RELAP5 BWR-4 model development and validation for NPP Muehleberg (KKM)

摘要

The RELAP5 thermal-hydraulics code was developed for best-estimate transient simulation of light water reactor coolant systems during postulated accidents. Muhleberg NPP (KKM) equipped with a BWR-4 reactor has been using the RELAP5 code from the early 90s for deterministic safety analysis. The KKM RELAP5 model has been undergoing an extensive project to upgrade and validate the plant model in order to provide the plant with the ability to: 1. independently review transient and accident analysis calculations that vendors perform to satisfy regulatory requirements or guidelines, e.g., ENSI Guideline A-01, 2. independently review transient and accident analysis calculations that vendors perform to validate the design of planned plant modifications. 3. independently perform scoping analyses in these same areas, in advance of vendor calculations. The KKM RELAP5 model development and validation consists of several interconnected stages. The validation work performed by KKM focuses mainly on the specific KKM model rather than on a generic validation of RELAP5. For the generic validation, KKM relies on the great amount of work performed by national labs, universities and research institutes all over the world. All the simulation results presented in this work were run with RELAP5 mod3.3 patch04 from the US NRC. The first stage of the KKM specific validation involves ensuring that the physical plant (e.g., geometry, materials, system and component design, and protection and control system logics and set points) is accurately represented by the RELAP5 model. In one sense, this is largely a documentation effort that requires showing traceability from the RELAP5 model back to a plant design or as-built document. The second stage, which aims to show that the model's predictions are consistent with actual plant behavior, requires benchmarking the model to actual plant operational data (steady-state and transient). These benchmark comparisons are evaluated to determine how accurately the model predicts plant behavior over a wide range of operating conditions. Results from these comparisons are used to validate the model. The third stage involves demonstrating that the model will simulate and yield physically reasonable results for postulated transient and accident scenarios that are analyzed by vendors. Additionally some separate test cases are run to validate specific parts of the model and to help assist the choice of some modelling options. Two examples are presented in this paper. This paper describes the KKM RELAP5 model, validation process and the transients that are used for model validation. The most interesting simulated plant events, including a total loss of feedwater (LOFW) test, are presented. Overall, the KKM RELAP5 model predictions match very well with the plant data.