Two simulation tests for a boiling water reactor large loss-of-coolant accident (LOCA), conducted in the two bundle loop, were analyzed using the current licensing code system. These tests were recirculation-pump suction-line double-ended break tests. One of these tests assumed failures for LPCS and 1 out of 3 LPCIs, and another test assumed HPCS failure. A main objective of these analyses is to confirm the conservativeness of the licensing analysis models. Conclusions reached from the analyses are as follows :
(1) Calculated heater surface temperature begins to rise much earlier than the measured temperature, due to the conservative GEXL model for a LOCA analysis.
(2) Calculated heat up rate is higher than the test data, mainly due to neglecting the steam cooling in the analysis.
(3) Calculated heater rewetting time is later than the test data, due to neglect of counter current flow limiting at the core inlet, when the measured ECCS flow rate is used in the analysis.
It has been confirmed that the current licensing analysis models give a conservative result for peak cladding temperature (PCT), due to the model conservativeness factors presented above, when the measured data are used in the analysis for the outflow from the system and the inflow to the system.
View full abstract