Emergency responders from the Department of Energy are trained regularly to assess the environmental consequences of a radiological or nuclear incident. While drills and exercises are highly effective tools in rehearsing for an emergency, the accidents at the Fukushima Daiichi Nuclear Power Plants presented real-world complexities that are difficult or impossible to simulate in such training. Customarily, the modeled hypothetical event used to create a drill or exercise data set is simple, well defined, and closely resembles conventional assumptions about the type of that event. Consequently, the modeling performed by players from the outset closely resembles the planner’s hypothetical event. This approach also entails idealized, uniform data in the simulated plume and ground deposition scenarios created for the drill that match the modeling closely. The real-world event that occurred in Japan sharply deviated from the simple picture ordinarily created for drills and exercises that typically involve a release of radioactivity that is of short duration, a single puff with constant meteorology, or simple deviation such as a wind shift to bifurcate the plume. In the very early stages, accurate plume and deposition model predictions were difficult to produce due to the lack of field monitoring data and other information. In contrast to drills and exercises where plant monitoring data is available, there was much less plant monitoring data, essentially no reactor state information, and the meteorological conditions and releases were much more complex. Inevitably, the measurements in Japan were not homogeneous, thus presenting technical challenges to assessors tasked with ensuring the quality of the finished assessments and data products for government officials, the responder community, and the public. In this paper, examples of some operational real-world complexities are considered. Procedures, measurements, or radiological assessments from the Fukushima response are not in the purview of this paper.
*Brookhaven National Laboratory, P.O. Box 5000, Upton, NY, 11973-5000; †Remote Sensing Laboratory, P.O. Box 98521, Las Vegas, NV, 89193-8521.
The authors declare no conflict of interest.
For correspondence contact: S. Musolino, Brookhaven National Laboratory, P.O. Box 5000, Upton, NY, 3-5000, or email at firstname.lastname@example.org.
(Manuscript accepted 21 January 2012)