The budget for the FBTR was approved by DAE as early as September 1971 and it was anticipated that the FBTR would be commissioned by 1976.4 The reactor finally attained criticality only in October 1985 and the steam generator began operating in 1993.5

Much of the first one and a half decades of the FBTR’s operations were marred by several accidents of varying intensity. Two of these are described below in some detail to illustrate the complexities of dealing with even relatively minor accidents and the associated delays, as well as the hazards posed to workers. When viewed in combination with similar experiences elsewhere, these circumstances suggest that it is unlikely that sodium-cooled breeder reactors will ever perform with the reliability that water-cooled reactors have demonstrated over the past two decades.

In May 1987 there was a major incident that took two years to rectify.6 This occurred as a fuel subassembly was being transferred from the core to the periphery.7 The problem began with the failure of a protective circuit involved in the rotation of the plug to move the selected fuel assemblies. For some reason, this protective circuit was bypassed and the plugs were rotated with a foot long section of one fuel subassembly protruding into the reactor core. This resulted in the bending of that specific subassembly as well as the heads of 28 reflector subassemblies on the path of its rotation. Various maneuvers to rectify the situation did not help and only resulted in one reflector subassembly at the periphery getting ejected as well as the bending of a sturdy guide tube by 32 cm. The last event has been described as the result of "a complex mechanical interaction" which seems to suggest that how it happened was never really understood.

Extensive repairs were required before the reactor could be restarted. First, the guide tube had to be cut into two parts using a specially designed remote cutting machine while ensuring that none of the chips produced during the cutting process fell into the core.8 Then the damaged reflector subassemblies had to be identified using a periscope. Finally part of the sodium had to be drained out and the damaged subassemblies removed using specially designed grippers. As might be expected, all of this took time and reactor operations commenced only in May 1989.9

The second accident described here is one that is common in fast breeder reactors — a sodium leak. That this occurred seventeen years after the reactor was commissioned underscores the generic nature of such accidents. The leak occurred in September 2002 inside the purification cabin, which houses the pipelines of the primary sodium purification circuit.10 The cause of the leak is said to have been "the defective manufacturing process adopted in the manufacture of the bellows sealed sodium service valves". By the time the leak could be confirmed and controlled, approximately 75 kilograms of sodium had spilled over and solidified on the cabin floor and various components in that cabin.

Removing this radioactive sodium was a major effort. To begin with, even to approach the cabin, the workers had to wait ten days to allow for a reduction in the radioactivity from the sodium, some of which had absorbed a neutron to become Na-24, a gamma emitter (15-hour half-life). Even then, in areas near the spilled sodium, the dose rate was as high as 900 millisieverts per hour (mSv/h).11 Another problem resulted from the whole cabin normally being surrounded by a layer of nitrogen so as to avoid sodium burning. At first, IGCAR tried to simply replace the nitrogen with regular air so that cleanup workers could breathe. But this led to sparks and fires involving the spilled sodium. These had to be put out with dry chemical powders — but then this led to lots of dust being suspended in the atmosphere and made visibility poor. Once again nitrogen had to be reintroduced. Workers were then sent in with masks that had tubes feeding them with breathing air. Much of the work had to be done remotely, which, while lowering radiation exposure, made it a very slow operation. In all, removing the 75 kg of sodium and bringing the cabin back to normal conditions took approximately three months.12

The FBTR has also seen several other accidents and unusual occurrences, such as unexplained reactivity transients.13 Overall, the reactor’s performance has been mediocre: it took fifteen years before the FBTR even managed fifty plus days of continuous operation at full power.14 In the first twenty years of its life, the reactor has operated for only 36,000 hours, i. e. an availability factor of approximately 20 percent.15 Despite this checkered history, IGCAR claims to have "successfully demonstrated the design, construction and operation" of a fast breeder reactor.16

The Prototype Fast Breeder Reactor

Even before the FBTR came on line, the DAE started making plans for a larger Prototype Fast Breeder Reactor (PFBR). In 1983, the DAE requested budgetary support from the Government.17 The first expenditures on the PFBR started in 1987-88.18 In 1990, it was reported that the Government had "recently approved the reactor’s preliminary design and has awarded construction permits" and that the reactor would be on line by 2000.19 Construction of the reactor finally began in October 2004 and was projected to be commissioned in 20 1 0.20 The PFBR will likely suffer from the two problems that have plagued breeder reactors elsewhere: the risk of a catastrophic accident and poor economics. It will also be a source of weapon-grade plutonium that might be used for the strategic program. See the discussion in chapter 2 of France’s use of its first demonstration breeder reactor Phenix to produce weapon-grade plutonium for France’s weapons program.