Been following this forum for a couple years, but just decided to join....
A question for folks closer to the industry than I. Suppose you were sent back in time 1 year or so. Your job is to provide information that there would be an earthquake and tsunami somewhere in the world that would cause a serious nuclear accident. You contact the US NRC, IAEA, TEPCO, GE, and whoever else has a pile of engineers that could develop procedures. You identify the equipment as GE-4, Mark I containment, and lay out the challenges related to the station blackout, keeping the core cool, hydrogen buildup and explosion, loss of secondary containment, sfp issues, etc.... You're not allowed to give the location of the plant to preclude shutting it down before the earthquake. Knowing the outcome, is there anything that could have been done differently as soon as the earthquake hit?
Many pages ago, someone tossed out the idea of bootstrapping the generator off residual steam. Is this something feasable, or just a "wouldn't it be nice if" Along those lines, how about a limited restart of one unit after the trip to "back up" the EDG's. What about running the HPCI turbine with the turbine exhaust routed to the atmosphere instead of the supression pool. Obviously a release, but would the tradeoff of being able to keep the core cool and suppression pool sub-saturated longer have improved the long term situation? Am I correct in assuming that if the core remained intact there wouldn't be any issues with hydrogen or containment integrity?
It looks as if the plant operators followed accepted procedures early in the situation. When working with a complex system like a nuclear power plant, you need to stay with proven, established procedures. I'm not suggesting that anyone should have responded differently to the situation, just looking to see if a better outcome could be provided if a similar situation is encountered in the future. My research suggests that earlier designs like the GE-4 have a great deal of their core damage frequency estimate attributed to station blackout events. How does that look going forward?