Determination of the plutonium content in a spent fuel assembly by passive and active interrogation using a differential die-away instrument

Abstract

In this paper, we present a novel approach to estimating the total plutonium content in a spent fuel assembly (SFA) that is based on combining information from a passive measurement of the total neutron count rate (PN) of the assayed SFA and a measure of its multiplication. While PN can be measured essentially with any non-destructive assay (NDA) technique capable of neutron detection, the measure of multiplication is, in our approach, determined by means of active interrogation using an instrument based on the Differential Die-Away technique (DDA). The DDA is a NDA technique developed within the U.S. Department of Energy's Next Generation Safeguards Initiative (NGSI) project focused on the utilization of NDA techniques to determine the elemental plutonium content in commercial nuclear SFA's [1]. This approach was adopted since DDA also allows determination of other SFA characteristics, such as burnup, initial enrichment, and cooling time, and also allows for detection of certain types of diversion of nuclear material. The quantification of total plutonium is obtained using an analytical correlation function in terms of the observed PN and active multiplication. Although somewhat similar approaches relating Pu content with PN have been adopted in the past, we demonstrate by extensive simulation of the fuel irradiation and NDA process that our analytical method is independent of explicit knowledge of the initial enrichment, burnup, and an absolute value of the SFA's reactivity (i.e. multiplication factor). We show that when tested with MCNPX™ simulations comprising the 64 SFA NGSI Spent Fuel Library-1 we were able to determine elemental plutonium content, using just a few calibration parameters, with an average variation in the prediction of around 1–2% across the wide dynamic range of irradiation history parameters used, namely initial enrichment (IE=2–5%), burnup (BU=15–60GWd/tU) and cooling time (CT=1–80y). In this paper we describe the basic approach and the success obtained against synthetic data. We recognize that our synthetic data may not fully capture the rich behavior of actual irradiated fuel and the uncertainties of the practical measurements. However, this design study is based on a rather complete nuclide inventory and the correlations for Pu seem robust to variation of input. Thus it is concluded that the proposed method is sufficiently promising that further experimentally based work is desirable.