Self-calibration Method for Dead Time Losses in Neutron Counting Systems

Most of the safeguards assay for quantitative characterization of SNM (mass, multiplication, random neutron contribution) are based on neutron measurements and rely exclusively on the counting information from very efficient, but slow He-3 proportional tubes. The response of neutron detection systems is inevitably affected by Dead Time (DT) losses that are generally caused by very complex and convoluted processes, which are difficult to take into account for corrections (for example, the DT losses for bipolar shapers differ from those of unipolar shapers). Therefore an empirical approach for calculating the DT losses assuming exponential (paralyzing ) DT using measurements with two Cf-252 sources with known activities was established as current practice for many safeguards neutron counting systems. The availability of a very wide range of such Cf-252 calibration sources becomes the limiting factor for extending the deadtime correction calibration over a sufficient dynamic range to reach the conditions of real measured material. In this paper we present a novel self-calibrating method for the determination and correction of deadtime losses that uses directly the neutron signal from real measured material. The count rate from the material is measured with two configurations of the preamplifiers: a standard configuration of the pre amplifier and tubes, corresponding to a nominal (100%) load per preamplifier and a second "deadtime measurement" configuration, where every two neighbouring clusters of He-3 tubes are connected together to a single preamplifier, corresponding to 200% load per preamplifier. A proof of principle DT calibration measurement over a wide dynamic range exceeding 106 reactions/sec using a 14 MeV neutron generator, demonstrated experimentally the viability of this method. The method produces the DT correction factor at every measured counting rate. The results show the very important observation that the correction factor does not fit with either fully paralyzing or fully non-paralyzing dead time models. Using either model could lead to substantial deadtime correction errors. Explanation of DT behaviour and implementation aspects of this method in typical safeguards neutron systems (already in use or to be built) such as differential die away, coincidence and multiplicity counting will be discussed.