Photopeak Efficiencies play an important role in the characterization of radioactive material. The presented approach coupling 3D sensors for geometry modelling and mathematical photopeak efficiency calculation offers an automated procedure for a wide range of applications, e.g. in drum measurement devices, clearance of bulk material on a conveyor belt, radiological sorting system as well as in-situ measurements on standing structures. For activity calculation, the modeling of an unknown object for a gamma spectrometric measurement to calculate the photopeak efficiency is highly individual and is currently mostly done by a human operator of the measurement facility. Thus, for more complex geometries, conservative assumptions are necessary, which sometimes oversimplify the geometry. Besides the high personnel expenditure for the modelling, this is associated with an overestimation of activity. The improvement of our developed nuclear simulation code TRACER (TRAck-based Calculation of Efficiency for Radiation) is based on generating the geometric models automatically from precise 3D point clouds recorded by a laser scanner for mathematical photopeak efficiency calculation. TRACER is a fast code for multicore photopeak efficiency calculation for a wide energy range from 50 keV to 10 MeV, primarily for High purity Germanium (HPGe) detectors. Thereby TRACER takes into account the objects surrounding the sample under investigation in an efficient hybrid approach combining Monte Carlo and deterministic methods including variance reduction. For the purpose of modelling the geometry of the objects, a 3D laser scanner is used which records position and dimensions of the sample and generates a 3D point cloud. From the point cloud data, TRACER automatically models the sample geometry, saving personnel expenditure and reducing uncertainty in the activity estimation. The additional information from the 3D point clouds offer e.g. the possibility to calculate the uncertainty for drum measurement systems more precisely and to detect errors in the measurement setup automatically.
In the R&D project VIRERO (VIrtual REmote RObotics for Radiometric Sorting), conducted by Framatome, FAPS and AiNT, a robotic system is developed, which uses the recorded 3D point cloud to create a 3D model of the measured material for radiological sorting. The choice of the model is adapted to the application purpose and offers the possibility to calculate a conservative prognosis activity as well as a realistic activity (best estimate). In the automatically generated model, geometric details for the efficiency calculation are retained without affecting the required computing power to a large extent. These models offer the possibility to create a 3D-resolved nuclide specific activity profile with the measurement data from the HPGe detector. The information, where the nuclide specific activity is located, is used to achieve much higher accuracy levels in radiological characterization of unknown objects.