Monju (CIS:GO JAPAN) / page 1

The following is a collection of clippings from the "Daily News" section of the Compuserve Japan forum as well as from messages from threads discussing issues related to the Monju Fast Breeder Reactor in Japan. These postings were made in 1994 and 1995, from the day Monju was started up to about two weeks after its first major accident.

The current state of Monju is that after almost 12 years of being idle, it was announced in August 2007 that the reactor would be restarted in October 2008, following some minor fixes but no major design changes.

If some of the following postings are too technical in nature, then skip those parts. If you have any questions, feel free to contact me.

Joe Wein

Subj: Monju goes critical
To: Gordon Housworth [ASysP], 72110,1666
Date: Tuesday, 5. April 1994 15:18:13
From: Joe Wein, 100142,3715

So Japan has finally gone ahead today (a beautiful April day in the cherry blossom season) and started up its first Fast Breeder Reactor (FBR), despite international concerns about the increased availability of Plutonium that FBRs will result in, a problem highlighted by the events around North Korea's suggested attempts to build a bomb.

Japan's official line is that reactors of this type will make Japan less dependent on energy imports as it will allow it to breed Plutonium (Pu 239) fuel from otherwise useless Uranium 238 (U238) that makes up 99.3% of all Uranium ore on Earth, thereby stretching usable Uranium supplies 140-fold. The general concern is that Pu is also the stuff that bombs are made from, and Monju will produce it in its purest form. Japan has of course denied any such intentions.

Yet it is interesting to note that Japan is the only country in the world that still runs a fast breeder program. The US has abandoned all plans to develop FBR technology, both because of fears about nuclear proliferation and because in the short and medium term it has been demonstrated not to be ecomonomically viable. The latter is certainly also true for the Monju FBR that reportedly cost about $6 billion to build, yet at 280 MW will produce about a fifth of the electricity of one typical 1300 MW block of a Light Water Reactor (LWR) type of nuclear power station that would actually be cheaper to build.

The same was the case with the German / Dutch / Belgian joint venture in Kalkar (Germany) that cost at least $4 billion to build and was supposed to generate only 300 MW of power. Because of environmental concerns and huge operating costs vs. collapsing fuel prices that reactor was never started up and now lies mothballed indefinitely, as a memorial to the shortsightedness of technocrat planners. The British plant up in Dounreay (Scotland) will soon have to be shut down after state subsidies were cut subsequent to the privatisation of the British electricity industry. The French Superphenix was shut down after a series of serious accidents.

In the late 1960s when electricity production was still doubling every decade and hundreds of nuclear power stations were planned worldwide, FBRs were seen as the only way of avoiding a future Uranium shortage. Since then Uranium exploration has been stepped up all over the world with large new deposits found in many politically stable countries. Any increase in Uranium prices caused by a shortage would now make vast deposits of medium and low grade ore economically viable long before a FBR could start paying for itself. Uranium is actually quite common in small quantities in many granite deposits around the globe. According to estimates from within the nuclear industry, it might even work out cheaper to extract the minute quantities of Uranium contained in sea water than to breed Pu in hundreds of FBRs.

When a Uranium (U235) or Plutonium (Pu239) atom is split by a neutron, 2 - 3 neutrons are released, one of which then splits another U atom. In an FBR most of the other neutrons (on average 1.3) are absorbed by U238 atoms in a blanket surrounding the FBR reactor core, turning them into Pu239, theretically producing more Pu239 in the blanket than is burnt up in the core. This is a rather delicate procedure. If too many neutrons start splitting other atoms, you have the chain reaction of an atom bomb!

In order to maximize this breeding effect, a FBR is more similar in construction to a bomb than to a LWR. LWR fuel only contains 5% U235 in oxide form, not 95% or more pure metallic Pu as the FBR or bomb. In an LWR ordinary water is used to slow down neutrons before they split more U235 atoms while in a FBR (or a bomb) they are not slowed down at all so they are more likely to be absorbed by U238 to produce fresh Pu or (in case of the bomb) to step up the reaction that leads to the blast.

Therefore, FBRs can't be cooled with water and use liquid sodium instead, a metal that doesn't slow down neutrons yet spontanously catches fire when exposed to air above room temperature and explodes when it gets into contact with water. Burning sodium can't be extinguised with anything but inert gas (nitrogen), so fire hoses spraying water are a BAD idea. Small leaks in heat exchangers where sodium has to pass its heat to water for generating steam to drive turbines can be quite dangerous. Because sodium is solid at room temperatures, the FBR has to remain electrically heated to almost the boiling point of water to keep the coolant from solidifying while the FBR is being refuelled, a frequent cause of fires. The French Superphenix had a number of such problems. In addition the sodium coolant does absorb some neutrons and gradually becomes radioactive.

Because fast neutrons are less efficient at splitting atoms the fuel in the FBR core has to be much more tightly packed than in an FBR to achieve a sustained reaction (criticality), which again makes it resemble a bomb more than a normal nuclear power station. That creates thermal problems and puts special strain on the building materials.

If an LWR overheats so the cooling water starts boiling or completely loses its cooling water its neutrons simply become too fast to be good at splitting atoms and the chain reaction comes to a halt (even though the reactor core might still melt at this point). In an FBR if the sodium boils at 883 C or coolant gets lost somehow, the reaction actually steps up. In other words, the only thing that might then prevent a minor nuclear blast are the cadmium control rods that can absorb extra neutrons. Because of the high packing density, inserting them becomes difficult once the fuel rods start deforming due to overheating. Metallic Pu has a much lower melting point than LWR oxide fuel. There have been such incidents in experimental FBRs, with molten Pu fuel collecting at the bottom of the reactor. Because LWR neutrons are produced partly from secondary delay 2-3 seconds after an atom has been split, response times for activating these neutron eating rods are actually almost two orders of magnitude shorter in FBR reactors than in normal LWRs, leaving less of a safety margin for mechanical malfunctioning or human error.

One particular concern is that excess neutrons (for example through blocked cooling ducts) in one part of the reactor could cause a minor blast that would deform the remainder of the core, leading to massive supercriticality in the rest of the FBR core (i.e. nuclear explosion). That secondary explosion might be minor compared to the Hiroshima or Nagasaki bomb but even the energy of several thousand kg of TNT predicted by computer simulations would be sufficient to destroy the containment and could release massive amounts of Pu into the atmosphere around central Japan.

On the other hand, it remains to be seen how well Monju will actually do at achieving a satisfactory breeding rate, i.e. the rate of blanket Pu bred vs. Pu fuel burnt up. Experimental FBRs consistently have problems even achieving 100%, that is they use more Pu than they breed, not much better than LWR which achieve some 20-30%. Since on average only 1.3 spare neutrons are released in the nuclear reaction (some of which inevitably get lost), the amount of fresh Pu bred in commercial fast breeders lies somewhere between 100% of input (i.e. the FBR produces just as much Pu as it uses) and 130%, but will never reach even 200% (i.e. the rate at which one FBR could supply an equally powerful LWR with fuel). Assuming the optimum value, it would take over a dozen copies of Monju ($6,000,000,000 apiece) to supply even one of the 20-odd LWRs in Japan with Pu. As a consequence, if Japan seriously wanted to become independent of Uranium imports, it would have to replace the vast majority of its LWRs with FBRs in the next couple of decades. The capital costs alone would be absurd.

Japan has been highly successful in export markets because it always delivered the quality goods the market required. Yet if the Monju Fast Breeder Reactor was a car, it would be a minicar that cost a million dollars and ran only on nitroglycerine fuel...


Next message

(Return to Monju problems)