Japan’s TV is reporting that GE is flying in several portable gas turbine generators to Fukushima, which would be to try to restore the on-site electric power, which be a really good thing.
I need to look more, but I think what I was told is the issue isn’t that that the diesel generators “failed” per se, but that water from the tsunami got into the room with the Power Distribution Units (PDU). An overly simplistic description of a PDU is the breaker box in your house, except 1000 times bigger. The incoming power arrives and is cleaned up and then spread out among all the circuits that need power – so everything in the place is dependent on the PDU. The PDU could easily become a “single point of failure”.
My first (and only) program at Duke before I was moved to the customer billing area was a program that prints out the Cable Routing database. Every scrap of wire in the entire plant is entered into the database (this is an NRC requirement). Each wire indicates which terminal it is connected to at each end, its purpose, what color the wire is, who made the copper in the wire, whether the wire is nuclear safety related (in which case it gets extra scrutiny), etc… every bit of information you could possibly ever need.
The Fukushima’s most positive thing is they are all GE designed reactors. One of the complaints as nuclear plants were starting to be built was that everyone was making up their own designs: See: Duke Power – each new design meant a seperate certification process, and the knowledge down that road that divergent designs would all develop different risks as the plants aged. A standard design might have other issues, but if you caught a flaw in one plant, you could then immediately apply that lesson to all reactors of the same design instead of every incident being a one-of-a-kind adventure.
I can’t begin to imagine the challenge of rewiring the circuits of a nuclear power plant while you’re being bombarded by radiation, but at least the technicians pretty much should know what is where – although people have been known to build stuff with slight deviations from the blueprints.
A diesel will “fail” if a non-compressible substance (like water) gets into the air intakes. As the diesel cylinder is essentially a positive displacement pump, this would cause catastrophic damage to the unit. Given all that, you are likely correct that the electrical distribution system was compromised by the waters of the tsunami.
The electrical distribution system is a nuclear safety-related system where it concerns the diesel generators. As such it has been “as-builted”, thoroughly documented, digitized and is maintained at the highest level of accuracy since construction. Unless there was structural compromise, it should be reasonable to believe that once a power source arrives, the distribution system will still be useable.
We can only hope that the lack of funding, interest and research in nuclear power hasn’t stripped GE of all of its top-level experts in the field.
I wandered onto an obscure web site making a pretty obvious point, but to lob up a softball…
The thinking of risk management goes something like this:
let’s say based on the evidence there is a one in 1000 chance when you push the butotn to start a diesel generator that it will fail.
So if you put in three diesel geneators, the risk of all three failing is 1/1000*1000*1000 or one in a billion.
The flaw to that thinking (which is painfully apparent now) is that you have to ask “is there something that could happen that would cause all 3 redundant systems to fail at the same time for the same reason?
The GE-Hitachi ABWR brochure mentions that in addition to the diesel generators, there is a gas powered turbine generator (probably very similar to the one they are flying in)
If it was just that the diesel generator was fouled by water, wouldn’t they have flown in a portable generator as fast as they could?
The unit down at Fukushima #2, 7 miles to the south reportedly has 1 out of the 3 generators that didn’t fail, so it is limping along for now.
When the outside power was lost, would it be normal for all 3 geneators to kick on, or only #1, and then #2 if #1 failed to start… would the generator still be ruined if it was temporarily under water but not running at the time? Do they start based on a battery? Is there a backup way to start up the backup generator if the battery is dead?
I follow the thoughts. I would only point out that the likelihood of 3 diesels failing may be very small while the lack of transportation, logistics, radiation levels and sheer panic causing staffers to miss an obvious solution is somewhat larger. Generally, nuclear plants are designed to be inherently safe. It is only when we add humans that the trouble begins.
The downstream load would determine when the diesels kick in and in what sequence (although it may be “hardwired”). The diesels fire a number of ways (depends on vendor) including explosives. If they auto-fired, they would most likely auto-stop when water was detected at the air inlet(s) which would render the site powerless. If all of that failed, then I also see the potential for the diesels to be broken beyond repair. I suspect the “portable” generators have negated the need for this assessment, however. I hope the wiring and circuitry is intact and will handle the minimal power that the generator(s) can supply.
They have also indiciated now that part of the problem is the circulator pumps were damaged by pumping sea water (they would know to put a filter on the inlet hose, right?)… So after they get power and repair the pumps, then they’ll resume pumping sea water through the pumps. Rinse and repeat.
In March 1975 there was a fire at the Browns Ferry Nuclear plant. This was a 3-unit GE-BWR operated by the TVA, a government utility (sound familiar?). The fire began in a cable spreading room as an assistant operator was using a candle (don’t laugh, it’s TRUE!) to check for air leaks in the foam used to seal the cable penetrations. BFN units 1 and 2 shared a common cable spreading room that was located beneath the common control room also shared by units 1 and 2. The fire got into the insulation and eventually damaged about 1600 circuits. Although nuclear safety was never compromised, the fire has far-reaching impact on nuclear plant design. From that point forward all U.S. plants were retrofitted to ensure positive physical separation between all safety-related components such that common-mode failures (fire, flood, pipe breaks, etc.) could not take out more than just one train of safety-related equipment. Cable rooms and penetration rooms were re-designed so that common points of failure were eliminated. There was an enormous body of rework done under 10CFR50 “Appendix R”, which provided the regulatory basis for the design changes.
I do not know whether the Japanese adopted Appendix R back-fits for their GE reactors. They certainly should have. If they did, there would be independent power distribution rooms for each diesel and they would be physically separated by fire and flood walls. It would be incompetent to specify three emergency generators with a common mode point of failure. Such an arrangement would never pass even rudimentary probabilistic risk management (PRA) review in the U.S.
From what you can see or find out – are they operating all 4 (or 6) units from a central control room or is it more seperated than that?
I’ve heard vague things so far that the NRC had warned Japan a couple years ago that they needed to do some retrofitting of the GE reactors.
Two different people in the chat room where I vanished to said that Japan had turned down the offer by GE for the gas turbine generators. The first question that hit my brain was “is there natural gas on the plant site?”…. if not, then I assume we’re talking Propane.
Another thing which my relative surfaced – Japan has a very peculiar issue – one side of the island is using 50 Hz power and the other uses 60 Hz – essentially creating two independent incompatible power grids. Shifting power around between the two means some unpleasant issues that you can probably explain better than me.
My assumption is that the plant was built to operate on the same power as the surrounding cities.
My relative also picked up on (and seems to be true) that the Fukushima is corneected to the ground by underground power cables, not overhead high tension lines. The oil filled underground power lines failed – which is why they are having to run cable on the ground to reconnect the plant. They started doing that today, apparently.