As part of the March 2, 2006, agreement between the United States and India on civil nuclear cooperation, India agreed to separate its nuclear facilities into civilian and military installations and to place the former under International Atomic Energy Agency (IAEA) inspection to verify that they are not being used to support the production of nuclear weapons.

On March 7, 2006, Indian Prime Minister Manmohan Singh presented the details of India’s “Separation Plan.” Singh stated that 14 of India’s 22 conventional nuclear power plants, now operating or under construction, would be placed on the civilian, IAEA-inspected list. He then gave particular emphasis to the fact that India would keep its fast breeder reactors, now operating or under construction, off the civilian list. (Breeder reactors are reactors that can create more fissile material than they consume.) On this subject, Singh declared:

We have conveyed that India will not accept safeguards on the Prototype Fast Breeder Reactor [now under construction] and the Fast Breeder Test Reactor [operating since 1985], both located at Kalpakkam. The Fast Breeder Program is at the R&D stage. This technology will take time to mature and reach an advanced stage of development. We do not wish to place any encumbrances on our Fast Breeder program, and this has been fully ensured in the Separation Plan. [1]

Later in his remarks, Singh again turned to India’s breeder program. Referring to declarations made suo motu (on his own initiative), Singh stated:

During my Suo Motu Statements on this subject made on July 29, 2005, and on February 27, 2006, I had given a solemn assurance to this august House and through the Honorable members to the country, that the Separation Plan will not adversely affect our country’s national security. I am in a position to assure the Members that this is indeed the case. I might mention:

i) that the separation plan will not adversely affect our strategic program. There will be no capping of our strategic program, and the separation plan ensures adequacy of fissile material and other inputs to meet the current and future requirements of our strategic program, based on our assessment of the threat scenarios. No constraint has been placed on our right to construct new facilities for strategic purposes. The integrity of our Nuclear Doctrine and our ability to sustain a Minimum Credible Nuclear Deterrent is adequately protected.

ii) The Separation Plan does not come in the way of the integrity of our three stage nuclear program, including the future use of our thorium reserves. The autonomy of our Research and Development activities in the nuclear field will remain unaffected. The Fast Breeder Test Reactor and the Prototype Fast Breeder Reactor remain outside safeguards. We have agreed, however, that future civilian Thermal power reactors and civilian Fast Breeder Reactors would be placed under safeguards, but the determination of what is civilian is solely an Indian decision. [2]

Singh’s comments suggested that India’s national security in the nuclear arena has two dimensions: sustaining a minimum credible deterrent; and implementing India’s “three stage nuclear program,” aimed at exploiting the country’s vast thorium reserves for energy purposes.

In an interview given on February 7, 2006, to the Indian Express roughly a month prior to the signing of the U.S.-India nuclear agreement, Dr. Anil Kakodkar, Chairman of the Atomic Energy Commission (AEC) and Secretary of the Department of Atomic Energy (DAE), made clear that India’s breeder reactor program does, indeed, have close links to the country’s nuclear weapons program:

Express: So categorically the breeder will not go under safeguards?

Kakodkar: No way because it hurts our strategic interest. You follow, no? There’s no way.

Express: The strategic interest of security or strategic interest of energy security?

Kakodkar: Both. It is linked through the fuel cycle.

Express: So will placing the fast breeder reactor program on the civilian list and hence under safeguards hurt India’s efforts at maintaining in perpetuity the “minimum credible deterrent” while hurting its need for long-term energy security?

Kakodkar: Yes, there can be no doubts on that. Both, from the point of view of maintaining long-term energy security and for maintaining the “minimum credible deterrent,” the Fast Breeder Program just cannot be put on the civilian list. This would amount to getting shackled and India certainly cannot compromise one [type of security] for the other. [3]

India’s breeder reactors were reportedly a contentious issue during the negotiations with the United States over the agreement. Ultimately New Delhi prevailed on this matter. [4] The prominent role of India’s breeder reactors in the consideration of the separation of Indian civilian and military nuclear facilities raises the question of what specific contributions these reactors, long justified as important to the future of the Indian nuclear energy sector, might make to its military capabilities.

Background: India’s Three Stage Breeder Program
In a breeder reactor, fuel containing a substantial proportion of fissile material, such as plutonium, is used to sustain a chain reaction. The chain reaction produces heat and excess neutrons. The heat is used to produce steam to turn electric turbines, and the neutrons are used to bombard “fertile” material in a “blanket” surrounding the core. In the blanket, new fissile material is created, traditionally, plutonium. The neutrons also create new fissile material in the core (in the non-fissile part of the fuel blend). Periodically the blanket is removed and a portion of the core is replaced. The blanket and the removed portion of the core are then processed to separate the new fissile material. Because, in total, more fissile material is created in the blanket and core than is consumed in the core, the reactor is said to “breed” fissile material. The bred fissile material is used as a portion of a subsequent breeder core, until enough bred fissile material is obtained to allow new cores to be made solely from bred material, and the cycle becomes self-sustaining. (The reactors are known as “fast” because they use neutrons that are not slowed by a moderating medium, such as water.)

India’s Fast Breeder Test Reactor (FBTR) and Prototype Fast Breeder Reactor (PFBR) are designed along these lines. Plutonium for the initial cores of these facilities comes from a number of India’s conventional nuclear power reactors, based on a Canadian design, known as the CANDU reactor. In these reactors, natural uranium is irradiated, transforming roughly 0.3 percent of the uranium into plutonium, which is then separated in a reprocessing plant. Because India has extensive deposits of thorium, but more limited deposits of uranium, India is working towards developing a breeder cycle in which plutonium is initially used as the fissile material in the core, thorium is used as the blanket, and uranium-233 is created in the blanket. Eventually, the uranium-233 will be used as the fissile material in future cores. Thus India’s three-stage breeder program begins with the conventional CANDU-style reactors (Stage 1), whose plutonium is used in the first generation of breeder reactors (Stage 2), until enough thorium has been transmuted into uranium-233 to be used in the core of a second generation of breeders designed to use this type of fuel (Stage 3). India is currently actively pursuing the second stage with the FBTR and the PFBR.

India’s fast breeder reactor program began with an agreement between France and India in 1969. Between 1969 and 1970, a team of Indian scientists and engineers traveled to Cadarache, France, home of the Rhapsodie breeder reactor, to finalize the plan for the FBTR, located at Kalpakkam near Chennai. [5] In 1974, Indian technicians, with the help of the French, began construction of the FBTR and completed it in 1984. Many of the key components for the FBTR were manufactured in India using French technology. [6] Currently the FBTR is testing a mixed plutonium-uranium oxide fuel, which will be used as the core material in the future for the PFBR. The plutonium for the FBTR’s initial operations is believed to have been extracted from the Madras nuclear power reactors in Kalpakkam and reprocessed at the Tarapur reprocessing facility. [7] In the 1990s, a working group began to design and develop India’s Prototype Fast Breeder Reactor (PFBR), a 500MWe pool type liquid metal fast breeder reactor located at the Indira Gandhi Center for Atomic Research in Kalpakkam. It is intended to be the first of a series of similar reactors to be constructed in the future. Construction of PFBR is believed to have begun in 1997 and is due to be completed by 2010. [8]

Potential Contribution of the FBTR and PFBR to the Indian Nuclear Weapons Program
The most immediate benefit that breeders could provide to the Indian nuclear weapons program would be to improve the quality of plutonium available to India for nuclear warheads. Plutonium with low concentrations of the isotopes plutonium-238, -240, and -242 and high concentrations of plutonium-239 is best suited for nuclear weapons. The presence of increased concentrations of the even-numbered isotopes makes nuclear weapon yields less predictable and requires special modifications in nuclear weapon designs. Usually the longer fuel is used in a reactor, the higher the concentration of these undesirable isotopes.

The plutonium that India is expected to introduce into the FBTR and PFBR comes from nuclear power reactors that have most likely been optimized to produce electricity, which means longer residence times for the fuel and increased presence of the unwanted plutonium isotopes. While less desirable for nuclear weapons, the material is quite suitable as fuel for fast breeder reactors. These reactors can then be used to produce plutonium in their blankets that has low levels of these isotopes -- less than 7 percent in total -- making it ideal for weapons. In effect, as one Indian author has noted, the breeders can act as a cleaner or “laundry” for contaminated plutonium. [9]

If the 500 MWe PFBR produced proportionally as much weapons-quality plutonium in its blanket as the now closed 1200 MWe French Superphénix breeder reactor, India could produce enough of the material for at least ten weapons annually (based on International Atomic Energy Agency standards, which specify that 8 kg of plutonium, sometimes called a “significant quantity,” is enough to construct one nuclear weapon). [10] Currently, using the CIRUS and Dhruva research reactors, at the Bhabha Atomic Research Center, in Trombay, India can produce only enough weapons-quality plutonium for 3-4 weapons annually. (These units are also being held off India’s civilian facility list.)

CANDU-style reactors, it should be noted, can also be optimized to produce excellent plutonium for weapons by moving fuel through the reactor more rapidly than would normally be the case if they were devoted to efficient production of electricity.

In his interview with the Indian Express, AEC Chairman Kakodkar denied that India planned to take plutonium produced in the breeders for weapons. Rather, he stated, elements of the fuel cycle supporting the breeder – implicitly, the eight CANDU-style power reactors India will keep on its military list, and the facility that reprocesses their fuel at Kalpakkam – are needed for this purpose.