ANALYSIS OF THE EFFECTS OF PIPELINE AND RAILROAD FIRES ON LEGAL WEIGHT TRUCK TRANSPORTATION CASKS

Year
2001
Author(s)
Carlos Lopez - Sandia National Laboratories
Jorman A. Koski - Sandia National Laboratories
File Attachment
33314.PDF310.66 KB
Abstract
Questions specific to the safety of rail transportation of spent nuclear fuel have been raised. The purpose of this paper is to examine the fire environments that could occur in rail accidents involving typical casks and to determine whether rod burst and/or seal failure represent a problem under such occurrences. The method used to address these issues was a detailed three-dimensional finiteelement computer simulation of torch and engulfing fire environments. The results are intended to provide details and information in a form that would be a useful reference for future risk-based studies of these accidents. The transportation cask that was studied has the overall dimensions of a typical legal-weight truck cask. Four different kinds of accidents were modeled for this study. These were: 1) a fire jet impinging at the center of the cask, 2) a fire jet impinging at the seal end of the cask (without impact limiter) from the side, 3) a vertical flare radiating to the side of the cask from about ten meters away, and 4) a fully engulfing fire. All fire scenarios were modeled assuming different fire temperatures to cover a wide range of fires that could be found in these accidents. All modes of heat transfer are included: radiation, conduction and convection. In Case 1, seal failure was not found to be a problem, whereas rod burst could be a problem after 2 hours and 20 minutes if the fire jet has a temperature of about 1200°C, after an hour and 30 minutes if the fire jet has a temperatures of about 1400°C, and after an hour for a fire jet temperature of about 1600°C. For Case 2, the temperature history plots showed that while seal failure could occur early, rod burst is not a problem. Case 3 did not represent a problem in terms of seal failure or rod burst during the simulated ten hours whereas Case 4 indicates that both seal failure and rod burst occur some time during the simulation.