No. 3 — January 26, 1996


Weekly Review

--- by Jon Choy

Japan's ambitious program to develop a self-sufficient nuclear energy industry suffered a significant blow in early December when a major leak occurred in the secondary cooling system of a prototype fast-breeder reactor and subsequent investigations revealed a series of errors by officials of the government-run Power Reactor and Nuclear Fuel Development Corp. both at the reactor site and at PNC headquarters. The gravity of the situation was hammered home in mid-January when one of the PNC officials investigating the chain of events committed suicide. Besides providing fresh ammunition to critics of the fast-breeder program, the entire affair has eroded the public's acceptance of plutonium-use policies and trust in nuclear agency officials. While the investigation continues and central government officials pledge to revamp procedures to prevent a recurrence of the leak and subsequent blunders, the leak's ultimate impact could be far-ranging.

Alleviating Japan's near-total dependence on imported energy supplies has been a long-sought goal of the Japanese government. Bureaucrats made nuclear power an important part of their efforts to reach this goal based on expectations that two key technologies would be developed: indigenous recycling of spent nuclear fuel and fast-breeder reactors. Career government officials envision commercial fast-breeder reactors generating a growing share of the country's electricity supply. Since such reactors produce more plutonium fuel than they consume, this "bred" plutonium is recovered through spent fuel recycling and used to produce more electricity and plutonium, thus closing the nuclear fuel cycle. (This is in contrast to the "once-through" fuel system where uranium or plutonium fuel is burned and then disposed of.) To realize this vision Tokyo constructed two experimental breeder reactors (Joyo and Monju) and is building a full-scale recycling facility.

The leak occurred December 8 in a stainless steel pipe of the Monju plant's secondary cooling system, allowing as much as three tons of liquid sodium to spew over the secondary cooling system and the floor of the compartment. The experimental reactor, which had only begun producing electricity in August 1995 on a test basis, was shut down and about 80 tons of sodium drained from the leaking cooling system. Although no radiation was released and no injuries were incurred, the incident could have ended very differently. The liquid sodium coolant, critics had warned, was itself dangerous because it is highly corrosive and explodes on contact with water. If the spilled sodium had corroded the stainless steel floor lining and seeped into the underlying cement foundation, there may have been enough moisture in the cement to cause an explosion; that, in turn, could have compromised the entire cooling system and led to a reactor core failure. As learned from the 1986 Russian experience at Chernobyl, the release of radioactive gas and particles into the air can cause widespread health problems.

The leak's severity, however, was increased by the poor emergency reaction of the on-site PNC staff. Despite plenty of sensors in Monju's fire alarm system, the plant's control room has only one display. Operators were aware that a fire sensor had been activated but could not determine where. Moreover, operators did not begin to shut down the reactor until several hours after the sodium leak had been identified. On-site PNC officials also failed to follow specified procedures to notify local officials immediately of any reactor incidents. Such notification procedures had been a condition of allowing the experimental reactor to be located at the Tsuruga, Fukui prefecture site.

The on-site PNC staff filmed the site of the leak on several videotapes both during and after the incident. PNC officials presented to prefectural officials and the media two videos — one running four minutes and the other a minute — in which only sodium-covered floors were shown. On December 13, however, PNC officials admitted that these videos had been edited in a deliberate attempt by some staff members to hide the extent of the leak. In addition, PNC representatives admitted that they had a third video, which had been taken immediately after the accident; they released it as well in mid-December.

The unexpurgated videos revealed to the public the true extent of the leak and the deception of the PNC staff. Officials of the Science and Technology Agency, PNC's oversight body, and prefectural authorities announced their own investigations of the leak and PNC's subsequent actions. In an effort to placate critics PNC reassigned the three managers who had admitted to the video cover-up. Subsequently, PNC workers were stunned — as was the entire nation — when the person in charge of investigating the incident committed suicide January 13. Top STA officials have promised to uncover all the details and are considering whether to appoint an independent investigator.

Investigators attributed as causes of the leak metal fatigue of the sensor housing, questionable pipe welds and unexpected vibrations resulting when very hot liquid sodium flowed turbulently around a temperature sensor that was installed in the cooling system pipe. PNC researchers are using a supercomputer to model the flow of the liquid sodium around the sensor, while other investigators are checking the integrity of the pipe welds. If the cause is an engineering defect, Monju will have to remain closed until the design problem can be corrected.

Beyond the immediate shutdown of Monju and the inevitable delay in its testing and operations the leak is having wider effects. The government has decided to delay beginning construction of a second prototype breeder reactor and a second spent-fuel reprocessing plant. In addition, the completion schedule for the facility already under construction likely will be stretched; it already has encountered cost overruns. Because Tokyo has pledged to match carefully the supply and the consumption of plutonium, Monju's shutdown will unbalance carefully laid plans for plutonium storage, reprocessing and shipment. Government emergency management officials have announced that they will add contingency planning for a nuclear power plant accident to their agenda. This seemingly is a belated action in a country that already has nearly 50 nuclear power plants in operation. Japanese companies that make and install nuclear power reactors have added the leak to their list of potential deal-wreckers; among these are shrinking domestic demand for new reactors and continued informal government restrictions on participation in foreign nuclear plant projects. The three top firms in this field — Mitsubishi Heavy Industries Ltd., Hitachi, Ltd. and Toshiba Corp. — recently have announced that they will shift resources and personnel away from their nuclear power departments.

An even bigger casualty of the leak is public confidence in PNC specifically and in nuclear power in general. Residents near the Monju facility are watching the investigation closely, but a common sentiment expressed to the Japanese press is that they will move if they are not convinced of the plant's safety. Opponents of nuclear power are capitalizing on the accident and subsequent cover-up to whip up public support for their positions. The incident also poses problems for the ruling Liberal Democratic Party/Social Democratic Party of Japan/New Sakigake Party coalition. The LDP traditionally has supported nuclear power development, while the SDPJ has opposed it. Even granting that Monju is an experimental design, advocates of nuclear power are concerned that the reactor experienced a major problem so soon after completion and that the problem may be an inherent design flaw. Government planners are resigned to the inevitable delays that the Monju leak will generate, but they remain convinced of the long-term viability of nuclear power in Japan.

The views expressed in this report are those of the author
and do not necessarily represent those of the Japan Economic Institute

Top aaaa Issue Index aaaa 1996 Archive Indexaaaa Search aaaa Subscriber Area aaaa Home