Transport and deposition behaviors of graphite aerosol in the steam generator building during an overpressure discharge accident of HTR-PM

Abstract

The water-ingress accident is one of design basis accidents of high-temperature gas-cooled reactors (HTGR). The rapid change of the flow field and the direct brush of the high-velocity gas flow may cause a large fraction of graphite dust to resuspend, which is further discharged into the steam generator building when the overpressure relief protection is triggered. Due to the coupling with the radioactive fission products (FPs), the behaviors of the discharged particles are crucial for the reactor safety analysis and source-term evaluation in the overpressure discharge scenario. In this work, we numerically investigate the transport, deposition and releasing behaviors of graphite aerosol in the full-scale steam generator building of HTR-PM by the Eulerian-Lagrangian method. The mixture-ideal-gas law is used for describing the mixture between helium and air, and the CFD-DPM scheme is employed for tracking the trajectory and velocity of individual particles. Particularly, the effects of the irregular shape on the particle-fluid and particle-wall interactions are included. The simulation results show the particles are quasi-uniformly distributed in the steam generator building and the deposition and releasing rates with the time follow the exponential relationship at the late-stage. The distribution of all the deposited, escaped and airborne particles are analyzed. Due to the strong thermophoretic effect, a large fraction of particles are deposited on the cold walls. The extrapolation by the exponential relation indicates a final releasing fraction of 18.94%. Compared to the current conservative assumption that all the dust particles in the steam generator building will eventually enter the environments, our results demonstrate that the steam generator building of HTR-PM still has certain capability to retain the graphite dust during an overpressure discharge accident scenario.