Response function for optical fiber based gamma thermometer in data analytic methodology for OFBGT array

Abstract

An optical fiber-based gamma thermometer (OFBGT) is being developed in a collaborative effort by The Ohio State University and Texas A&M University. With an array of OFBGTs within a reactor core, one can infer the power distribution among the fuel assemblies using a data analytic methodology. With this data analytic methodology, both the code Monte Carlo N-Particle Transport (MCNP) and experimentally determined parameters are used as input, and energy balance is employed to determine estimates of power for each fuel assembly segment, from each OFBGT segment. Then, these estimates are averaged together using a weighting scheme, and an iterative solution is used so that the inferred power distribution is self-consistent. The data analytic methodology is intended to be used to determine the power distribution in power reactors at steady state, utilizing MCNP data. However, the OFBGTs that we have constructed are being tested using the Ohio State University Research Reactor (OSURR), and the OSURR is not operated for intervals that are long enough to establish equilibrium concentrations of gamma emitting fission and activation products. The impact of the transient nature of the reactor power on OFBGT measurements and the data analytics that is associated with the OFBGT is investigated in this paper using MCNP. The results of our analysis show that if an OFBGT's output is processed using response functions that are appropriate for a steady-state distribution of fission products, then the OFBGT can significantly underpredict the power in the reactor. The magnitude of the underprediction is largest for a clean core, which operationally means for measurements that are made early in the day for a research reactor that is operated intermittently. For a clean core, the underprediction may be as large as a factor of 0.8 for fuel assembly segments that are r=5cm from the OFBGT thermal mass segment.