FAULT IDENTIFICATION OF A SMALL MODULAR REACTOR BASED ON PCA METHOD

Peiwei Sun,Huasong Cao,Bingwang Li
DOI: https://doi.org/10.1299/jsmeicone.2019.27.1254
2019-01-01
Abstract:Small Modular Reactor (SMR) usually adopts an integral design. The reactor core, pressurizer, coolant pump and steam generator are placed inside the reactor vessel. The responses of the SMR to accidents are different from those of traditional pressurized water reactor. Therefore, it is necessary to investigate the accident and find a suitable method to identify them. An SMR is modeled using RELAP5. The model includes the reactor core, coolant pump, pressurizer, and once-through steam generators. To simulate the steam generator tube rupture accident (SGTR), one tube of the steam generator is modeled separately from the others. Valves are connected to the tube and the SGTR accident is triggered by opening the valves. Different size of breaks are simulated and the accident process is investigated. To identify the SGTR accident, principal component analysis (PCA) method is applied. The values of critical parameters are recorded. The PCA method is used to analyze the data during the accident process. The changes of the principal component score, the T2 statistics and the Q statistics with time are calculated. The dynamic relationship between the break sizes of the accident and the three variables is derived. The occurrence of the accident is detected by the change of the three statistical results, and the relationship between the different breaks and the statistics is established to determine the size of the accident. From the results, it can be concluded that the PCA method can be used to identify the SGTR of an SMR and the size of the break can be quantified.
What problem does this paper attempt to address?