The operator of a Japanese prototype fast-breeder reactor which was hit by a sodium leak had breached regulations by not immediately closing air ducts after smoke alarms went off, the Kyodo news agency said on Monday.Governmental Power Reactor and Nuclear Fuel Development Corp (PNC) which runs the reactor, Monju, said it did not close air ducts until 3 1/2 hours afteralarms went off. Regulations in the construction permit for Monju, submitted to the government, stipulate air ducts to be immediately closed when smoke alarms go off, the Kyodo news agency said.
>>Governmental Power Reactor and Nuclear Fuel Development Corp (PNC) which runs the reactor, Monju, said it did not close air ducts until 3 1/2 hours afteralarms went off. Regulations in the construction permit for Monju, submitted to the government, stipulate air ducts to be immediately closed when smoke alarms go off, the Kyodo news agency said.<<
For anyone wondering why shutting down the air conditioning is so significant in an accident like that, some more background. Burning Sodium can not be extinguished using water as it violently reacts with water, producing explosive hydrogen gas in the process. Halon and carbon dioxide (dry ice) based fire extinguishers can't be used for a similar reason. The only way of putting out a sodium fire therefore is to starve it of all oxygen by cutting off the air supply and waiting for the fire to use up all oxygen.
Because of the danger of sodium fires, the primary coolant cycle where the sodium is highly radioactive is surrounded by inert nitrogen gas to reduce the chance of fires. Not so the secondary coolant cycle where the sodium fire occured. By leaving the air conditioning on the operators of Monju kept pumping fresh oxygen into the room, fanning the flames of the burning sodium. What they were doing was like pouring oil into a fire (I mean a conventional fire -- flooding the room with oil actually could have put out the sodium fire!). As a result it took 12 hours before the dense smoke had cleared enough for the operators to be able to see what had actually happened in there!
One of the weak points of sodium cooled reactors is the risk of reactions between sodium and water that produce hydrogen gas. It was a hydrogen gas explosion that blew the roof off the Chernobyl reactor. Running air conditioning systems could spread that gas into other parts of the reactor. Last but not least, if the accident that caused the fire had involved any radioactive substances (which luckily it didn't) then the running airconditioning system could have spread radioactivity to other parts of the reactor or even the outside environment.
That this gross operator error occured at all is reason for concern (remember, the Chernobyl disaster was also caused by an operator error). It makes you wonder how well trained the crew operating the reactor really was. It has been reported that in the three years since it was first readied for startup, Monju has never had a single fire drill specifically for sodium fires, only for conventional ones, despite the peculiar fire risks associated with that substance (how often have you seen a bucket full of water fan the flames of a fire?)
DAILY NEWS, 19. Dezember 1995
Panicked workers afraid to act without permission worsened a leak at a plutonium-producing reactor by waiting 90 minutes to shut down the overheated system, officials admitted Monday.
Two to three tons of corrosive sodium coolant leaked from the government-funded reactor in western Japan on Dec. 8, after smoke expanding in a secondary cooling facility nearly got out of control.
The leak would have been less severe if employees of the Power Reactor and Nuclear Fuel Development Corp. had shut it down sooner, officials said.
>>The leak would have been less severe if employees of the Power Reactor and Nuclear Fuel Development Corp. had shut it down sooner, officials said.<<
I already pointed out that long delay one day after the accident. The leak could only be stopped by lowering the coolant level in the affected coolant cycle and that requires first shutting down the reactor. Having to restart it from scratch after the cleanup and repairs would cause a delay of several months in the testing program of this highly prestigious reactor project. That's why the operators were so reluctant to take the necessary actions.
On the other hand, no one knew exactly what was going on or how severe the leak was/had been until HALF A DAY AFTER the accident. They had no idea where exactly the leak was and they didn't have a clue how much sodium was leaking per minute. All they knew at the time was that a heat sensor on a coolant pipe had triggered a fire alarm, that there was a problem with the coolant temperature (which could actually mean the reactor core would overheat, which is extremely dangerous on Fast Breeders) and that there was dense white smoke indicating a sodium coolant fire. Continuing running a Fast Breeder reactor under these conditions for 93 minutes is like keeping on driving a fuel truck after you notice the engine is on fire, except that a resulting explosion could be far more serious...
What that aspect of the incident indicates most clearly is where the real priorities are in this project: Certainly not on safety!
>><<What that aspect of the incident indicates most clearly is where the real priorities are in this project: Certainly not on safety!>><<
>>I really can't agree with that, Joe. Even if the priorities of the Monju project were so flawed, I doubt that they'd risk a home-made Hiroshima. I think it's much more likely to have been a bad case of "manual syndrome" on the part of the workers; that well-publicised panic reaction to new situations said to be instilled by some sections of the Japanese education system.<<
I'm afraid I have to call your "home-made Hiroshima" comment naive. Do you think anybody in the former USSR *wanted* the Chernobyl disaster? And yet it happened.
I wouldn't disagree about the "manual syndrome", it certainly was a major reason for the lack of action taken when the alarm went off, but I think they wouldn't have been so scared of shutting down the reactor if they had been explicitly assured that that was the right thing to do whenver they weren't sure how serious a problem was. That doesn't appear to have been the case.
Think about it: If there's a bomb threat against a jetliner then the plane is grounded and searched, even if the threat might turn out to have been a hoax. Safety of the passengers always comes first. Why were much greater risks taken in Monju?
A study on the US nuclear industry in the late 1970s investigated a few hundred incidents and traced about half of them to operator errors. To err is human, unfortunately. You can't stop that. What you can do is train people well and make sure they've got their priorities right. In an industry where accidents can have effects as disastrous as the Chernobyl accident the first priority should always be safety. Shutting down the reactor would have been the safest thing to do as long as the seriousness of the problem was unknown. If the emphasis had been put on safety during the training of the reactor crew I doubt they would have responded the way they did.
A second aspect of this question concerns the technology of the reactor itself. Most nuclear reactors in earthquake prone areas have auto shut-off systems that automatically shut down the reactor in case of a major earthquake. Your average family saloon has a fuel injection system that automatically shuts down as soon as the airbag triggers in a crash, to reduce the risk of subsequent fires. How come a $20,000 car with 10 gallons of fuel on board automatically shuts down its engine in an accident but a $6,000,000,000 nuclear power station with 1600 kg of the deadliest substance known to man inside its reactor core has to be switched off MANUALLY ninety-three minutes after it catches fire?
DAILY NEWS, 21. Dezember 1995
Angered by a nuclear plant operator's cover-up of videotape showing the full extent of a recent accident, Japan's government today stepped up its investigation of the mishap.
The Science and Technology Agency said it has lost trust in the operator's ability to probe its own actions, so agency investigators will now be empowered to give orders and control the inquiry.
The move comes a day after the Power Reactor and Nuclear Fuel Development Corp., known by its Japanese acronym of Donen, admitted that it purposely concealed graphic videotape of the Dec. 8 accident, in which two to three tons of sodium leaked in a secondary cooling system of the plutonium-powered plant.
The head of the Power Reactor and Nuclear Fuel Development Corporation that runs the Monju reactor in Fukui pref., that became the scene of the most serious coolant leak in the history of Fast Breeder nuclear reactors on Dec 8, had to publicly apologize on Wednesday for what appears to have been a deliberate coverup of evidence.
The PNC (called Donen in Japanese) which is controlled by the Japanese Science and Technology Agency had initially shown a 1 minute video of the accident site, later followed by another 4 minute clip and claimed that was all the footage that was available. None of this showed the actual coolant pipe that had leaked, only the floor of the site where the leak had occured. Later it was revealed that a total of 15 minutes of video had been taken, including footage of the affected pipe. The one minute and four minute clips were only an edited version of the original clip. The existence of the unedited tape was kept secret for several days. It appears that the original tape had been deliberately edited to remove the footage of the leaking pipe. Asked for an explanation at a press conforence of why the PNC had withheld information, Takao Takahashi, a board member of the PNC could only say: "There is none." It also turned out that the PNC had inspected the site of the accident before the time given to the press.
The handling of this incident has considerably irritated the Science and Technology Agency, causing it to switch to a stricter regime of investigation under which it can fine the PNC for withholding information or otherwise blocking the investigation, an unprecedented step in the history of the Japanese nuclear industry. The PNC is not only in charge of the Monju reactor but also of the nuclear reprocessing plant under construction in Rokkasho, Aomori prefecture that will extract plutonium from spent nuclear fuel from Japan's 49 nuclear power stations once it will be completed.
DAILY NEWS, 22. Dezember 1995
Investigators found a second videotape on Friday of the scene of an accident at Japan's prototype fast-breeder nuclear reactor, an official at the Science and Technology Agency said.
The agency also said the reactor's operator had falsified a report on the time when workers first entered the scene.
Science Minister Yasuoki Urano expressed anger at the falsified report and said the operator could be penalised.
The operator, the Power Reactor and Nuclear Fuel Development Corp (PNC), said its director, Yasuhito Ohmori, gave orders for a first video to be edited "in order to make the tape easier to understand."
A PNC spokesman said: "The deputy director told some workers to keep quiet about the video editing."
PNC admitted on Thursday that it deliberately withheld some video footage of the accident on December 8, involving a leak of sodium coolant
>> The operator, the Power Reactor and Nuclear Fuel Development Corp (PNC), said its director, Yasuhito Ohmori, gave orders for a first video to be edited "in order to make the tape easier to understand."<<
Isao Sato, the deputy chief of the Monju office of the PNC was a little more open about this at a press conference. He admitted to giving orders to hide the master tapes of the video and to issuing gag orders to employees about the existence of the unedited originals. He was quoted as saying: "I was determined to cover up the case."
DAILY NEWS, 23. Dezember 1995
Three officials were relieved of their posts today because of attempts to cover up the seriousness of a recent accident at an experimental breeder nuclear reactor.
No one was reported injured in the accident, which has triggered renewed opposition to Japan's ambitious plans to depend on breeder reactors for its electric power needs. Breeder reactors are capable of producing more plutonium than they use. The government agency that runs the $5.6 billion Monju reactor in western Japan said the men were being disciplined for their handling of the Dec. 8 accident, when tons of caustic liquid sodium coolant leaked. Officials at the plant heavily edited a videotape taken after the accident to conceal the most serious damage, and hid a second videotape.
The cover-up angered Japan's government as well as local residents, and the Science and Technology Agency took over the inquiry into the accident from the governmental Power Reactor and Nuclear Fuel Development Corp., which runs the reactor. Officials of the corporation, called Donen in Japanese, said the three executives will be reassigned to other Donen ventures after the government inquiry into the accident is completed.
They said the Donen director in charge of Monju's operations will also be replaced pending approval by Prime Minister Tomiichi Murayama.
According to a source familiar with the Monju Fast Breeder reactor quoted by Kyodo news agency, the fire monitoring system of that nuclear power plant was shut off as soon as it set off a fire alarm after a sodium coolant leak that started a fire on Saturday, Dec 8, to keep it from ringing alarm bells.
While many fire monitoring systems automatically switch themselves back on again after they have been shut down, the Monju system staid disabled until manually re-enabled. Operators at the plant's control center can monitor the locations of hundreds of fire alarm sensors via video cameras but according to that source, a vital system that could indicate which of the many sensors was causing the alarm was missing in Monju. This deficiency in the fire alarm system made it difficult to locate the leak and may have contributed to the long delay before the reactor was finally shut down. Without the ability to locate the sensor that triggered a fire alarm the system was unable to serve its purpose while the original fire was still on, which is why it was switched off altogether.