PDF Version

1.0 OBJECTIVE

The objective of this transportation risk assessment is to determine the impacts of the transportation of unirradiated uranium in the form of metallic billets, UO₃ powder, and finished and unfinished N Reactor fuel elements from the Hanford Site, Washington, to Portsmouth, Ohio. The radiological risk is determined for both incident-free transport and transport involving potential accidents. The toxicological consequences are determined for the case in which a credible accident is assumed to occur, without regard for the frequency, and thus the risk, of such an accident.

2.0 REFERENCES

ANS, 1991, Neutron and Gamma-Ray Fluence-to-Dose Factors, ANSI/ANS-6.1.1-1991, American Nuclear Society, La Grange Park, Illinois.

Craig, D.K., 1999, ERPGs and TEELs for Chemicals of Concern: Rev. 15 Abbreviated (January 4, 1999), WSMS-SAE-99-0001, Westinghouse Safety Management Solutions, Aiken, SC.

DOE, 1994, Airborne Release Fractions/Rates and Respirable Fractions for Nonreactor Nuclear Facilities, DOE-HDBK-3010-94, U. S. Department of Energy, Washington, D. C.

DOE, 1994b, Primer on Spontaneous Heating and Pyrophoricity, DOE-HDBK-1081-94, U. S. Department of Energy, Washington, D. C.

DOE, 1997, Introduction to the Emergency Management Guide, Volume I, DOE-G-151.1-1, U. S. Department of Energy, Washington, D. C.

FDH, 1999, Safety Analysis Report for Packaging, Steel Banded Wooden Shipping Containers, HNF-SD-TP-SARP-019, Rev. K, Fluor Daniel Hanford, Richland, Washington.

Ferrell, P. C., 1999, Personal Communication, 8/17/99.

Green, J. R., 1995, Transportation Impact Analysis for the Shipment of Low Specific Activity Nitric Acid, WHC-SD-TP-RPT-015, Westinghouse Hanford Company, Richland, WA.

Herrmann, O. W., and R. M. Westfall, 1997, ORIGEN-S: Scale System Module to Calculate Fuel Depletion, Actinide Transmutation, Fission Product Buildup and Decay, and Associated Radiation Source Terms, NUREG/CR-0200, Rev. 5, U.S. Nuclear Regulatory Commission, Washington, D.C.

Hey, B. E., 1993a, GXQ Program Users' Guide, WHC-SD-GN-SWD-30002, Rev. 0, Westinghouse Hanford Company, Richland, Washington.

Hey, B. E., 1993b, GXQ Program Verification and Validation, WHC-SD-GN-SWD-30003, Rev. 0, Westinghouse Hanford Company, Richland, Washington.

ICRP, 1991, “1990 Recommendations of the International Commission on Radiological Protection,” ICRP Publication 60, Annals of the ICRP, 21 (1-3), Pergamon Press, New York, N. Y.

Johnson, P.E., D.S. Joy, D.B. Clarke, and J.M. Jacob, 1992, INTERLINE 5.0 – An Expanded Railroad Routing Model: Program Description, Methodology, and Revised User’s Manual, ORNL/TM-12090, Oak Ridge National Laboratory, Oak Ridge, Tenn.

Johnson, P.E., D.S. Joy, D.B. Clarke, and J.M. Jacob, 1993, HIGHWAY 3.1 – An Expanded Highway Routing Model: Program Description, Methodology, and Revised User’s Manual, ORNL/TM-12124, Oak Ridge National Laboratory, Oak Ridge, Tenn.

Lawson, K.A., 1987, T-Hopper Shipping Container Handling Procedures and Container Criteria, FMPC-2066, Feed Materials Production Center, Cincinnati, Ohio.

Lide, D. R., Ed., 1993, CRC Handbook of Chemistry and Physics, 74^th Ed., CRC Press, Boca Raton, Fla.

McCoy, J. C., 1998, WMNW Computer Program Verification for SCALE 4.3, EBU-SQA-002, Revision 1, Waste Management Federal Services, Inc., Northwest Operations, Richland, Wash.

Neuhauser, K. S. and F. L. Kanipe, 1989, RADTRAN 4: Volume 2 Technical Manual, SAND89-2370, Sandia National Laboratories, Albuquerque, New Mexico.

Neuhauser, K. S. and F. L. Kanipe, 1992, RADTRAN 4: Volume 3 User’s Guide, SAND89-2370, Sandia National Laboratories, Albuquerque, New Mexico.

NRC, 1977, Final Environmental Statement on the Transportation of Radioactive Material by Air and Other Modes, Volume 1, NUREG-0170, U. S. Nuclear Regulatory Commission, Washington, DC.

NRC, 1982, Atmospheric Dispersion Models for Potential Accident Consequence Assessments at Nuclear Power Plants, Regulatory Guide 1.145, U.S. Nuclear Regulatory Commission, Washington, D.C.

NRC, 1987, Shipping Container Response to Severe Highway and Railway Accident Conditions Main Report, NUREGCR-4829-V1, U. S. Nuclear Regulatory Commission, Washington, DC.

Rittmann, P. D., 1995, ISO-PC Version 1.98 - User's Guide, WHC-SD-WM-UM-030, Westinghouse Hanford Company, Richland, Wash.

Rittmann, P. D., 1996, “Summary of Changes to ISO-PC Version 2.1,” CCC-636, ISO-PC 2.1: Kernel Integration Code System for General Purpose Isotope Shielding Analyses, Radiation Safety Information Computational Center, Oak Ridge, Tenn.

Saricks C. and T. Kvitek, 1994, Longitudinal Review of State-Level Accident Statistics for Carriers of Interstate Freight, ANL/ESD/TM-68, Argonne National Laboratory, Argonne, IL.

UO-555-001, 1985, Plant Operating Procedure, Load T-Hopper onto Flatbed Rail Car, Rockwell Hanford Operations, Richland, Wash.

WHC, 1992, Finished Fuel Assembly Shipping Package Safety Analysis Report for Packaging (Onsite), WHC-SD-NR-SARP-001, Westinghouse Hanford Company, Richland, Wash.

3.0 ASSUMPTIONS, RESULTS, AND CONCLUSIONS

The following assumptions were made in the risk calculations for all payloads.

• Risk calculations were made with the computer code RADTRAN version 4.0.19.SI (Neuhauser and Kanipe 1989 and 1992). Assumptions for specific parameters in the RADTRAN code are given in Section 4.0. Input files are given in Section 5.0.

• Routes were obtained using the computer code Highway version 3.3 (Johnson et al. 1993) for the truck routes, and the computer code Interline version 5.0 (Johnson et al. 1992) for the rail routes. Output files are given in Section 5.0

• Mileage through each zone of population density (rural, suburban, and urban) was aggregated along the entire route, and national average accident rates from Saricks and Kvitek (1994) were applied to each zone.

• Eight accident severity categories and the corresponding severity fractions for truck and rail transport were taken from NRC (1977).

• The shipments were exclusive use based on calculated dose rates.

The following assumptions were made specifically in the risk calculations for the uranium billets.

• Release fractions for Category 1 accident severity were assumed to be zero, and 1.0 for Categories 2 through 8. The Category 2 and 3 release fractions are conservative by a factor of 100 and 10, respectively, compared to values for Type A containers given in NRC (1977).

• Aerosol fractions and respirable fractions were taken from DOE (1994a) for the complete oxidation of uranium metal in a fire.

• The conveyance was a truck, with a trailer of width 3 m.

• The container was assumed to be the G-4255 Wooden Box (FDH 1999), with interior dimensions 8 in. x 24.125 in. x 30.75 in.

• A total of 75 shipments for the campaign of billets was used, based on a total of 234 MTU, 175 kg U per billet, 3 billets per box, 6 boxes per shipment.

• A dose rate of 0.086 mrem/h at 1 m from the edge of the conveyance was used based on the shielding calculation included in the Appendix in Section 5.1.

The following assumptions were made specifically in the risk calculations for the UO₃ powder.

• Release fractions for Category 1 accident severity were assumed to be zero, 0.1 for Category 2, and 1.0 for Categories 3 through 8. The Category 2 and 3 release fractions are conservative by a factor of 10 compared to values for Type A containers given in NRC (1977).

• Aerosol fractions and respirable fractions were taken from DOE (1994a) for powder with particle diameter less than 2 mm in metal containers.

• Both truck and rail conveyances were modeled.

• Two routes were considered, a direct route and an indirect route through Paducah, Kentucky.

• The container was assumed to be the T-Hopper (Lawson 1987), a cone-shaped container enclosed in a 5 ft x 5 ft x 6 ft steel frame.

• A total of 5 shipments for the campaign of powder via rail were modeled, based on a total of 147 T-Hoppers, 10 T-Hoppers per rail car, 3 rail cars per shipment.

• A total of 49 shipments for the campaign of powder via truck were modeled, based on 3 T-Hoppers per truck.

• A dose rate of 0.73 and 0.44 mrem/h at 1 m from the edge of the railcar and truck trailer, respectively, was used based on decay and shielding calculations. A discussion of the shielding calculation is included in the Appendix in Section 5.1.

The following assumptions were made specifically in the risk calculations for the finished and unfinished fuel elements.

• Release fractions for boxes of finished fuel were those recommended for Type A containers. For unfinished fuel, the release fractions were the same as for the UO₃.

• Aerosol and respirable fractions were the same as for the billets.

• Only a direct route by truck was modeled.

• The container was assumed to be the G-4214 Wooden Box (FDH 1999), with interior dimensions 30 in. x 14.125 in. x 8.375 in.

• The campaign of finished fuel was assumed to require a total of 537 shipments; 94 shipments for the campaign of unfinished fuel. Note that these numbers are based on preliminary, unpublished criticality-based shipment limits (Ferrell 1999) for each ²³⁵U enrichment content.

• Dose rates at 1 m from the vehicle edge of 0.023 - 0.052 mrem/h for the various ²³⁵U enriched fuels were calculated based on an assumed box arrangement, assumed box loadings, box capacity, and shipment limits. The shielding calculation is addressed in Section 5.1.

A small amount of UO₂ is also to be transported. The UO₂ consists of 4.86 metric tons uranium enriched in ²³⁵U to levels between 0.2 and 4.31%, with a weighted average of 1.12%. Because a shipping container for this material has not been identified, this payload is not analyzed.

Table 1 gives the total radiological risks from the shipping campaigns of the billets, powder, and fuel payloads. The total radiological risk is broken into contributions from incident free transport, i.e., during which no accidents occur, and from accidents during transport, which account for the probabilities and content releases of accidents of various severity. The total detriment is the number of fatal cancers, non-fatal cancers, and severe hereditary effects weighted by the severity of that effect. Fatal cancers are given the maximum weight of 1.

Table 2 gives the toxicological consequences from a potential accident involving a single shipment. As these values are consequences rather than risk, they cannot be compared directly to the radiological risk values in Table 1, because a risk assessment weights the consequences by the frequency (or probability) of occurrence of the release.

Table 1 Radiological Risk from Uranium Shipments

4.0 EVALUATION

4.1 Methodology

The RADTRAN 4 computer code (Neuhauser and Kanipe 1992) was used to analyze the risks of transporting unirradiated uranium in the form of metallic billets, UO₃ powder, and N Reactor fuel elements from the Hanford Site in Washington State to the DOE site near Portsmouth, Ohio. RADTRAN was originally developed by Sandia National Laboratories (SNL) in conjunction with the preparation of NUREG-0170, Final Environmental Statement on the Transportation of Radioactive Material by Air and Other Modes (NRC 1977). Since then the code has been expanded and refined several times.

4.2 Source Term

Three forms of uranium are considered in this analysis: metallic billets, UO₃ powder, and finished and unfinished N Reactor fuel. The uranium is unirradiated and slightly enriched in ²³⁵U. The source terms used in this analysis are listed in Tables 3-5, respectively.

234 metric tons of uranium are in the form of forged billets, each about 175 kg and containing 1.25% ²³⁵U. Billets of this enrichment are in the shape of an annular cylinder, 17.73 cm OD, 7.1 cm ID, and 40.64 cm long (FDH 1999). The billets are shipped by truck in the Model G-4255 wooden box, which has a capacity of 3 billets and, when loaded with 1.25% ²³⁵U billets, may be shipped six at a time (FDH 1999). This gives a total of 75 shipments [75 shipments = 234,000 kg / (175 kg/billet x 3 billets/box x 6 boxes/shipment)].

669 metric tons of uranium are in the form of UO₃ powder, enriched to 0.87% ²³⁵U. The powder is currently stored in 147 T-Hoppers, each of which has a capacity of 4.5 metric tons of uranium. T-Hoppers are to be shipped either by truck three at a time or by rail, ten per railcar, three railcars per shipment. This would require a total of 49 shipments by truck or 5 shipments by rail.

Table 3 Source Term for the Billets

Table 4 Source Term for the T-Hoppers

The N Reactor fuel consists of finished and unfinished inner and outer fuel elements of five different ²³⁵U enrichments. Both elements are annular cylinders, the outer element has dimensions of about 2.4 in. OD, 1.8 in. ID; the inner element is about 1.2 in. OD, 0.5 in. ID, with lengths varying between 15 and 26 in. (WHC 1992). A total of 957.3 metric tons of uranium as fuel are to be shipped in the Model G-4214 wooden box, which has a capacity of 544 kg. The unfinished fuel elements are differentiated from the finished fuel in that they do not have the end caps welded on. The enrichment levels of ²³⁵U consist of 0.71, 0.95, 1.03, 1.15, and 1.25%. Due to the possibility of forming a critical configuration in the event of an accident, preliminary limits on the total uranium mass in a shipment of the 0.95% and 1.25% enriched fuel have been derived (Ferrell 1999). Mass limits for the 1.03 and 1.15% enriched fuel were interpolated from these limits. The fuel with a ²³⁵U content of 0.71% is considered to be natural uranium and is not considered to be fissile material. The criticality based shipment mass limits, total mass of both finished and unfinished fuel to be shipped, and calculated number of shipments of fuel of each ²³⁵U content are included in Table 5.

Table 5 Source Term for the Fuel

²³⁴U and ²³⁶U were not included in RADTRAN’s library of radionuclides, so the isotopes were defined in the input file. These isotopic definitions were taken from RADTRAN input files in Green (1995), which used the sources referenced in RADTRAN (Neuhauser and Kanipe 1992) to obtain the required isotopic properties.

4.3 Incident-Free Transportation

The RADTRAN 4 User Guide (Neuhauser and Kanipe 1992) defines incident-free transportation as transportation during which no accident, packaging or handling abnormality, or malevolent attack occurs. The consequence due to incident-free transportation is the dose received by people in the vicinity of the package due to external exposure. These people may include passengers, transportation workers (crew, inspectors, etc.), handlers, population off-link, population on-link, population during stops, and population during storage. The probability of the afore-mentioned consequences is always set to unity, as the probability of an accident is much less than unity. Thus, the risk due to incident-free transportation is numerically equal to the consequences.

Table 6 lists the input parameters common to all shipments made by truck or rail that are used by RADTRAN 4 in the calculation of population dose for incident-free transportation. Many of the values used for these parameters are defaults recommended by the RADTRAN User Guide (Neuhauser and Kanipe 1992). Others are either calculated or assumed and are discussed below. Parameters dependent on the package transported are listed in Table 7.

Table 6 Input Parameters for Incident Free Transport by Truck and Rail

^a Default values taken from RADTRAN 4 User Guide (Neuhauser and Kanipe, 1992)

Table 7 Package-Specific Input Parameters for Incident Free Transport

^a Default values taken from RADTRAN 4 User Guide (Neuhauser and Kanipe, 1992)

One of the calculated parameters in Table 7 is the characteristic package dimension (CPD). This is usually the largest dimension of the package. However, when arrays of similar packages are shipped, the RADTRAN User Guide (Neuhauser and Kanipe 1992) suggests treating the array as a single package. The CPD selected for the array of six G-4255 wooden boxes transporting billets was the length of the array, i.e., six box widths, calculated to be 3.91 m (= 6 x 25.625 in.). The lengths of the array of three T-Hoppers by truck, 4.57 m (= 3 x 5 ft), and of the array of ten T-Hoppers by rail, 15.24 m (=10 x 5 ft), were used as the CPD for the powder shipments. The CPD for the G-4214 wooden boxes used to transport the fuel was a multiple of the box width, 16.375 in., and the number of boxes depended on box capacity and the mass limit imposed by criticality constraints for each enrichment. Schematics of the wooden boxes and T-Hopper are included in Section 5.13.

Another parameter in Table 7, the source to crew distance, is calculated for the transport of billets and T-Hoppers by truck. RADTRAN calculates dose rates to the crew by extrapolating the dose rate at the side of the array, without accounting for the fact that the crew is not at the side of the array but at the head of the conveyance. Because the dose rates on the side are larger than at the head of the array, the crew dose rate is overestimated. The RADTRAN User Guide (Neuhauser and Kanipe 1992) suggests fixing this by inflating the source to crew distance. The shielding calculation in Section 5.1 determined the dose rate from the various package arrays for an estimated source to crew spacing of 3.1 m. The equation for the dose rate to the crew given in the RADTRAN Technical Manual (Neuhauser and Kanipe 1989) is

where DR_c = dose rate in the crew compartment
     r_c = source to crew distance, (m)
     PPS = number of packages per shipment
     DR_p = dose rate at 1 m, and

de = effective package dimension =

The effective package dimension is a function of the characteristic package dimension (CPD). The CPD of the array of 6 G-4255 boxes is less than 4 m, so d_e is equal to the CPD. The CPD of the array of 3 T-Hoppers is greater than 4 m, so the d_e is calculated to be 4.33 m using the above formula.

Rearranging for the effective source to crew distance gives

The parameter values and resulting effective source to crew distances are

Array	de	PPS	DRp	DRc	rc
6 G-4255 boxes	3.91	1	0.086	0.011	8.27
3 T-Hoppers	4.33	1	0.44	0.074	7.71

An effective source to crew distance was not calculated for the array of T-Hoppers transported by rail, as the RADTRAN default value for rail shipments is sufficiently large to account for the massive shielding provided by the locomotive. Shipments of fuel also did not require a calculation of the effective source to crew distance, as more of a square footprint was assumed for the arrays of fuel enriched to 0.95 and 1.25% ²³⁵U. Consequently, the use of the lateral dose rate was not overly conservative.

Two other parameters in Table 6 for which derived values were used are the stop time per kilometer traveled and the minimum stop time per trip. The computer code HIGHWAY (Johnson et al. 1993) assumes that a two-person truck driving team will move for 4 hr and then stop for a 0.5 hr break, repeating this cycle until the destination is reached. This approach is considered more realistic than the defaults provided in the RADTRAN 4 User Guide (Neuhauser and Kanipe 1992), in which the drivers are assumed to stop for an hour after every 90 km. However, to be conservative, the stop time using the HIGHWAY approach is multiplied by a factor of 2. The HIGHWAY output files in Section 5.0 give a total road time of 43.3 hr by the direct route and 48.8 hr by the indirect route. The stopover time in Paducah in the indirect route is not included in the total stop time, as the T-Hoppers are removed from the transport vehicle for maintenance. Thus, the total stopover time is 10.8 hr (= 43.3 / 4 x 0.5 x 2) by the direct route and 12.2 hr (= 48.8 / 4 x 0.5 x 2) by the indirect route.

4.4 Transportation Accidents

Accidents occurring during transportation may cause damage to the package’s shielding or cause a release of radioactive material from the package. The consequence of an accident during transportation is the dose received by the nearby population from this release by any of six potential exposure pathways considered in RADTRAN. These pathways are direct external irradiation, cloudshine, inhalation, groundshine, resuspension, and ingestion (Neuhauser and Kanipe 1989). The probability of an accident is based on the total distance traveled and on tabulated accident frequencies per unit distance. Thus, knowledge of the transportation route is required for calculating the risks from transportation accidents.

The truck transportation routes between the Hanford Site in Washington and the Portsmouth Site in Ohio were generated using the computer code HIGHWAY 3.3 (Johnson et al. 1993) via the TRANSNET network at Sandia National Laboratories. Two distinct truck transport routes were calculated. One route, which stops in Paducah, Kentucky, is used for the shipment of the T-Hopper packages. All other packages will be shipped via a direct route between the origin and destination. The rail transportation route was generated using the computer code INTERLINE version 5.0 (Johnson et al. 1992), again via TRANSNET. As before, a direct route and an indirect route were obtained. Weighted population densities in the rural, suburban, and urban zones were calculated by HIGHWAY and INTERLINE for the specific routes traveled and used in the RADTRAN input files. The total distance and fraction of distance traveled in each population zone are given in Table 8 for the rail and truck routes. Maps of the routes obtained from HIGHWAY and INTERLINE are included in the Appendix in Section 5.13.

Table 8 Population Breakdown of the Truck and Rail Routes

Nationwide average accident rates were taken from Saricks and Kvitek (1994) for truck and rail shipments. The accident rates per km for rural and urban/suburban truck shipments are 2.03E-7 and 3.58E-7, respectively. The accident rate on mainline railroads per km per railcar is 2.66E-8. Because three railcars will be transported at a time, that rate is multiplied by three. The mainline accident rate is used since the vast majority of the distance traveled is on mainline routes.

Because accidents may vary in terms of their severity, an accident severity classification scheme is required that groups accidents of similar severity together. A scheme of eight severity categories of increasingly severe accidents, defined in terms of mechanical and thermal (fire) loads, for different transportation modes is provided in NUREG-0170 (NRC 1977). Also reported in NUREG-0170 are the fractional occurrences of accidents in each severity category, further subdivided by the fractional occurrence in each of three zones of population density. Accidents of Category 1 are defined to be less serious than the accident performance capabilities of a Type A packaging and are not expected to result in the release of the radioactive material. Similarly, a Type B packaging is expected to survive a Category 2 accident with no release. The probabilities of occurrence of accidents of each severity category and in each population zone are given in Table 9 for truck and rail transportation. Table 10 gives the same data after normalizing the accidents according to population density zone.

Table 9 Fractional Occurrences for Rail and Truck Accidents by Accident Severity Category and by Population Density Zone

Table 10 Fractional Occurrences for Rail and Truck Accidents Normalized to Population Zone

With the total distance and the frequency of accidents occurring in each severity category known, the probability of an accident occurring is established. The other half of the risk equation, the consequences of an accident, must now be determined.

The response of a package to an accident of a particular severity is given by the release fraction parameter in RADTRAN 4. The release fraction as used in RADTRAN is the amount of material available for dispersal or exposure in an accident expressed as a fraction of the amount of radioactivity present in the package. NUREG-0170 (NRC 1977) recommends the following release fraction model for Type A containers and LSA drums: 0 release for Category 1, 0.01 for Category 2, 0.1 for Category 3, and 1.0 for Categories 4-8. The Model G-4255 and G-4214 wooden boxes are certified Type AF packagings (FDH 1999), and the T-Hopper is a strong, tight packaging used since the 1950s to transport LSA quantities of materials; therefore, the use of the release fractions in NUREG-0170 is justified. This analysis uses the recommended release fractions for Categories 1 and 4-8 for all payloads. However, to be conservative, larger release fractions are used for Categories 2 and 3 for the billets, powder, and unfinished fuel payloads. The recommended release fractions for all categories are used for the finished fuel payload, as this fuel has a zirconium cladding as an additional containment boundary. For the G-4255 box containing billets, a value of 1.0 is conservatively used for categories 2 and 3. For the T-Hopper and the G-4214 box containing unfinished fuel, release fractions of 0.1 and 1.0 are used for Categories 2 and 3, respectively. Although a detailed structural and thermal evaluation of the various accident scenarios could justify the use of lower fractional releases within Categories 2 through 8, it was not felt to merit the additional time required.

Once the material is released from the container and available for dispersal, it must be in the form of an aerosol to present an inhalation hazard. An accident, such as an impact or fire, will cause a fraction of the contents to form particulate material. This fraction is known as the aerosol fraction. The particulate material that is less than 10 mm aerodynamic equivalent diameter (AED) is assumed to be capable of being inhaled into the human respiratory system. This fraction is known as the respirable fraction. The aerosol and respirable fractions depend on the severity of the accident and the physical characteristics of the material. The respirable fraction should not be less than the respirable fraction of the pre-accident material. The release, aerosol, and respirable fractions used for the billets, powder, and fuel payloads are summarized in Table 11.

Table 11 Release, Aerosol, and Respirable Fractions for Accident Conditions

The aerosol and respirable fractions are set to zero for Category 1 accidents because no release is anticipated. The fractions for Category 2 accidents are conservatively based on the maximum credible accident scenarios discussed in the toxicological consequence assessment in Sections 4.6.1 and 4.6.2. The aerosol fractions used for Categories 3 through 8 are a factor of 10 higher than for Category 2 for the billets payload; these values represent bounding values from DOE (1994) for the billets payload in the fire scenario described in Section 4.6.1. The aerosol fractions used for Categories 3 through 8 are a factor of 100 higher than Category 2 for the powder payload; these values are conservatively higher than the bounding values for the powder in the impact scenario described in Section 4.6.2. Because the fuel is in the same physiochemical form as the billets, the same aerosol and respirable fractions are used for both payloads.

4.5 Health Effects

Deleterious health effects ranging from minor to severe arise from exposure of individuals and populations to ionizing radiation. These effects have been correlated to doses by the International Commission on Radiological Protection (ICRP) based on historical exposures and summarized in conversion factors that consider both the probability of occurrence and a judgment of the severity of that effect (ICRP 1991). Values are given in ICRP for the estimated probabilities of a fatal cancer, of a non-fatal cancer, and of a severe hereditary effect per unit effective dose. The total detriment is the sum of these three probabilities. These values are listed in Table 12.

Table 12 Health Effect Conversion Factors (ICRP 1991)

4.5.1 Results of the Radiological Risk Assessment

Table 13 lists the results of the radiological risk analysis. These results are for the total number of shipments made of a particular payload. Four different shipping scenarios were considered in the shipment of UO₃ powder: the combinations of rail vs. truck, and direct route vs. indirect. The risk from each fuel type, i.e., unfinished vs. finished, for each ²³⁵U enrichment, is listed separately, as well as a summed risk from all fuel types. The values given for incident-free transportation are the consequences that result from the normal shipment of these radioactive materials. Because the probability of incident-free transportation is unity, the risks of these shipments are also the consequences in person-rem, number of latent cancer fatalities, and total detriment. The values given for accidents in transportation are risk values, as they are the product of the radiological consequences and the probability of occurrence for accidents of various severity. The sum of the risks from incident-free transportation and from accidents in transportation represent the total radiological risk. The summed risk for the entire shipping campaign of all payloads, assuming the worst-case scenario for shipping the UO₃ powder, is 1.92 person-rem, 8.55E-4 latent cancer fatalities, and 1.22E-3 total detriment.

Table 13 Radiological Risks from Uranium Shipments

4.6 Toxic Chemical Consequence Assessment

This section evaluates the consequences due to the chemical toxicity of uranium that could result from an accidental release during transport of the metallic billets, UO₃ powder, and the worst-case shipment of fuel. The toxicological consequences are given in terms of the concentrations of airborne uranium particulates at various receptor locations. The calculated concentrations are then compared to various exposure limits to evaluate the effects of the release on the public.

According to DOE (1994), for natural or depleted uranium or uranium enriched < 10% in ²³⁵U, the toxicity of uranium as a heavy metal is of greater concern than the radiological hazard. The toxicological hazard results from the accumulation of uranium in the kidneys due to the transport of inhaled, soluble uranium compounds or non-soluble particulates. For non-soluble materials to be an inhalation hazard, the size of the particles/aggregates must be 10 mm AED (more probably 3 mm AED) or less (DOE 1994).

The maximum credible release depends on the physical and chemical form of the payload. Powder and large solid masses respond differently to a given accident scenario; the same applies to oxides and metals. The maximum credible accident scenario for the UO₃ powder is an energetic impact event which damages the T-Hopper container and nearly instantaneously creates a puff of particulates that is released to the atmosphere and transported downwind. On the other hand, an impact event is not expected to significantly damage the solid metal billets or fuel. A fire event is postulated as the maximum credible accident scenario for the billets and fuel, which are engulfed in flames due to the combustion of an external fuel, e.g., the diesel fuel from the truck’s fuel tank. The duration of the fire is assumed to last 2 hours. The billets and N Reactor fuel elements are treated together, as they are both uranium metal.

4.6.1 Uranium Billets/Fuel Release Rate

According to DOE (1994), no significant airborne release is postulated for solid metal in an impact event; however, particulates are released during the oxidation of the metal in a fire. Therefore, the maximum credible release is calculated for a fire event.

Massive uranium metal is difficult to ignite, as large amounts of external heat must be supplied and serious heat loss prevented (DOE 1994). This external heat is assumed to arise from the combustion of diesel fuel from the transportation vehicle. DOE (1994, p. 4-3) provides median values of 1E-4 for the airborne release fraction and 1.0 for the respirable fraction for uranium metal subjected to a fire. These values correspond to the complete oxidization of the metal; experimental values reported for a 2 hour burn produced smaller release fractions. Thus, the use of the median release fractions is conservative.

An additional conservatism is introduced by using the two hour fire duration as the duration for the release. Although the uranium will likely not completely oxidize in two hours, assuming this smaller release time increases the release rate. Regardless of the speed at which uranium oxidizes, it is likely that the fuel source will be exhausted before that time. The efforts of emergency responders in mitigating a fire during the assumed burn time is also conservatively neglected.

This analysis conservatively does not consider any removal mechanisms of the particulates, e.g., washout, gravitational settling, or removal through contact with vegetation or buildings. This assumption maximizes the airborne concentration and is conservative. Because the molecular weight of uranium oxide is an order of magnitude greater than air, significant settling would be expected, as DOE (1994b) states that in the absence of strong drafts, uranium oxide smoke tends to deposit in the immediate area of the burning metal.

The worst-case shipment of fuel consists of 0.71% ²³⁵U unfinished fuel, as this gives the largest uranium loading. The zirconium cladding on the sides of the unfinished fuel is neglected. The entire truckload of 18 billets or 3264 kg U of fuel is assumed to be engulfed in the fire. A 0.044 g/s release rate of aerosolized, respirable particles from burning uranium billets is calculated. Using the same logic for the fuel gives a 0.045 g/s release rate.

0.044 g/s = (10^-4)(175 kg/billet)(1000 g/kg)(3 billets/box)(6 boxes/truck) / (2 hr (3600 s/hr))
0.045 g/s = (10^-4)(3264 kg/shipment)(1000 g/kg) / (2 hr (3600 s/hr))

For simplicity, the 0.045 g/s release rate will be used for both fuel and billets.

4.6.2 UO₃ Powder Release Rate

Powder can be made airborne by either a fire or an impact event. The airborne release of powder during a fire is due to entrainment caused by the air turbulence induced by the fire. Similarly, during an impact, powder entrainment may be caused by the mechanical disturbance during the dynamics of the impact. Both stresses are considered in the accident scenario involving the T-Hoppers.

The maximum credible accident considered for the UO₃ powder transported by truck is expected to arise from an impact due to a collision. The impact is assumed to cause moderate damage, consisting of rupture of the wall and failure of the gasket, to all the T-Hoppers on the trailer.

For rail transport, the maximum credible accident is due to a collision with a vehicle at a crossing. This scenario would most likely cause considerable damage to the offending vehicle, while the train would sustain minimal damage. However, this analysis conservatively considers the more unlikely scenario in which a vehicle rams the side of a rail car transporting the T-Hoppers. Although the offending vehicle is still likely to sustain the majority of the damage incurred in this accident, the same release rate as calculated for the truck accident is used for simplicity.

DOE (1994, p. 4-87) provides values for the airborne release of powder contained in metal enclosures. The release fractions are dependent on the particle size of the powder. The fraction of the UO₃ powder as a function of particle diameter is not yet known, but it can be assumed that the powder contains particles greater than 0.5 mm in diameter. Experimental measurements involving an impact on steel cans without lids containing powder less than 2 mm in diameter produced an airborne release fraction of 3E-4 and a respirable fraction of 1E-2. If the respirable fraction of the original powder is less than this value, the respirable fraction of the source powder should be used (DOE 1994).

The leakage of aerosolized powder is inhibited by the damaged T-Hopper. Although the rupture from the impact event provides an escape route for the powder, the bulk of the container still encloses the contents. Thus, the amount airborne is reduced by a factor representing the presence of the damaged container. This factor is the leak path factor. This analysis assumes that 10% of the surface area of the container has been compromised in the impact event; thus, 90% of the aerosol undergoes filtration and deposition by the damaged T-Hopper.

The total release of aerosolized, respirable UO₃ particulates is then 4.1 g.

4.1 g = 3*(4.5E6 g U)*(3E-4)*(1E-2)*(0.1)

4.6.3 Concentration Calculation

The concentration is related to the release rate in the fire event, or total release in the impact event, by the atmospheric dispersion parameter, c/Q. This parameter is a function of the receptor location, wind speed, and atmospheric turbulence. c/Q is normalized either to the release rate of a sustained release (in which case the Q is primed Q’) or to the total release of a nearly instantaneous “puff” release. This analysis will determine the uranium airborne concentration at three downwind receptor locations: 100 m, 200 m, and 1000 m. The 100 m distance was assumed to be a reasonable estimate of the distance between an interstate highway and the nearest resident, while the further distances show how the concentration falls off.

Two sets of meteorological conditions are examined. The first consists of worst case conditions of wind speed (1 m/s) and atmospheric turbulence (Pasquill stability class F) that cause a maximum concentration. These conditions tend to disperse the released material very slowly, resulting in the highest possible downwind concentrations. However, these conditions are rarely encountered, except perhaps for night conditions, and tend to overstate the actual impacts. The second case consists of more likely, but still relatively rare, conditions of a wind speed of 2 m/s and neutral stability (Pasquill stability class D). The latter set of conditions will be used to calculate the concentrations at all three potential receptor locations and the former will be used to calculate the worst-case conditions at the shortest distance (100 m).

Green (1995) calculated c/Q’ for the weather conditions and receptor locations described above using the methods of NRC (1982). These values are given below.

• 2.85E-2 s/m³: c/Q’ for 100 m receptor, Pasquill F, and 1 m/s wind speed
• 3.76E-3 s/m³: c/Q’ for 100 m receptor, Pasquill D, and 2 m/s wind speed
• 9.68E-4 s/m³: c/Q’ for 200 m receptor, Pasquill D, and 2 m/s wind speed
• 6.63E-5 s/m³: c/Q’ for 1000 m receptor, Pasquill D, and 2 m/s wind speed

The computer code GXQ version 4.0 (Hey 1993, 1994) was used to calculate c/Q for the puff releases for the same meteorological conditions as for the sustained releases. These values are given below. The GXQ output file is given in the appendix.

• 2.65E-3 m^-3: c/Q for 100 m receptor, Pasquill F, and 1 m/s wind speed
• 3.14E-4 m^-3: c/Q for 100 m receptor, Pasquill D, and 2 m/s wind speed
• 4.74E-5 m^-3: c/Q for 200 m receptor, Pasquill D, and 2 m/s wind speed
• 7.10E-7 m^-3: c/Q for 1000 m receptor, Pasquill D, and 2 m/s wind speed

The release rate from the billets in the fire event is multiplied by c/Q’ to obtain the downwind uranium concentration. Similarly, the total release from the UO₃ powder in the impact event is multiplied by c/Q. Table 14 summarizes the results of the toxic chemical consequence analysis.

The results in Table 14 are then compared with Temporary Emergency Exposure Limits (TEELs) for uranium oxide established by the Department of Energy Subcommittee on Consequence Assessment and Protective Actions (SCAPA) (Craig 1999). Uranium oxide is used because the billets and fuel will oxidize during the fire; also, the limits for oxide are the same or more conservative than for metal. The DOE Emergency Management Guide (DOE 1997) calls for the use of TEELs when Emergency Response Planning Guidelines (ERPGs) are not available. Although ERPGs are the standard community exposure limits approved by the American Industrial Hygiene Association, less than 100 chemicals have been assigned ERPGs, and none of those include compounds of uranium. The definitions of the TEEL limits are as follows:

• TEEL-0: The threshold concentration below which most people will experience no appreciable risk of health effects. The TEEL-0 for uranium oxide (insoluble compound) is 0.05 mg/m³.

• TEEL-1: The maximum concentration in air below which it is believed nearly all individuals could be exposed without experiencing other than mild transient health effects or perceiving a clearly defined objectionable odor. The TEEL-1 is 0.6 mg/m³.

• TEEL-2: The maximum concentration in air below which it is believed nearly all individuals could be exposed without experiencing or developing irreversible or other serious health effects or symptoms that could impair their abilities to take protective action. The TEEL-2 is 0.6 mg/m³.

• TEEL-3: The maximum concentration in air below which it is believed nearly all individuals could be exposed without experiencing or developing life-threatening health effects. The TEEL-3 is 10 mg/m³.

Using these definitions and the results in Table 14, at distances of 200 m and greater from an accident involving any payload, the results are either mild transient health effects (TEEL-1) or nothing at all (TEEL-0). At a distance of 100 m, an accident involving powder could result in an airborne concentration at which irreversible or other serious health effects could occur (twice the TEEL-2). This is about 13% of the level at which most people could be exposed without experiencing life-threatening health effects. At the same distance involving an accident with the fuel or billets payload, only mild transient health effects are expected to occur (TEEL-1). It should be noted that for the very rare weather conditions at 100 m, the TEEL-3 limit is exceeded for an accident involving powder, while for the billets and fuel payloads under the same worst-case meteorological conditions, the downwind concentrations do not exceed TEEL-1.

Table 14 also indicates the dilution of the uranium aerosol with distance. The airborne concentrations of uranium drop by about an order of magnitude from 100 to 200 m, and again from 200 to 1000 m. Although the concentrations at 100 and 200 m are about an order of magnitude less for the fuel or billets payloads than for the powder payload, the concentrations are nearly equal at 1000 m, despite the difference in the releases.

Note that these values are the consequences from an accident, and do not reflect the frequency of occurrence of an accident or the assumed meteorological conditions. As such, they cannot be compared directly to the radiological risk values in Table 1. A risk assessment weights the consequences by the frequency (or probability) of occurrence of the release. The toxicological consequences have not been weighted by the probability of the release.

5.0 APPENDIX

5.1 Dose Rate Calculations for Billets, UO₃ Powder, and N Reactor Fuel

The RADTRAN v. 4 computer code requires as input the dose rate at 1 m from the vertical planes projected by the outer lateral surfaces of the transportation vehicle for exclusive use shipments. This dose rate is then used to extrapolate the dose rate at further distances using the method described in Neuhauser and Kanipe (1992). Shielding calculations were done to estimate the dose rates at 1 m from the outer lateral surfaces of the transportation vehicle loaded with the Model G-4255 Wooden Box, the Model G-4214 Wooden Box, and the T-Hopper.

The billets are transported in the G-4255 wooden box. Six boxes are shipped by truck per shipment, each holding three 175 kg billets in an unknown arrangement. For simplicity, the billets were assumed to be smeared over the entire interior volume of the box. The interior dimensions of the wooden box are taken from FDH (1999), consisting of 30.75 in. L x 24.125 in. W x 8 in. D interior, with a minimum plywood thickness of 0.75 in. A schematic of the box is shown in Section 5.13. The box rests on top of support skids attached to its largest side, while a smaller side faces the lateral surface of the transport vehicle, assumed to be 3 m wide. For this calculation it was assumed that the 30.75 in. x 8 in. side faced the front of the trailer, and the 24.125 in. x 8 in. side faced the lateral side. The six boxes were assumed to be aligned one behind the other, neglecting the shielding between boxes. The dose rate was calculated at 1 m from the edge of the transport vehicle at the midpoint of the lateral surface of the array.

The UO₃ powder is transported in T-Hoppers. An array of three T-Hoppers is shipped by truck, while an array of ten is shipped by rail. The T-Hopper consists of a frame that encloses a conical structure that is widest at the bottom. A schematic of the T-Hopper is shown in Section 5.13. The dose rate was calculated at 1 m from the lateral surface of the transport vehicle, assumed to be 3 m wide. The UO₃ powder is contained in the cone-shaped structure, with a 5 ft diameter cylindrical base at the bottom. This geometry was approximated for simplicity as a cylinder of the height of the T-Hopper (6 ft), with a radius calculated from the powder mass m, density r, and height h. Using the equations for density and volume, the radius r was calculated as

The density of UO₃ powder is 7.29 g/cm³ (Lide 1993). Assuming the interstitial void space of the powder results in a packing fraction of 0.68, the bulk density of the powder is 4.96 g/cm³. For a powder mass of 5454.5 kg, the calculated radius is 43.75 cm.

The fuel elements are transported in the Model G-4214 Wooden Box. The interior dimensions of the wooden box are taken from FDH (1999), consisting of 30 in. L x 14.125 in. W x 8.375 in. H interior, with a minimum thickness of the plywood container of 0.75 in. A schematic of the box is shown in Section 5.13. To prevent the formation of a critical configuration in the event of an accident, limits on the total uranium mass in a shipment of the 0.95% and 1.25% enriched fuel have been derived (Ferrell 1999). Mass limits for the 1.03 and 1.15% enriched fuel were interpolated from these limits. These limits are 1628, 1375, 996, and 680 kg, for fuel containing 0.95, 1.03, 1.15, and 1.25% ²³⁵U, respectively. The number of boxes per shipment was assumed based on these criticality based shipment mass limits and the 544 kg capacity of the boxes. Fuel containing 0.71% ²³⁵U, the same amount found in natural uranium, is not limited by criticality, in which case 6 boxes per shipment were assumed. The array was assumed to be arranged in a similar fashion as the array of boxes containing billets, i.e., the side with the skids on bottom, the largest lateral face of the box toward the front, and the boxes of the array adjacent to each other centered on the trailer. The dose rate was calculated at 1 m from the lateral sides of the vehicle edge.

The source terms for the billets and fuel are taken from Table 5.2.1-1 of FDH (1999), which gives the photon production in eighteen energy groups at a decay of 1 year. Because of the similarity between the source terms for the 0.95%, 1.10%, and 1.25% ²³⁵U enriched fuels, the same source term is used for each enrichment. There are a couple of small differences in the source terms in Table 5.2.1-1 of FDH (1999), reproduced in Table 15; the source term used in the calculations conservatively took the highest photon production of each energy group of the source term of the three enrichments.

Table 15 Billets and Fuel Source Term from FDH (1999)

The source term for the powder is derived from Table 4, decayed 10 years using the computer code ORIGEN-S (Hermann and Westfall 1997) of the SCALE v. 4.3 code package (McCoy 1998). This decay time allows some buildup of daughter products that are part of the long decay chains of uranium and is a conservative estimate of the time since the powder was processed. The most important daughter product in this inventory from a shielding standpoint is ^234mPa, with several low-intensity, high-energy gamma rays and a 2.28 MeV endpoint energy beta particle at 98.6% intensity. The very short-lived daughter products ²¹⁰Po, ²¹¹Po, ²¹²Po, ²¹⁵Po, ²¹⁶Po, ²¹⁸Po, and ²²³Fr included in the ORIGEN-S output are not included in ISO-PC’s data library, but all are of very low activity, energy, or intensity, and so have no effect on the shielding analysis.

The computer code ISO-PC version 2.1 (Rittmann 1995, 1996) was used to calculate the dose rates, summarized in Table 16. Dose rates were calculated at 1 m and 2 m from the vertical plane projected by the outer lateral surface of the transportation vehicle, and at the crew location, assumed to be 3.1 m from the front of the array, which is the RADTRAN default value for trucks (Neuhauser and Kanipe 1992). The anterior-to-posterior flux-to-dose-rate conversion factors from ANS (1991) were used, which are the most conservative and represent radiation entering the front of the body. Buildup was calculated in the uranium source material for all shipments.

Table 16 Dose Rates (mrem/h) from Uranium Payloads

(this page reserved for shielding check sheet)

Continue

For questions or comments, please send email to R_L_Leo_Guillen@rl.gov
Document Number: DOE/EA-1319
URL: http://www.hanford.gov/docs/ea/ea1319/ea1319.html