Improvement and validation of OSCAR-3 usage in SAFARI-1 core modelling : focus on cross section library creation and target

Abstract

The SAFARl-1 nuclear reactor, an acronym for South African Fundamental Atomic Reactor lnstallation-1, is a 20 MW tank in pool Oak Ridge type Material Testing Reactor at Pelindaba near Pretoria in South Africa. SAFARl- 1 is a research reactor built primarily to supply neutrons for basic research and radioisotope production. Various calculational tools and computer codes are essential for the estimation of safety parameters, core reload planning, and effective fuel and product utilisation support. The ever increasing demand for isotope production, specifically Mo-99 fission product, produced by in-core irradiation of target plates, has led to the development of various methodologies of including target plate irradiation models in SAFARl-1 calculations. These studies are essential for optimisation of isotope production yields, while ensuring that safety limits are met. The methods under consideration for modelling the on-power loading and unloading of the target plates use the core calculational code system, OSCAR-3. Results are evaluated based on comparison to plant data, results from existing methodologies, and past experience of reload planning operation. Effects on the reactivity, flux and power distributions, fuel economy and safety, as a result of target plate irradiations were quantified for each of the various modelling methods using core cycle calculations. The inclusion of target plates in reload calculations, showed a significant increase in predicted reactivity of up to 2000pcm. A core-wide power redistribution and tilt of up to 2% along with much improved cycle length prediction was found. It is concluded that all considered methods of including target plates will find application, some important for safety, others for planning and core-follow calculations. Independent of which methods will eventually be applied, it is clear that the inclusion of target plates is essential for accurate reload planning and should be applied in future.