Normal transportation environments do not produce environments that threaten the integrity of the packaging system. However, the environments produced from very severe traffic accidents could exceed the capabilities of the packaging system and cause a release of radioactive material.

The risk assessment implemented in the ADROIT code considers impact, puncture, crush, and thermal environments. In traffic accidents, impact, puncture, and crush environments are associated with collision and rollover events; thermal environments are associated with fires involving the fuel system, cargo, or other elements of the vehicles and/or objects involved in the accident. The response of the packaging system to these environments is interdependent. For example, the response of the packaging system to a fire reflects damage to the packaging system caused by collision and rollover.

The accident data needed to define the probability of packaging system failure include probabilities of various accident types and distributions of collision, rollover, and fire severity. The response of the packaging system and the collision, rollover, and fire severity depend on the type of accident. There are a large number of accident variables that can be used to characterize the accident. In the ADROIT code, the emphasis is placed on defining the minimum number of accident characterizations that would provide sufficient definition of the accident environment to analyze the response of the packaging system and also provide reasonable differentiation of the collision, rollover, and fire severities. As a result, the variables used to characterize accidents were selected based in part on specific features of the packaging system used by DOE to transport radioactive material.

The packaging system used by DOE to transport radioactive material includes the cargo, containers, and the SST. The most important features of the SST that protect the cargo in accident environments include the walls of the trailer, which provides significant thermal protection in fire environments, and the cargo tie-down system, which provides a means of holding the cargo in place. In general, damage to the packaging system in an accident depends on a number of factors:

If the accident involves a collision, accident characteristics that may affect the response of the packaging system include the location of the principal impact, the impact direction, the collision energy absorbed, the peak contact velocity, and the collision duration. Specifically, damage to the trailer walls depends on the location of the principal impact, the impact direction, and the collision energy absorbed. The damage to the container and cargo depends on the peak contact velocity and the collision duration.
If the accident involves a rollover, the primary characteristic of the rollover used to evaluate the response of the packaging system (primarily damage to trailer walls) is the skid distance.
If the accident involves a fire, additional accident characteristics that affect the response of the packaging system include the effective fire temperature, the size of the fire, the separation between the fire and the trailer, and the duration of the fire.

The types of vehicles and/or other objects involved, collision occurrence, angle of impact, location of principal impact, rollover occurrence, and fire occurrence are factors that define the types of accidents. The peak contact velocity, skid distance, effective fire temperature, fire size, fire separation, and fire duration are factors that define the severity of the accident.

The branches of Questions 1-4 and 8 in the ADROIT event tree define the factors used to characterize the type of accident. Questions 9-11 are used to describe the fire separation, fire size, and fire temperature. The peak contact velocity, skid distance, and fire duration are used in the evaluation of the branch probabilities for Questions 6, 7, 12, and 14. Details of the statistical distributions used to describe these input variables can be found in A Statistical Description of the Types and Severities of Accidents Involving Tractor Semi-Trailers (SNL 1994b).

F.6.3.4 Response of Packaging System to Accident Environments

The response of the packaging system to the accident environments is addressed in Question 5-7 and 12-14. The possible responses to collisions, rollovers, and fires are considered. The response states address mechanisms that lead to release of radioactive material (e.g., violent reaction of HE by either mechanical or thermal initiation) as well as damage to the SST walls and cargo that may affect subsequent thermal response to a fire.

Question 5 is used to define the type of mechanical environments to which the cargo may be subject. Question 6 addresses the damage to the packaging system associated with collision events. Twenty-five collision damage states, which are determined from combinations of SST wall damage and cargo damage, are represented in the event tree. The damage states are hierarchical and are listed from the most severe to no significant damage. For scenarios in which a collision occurs, the branch probabilities are obtained by the split fraction method, which entails evaluating the probability that the peak contact velocity is greater than a calculated threshold value for the damage state of interest.

Question 7 is used to define the SST wall damage from rollover. Damage due to rollover is limited to the outer skin and insulation; the probability of opening in the SST walls or damage to the cargo from rollover is considered negligible. For scenarios in which a rollover occurs, the branch probability is obtained by evaluating the probability that the skid distance is greater than the threshold value for the damage state of interest.

The response of the cargo to fire environments is addressed in Questions 12-14. For scenarios in which a fire occurs, the probability that the fire duration is greater than the minimum fire duration for HE ignition t*, which depends on the cargo of interest, cargo damage, total wall damage, effective fire temperature, fire separation, and fire size. Since collision and rollover can occur in the same accident, the total wall damage is obtained from the combination of the wall damage from collision with that from rollover. The computer code MELTER was developed to calculate t* (SNL 1994).

Question 13 is used to define the probability of a thermally initiated violent reaction of the HE given ignition. If a violent reaction of the HE does not occur, Question 14 is used to assess the probability that aerosol is generated by oxidation of the radioactive material.

F.6.3.5 Consequence Scenario

For a given release mechanism, Questions 15-18 provide the remaining conditions needed to define a consequence scenario. For a given scenario (which defines the operating environment and the route), specific locations are sampled randomly from that segment of the route and operating environment considered. The sampling density is higher in urban areas than in rural areas. The location of the accident affects both the distribution of meteorological stability and the exposed population. The probabilities of the meteorological stability classes depend on the accident location and are obtained from data recorded at stations operated by the National Climatic Data Center.

F.6.3.6 Consequence Assessment

A consequence assessment estimates the health and environmental effects from radioactive materials transport. Health consequences are given in terms of the expected number of excess LCFs produced in the exposed population. The number of excess LCFs is determined from the collective committed effective dose and the latest dose-to-risk conversion factors. The ADROIT code performs dispersal analysis using the Explosive Release Atmospheric Dispersion (ERAD) code and exposed populations determined from route characterization and population counts obtained from 1990 Census data. The dispersal analysis depends on the dispersal mechanism, meteorological stability, and the cargo of interest. The exposed population depends on the accident location and wind direction.

F.7 OCCUPATIONAL RADIATION IMPACTS FROM MATERIAL HANDLING

Occupational exposure to radiation from interzone material movements at Pantex Plant and alternative sites was estimated using historical annual exposure information for transportation and staging workers. This information was correlated with the number of nuclear explosive and pit movements to obtain a dose per material movement. This determination was performed to ensure a bounding estimate that at the same time provides a realistic estimate of future impacts. Historical information was used instead of time and motion estimates to preclude the introduction of a large number of uncertainties in the estimates. Historical information from similar activities provides a more accurate estimate of future exposures.

Historical dosimetry information indicates a conservative dose estimate of 6.5 x 10^-4person-rem per handling operation for the transportation and staging department. Handling operations include interzone transfers of weapons and weapons components with associated vehicle loading/unloading as well as SST loading/unloading. For the 10-year period under evaluation in this EIS, a conservative estimate of 94,000 handling operations was made resulting in a cumulative dose of 61 person-rem to the transportation and staging workers.

The bounding estimate of handling operations for the Proposed Action includes the transfer of 20,000 pits from Zone 4 to Zone 12 with a subsequent return to Zone 4 for the pit repackaging project. The estimate also includes the transfers of weapons and weapons components at an operational level of 2,000 weapons per year. The conservative estimates for both dose per handling and quantity of handling events were made to bound the impacts expected from Zone 4 operations with associated material handling and interzone transfers.

Because there were no identifiable historical doses exclusively relating to large numbers of SST pit container loading or unloading operations, these radiological exposures were obtained using time and position estimates for placing and securing pit containers within an SST. Figure F.7-1 shows the required restraint for pit containers within an SST. The following assumptions were used to estimate worker exposures for pit container loading/unloading operations:

To position and restrain a pit container within an SST requires 15 minutes.
An AT-400A pit container dose rate of 1.5 millirem per hour at 1 meter (3 feet) was used.
Two personnel are required for the pit container restraint operation.
During the loading operation, the workers spend 5 minutes at a distance of approximately 1 meter (3 feet), and 10 minutes at approximately 0.3 meter (1 foot) from the pit containers.
During the loading operations, the workers are exposed to the pit container being loaded and the nearest seven pit containers.
Impacts from unloading operations are the same as for loading operations
No credit is taken for protective gear (e.g., lead aprons).

F.8 REGULATIONS

This section briefly describes the regulatory environment for transporting radioactive materials and wastes out of Pantex Plant. These materials must comply with both DOE requirements and DOT regulations governing offsite shipments.

DOT has the responsibility for establishing transportation regulations, but the enforcement of these regulations is shared with the Federal Highway Administration, the Federal Aviation Administration, the National Highway Traffic Safety Administration, the Native American tribes, and the states. Other aspects of radioactive and mixed materials or wastes transportation are regulated by NRC, the Environmental Protection Agency, and the Occupational Safety and Health Administration. The Environmental Protection Agency is responsible for hazardous wastes, and Occupational Safety and Health Administration is concerned with the safety of workers. The Federal Emergency Management Agency has the responsibility of coordinating the Federal assistance, planning, and training for any and all types of emergency response situations (including transportation accidents) with local, tribal, and state governments.

The DOT and NRC share the regulatory responsibility for radioactive materials and wastes containers. The NRC regulates, reviews, and certifies Type B containers. The DOT, with the NRC's consultation, regulates all other radioactive materials packaging. DOE has the authority of DOT (under 49 CFR section 173.7) to give approval for certain packaging and certain operational aspects of its research, defense, and contractor-related shipments of materials requiring Type B containers. However, in this process the DOE must use NRC-equivalent standards and procedures.

Because of the stringent regulations on transport packaging manufacture, shipment identification, package and vehicle inspections, and routing and driver training, there has never been a documented death or significant injury associated with radioactive materials transport in the DOE complex for more than 40 years.

Table F.8-1 is a list of regulations and standards that govern the transportation processes at Pantex Plant. (.pdf)

REFERENCES

References for Appendix F (.pdf)

NEWSLETTER

Join the GlobalSecurity.org mailing list

Weapons of Mass Destruction (WMD)

REFERENCES