Preliminary reactor design for nuclear thermal propulsion

In the last years, new interest in nuclear space propulsion has been shown by space agencies. NASA defines Nuclear Thermal Propulsion (NTP) as a key technology for long-range spaceflight. Many HEU (Highly Enriched Uranium), fast reactors design for space applications were proposed during and after NERVA and ROVER projects. HEU reactors have many advantages: they are usually more compact than LEU (Lowly Enriched Uranium) reactors, they show a longer operating lifetime and safety in accidental scenarios can be easily ensured. Their main drawback is the high content of fissile materials, which is a primary security concern. The present work investigates the possibility of a LEU reactor design with CERMET (Ceramic Metallic) fuel and its optimization. Due to the lower U-235 content, LEU reactors are preferable from the nuclear proliferation viewpoint, while CERMET fuel may provide better performances than NERVA derived fuels, such as (U,Zr)C. Neutronics, burnup, safety and thermal-hydraulics simulations were performed by Serpent and MATLAB: the model was first tested against an available case in literature, then applied to the reactor of interest. Criticality eigenvalue in both operating and fault conditions, peaking factors, reactor mass, burnup time, coolant average temperature and fuel maximum temperature were chosen as figures of merit to lead the design. It is found that safety is much harder to ensure in a LEU reactor than in a HEU reactor, leading to the necessity of exploring a large number of configurations: for this reason, a Python code was developed; the code can handle the whole process, from pre-processing to post-processing, including the integration between Serpent and MATLAB. Thus, the following procedure was implemented: a group of new reactor configurations were built and tested; checking the figures of merit, only the most promising configurations are selected for the optimization and for a safety analysis that simulates reactor’s behaviour in four accidental scenarios; finally, the cores that satisfy safety requirements undergo a thermal-hydraulic analysis, which verifies that thermal limits are not exceeded and evaluates rocket performances. In that way it was possible to simulate about 1000 configurations, which differ in core materials, core geometry, fuel element enrichment and enrichment zones. Results show that radial reflector thickness heavily affects neutronics and safety, while coolant mass flow rate needs a precise tuning to lower fuel temperature and to ensure competitive rocket’s performances, making orificing a desirable optimization. In addition, spectral shift absorbers look not as effective as in fast reactors, owing to their massive absorption of thermal neutrons during normal operations. Therefore, an advanced control drums system is proposed to ensure subcriticality even in the worst accidental scenario. It is found that few configurations fulfil both safety requirements and NASA requirements for rocket thrust, specific impulse and mission time. Thus, a LEU, CERMET fuelled reactor design is challenging but feasible, and both the developed code and the optimization procedure may be applied successfully to find the best configuration.