Comparison of the U.S. NRC PARCS Core Neutronics Simulator Against In-Core Detector Measurements for LWR Applications
The safety analysis of a Nuclear Power Plant (NPP) is based on the application of complex computer codes that are able to simulate the physical behaviour of the system under normal operations and abnormal conditions. Therefore, such codes must be extensively and continuously verified and validated in order to demonstrate their reliability.
In this context, the current report presents an assessment study for the U.S. NRC 3-D neutronic core simulator PARCS, and it includes an evaluation of the performances of the code for LWRs applications. For this purpose, the cores of the Swedish Ringhals-3 Pressurized Water Reactor (PWR) unit and the Forsmark-2 Boiling Water Reactor (BWR) unit were modeled with PARCS. As regards the cross-sections needed for this kind of calculations, they were prepared by following a special procedure developed by the present authors since core material data were only available in the format of library and restart files created by the SIMULATE-3 neutronic core simulator. Correspondingly, a new cross-section interface was developed and verified by the Division of Nuclear Engineering, Chalmers University of Technology, in order to convert the SIMULATE-3 data into data suitable for PARCS. Thereafter, the PARCS models developed for Ringhals-3 and Forsmark-2 were used for neutronic core analyses, at different operating conditions, along several fuel cycles. The results achieved from these simulations were then compared against the axial power and the radial power distribution estimated from the measurements that were provided by the owners of the plants.
In the PWR case, the PARCS simulations predict satisfactorily both the core axial power profile and the core radial power distribution, although, in some cases, the deviations between calculated and measured data exhibit trends that need further investigations. For instance, the PARCS simulation at the beginning of a fuel cycle seems to overestimate the power in the center of the core and to underestimate the power at the periphery, whereas, at the end of a fuel cycle, the situation is opposite.
In the BWR case, the core axial profile was predicted in a reasonable manner, but quite significant discrepancies for the radial power distribution was found. The current work suggests that such a disagreement might be due to the inability of PARCS to properly model multiple composition control rods. In fact the largest deviations in the computed power from the measurements were observed for those fuel assemblies placed in the neighborhood of control rods.