Year
2011
Abstract
This paper will present case studies of undeclared activity at the mock feed and withdrawal (F&W) facility at the Oak Ridge National Laboratory (ORNL) using A MATLAB software tool designed specifically for process load cell monitoring of the mock F&W facility. The mock F&W facility mimics a continuous batch operation process such as a gas centrifuge enrichment plant (GCEP). The facility has an accountancy scale, three feed stations, two tail stations, and three product stations. Each station has a process scale similar to the accountancy scale. The data from all the scales is sampled at 1 hertz and stored in a database. During normal operation a process tank is weighed on the accountancy scale, placed on the appropriate process station, and then reweighed on the accountancy scale after processing. The total amount of material processed is declared from the accountancy scale weights. In one undeclared activity scenario the process tanks are never weighed on the accountancy scale and the amount of material processed is never reported to the IAEA. Using the MATLAB software the data from the process scales is used to make a declaration sheet similar to declaration sheet generated from the accountancy scales. Discrepancies between the process scale declaration sheet and the accountancy scale declaration sheet quickly identify undeclared activity. In a second undeclared activity scenario product material is diverted during processing with the possibility of adding substitute material at the end of processing to rebalance the mass inventory difference. The MATLAB software includes the ability to calculate the facility inventory difference versus time plots. During a diversion these plots should show a decreasing residual correlated to the diversion of material instead of a steady value around zero. By implementing load cell monitoring of the process scales, undeclared activity scenarios that would remain unknown using only accountancy scale data can be quickly and efficiently identified.