OPTIMAL MEASUREMENT UNCERTAINTIES FOR MATERIALS ACCOUNTING IN A FAST BREEDER REACTOR SPENT-FUEL REPROCESSING PLANT

Year
1982
Author(s)
Jack T. Markin - Los Alamos National Laboratory
Hassan A. DAYEM - Los Alamos National Laboratory
Edward A. Kern - Los Alamos National Laboratory
Abstract
Optimization techniques are used to calculate measurement uncertainties for materials accountability instruments in a fast breeder reactor spent-fuel reprocessing plant. Optimal measurement uncertainties are calculated so that performance goals for detecting materials loss are achieved while minimizing the total instrument development cost. The optimization methodology is useful in answering the following questions . (1) Given limited development resources, what measurement technology improvements provide the maximum increase in accounting system performance? (2) Which measurement uncertainties dominate the materials balance variance? (3) What values of measurement uncertainties are required to meet a given performance goal? Improved materials accounting in the chemical separations process (111 kg Pu/day) to meet 8-kg plutonium abrupt (1 day) and 40-kg plutonium protracted (6 months) loss-detection goals requires: • process tank volume and concentration measurements having precisions .5.1%; • accountability and plutonium sample tank volume measurements having precisions 5.0.3%, short-term correlated errors 5.0.04%, and long-term correlated errors 5P'04%; and • accountability and plutonium sample tank concentration measurements having precisions 50.4%, short-term correlated errors ^Q.1%, and long-term correlated errors