Comparison of an integral and a stand-alone calculations of hypothetical severe accident in a generic LW SMR with ASTEC code

This paper investigates a Beyond Design Basis Accident (BDBA) simulation with ASTECv3.1.1 computer code in a generic design of Light Water SMR. This design is an Integral Pressurized Water Reactor (iPWR) with dry containment and electrical power 300 MW (thermal power 1000 MW). The investigation concerns the efficiency of applying the In-Vessel Melt Retention strategy (IVMR) with external vessel water cooling in the considered SMR. The initial event that actuates the accident is double-ended guillotine break in the Direct Vessel Injection Line (DVI) and the postulated failure of all passive safety systems. The lack of all safety systems leads to impossibility the reactor core to be cooled. It creates conditions for accident progression with core overheating and slumping in the lower head of the reactor vessel - Lower Plenum (LP). At the moment of corium pouring in the LP the cavity is full with water from the Pressure Supress System (PSS) containers. So, there are conditions for investigation of all important phenomena that could be seen during the applying IVMR strategy with external vessel water cooling in the considered integral light water SMR. An integral ASTEC input model prepared by ENEA (Italy) and modified by INRNE was used to simulate this severe accident. The integral input model represents all basic systems of the light water SMR including all safety systems. The simulation finishes with LP vessel failure due to water depletion in the cavity that trouble significantly reactor cooling. This simulation takes a month and half computation time. And it is not so suitable for uncertainty and sensitivity investigations. That’s why, using the results from the integral calculation a stand-alone ASTEC input model was developed by INRNE to represent only the IVMR phase of the transient when the melted corium pool is slumped in the LP and the cavity is still filled with water. The initial and boundary conditions for the stand-alone input model have been taken from the integral calculation. The stand-alone input model has been validated against the integral input model. The calculation results of the stand-alone model are enough closer to the results from the integral input model and last for a few hours. In this way they are more appropriate for uncertainty and sensitivity investigations.