First I want to thank the many contributors to this forum especially the moderators/mentors. I have read it all from the beginning and understood more than I would have thought. Thanks for the clear and concise explanations. Keeping up with it has been a bit like chasing the pot of gold at the end of the rainbow though.
As a teen I worked on the construction of the Pilgrim 1 plant in Plymouth, Mass (1970). I had the run of the place and recognize the Daiichi 1-4 designs. I later worked building construction and as an aircraft mechanic for a major airline. Part of the time with the airline, ~Y2K, was in the department that maintained the physical plant; emergency generators, boilers, fire suppression, electrical, plumbing, lighting, PLC’s, etc. The focus then was contingency planning and mitigation.
What I will contribute will be from a practical experience perspective. This will be long, typed with two fingers, copied and pasted and will cover some areas already addressed so skip it unless or until you have the time or interest to read it.
Emergency diesel generators in this application range from 1-25Mw and even in the smaller end
http://www.fairbanksmorsenuclear.com/engine_alco_251f.php range from 26-43 tons. As part of the earthquake design specification of the original plant they would have to have been anchored to the concrete floor of the building designed to hold such weight in the specified earthquake. Without witnessing it I have high confidence that they are still where they were installed (allowing for the fact of the whole nation moving a few meters). With the air intakes near the top of the (3+ meters, see dimensions) generators, inside an intact building 10-13meters above sea level I am confident that they did not draw water into the intakes and liquid lock. In addition the basements of the T/B and R/B remained dry until ~ 24 March after the tsunami and after much seawater had been pumped into the building and the generators ran ~30 min which is 20 min past the arrival of the after the tsunami.
I believe we will eventually find that the diesel generator problem originated with the fuel supply. The fuel tanks are outside of the buildings on a concrete pad separate from the building itself. Five minutes of earthquake shaking independent of the building, with fuel sloshing and adding to the forces on the tank/foundation, would stress the fuel supply lines and connections. If the break happened low enough to draw in water it would result in damage to filters, injector pumps and injectors that would consume more time than the golden 8 hours
http://www.osti.gov/bridge/servlets/purl/6124656-R8y05j/6124656.pdf might allow even if materials were available on site and undamaged. This fits with the ~30 min run time as generators installed inside of buildings typically have a small “day tank” that they actually run from. The tank calls for replenishment from the main fuel source as it draws down and holds a small amount of fuel. If a break drew in air the generators would run out of fuel and the fuel system would have to be purged of air which requires climbing all over a large engine. It is not as simple as filling your empty car.
Adding to the problem is the probability that emergency generator maintenance is probably done by an outside company on a contract basis. It is unlikely they would have had someone on site, and they would have needed mere than one. This is not a criticism. Elevator maintenance and air conditioning (HVAC) are also specialty occupations that are required only on an intermittent basis and not part of the skill sets needed to run a reactor.
As the NRC recognizes the critical nature of backup power many plants now have more than the minimum two generators in order to comply with new regulations.
http://www.nrc.gov/reading-rm/doc-collections/cfr/part050/part050-0063.html
I will add my opinion to the excellent suggestions already offered here in hindsight. Also, because I have just pointed out problems, it will make more sense to offer the solutions now.
In a tsunami risk area at least one new diesel generator needs to be placed in an elevated area outside of the risk zone. It should supply an electrical buss available to all reactor plants at its installation. All new or existing diesel generators at nuke plants should supply an electrical buss available to each plant in its location. All existing diesel generators in seismically active areas should be retrofitted with flexible fuel supply lines between the main fuel supply and the generator building. (Existing aircraft jet engine flexible fuel lines should easily gain approval for this application.) New construction of emergency power generation for nuclear power plants in seismically active areas should have the main fuel supply tank foundation integral with that of the generator building to minimize independent movement. The size of the day tanks inside the building should be increased to allow several hours use set to maintain 2/3 to 3/4 full and alarmed for any level below 1/2. This may require a waiver and additional fire suppression equipment. It will buy time however.
Daiichi 4 fires explosions while in maintenance shutdown have been difficult for some to understand. After all it was shut down safely. However it was not in its normal configuration due to it undergoing maintenance. This does not mean that it was in an unsafe condition. Just that the process of undergoing maintenance is not a normal configuration. By design, the systems that must be shut off and locked out, disassembled, removed for replacement or repair are placed in a condition that is not normal for operation. They are brought to this state by a safe, prescribed process. But there still remains a great deal of stored energy. The spent fuel storage heat for example. All tools, parts and assemblies at any vertical height have stored kinetic energy and are potential sources of sparks if they fall. Five minutes of earthquake, tsunami and aftershocks may account for fire origination. There may have been hot tools or torches in use at the time of the earthquake and flammable materials, like boxes containing new parts and supplies present that normally would not be present. Oxy/acetylene is commonly used in maintenance and presents the potential for another source of flammable gas before hydrogen might have become present. Portable, as well as normal, lighting present a potential heat source to flammable materials not normally present. If the spent fuel pond, or its piping, were to leak the potential for hydrogen formation in falling water levels has been well addressed by others. I think this is a sufficient, if not complete, list of potential sources of trouble for a shut down reactor going thru an earthquake. It may be a long time, if ever, before we know cause exactly. The information may not be worth the danger it presents to persons trying to discern it.
I have great admiration for those present at the Fukushima plants. They endured an earthquake, a tsunami, knew that their families (most are reported to have lived nearby) were in danger or worse, had their plants lose all electrical sources, were in a location inaccessible to rescue, relief or re-supply, lost heat, faced winter conditions outside and had limited food and water. Afterwards, things got bad as aftershocks continued, explosions and fires occurred, tsunami warnings repeated while they were trying to get a grip on things and then the batteries ran down leaving them with no instruments to tell them how bad things were becoming. They knew that the people and families nearby would be affected by what they did to control their plants at the same time that they continually had fewer and fewer tools to do anything about their deteriorating situation. The situation became different as well with each passing event (tsunami, aftershock, explosion, power loss).
I have fewer good feelings toward the utility management outside of the area. They also had loss and stress and fear but they also had distance from the immediate danger. They had greater means and responsibility to mitigate the damage that was sure to come. It seems that there was no one with a voice in the back of their head saying “autodestruct engaged, autodestruct will commence in eight hours” and then repeating with diminished time periods. The absence of practical actions taken is not in keeping with my understanding of the Japanese character. I do not know what the source of the inertia was. Perhaps in another thread when we have some facts.
) related to the hypothesis that it seems there is now some plutonium in the environment of the plant: when we talk about "millisieverts" for example, this is a measure of the dose equivalent radiation, which tries to quantitatively evaluate the biological effects of ionizing radiation. 