Performance Estimate for the Detection of Plutonium Separation by 85Kr

This paper gives new estimates of 1) the lower limit of detection and 2) the probability of detecting plutonium separation using atmospheric krypton-85. The analysis is based on two 6.5 year datasets. Results are given for di®erent separation rates and di®erent false-alarm rates under conditions of low background variabil- ity. The data were acquired under su±ciently generic conditions that these results can be generalized to other locations and used in future studies. For a false-positive rate of 0.05%, the continuous separation of 100, 300, and 1000 grams equivalent weapon-grade plutonium per day were found to correspond to 10%, 50%, and 80% probability of detection respectively at a distance of 60 km. The smallest detected concentration was equivalent to the continuous separation of 44 grams per day, with a probability of detection of about 2%. It was further found that increasing the false-positive rate does not usefully increase the detection rate. A second analysis of the 85Kr-background variation at Freiburg, Germany, suggests that a similar arrangement would be 15 to 20 times less sensitive in this area, which is closer to commercial reprocessing plants.