Simulation of AP1000's passive containment cooling with the German Containment Code System COCOSYS

In this paper the applicability of the German containment code system COCOSYS to the simulation of AP1000's passive containment cooling is presented. For this purpose calculation results of an adopted nodalisation from Westinghouse's WGOTHIC code and a nodalisation featuring a plume modelling are validated against experimental data of the large scale test facility. This facility is used to experimentally investigate AP1000's passive containment cooling system. The gained experience helps to perform conservative and best estimate calculations of AP1000's containment response during a large break loss of coolant accident. Results of the validation show reproducibility of WGOTHIC results with COCOSYS and good agreement of the calculated values using a plume nodalisation with the experimental values. The loss of coolant accident peak pressure of AP1000 containment conservatively calculated by COCOSYS only differs about 0.1–0.2 bar from results obtained by using the codes MELCOR and WGOTHIC. A best estimate calculation considering plumes above the compartments of the steam generators results in a peak pressure 0.8 bar lower than the WGOTHIC calculation. The performed validation shows the applicability of COCOSYS’ present heat transfer models to a new application field like the passive containment cooling. In consequence COCOSYS is able to deliver reasonable results for both conservative and best estimate AP1000 PCS calculations.

► Simulations of passive containment cooling with thermal–hydraulic codes are benchmarked.
► WGOTHIC and MELCOR results for experiments and AP1000 can be matched by COCOSYS.
► Validation against experiment indicates that COCOSYS is suitable for PCS simulation.
► Best-estimate calculation results in about 1 bar lower peak pressure during DBA LOCA.