STUDY OF THE COMPONENTS OF EVACUATION TIMES

Year
1998
Author(s)
G. S. Mills - Sandia National Laboratories
K. S. Neuhauser - Sandia National Laboratories
JD Smith - Sandia National Laboratories
File Attachment
921.PDF890.18 KB
Abstract
The magnitudes of accident dose-risks calculated by the RADTRAN code depend directJy on the time span between an accidental release and evacuation of the affected area surrounding potential radionuclide releases. In a previous study of truck and rail transportation accidents, and other incidents requiring evacuations (Mills et. al., 1995), a lognonnal distribution of evacuation times (time span from decision to evacuate until complete) was developed, which provided a better model for this parameter than the practice of using a highly conservative value of 24 hours. However, the distribution did not account for time required for responders to arrive on the scene, to evaluate the hazards to surrounding population and to initiate an evacuation. Data from U.S. Department of Transportation (DOT) accident statistics have been collected and their distribution functions determined. The separate distribution functions were combined into a single, comprehensive distribution which may be sampled to supply values of the RADTRAN input parameter, EVACUATION. A sample RADTRAN calculation illustrating the effect on risks of using the distribution versus the original (24 hour), conservative point-estimate are also presented ..