DAILY NEWS, 9 December 1995
Japan faced demands for sweeping changes in its ambitious nuclear power program on Saturday as officials began investigations of a massive coolant leak at a prototype fast-breeder nuclear reactor. Japan's most advanced reactor, Monju, was manually shut down on Friday after liquid sodium leaked from the cooling system, reactor operators said. Heavy smoke caused by a reaction between the leaking sodium and air delayed technicians' efforts to investigate the leak, officials of the governmental Power Reactor and Nuclear Fuel Development Corporation (PNC) said. Politicians from the western coastal prefecture of Fukui, where Monju is located, blasted the PNC for its delay in announcing Friday's accident. Fukui Governor Yukio Kurita criticised the corporation for waiting nearly one hour after the shutdown before contacting his government. "Along with the delay in the advisory bulletin, subsequent reports were insufficient," the governor said in a statement. "This is an extremely regrettable loss of the prefecture's trust in the overall safety of Monju." Kurita demanded a thorough investigation of Friday's leak, a "change of consciousness" about accident public relations and a complete revision of Monju's operating schedule.
Subj: Monju problems
To: all
Date: Sunday, 10. December 1995 03:21:17
From: Joe Wein, 100142,3715
A bit of background information on the significance of the recent accident (see 12-9 News section):
1) While the more common Light Water Reactor (LWR) type of nuclear power station that supplies 30% of Japan's electricity uses water as coolant, the Monju reactor uses liquid sodium, a chemically very agressive metal. Sodium is solid at room temperature and has to be heated to at least 97C (207F) in order to melt. At that temperature sodium spontanously catches fire when coming into contact with air or water. If it were ever let to cool off below that temperature it would solidify and expand, bursting the cooling pipes of the reactor like water in a car radiator without anti-freeze in a cold Chicago winter night. A major Russian FBR had to be shut down after a severe sodium fire last year. The French Superphenix FBR was closed after an explosion involving 100 kg of sodium that killed several workers. This accident only proves that the same laws of physics and chemistry apply in Japan.
2) According to one report, the reactor lost several thousand kilograms of coolant, which would be several percent of its total coolant volume. Despite this, the reactor was not shut down for over an hour. While an LWR has an emergency core cooling system (ECCS) that should cool the reactor core even if the normal cooling system fails (e.g. due to loss of coolant), the Monju FBR has nothing of that kind. Its regular cooling system is the only thing preventing the reactor from overheating.
3) While overheating in a conventional LWR will stop the chain reaction, it will actually accelerate it in an FBR!
4) The only way to shut down an FBR is by inserting control rods. These control rods may jam once the reactor overheats. (For reasons inherent in the design of FBRs, the reaction times available while operating these control rods are about ten times shorter than in LWRs).
5) While LWRs use oxide fuel with a comparatively high melting point, FBRs use metalic fuel that melts at much *lower* temperatures. Despite this fact the core of an FBR is actually operated at a temperature 260C (470F) *higher* than an LWR. If due to overheating molten plutonium collects at the bottom of the reactor vessel or even if rods just start bending the reactor will become supercritical, leading to a small nuclear detonation. With "small" I mean small by atom bomb standards but large for a power station. There have been a couple of computer simulations by experts in the US and they came up with something like the bomb that blew up the Federal Building in Oklahoma City. An explosive force of that order of magnitude could rupture the reactor vessel, allowing radiactivity to escape, or worse, the explosion itself or a collapsing reactor after structural damage has occured could compress the remaining plutonium, leading to an even bigger blast. The possibility of this scenario is one of the side effects of the FBRs sharing several critical design principles with the atom bomb (unlike the more common LWRs). The bottom line is, an FBR must NEVER EVER be let to overheat by loosing coolant.
6) Despite this the walls of the pipes in Monju (which vitally depends on them as it lacks an emergency cooling system) are only 9.5 to 11 mm (about 0.4 in) in thickness. By contrast, the coolant pipes in LWRs (which do have emergency cooling systems) have steel walls that measure 69 to 78 mm (3 in) in thickness.
7) While the pipes used in the experimental Joyo FBR in Ibaraki that preceeded Monju still had double walls (inner pipes within outer pipes) to prevent coolant leaks, Monju was designed entirely with single walls, in order to cut costs. Remember, even with that saving it already cost $6,000,000,000 (roku sen oku en) to build.
8) The fuel core of an LWR contains on average about 0.5% (1 in 200 parts) of Plutonium. 1 gram of Plutonium can cause cancer in between 20,000 and 1,000,000 people (depending on whose figures you want to believe). The fuel core of Monju consists of pure Plutonium, 1500 kg of it altogether (that's 1,500,000 grams to all non-metric people reading this).
9) While LWRs use lowly enriched uranium for fuel, a material that is unsuitable for building nuclear weapons from, the kind of fuel that Monju breeds and that eventually will provide its own fuel, is *identical* to weapons grade plutonium, the material nuclear warheads are made from.
10) The latest LWR planned in the UK (Sizewell C) will be significantly cheaper to build than Monju but will provide almost five times more power. Even then it is expected to produce electricity at twice the cost per KWh as a modern gas fired power station, making Monju's electricity over ten times as costly as conventional energy. The people in charge of FBRs in Japan hope that they can cut costs by one third on the next FBR (I wonder what they will leave out this time...). So maybe it would only be seven times more expensive then.
11) To make Japan independent of uranium imports from foreign countries (mostly from the USA, Canada, Australia and Zaire), Japan would have to build over 100 reactors the size of Monju just to replace the LWRs that currently depend on uranium fuel. Even if Japan could find the money to build all these FBR, that huge investment would only save the country what amounts to about 15% of its total energy imports. Their fuel rods on the other hand would contain several times more weapons grade plutonium than all weapons held by the US and the successor states of the USSR taken together.
Kowai desshou! [Scary, isn't it]
Joe Wein
Sakado-shi
Japan
DAILY NEWS 11 December 1995
The director of Japan's Science and Technology Agency said Monday that a two-to-three ton coolant leak at a reactor will not halt the country's controversial nuclear power program.
"We are not considering giving up the full operation of the plant at this moment," said Science and Technology Agency Director Yasuoki Urano.
An emergency shutdown on Friday halted the Monju fast-breeder reactor in central Japan after leaking sodium coolant created clouds of opaque fumes.
An alarm sounded around 7:30 p.m. Friday, switching the system over to manual operations. After fumes from the leaking coolant pipe were spotted a full operational shutdown was ordered around 9:00 p.m., the Power Reactor and Nuclear Fuel Development Corp. said Monday.
The leak in the plant's secondary cooling system did not release radioactive sodium, but some two to three tons of solidified sodium were found when investigators battling the thick fumes finally located the source of the spill, authorities said. The $6 billion Monju reactor has been the subject of controversy since construction started on it 10 years ago.
Subj: 12-11 news
To: Bob Eichelberger, 75500,3413
Date: Tuesday, 12. December 1995 15:28:16
From: Joe Wein, 100142,3715
>>Thanks to you and to Mike Ross for telling me where to read Joe Wein's comments on Monju. They are posted in the right section, "Politics and Opinion," as there are significant errors of fact in his report, both regarding the design of Monju and the properties of sodium. An easy example: Liquid sodium does not expand when it solidifies. Water and gallium are the only two substances I know of which do--other materials have smaller volume when solid than when liquid, at the melting point. [1] <<
I wouldn't mind if you would address your comments about my message to me *directly*, rather than just making negative comments to others, in a different section. I welcome someone in this forum who could contribute to a healthy debate on this important issue.
The reference for solidifying sodium bursting pipes was a quote by Keiji Kobayashi, a nuclear engineering expert at Kyoto university's Nuclear Research Institute, published in an article in "Japan Times" of April 6, 1994 (the day Monju went critical for the first time). He claimed that the coolant pipes carrying the sodium were surrounded by heating coils because solidifying sodium could rupture the pipes. Is it not correct that the pipes are being heated? Is there a reason for that other than the one given by Mr. Kobayashi?
>>A large sodium leak, such as this one apparently was, is a serious event, but recovery from it should not be devastating to the Japanese sodium reactor development program. It will provide important data for the commercial plants which will be built in the next century.<<
Commercial Fast Breeder Reactors will never be built, not in Japan and certainly not anywhere else. FBRs were killed in the US by President Carter because of the danger of nuclear proliferation (no FBRs without weapons grade plutonium production) as well as by falling nuclear fuel prices. The German / Dutch / Belgian joint venture FBR in Kalkar Germany was completed at a cost of $5 billion and then mothballed forever without once being started up, because there was no future for it. The French Superphenix, the largest FBR in any western country is no longer used for breeding. The British FBR in Dounreay / Scotland will shut down because the government cut its funding. The Russian FBR is a nightmare amongst nightmares.
The current Japanese plan is for commercial Fast Breeders plants not to be built before the year 2030 (by when both you and I will have long retired), which is a significant postponement compared to earlier plans. Fast Breeder Reactors are way too expensive to compete with LWRs until uranium resources are virtually depleted which, given the abundance of low grade ores and the stagnating demand for new nuclear power stations in much of the western world, is unlikely to happen anytime soon. Even in Japan FBRs will eventually be killed by economic forces, assuming they survive ecological criticism and the political risks of a large-scale plutonium economy and the associated danger of nuclear non-proliferation. Monju is a huge waste of taxpayers money as well as a ticking time bomb.
I'd welcome your response as a reply to my "Monju problems" message in the "Politics and Opinion" section.
Talk to you later!
Joe
Subj: Monju problems
To: Timothy Morgan, 100535,475
Date: Wednesday, 13. December 1995 22:08:21
From: Joe Wein, 100142,3715
>>I must admit to being a little frightened by the idea of a nuclear reactor built on the cheap!<<
"On the cheap" sounds like an odd expression for something that cost $6,000,000,000 to build: Kim Jong Il will get two South Korean reactors (based on American designs [Westinghouse]) with a total power output of 2000 MW for _one third_ of the price tag of Monju with its 280 MW of electrical power output. But nevertheless, they must have cut corners somewhere. I was somewhat irritated by the fact that Monju cost only 20% more to build than the German FBR in Kalkar which is of virtually identical size (300 MW), with Japan being an earthquake country, local land prices, transport costs, etc. being as they are over here. I somehow expected something twice the price of the German design so there has to be a catch somewhere...
Today we were told that the amount of sodium escaped in the recent Monju accident was about two to three times greater than the previously largest coolant leak in the history of Fast Breeder reactors (that one actually happened in a Russian reactor twice the size of Monju).
The clearup operation will have to proceed very carefully. Underneath the burnt up sodium ashes (sodium peroxide, sodium hydroxide) in the piles under the leakage site there still could be sizeable quantities of metallic sodium that could catch fire again as soon as the ashes are removed, admitting air to the unburnt metal.
Have you ever seen sodium react with water? It is quite a sight, especially if it doesn't happen in a nuclear power station down the road
Well, the fun part is that in the heat exchanger between the secondary cooling cycle and the water/steam cycle that drives the power turbines about 200 tons of liquid sodium at temperatures of up to 500 C (about 900 F) will be separated from comparable amounts of water only by a few millimetres of stainless steel. Any leak in the welding joints or caused by chemical corrosion will cause a violent reaction involving large quantities of these two substances. There is quite a history of sodium leakage problems in heat exchangers of Fast Breeders and they usually involve explosions.
Because unlike in the primary coolant cycle the sodium in the secondary coolant cycle is not radioactive that is not a disaster by itself, even though it could put the reactor out of action for several months. The problem is more that a problem in the secondary coolant cycle makes it necessary to quickly shut down the reactor to prevent the primary cycle from overheating. Unfortunately nuclear reactors are not designed to change output very rapidly, they prefer being stepped up and down very sloooowly (one reason why they can never supply daytime peak power). Any rapid change in temperature puts thermal strain on the pipes and makes them more brittle, expanding cracks and straining welding joints... I suspect that this is one of the reasons why the operating crew waited 93 minutes before they manually shut down the reactor.
The primary reason the pipes in Monju are so thin, by the way, is the fact that water boils at much lower temperature than sodium, making it necessary to use thicker steel in Light Water Reactors (LWRs) that is strong enough to withstand steam pressures of 200 bar (about 2800 psi). Sodium cooled reactor such as Monju don't have that problem since sodium would not boil at *normal* operating temperatures, but the downside is that if cracks do develop in the steel they don't have to travel very far before a crack turns into a leak. In LWRs X-ray and ultrasonic scans are used regularly to check for cracks. Naturally it is easier to spot a dangerous crack that has to grow to a full 78 mm (3 in) before the pipe starts leaking than one that will turn into a leak after only 10 mm (0.4 in)!
The strength of the pipes also becomes an issue during earthquakes. Thinner walls will make it much more likely that pipes would shear off in a massive jolt caused by a quake, instantly leaving the reactor without cooling (remember, unlike with an LWR there is no Emergency Core Cooling System in Monju!). Furthermore, the fact that sodium coolant pipes have to be surrounded by heating coils to ensure that sodium never solidifies in the pipes in the event of a shutdown makes it much more difficult to regularly check the pipes for cracks than it is in LWRs. Even there the track record on spotting cracks has not been very good and we've had a scandal in Germany where the company doctored or substituted x-ray photos to cover up cracks that would have made it necessary to shut down the reactor for major repairs.
Joe
Subj: Monju problems
>>if I understood your message right, you are saying that a FBR
- costs more than a LWR (and even more than a gas plant) - produces less energy, - is inherently less safe - is potentially more dangerous - produces weapon grade plutonium (which no Japanese seems to believe me!)
All I can say is "Hmmm...where's the merit?"<<
In the 1970s when electricity demand was growing at 7% a year in most western countries and nuclear energy was still seen as they way of the future there was significant concern about a shortage of nuclear fuels as uranium prices started to rise, which prompted first research into Fast Breeder technology.
Regular nuclear power stations (Light Water Reactors = LWRs) burn U235, an isotope which constitues only about 0.7% of all natural uranium. Fast Breeder Reactors (FBRs) on the other hand burn almost pure U235 or plutonium, or a mixture thereof. They use spare neutrons from the chain reaction to turn otherwise unusable U238 (the other 99.3% of natural uranium) into fresh plutonium. This is called "breeding". The interesting point is that an FBR can breed slightly more fresh plutonium from U238 than it burns as fuel. It should therefore eventually be self-sufficient, producing all its own fuel from otherwise unusable and plentiful U238, with even some spare fuel being produced for fuelling LWRs. If you could use all uranium (U235 and U238) instead of just U235 this should increase usable nuclear fuel stocks by a factor of 140. However, because there is a certain amount of breeding in even LWRs the nuclear industry claims only a factor of 60 by which an FBR stretches fuel reserves over LWRs.
There are a number of problem with this whole scenario. The first is economy:
A fast breeder only makes commercial sense if it can produce plutonium cheaper than it costs to mine and enrich uranium. The shortage of uranium on the world market proved temporary and quickly evaporated at the end of the 1970s, as new mines flooded the market while orders for new power stations were cancelled after the Three Miles Island (Harrisburg) accident and later the Chernobyl disaster (it has been pointed out that the uranium supply contracts with France that Australia threatened to cancel after the recent tests were actually signed at a price that was still twice the current world market price, meaning France would save millions from such a move!). Demand for electricity levelled off as economies were growing more slowly and energy use became more efficient. Prices for new power stations spiralled as safety and environmental rules were tightended.
By the time this took place, several countries had already gone ahead and had started building prototype Fast Breeders. When they were completed, the end product of these plants, plutonium, was way too expensive to compete with by then cheap and abundant uranium from mines. Even if uranium prices were to rise again (highly unlikely given the stagnating nuclear industry), that would only help to make large quantities of lower grade uranium ores economically viable, long before plutonium from FBRs were to become economical. Uranium is actually more common than silver or mercury, has been mined commercially on all five continents and there are large quantities of it in low concentration in sea water (according to one study, it would be cheaper to extract it from there than to reprocess plutonium from nuclear waste).
A second problem is to do with the size of FBRs. While LWRs have been scaled up to huge generating capacities (up to 1400 MW per plant), primarily to lower cost per KW of installed capacity, there is a technical limit to this with FBRs: beyond a certain point, the larger you make an FBR the lower its breeding rate becomes. This is because the U238 (blanket uranium) surrounds the fuel core and a larger fuel core (to allow for more power) means a smaller surface area per core volume, hence fewer neutrons that can be absorbed by U238 to turn it into fresh plutonium. The result of that is that if you want to produce more plutonium than you burn you have to keep the reactor below a certain size, making it significantly less powerful than a regular LWR. Because of it's more challenging technology (metallic fuel, sodium cooolant, fast neutrons, etc.) it is impossible to build an FBR for less than the cost of an LWR, even at much reduced power output. No wonder all countries except Japan have abandoned all Fast Breeder plans.
Which brings us back to your question: Why hasn't Japan abandoned Monju yet?
I think, essentially we are dealing with a political problem here. Monju is run by the Power Research and Nuclear Fuel Development Corp., which is affiliated with the Science and Technology Ministry. It is paid for by taxpayers money. It is not run by a commercial power company. It does not have to make a profit. Its main purpose is to develop a technology that certain Japanese politicians in power twenty years ago thought essential then, even though its economic outlook has changed completely. Nobody has told the bureaucrats in charge to stop this giant exercise at throwing good money after bad money. It is in the nature of bureaucrats to follow established plans and not to question them. And you don't get promoted as a bureaucrat by questioning the decisions made by the people who hired you.
It is therefore not for any great technical achievements or scientific breakthroughs made in Japan in recent years that Monju went into operation. The truth is, it is only because there currently are no politicians powerful enough to to fight the bureaucratic momentum behind outdated decisions and to turn the ship around that Japan is still pursuing its insane policy on Fast Breeder Reactors. I just hope that will change before the first major accident occurs.
Joe
References:
[1] Bob Eichelberger was right that liquid sodium is less dense (927 g/l) than solid sodium (951 g/l), so solidifying sodium would not expand and would not rupture the pipes, unlike stated by Keiji Kobayashi.
To: Christian Vitroler, 100042,662
Date: Tuesday, 12. December 1995 03:22:09
From: Joe Wein, 100142,3715