Photofission delayed gamma-ray measurements in a large cemented radioactive waste drum during LINAC irradiation

Abstract

Nuclear materials characterization in 870 L cemented waste drums is challenging due to significant neutron and gamma ray attenuation. Among the existing non-destructive approaches, photon active interrogation is of particular interest. A LINAC accelerates electrons to energies greater than 10 MeV, leading to highly penetrating interrogating bremsstrahlung photons that can reach nuclear materials in the depth of the 870 L waste drum and allow the detection of delayed gamma rays emitted by photofission products. Because the interrogating photon and neutron fluxes are extremely high during irradiation, measurements with high purity germanium detectors are usually performed after a long irradiation time, for example by moving the detector or the waste drum. Therefore, such measurements are only sensitive to delayed gamma rays associated to photofission products having sufficient long half-life of at least tens of seconds, which cuts a valuable signal coming from fission products with a shorter half-life. In this paper, we report gamma rays measurements of very short lived photofission products using a 3 in × 3 in cylindrical LaBr3(Ce) fast scintillation detector. A depleted uranium sample was irradiated by a pulsed bremsstrahlung source with maximum energy 17.5 MeV in a macro pulsing mode (1 s irradiation, 2 s cooling). This allowed to acquire a strong delayed gamma signal within a time window of 2 s.A clear signature of nuclear materials was observed even in the depth of a concrete mock-up, with an excellent signal-to-noise ratio beyond 3.5 MeV where the background of nonnuclear activated materials is negligible. These results are compared to a semi analytical model that qualitatively agrees but underestimates the measurements by a constant factor 1.7, probably caused by a wrong beam intensity normalization. Indeed, in this work the LINAC pulse time structure (macro pulsing) was different from its nominal working point used to characterize the beam (usual pulsing mode). Since it allows direct calculations of the detector response to delayed gamma-ray photofission measurements, this model is a fast calculation alternative to time-consuming Monte Carlo simulations, in view of further studies that will follow this feasibility demonstration.