An incident at Kalpakkam

India's National Magazine from the publishers of THE HINDU

A heavy water leak in the second unit of the Madras Atomic Power Station (MAPS) on March 26 causes widespread concern.

HOW much heavy water did leak into the reactor vault of the second unit of the Madras Atomic Power Station (MAPS) at Kalpakkam, about 60 km from Chennai, on March 26? Consequently, how much radioactive tritium was released into the reactor building? What was the level of radioactivity to which the people who stopped the leak and mopped up the heavy water that had spread on the floor were exposed? Opinion is divided on these questions, which are at the heart of a new debate on the safety of Indian nuclear reactors. However, there is no uncertainty about why the leak occurred: a sealing plug of a coolant channel of the second unit was not positioned properly after the completion of maintenance work which was undertaken during the reactor's biennial shutdown.

Top MAPS officials maintained that the leak was an insignificant and anticipated incident which occurred during a planned operation and that the level of tritium released was within permissible limits. Knowledgeable sources, however, asked why then a plant emergency was declared. These sources said that the declaration of a plant emergency was just a step short of ordering the evacuation of plant personnel. They challenged officials of the Nuclear Power Corporation (NPC) to release the figures regarding the heavy water leak and tritium release. (All nuclear power stations in the country come under the umbrella of the NPC.)

Officials at MAPS countered these arguments and said that a plant emergency was a part of routine procedure to insulate the plant and bring safety mechanisms into play, besides regulating the entry of persons into the plant in order to prevent the spread of any contamination. A plant emergency, they said, did not mean that the reactor was going to "explode like a nuclear bomb", as was made out in a section of the press.

MAPS has two units of Pressurised Heavy Water Reactors (PHWRs), each with a capacity of 170 MW. They use ordinary uranium as fuel and heavy water as both coolant and moderator. According to K. Hariharan, Station Director, MAPS, both units had performed very well in 1998-99, with an average capacity factor of 75 per cent. The second unit was shut down from February 14 to carry out maintenance work. It was scheduled to be restarted this summer when the demand for electricity peaks, he said.

Maintenance work, including the overhauling of the turbine generator, was almost complete and the MAPS authorities planned to synchronise the unit with the Tamil Nadu grid by May 1. The inspection of coolant channels to assess their health is the main component of the maintenance work. Remote-inspection through ultrasonic methods reveals the channels' metal integrity.

In a reactor, the coolant tubes (also called pressure tubes) are situated inside the calandria tubes (calandria is the reactor vessel). Both coolant tubes and calandria tubes carry heavy water, and between them are garter springs. Coolant tubes with end-fittings on either side are called coolant channels. Seal plugs (or sealants) form part of the end-fittings. The 306 coolant channels in the calandria tubes are arranged within a circular space and together they resemble a lattice. The flow of heavy water in the coolant tubes results in vibrations, as a result of which the garter springs tend to move. This in turn causes the coolant tubes to sag and come in contact with the calandria tubes. Since there is a temperature difference - it is 295 C inside the coolant tubes and about 70 C in the calandria tubes - the temperature of the coolant tubes comes down at the point of contact. This results in the formation of zirconium hydride, which leads to the coolant tubes, made of zircaloy, becoming hard and brittle. This in turn causes micro-cracks in the coolant tubes. (Such cracks have been noticed in the reactors at the Rajasthan Atomic Power Station, or RAPS, at Rawatbhatta and at MAPS.) If the micro-cracks cut across the thickness of the tubes, they can lead to loss of coolant from the primary system and result in inadequate cooling of fuel in the core. Such a "loss of coolant accident" (LOCA) may lead to a fire in the reactor.

The Bhabha Atomic Research Centre (BARC) in Mumbai has developed a technique to carry out a vibration probe near the face of the end-fitting of the coolant tube to help determine the extent to which the coolant channels will sag and the possibility of their coming in contact with the calandria tubes. These channels are defuelled and subjected to a BARC Channel Inspection System (BARCCIS) to determine the position of the garter springs. All the 306 coolant channels in the second unit of RAPS were replaced and the unit attained re-criticality on May 27, 1998. The new coolant channels are made of zirconium niobium instead of zircaloy.

BARCCIS was under way at the second unit of MAPS as part of maintenance work. The coolant channels are ultrasonically tested to assess their condition. These checks are conducted through remote-handled semi-automatic tools. The data collected help assess the life of the tubes and the unit's potential for continued operation. During such inspections, when new seal plugs are positioned and operated, a certain amount of information is available to the control systems.

According to Hariharan, the inspection of the coolant channels of the second unit revealed that they were healthy and that there was "no deterioration", and that they would continue to perform for years to come. A seal plug operation was also undertaken during the inspection. Whenever a seal plug is not positioned properly, indications to that effect are available and the position is corrected. When a seal plug operation was undertaken in one of the coolant channels at the second unit of MAPS, the control systems received information that the plug was not seated properly. It was discovered on March 26 that heavy water had leaked out into the reactor building because the seal plug was not positioned properly.

Hariharan said: "We undertook the operation of re-seating the plug in the form of replacement which involves planned escape of heavy water from the channel inside the fuel machine vault." According to him, such situations are not unusual and there are adequate built-in provisions in the unit to tackle them. These included provisions to collect heavy water and heavy water vapour through recovery systems and make an adequate inventory of heavy water available. He said: "The March 26 incident thus did not involve an unusual situation, especially since the unit was under shutdown and the whole activity was planned." During such operations, it is customary for nuclear stations to ensure that the environment is not affected. Hariharan said that there was "no such release into the environment" around MAPS.

On the quantity of heavy water that had leaked, Hariharan said that it was of "an insignificant quantity". On the amount of radioactive tritium released, he said: "The release was well within AERB limits." (The AERB, or Atomic Energy Regulatory Board, is the watchdog for safety at the country's nuclear facilities.)Hariharan said that a plant emergency was declared only to regulate the movement of personnel inside the plant and to prevent non-essential persons from entering it. He said that normal operations had continued at the plant during this period. B.K. Bhasin, Director (Operations), NPC, who visited MAPS on April 1, said that the situation had been handled "reasonably well". "Systematic work was done because you just cannot take any chances," he said. "I sent three experts immediately from the NPC headquarters."

HOWEVER, Dr. A. Gopalakrishnan, former Chairman of the AERB who has been closely monitoring the safety status of the Department of Atomic Energy's (DAE) facilities and who has found them to be "much below international levels", said that plant emergency "ranks only a little below evacuation". Responding to the stand taken by MAPS officials that it was an insignificant leak, he asked: "Why was a plant emergency declared (during this period the reactor was shut down)? If the leak was only like that from a tap, why declare a plant emergency?" He further said: "I feel that a plant emergency was declared because a lot of radioactivity was released into the plant. It could have already spread, depending on how long it took to plug the leak." According to him, if the escape of heavy water was confined to, say, only one room with 10 workers in it, a plant emergency would not have been declared. Gopalakrishnan said that the magnitude of the leak and the rate of discharge may have been so large that it must have become evident to persons concerned that a plant emergency would have to be declared within two or three minutes.

Gopalakrishnan said that going by the initial estimates given by MAPS, several tonnes, possibly 14 tonnes, of heavy water leaked out. "For this number the declaration of a plant emergency makes sense," he said. "You will not declare a plant emergency if just one bucketful of heavy water leaks or about 10 workers are concerned. A very large amount must have gone into the drain and spread out. The (tritium) release must have also been sucked into the ventilation system."

Gopalakrishnan explained that there were two separate systems - primary and secondary - that carried heavy water in the reactor. Both are radioactive. If the pressure tube, that is, the coolant tube, had ruptured or its capping had come off, then it was the primary coolant heavy water that leaked out. On the other hand, if a calandria tube, which is larger than a coolant tube, had ruptured or its cover had come off, it was moderator heavy water that spilled. Gopalakrishnan said: "I think it was the former (coolant heavy water) that escaped."

He pointed out that the heavy water that ran through the primary system was more radioactive because it ran close to the fuel in order to keep the reactor cool. This heavy water is in circulation from the time a reactor attains criticality. In the process, tritium forms in the heavy water. Tritium is highly radioactive, with a half-life of over 12 years. Tritium content in heavy water increases with the number of years the reactor operates. Almost all the tritium continues to be in the heavy water even if the reactor is shut down for a couple of months. When there is a spillage of heavy water, the tritium affects the plant personnel.

The former AERB Chairman said that if the heavy water had leaked into the measured tank, the quantity of heavy water that had leaked could have been estimated (from the increased level of the tank's contents). He said that heavy water leak had occurred in other countries too. "But there was openness there," he said. "The workers and the environment were taken care of." He demanded that personnel in MAPS be given the data on the release of radiation and be told how much radiation they had ingested. He said that he was deeply concerned about the use of contract labour for cleaning jobs because they were often overexposed to radiation.

TOP NPC officials repeatedly stonewalled queries about the leak. They declined to give any figures and accused the press of overplaying the incident. They said that more than 120 tonnes of heavy water had leaked out of reactors in Canada and the United Kingdom, compared to which the leak at Kalpakkam was "nothing".

According to Hariharan, there were 65 tonnes of heavy water in the second unit of MAPS and "it was only an insignificant quantity" that had escaped, and "that too in a planned operation". He maintained that it was only the coolant heavy water that had leaked.

T.S. Rajendran, Chief Superintendent of MAPS, said that every nuclear power station had a safety committee headed by the Station Director and included the Chief Superintendent, the Operations Super-intendent, the Maintenance Superinten-dent and others. They discussed every incident and came up with suggestions. A Root Cause Analysis Committee is constituted if the need arises.

Meanwhile, in response to a short questionnaire sent by Frontline, Dr. K.S. Parthasarathy, Secretary, AERB, sent the following statement: "Nearly four tonnes of heavy water leaked from the second unit of MAPS. However, most of it has been mopped up and will be upgraded and reused. For the MAPS units, the amount of tritium permitted to be released through the gaseous route is 300 Ci per day per reactor or 1,10,000 Ci per year per reactor."

He said: "A discharge limit ten times the daily limit is permitted, but the annual discharge shall not exceed those specified above. The releases of tritium during and after the incident were within the above limits. A large group of 42 persons was mobilised to stop the heavy water leak. The collective dose to these people was 0.25 person Sievert. Thus the safety of the reactor was ensured. The dose to the public from the incident is very much below the limit prescribed by the AERB and the dose to the station staff was not unacceptably high."