Modeling Pebble Bed Type High Temperature Gas Cooled Reactors Using Particle Transport Code
The accurate modeling of nuclear reactors is essential for design, regulation, safety analysis, operations and forensic analysis. In the most demanding case, the calculations must be sufficiently rapid that they are much faster than natural time for burn and decay processes while still being sufficiently accurate. There are two classes of approaches to modeling the neutronics of nuclear reactors. The first is deterministic, where the neutron transport equations are solved using a combination of approximations and numerical methods based on a space-time discretization. Usually, compromises are made in the accuracy of the model in terms of the level of approximations used, the scale of the discretization and the level of geometric and material detail, in order to achieve a sufficiently fast speed of calculation. The error induced due to the compromises made is offset with compensating modifications of the physical description or processes. The second class of models is stochastic in nature. Here many statistically independent histories for each neutron event and all secondary events related to its interactions are tracked and various physical data is stored for later statistical analysis. The physics content manifests through the microscopic cross sections (probabilities) for each interaction. The thermal and material information is encoded into thermally averaged macroscopic cross sections, which are assembled together with the microscopic cross sections. The geometrical information is reconstructed using technical drawings of the system, essentially to the desired precision. The full physics of the reactor are in this way encoded into probability distributions to be sampled on an event-by-event basis in the particle tracking and data extraction process, as various objects tracked are manifested in different regions of the reactor. The particle transport code, Geant4, developed for high energy physics, has recently been extended to accommodate lower energy nuclear physics and energy scales down to the far sub-eV region. In this work, the application of the Geant4 framework to a new area, that of gas-cooled reactors, is investigated. This approach promises the capacity for a much higher degree of detail in the geometry and material definitions as well as the physics modeling detail. So far, some microscopic cross-sections, thermal macroscopic cross-sections and thermalisation has been tested. In addition, fission, burn and decay are modelled in the context of a scalable pebble bed HTGCR scenario. Wall-clock time stepping in the millisecond regime, as separate Geant4 runs, has been implemented, with complete handover of parameters to the next iteration. This allows subsequent time intervals to be continuous to previous ones, and also allows for later workflow scheduling with a thermal hydraulic code, which is designed to share pre- and post- time step data, in a joint simulation. Two new classes of physics for Geant4 have been implemented and the messaging machinery for the materials and geometry updates between runs (steps) has been used to achieve the current results. A criticality calculation for the neutron population of the system has been defined. The advent of massively parallel high-performance computing is expected to make it feasible in terms of the speed of the calculations.
Professor of physics at the University of Johannesburg, Faculty of Engineering and the Built Environment, Department of Mechanical Engineering Science
Simon Connell is professor of physics at the University of Johannesburg within the Faculty of Engineering and the Built Environment in the Department of Mechanical Engineering Science. He has research interests in Particle Physics, Nuclear Physics, Nuclear Energy, Materials Science, Quantum Physics, High Performance Computing and Applied (innovation) Physics. His research spans the fields above as does his teaching. He has participated in developing a 10 Module accredited course at the 4th year level in Nuclear Energy. He has published over 150 papers in International Journals and is an ATLAS author. His rating by the SA Research Funding Agency (NRF) cites him as having “considerable international recognition”. He is a past president of the South African Institute of Physics.
He is the founding member of the South African participation in High Energy Physics at the ATLAS Experiment at CERN, where with his group he participates in a Beyond Standard Model searches as well as engineering and technical activities. Specifically, they are part of a team that searches for the force carrier particles associated with dark matter, known as the dark vector boson(s). They use the discovered Higgs as a portal, but they extend the search to include also possible undiscovered Higgs’s as further portals. Their current publications have shown two dark vector boson candidates. The work focuses on finding further candidates leading to discovery or finding a background process, which would otherwise explain these events. The group is also working on Radiation Hard Humidity Sensors for the upgraded Inner tracker, and here there is a collaboration with colleagues in the Department who are experts in Optical Fibre based sensors. A further contribution to ATLAS is research on containerization and virtualisation of High-Performance Computing based on clusters as a new solution for the main trigger and data acquisition (TDAQ) farm. He has also worked for many years at the European Synchrotron Research Facility (ESRF). The current research focus is to develop a gamma ray laser based on aligned charged particle beam incidence on a specially fabricated diamond superlattice functioning as a crystalline undulator. This is done as part of the PEARL EU collaboration. He is interested in technology for competitive industry and in innovation and has a project on the intelligent sensor-based sorting of diamond in kimberlite, which is now being commercialized. In this project, they have shown at Lab-scale the capacity to detect diamond within kimberlite at a statistical significance that indicates the process would work scaled to run-of-mine online sorting at 700 tons per hour for kimberlite rock sizes up to 10cm with enclosed diamonds of down to 5mm. The high-rate, high-sensitivity detectors for this project have been developed also as entrained micro-source trackers to study fluid-flow in hydro-cyclones. In another inter-departmental collaboration this is aimed at benchmarking CFD calculations to optimize the hydro-cyclones as mineral separators for the mining industry. He is also involved with other colleagues in the Department to build the National Case through research for South African Advanced High Temperature Gas Cooled Reactors. He is interested in the combination of Monte Carlo and other beyond-deterministic methods with advanced computing solutions for the modelling of neutronics in the nuclear reactor core. A current major activity in the service of the discipline is the development of the South African user base for Light Sources, (these are premier international multi-disciplinary research tools) and the implementation of the roadmap towards the African Light Source.