The reactor building of Units 1, 3 and 4 were destroyed by hydrogen detonation. The building of unit 2 was not destroyed, because the blowout panel of the reactor building dropped down by the detonation of Unit 1. The hydrogen detonation of Unit 1 building scattered the debris on the site and made preparation of securing activities for Units 2 and 3 difficult.
The primary containment vessels (PCVs) were inerted by nitrogen. Recombiners of hydrogen were equipped. The temperature and pressure of PCVs became high above the design conditions. The leakage of hydrogen from PCVs occurred at the penetrations and the gasket seals of the flange. It accumulated within the reactor buildings. Venting of PCVs caused hydrogen leakage to the piping connected to the stack.
The detonation of the reactor building of Unit 4 was thought to be caused by the leakage through the piping of stand-by gas treatment system (SGTS) connected to the common stack. The air operated valve of the SGTS piping failed open by the loss of power as well as the loss of air driving the valve as the backup. It caused the leakage of hydrogen from Unit 3 to Unit 4 that was not in operation at the accident .
The provision against hydrogen leakage at severe accidents should be elaborated and the respective measures should be performed.
Measurement at Severe Accidents
Important reactor parameters such as water level, pressure and temperature could not be measured due to the loss of DC power after the tsunami. The water level, the most important safety parameter of LWRs was measured erroneously after core melt down because of the change of the reference water level by evaporation due to the high containment temperature. It erroneously showed that the water level existed in the middle of the core. The wrong information confused the actions and harmed the reliability of the TEPCO information to the public. Mental bias of the specialists hoping the survival of the plants also decreased the reliability. It should be noted that the water level monitor did not work well at TMI-2.
Habitability of the main control room (MCR) was deteriorated at the accident. The air ventilation system of the MCR with charcoal filters lost the function.
Important reactor parameters as well as radiation level, radioactivity and hydrogen concentration in PCV need to be measured for management of severe accidents.
Management of Severe Accident
The employees and workers at site had to conduct accident managements under extreme circumstances such as darkness, high radiation, loss of reactor monitoring and communication ability, scattered debris by earthquake, tsunami, and hydrogen explosions. Working under such conditions was not prepared at all. The command of TEPCO headquarters also suffered from difficulty in understanding the situations and making decisions.
The containment venting procedure is written in the manual that the director of the plant orders it. But it was negotiated with the central government and took time to be conducted. The seawater injection was halted by the order of the TEPCO headquarter, but it was continued by the decision of the plant director. There was confusion of command.
There should be a clear definition of information, decision responsibility and actions dedicated to the organizations involved during the management process in case of extreme situations or a severe accident.