HOW FUKUSHIMA DAIICHI SEVERE ACCIDENTS COULD BE AVOIDED

摘要

As a result of the very serious consequences of the Fukushima Daiichi core melts at Units 1F1,1F2, and 1F3 and of their significant implications to worldwide nuclear power generation, numerous assessments have been published to-date. There is general agreement about their three principal causes: the assumed tsunami design basis which was low by a factor of 2 to 3; the very difficult post-tsunami impact which led to quite poor site conditions and overstressed plant personnel and management; and the Japanese regulators decision to not consider beyond design basis station black out (SBO). It left Japan government, its regulators, plant owners, plant management, and operating personnel without necessary plans and procedures to deal with the extreme SBO at Fukushima. During SBOs, all water nuclear power plants require early cooling of the reactor cores to avoid core melts. In Boiling Water Reactors (BWRs), this need is satisfied for limited periods of time by isolation condensers (IC) as installed at 1F1 and by reactor core isolation cooling (RCIC) systems utilized at 1F2 and 1F3. If they stop working, provisions are installed to add fire water to the depressurized reactor pressure vessels (RPV) to insure that two critical conditions are avoided: (a) keeping the reactor water level above the middle of the reactor core to prevent zircaloy fuel cladding chemical reaction with steam producing hydrogen, accelerating decay heat, and invalidating reactor water level data; and (b) not allowing the containment pressure to reach excessive values in order to avoid radioactivity leakage to the environment and to permit the reactor to be depressurized to allow adequate fire water injection. It is the purpose of this presentation to show that those two critical safety conditions during SBO could have been predicted readily by an integral decay heat method which would have helped avoid the 3 core melts at Fukushima. At Fukushima, not enough priority was given to assure early and safe reactor cooling by adding fire water to reduced reactor pressure of their BWRs. In the case of Unit 1, the safe strategy would have been to restart IC after its loss of DC power, and, if that was not possible, to use the installed fire water system to provide water to the depressurized reactor. That action had to be carried out within 3 hours as determined by the proposed simplified integral decay heat method to avoid the critical reactor water level falling below its core midpoint. In the case of Units 2 and 3, RCICs were allowed to run too long without the foresight and a clear priority given to implement reactor depressurization and fire water addition. Unfortunately, also, it happened that preparations for 1F1,1F2, and 1F3 to carryout RPV depressurizations and fire water additions were disrupted by hydrogen explosions at 1F1 and 1F3 as well as by other unanticipated circumstances. It is important to note that the proposed strategy to avoid core melts would have required full delegation of authority to the sites with no time available for necessary approvals for fire water addition or containment venting as well as intensive training and very strong safety culture of plant managers and personnel. That necessary condition may not have been prevalent at Fukushima Daiichi and not in agreement with Japanese usual reliance upon consensus.