OpenMC Analysis of TRIGA Mark II Reactor Void and Temperature Reactivity Coefficients

Due to the very stringent safety requirements of nuclear facilities, there is a need of developing precise and accurate computational tools for the reactor’s safety analysis both during the licensing process and standard operation. Achieving this goal requires verification by other state-of-the-art neutronic codes and validation through comparison with experimental data of the simulation tools. The recent development of open-source platforms has increased the interest in adopting these technologies, which, compared to proprietary software, offer continuous exchange between developers and users and direct access to the source code. In safety analyses, studying feedback coefficients is crucial for evaluating the reactor dynamic response tocontrol and accidental scenarios. This kind of analysis provides an in-depth understanding of reactor behaviour under different operating conditions (i.e., different power levels) and experimental settings. Existing codes are suited for commercial reactors and may not offer the capabilities for simulating future generation systems. In this framework, this paper analyses the thermal feedback coefficient and void coefficient of the TRIGA Mark II reactor using a Monte Carlo model developed with the Python-based open-source OpenMC code. The reason behind the choice of this particular reactor is twofold: first, this reactor represents a landmark in nuclear research due to its unique asymmetrical configuration and the presence of UZrH fuel, and in particular, for its passive safety feature, made possible by its highly negative feedback coefficients. Second, a large number of available experimental data are readily available. This work considers two different scenarios for the validation: the first case is the insertion of positive reactivity through a control rod extraction, allowing the temperature to increase along with the power; the second case simulates the reactivity perturbation coming from the presence of a void volume (e.g., in sub-cooled boiling regime) through the placement of a sample made by aluminium and filled by air or water in the central channel. The experimental scenarios related to the evaluation of the feedback coefficients are accurately reproduced: the tracking of the change in the criticality level (k-eigenvalue) compared to some physical quantities (i.e., the temperature or the void level) allows for the calculation of the feedback coefficients, and the results obtained from the OpenMC simulation are compared to both the experimental results as well as the outcomes from the SERPENT Monte Carlo code, showing a good agreement