Abstract
On 2011 March 11, a strong earthquake induced several large waves of tsunami hit the east coast of Japan. The tsunami caused extensive damage to the site of Fukushima Daiich Nuclear Power Plant (NPP). Fukushima Daiichi NPP employed BRW4 reactor with MARK I containment, which is identical to the design of Chinshan NPP of Taiwan Power Company. In the present study, an input deck of MELCOR based on the design of Chinshan NPP is construct to simulate the Fukushima Daiichi Unit 3 accident. In the simulation, the timings of water injection, depressurization, and containment venting are set as specified in the official Japanese government report about the accident. The results of simulation demonstrate that the hydrogen burns observed during the accident are not predicted by the code. The the reactor pressure vessel and the containment are remain intact throughout the accident. As predicted in the MELCOR simulation, there is 67 kilogram of hydrogen leak into the reactor building. Nevertheless, the concentration of hydrogen is still too low to induce hydrogen burn. There is no radionuclide release to environment in the base case simulation due to that the reactor building is kept intact up to the end of the simulation. In a sensitivity study, it is assumed that the reactor building breaches at the time specified in the official government report, The results show that 1.3% of Xe is released to the environment. The releases of other radionuclides are very minor. The results of MELCORE simulation of the accident are compared with the results of MAAP5 simulation. The differences are very significant. The results of MAAP5 simulation indicated that the reactor building breached due to the pressure spike predicted at 68.2 hours after scram, which was caused by the relocation of molten core from core region into vessel lower plenum.