In an era of natural gas combined cycle stations, distributed generation, and privatization it may come as a surprise to some to find there are still a few nuclear reactors being built around the world – around 40 if you include half a dozen projects in Russia and the Ukraine where funding is uncertain. But well over half of the current nuclear construction projects are in Asia. By the end of 1999 construction will have started on a further six units worldwide, and these will all be in Asia: two in Japan, one in China, two in India and one in Taiwan. When it comes to planned nuclear units there are several in Korea, Japan, India and Taiwan but essentially nothing outside Asia.
In Japan proposals exist for building about 20 more reactors before 2010. Against this background the primary coolant leak incident at Japan Atomic Power Company’s 1200 MWe Tsuruga 2 pressurized water reactor on 12 July must raise further concerns for hungry nuclear vendors looking for resumption of nuclear power expansion in the Asia Pacific countries, now that the worst of the economic downturn appears to be behind them.
Admittedly the Tsuruga 2 incident was only given a preliminary rating of Level 1 on the International Atomic Energy Agency’s nuclear event scale, which denotes merely an “anomaly beyond the authorized regime but with significant defence in depth remaining.” However this rating may be subsequently revised upwards when root cause analysis and evaluation is completed because of the time taken to locate and stop the leak – about 14 hours.
It is worth recalling that the leak in 1995 at Japan’s Monju reactor, which involved the loss of 700 kg of highly reactive sodium rather than 51 t of water, as at Tsuruga 2, was also given a Level 1 rating. At Monju the damage was localized, the plant was safely shut down and there was no leakage of radioactive material. But the Monju incident and the way it was handled – which included a totally misguided and ill-fated attempt at a cover-up and the suicide of one of the managers investigating the event – had serious and far-reaching negative repercussions throughout the Japanese nuclear industry, leading to delays and cancellations of planned plants. The Monju fast breeder reactor itself remains shut down with little prospect of restart for several years. The problem was traced to high cycle fatigue in a thermocouple thermowell.
At Tsuruga fatigue has also been mentioned as a possible factor in the 8 cm long crack which caused the leak, in an elbow of the outlet piping of one of three regenerative heat exchangers. These form part of what is known as the Chemical and Volume Control System (CVCS), which is located inside the reactor containment and regulates the boron content and the volume of the coolant. The leak was detected at 6.05 am on 12 July when fire alarms near coolant pipework in the containment sounded while the plant was running at full power. At the same time other alarms began indicating high levels of sump water in the containment and showing increased radioactivity within containment. There were also indications of increased flow for coolant replenishment. At 06.24 am the operators recognised that they had a leak of primary coolant water and started shutdown procedures. Shutdown was completed at 06.48 am.
Some 12 hours later a team of engineers was able to enter the reactor building, during which time make-up coolant water was continuously injected to keep the core cool. The exact position of the leak was found at 7.20 pm and the flow was finally stopped at 8.29 pm – by which time some 51 m3 had been lost (compared with a total inventory of about 345 m3). In the first few hours of the leak, radioactivity levels in containment had reached about 23 times normal levels.
One immediate question is why the precise location of the leak took so long to track down. At the very least the incident is likely to result in some re-evaluation of leak detection and location capabilities at nuclear plants and possibly requirements for backfitting of improved systems.
Nevertheless, no changes in radiation levels were picked up by environmental monitoring systems around the plant or in the plant’s off-gas stack. Although the safety significance of the July event in terms of risk to the public was therefore minimal it is, to say the least, embarrassing in public relations terms for an industry that must be as close to error-free as is humanly possible and for a plant considered to be one of Japan’s best, and among the most advanced in the world.
One problem the Japanese nuclear industry faces is that it has essentially set itself the goal of being incident-free, which is of course impossible. The July Tsuruga 2 event is the third significant PWR primary coolant leak incident experienced so far in Japan. In 1979 there was a 79t leak at Takahama 2 and in 1991 there was a 55t leak due to a steam generator rupture at Mihama 2. Indeed Tsuruga 2 itself has had a previous primary leak, albeit relatively small. This was in 1996 and the root cause was surface impurities introduced during manufacture. The position of the latest leak, in a pipe elbow rather than a weld also raises questions about the production history of the pipe involved, which will need to be investigated.
In a complex industrial undertaking like nuclear power it is unrealistic to expect no mistakes. The best you can hope for is to contain the effects of any mistakes that do occur so that there is negligible impact on the public at large. The problem is that not even that is good enough in the hostile political climate that nuclear power finds itself in. Already there are calls for cancellation of two new PWRs that were planned for the Tsuruga site.
