International Network of Engineers and Scientists Against Proliferation

India’s Uranium Enrichment Program

M.V. Ramana

The last two years have seen revelations about secret uranium enrichment programs in Iran and South Korea. An older uranium enrichment program about which little is known publicly is India’s. Details of uranium enrichment activities have been kept largely secret in India, even more so than its other nuclear activities.^[1] Here I summarize what is known about the program, estimate its capacity and how much weapons-grade uranium could be produced at this facility should it be used to do so.

India’s interest in uranium enrichment dates back to the early 1970s.^[2] But it was only in 1986 that Indian Atomic Energy Commission Chairman Raja Ramanna announced that uranium had successfully been enriched.^[3] According to one report, a pilot scale plant has been operating in the Bhabha Atomic Research Center complex since 1985.^[4] A larger centrifuge plant has been reportedly operating at Rattehalli, Karnataka, since 1990.^[5] This is India’s main uranium enrichment facility. There is also an experimental laser enrichment program.

Construction of the Rattehalli plant started in the mid 1980s.^[6] During the initial years of operation, the plant reportedly had “frequent breakdowns as a result of corrosion and failure of parts.”^[7] It is therefore not surprising that many leaders of the Indian Department of Atomic Energy (DAE) held that uranium enrichment was very difficult and were skeptical of Pakistani claims that they had succeeded in enriching uranium to weapons-grade levels. Indeed, at least one chairman of the Indian DAE has privately argued that Pakistan did not have a nuclear weapon capability because it could not have succeeded in enriching uranium to weapons-grade levels because of the difficulties in constructing centrifuges of the required capacities.

The Rattehalli Plant and the Nuclear Submarine Program

The primary purpose of the Rattehalli plant appears to be to enrich uranium for the Indian nuclear submarine, officially termed the Advanced Technology Vessel (ATV), program. It is also possible that enriched uranium from this facility was used in the hydrogen bomb tested on 11 May 1998.^[8] Highly enriched uranium (HEU) is used in U.S. and Russian thermonuclear weapons.

According to a report from the early 1990s quoting unnamed official sources, the Rattehalli facility consists of “several hundred operating centrifuges made of domestically-produced maraging steel” with “a likely design throughput of under three separative work units (SWU) per machine per year.”^[9] One could take this to mean a total enrichment capacity of about 1,000-2,000 SWU/year. In 1997, it was reported that the DAE was planning to “build and install rotor assemblies of improved design.”^[10] Different enrichment levels for the output of the facility and the fuel used in the ATV submarine reactor have been reported these range from 6%-45%, but we will assume a range from 30-45%.

If, as Indian officials state, the primary purpose of the Rattehalli facility is to produce fuel for the ATV program, an analysis of the submarine reactor requirements could help estimate the uranium enrichment capacity.^[11] Strictly speaking this would only provide a lower bound. However, since the program has had operating difficulties, it is quite likely that the actual capacity is reasonably close to this lower bound.

Despite relatively prolific coverage of the Indian nuclear submarine program in the media, there is considerable confusion about its technical characteristics. In part this reflects the fact that the program started over 25 years ago and has evolved considerably over the decades.^[12] After numerous setbacks and failures, by the late 1990s a reactor design was finalized, and testing of a prototype commenced at the Kalpakkam nuclear complex in southern India.^[13] This implies that between 1990 and the late 1990s, the Rattehalli complex should have produced at least sufficient enriched uranium to fabricate the reactor core.^[14]

HEU for the Submarine Reactor

The amount of enriched uranium needed for a nuclear submarine reactor depends on a number of factors. These include the reactor power rating, the level of enrichment, the time intervals between core refuelings, the burn-up (which determines the fraction of the initial U-235 that is consumed before re-fueling [by conversion to U-236 as well as fission]), the reactor design (including factors such as fuel geometry, the use of burnable absorbers, and so on), and the use pattern (the average number of effective full power days per year or the average power the reactor produces through its lifetime). None of these are definitively known, and there are contradictory reports on some of these quantities (such as the power rating of the reactor).

One can set a lower bound on the power required by a submarine by estimating what is needed to overcome drag at the maximum design velocity. The maximum velocity of the ATV is usually given as about 30-35 knots, which corresponds well to maximum velocities reported for other nuclear submarines.^[15] At the lower value of 30 knots (15 m/s), the ATV would take about 36 hours to go from Thiruvananthapuram, a likely site for India’s strategic nuclear command center, to Karachi, Pakistan’s chief port and a likely site of naval blockade by India in the event of a war.

Assuming this speed and reasonable values for the other physical variables, the reactor power needed is about 112 MWth. However, this estimate is sensitively dependent on the assumptions made about various quantities. Therefore, it may be safe to assume that the submarine reactor power is somewhere between 90 MWth and 150 MWth.

The amount of uranium that is required for the submarine reactor core to operate at this power level depends on the reactor design, the time between re-fuelings, as well as the operational procedures and patrol routines followed by the submarine. For the same power rating and time between re-fuelings, the uranium inventory for a submarine that has a more demanding patrol routine would be higher. The core of the ATV is reported to have a design lifetime of ten years.^[16]

Estimates of the U-235 requirements for U.S. submarines are about 0.6-0.7 g/shp-year,^[17] while the requirements for Russian submarines are likely to be about 0.315–0.35g/shp-year.^[18] The propulsive power rating for Charlie Class submarines is 20,000 shp.^[19] Similarly, the Sevmorput nuclear icebreaker/cargo ship requires about 0.375 g/shp-year of U-235.^[20] Due to the smaller distances that the ATV is likely to traverse, one could assume that the ATV will require about 0.3 g/shp-year of U-235. Based on all these different assumptions, I estimate that the core of the ATV might use about 40-160 kg of U-235, with a median estimate of 90 kg. The actual amount of uranium used will also depend on the level of enrichment.

HEU for Nuclear Weapons

An additional demand for enriched uranium might be to test or manufacture thermonuclear weapons. In two-stage thermonuclear weapons, enriched uranium can be used in the primary, in the “spark plug” to aid in initiating the fusion reaction, or in the “pusher” encasing the fusion fuel.^[21] Enriched uranium may also be used to replace the “blanket” surrounding the warhead, which is usually made of depleted uranium, so as to increase the yield of a thermonuclear weapon.^[22] However, according to the official announcement that followed the 1998 tests, “the yield of the thermonuclear device tested on May 11 was designed to meet stringent criteria like containment of the explosion and least possible damage to building and structures in neighboring villages.”^[23]

If this were indeed the case, it is likely that the blanket may have been made of inert material. The requirement for enriched uranium may only be a few kilograms used in the spark plug in this case.^[24] We will assume this figure to be 5 kg and that about 10 kg of U-235 was produced for the test carried out on 11 May 1998. Apart from the amount actually used in the device exploded, this would also include processing losses and material used for conducting laboratory experiments.

In this scenario, the Rattehalli plant should have produced at least 100 kg of U-235 by 1999 for the submarine core and the thermonuclear device tested on 11 March 1998. Depending on whether India decided to stockpile thermonuclear weapons, there may or may not be a continuing demand for enriched uranium for weapons.

Estimated Enrichment Capacity

Enrichment capacity is measured in Separative Work Units (SWU), or more precisely kilogram SWUs, and is the quantity of separative work (indicative of energy used in enrichment) when the quantities of the feed (usually natural uranium), the enriched product, and the remaining tails are expressed in kilograms. The enrichment capacity requirements, the feed requirements, and the amounts of uranium enriched to different levels containing 1 kg of U-235 are summarized in Table 1. For a given tails enrichment level, the SWU requirements per unit mass of U-235 does not depend strongly on the enrichment level. This SWU requirement may be lowered by using a higher enrichment level for the tail, but that would significantly increase the amount of uranium used as feedstock. We therefore assume that the tails enrichment level is 0.3%, which means that that it would take approximately 200 kgSWU to produce a kilogram of U-235. Thus, manufacturing the 90 kg submarine core would require 18,000 kgSWU of enrichment capacity. A 40 kg core would require 8,000 kgSWU of enrichment capacity and a 160 kg core would require 32,000 kgSWU of enrichment capacity.

Based on reasonable assumptions about when enrichment started and the tail enrichment level, our estimates of enrichment capacity are listed in Table 2. The best estimate of current (2004) capacity from our analysis is 4,800 kgSWU/year. However, we emphasize that there are significant uncertainties and so it would be more reasonable to quote a range of enrichment capacities, namely 3,900-10,400 kgSWU/year. The amount of enriched product produced would depend on the enrichment level. For example, if the facility were producing weaponsgrade uranium (93% enrichment), then with this range of capacities, the facility could produce about 20-50 kg every year. If the facility were producing 30-45% enriched uranium, then it could produce about 40-175 kg of enriched product and 4,500-12,400 kg of depleted uranium every year.

There is no indication that India is seeking to fuel any of its Light Water Reactors with indigenous enriched uranium. This is also borne out by the above estimates of the capacity. A facility with a capacity of about 10,000 kgSWU/year could produce about 2.6 tons of 3.3% enriched uranium each year. This is to be compared with the initial core loading of 66 tons for the VVER-1000 reactors that India is importing from Russia.^[25] Thus, at the estimated range of current capacities, it would take decades for India to enrich sufficient uranium for even one reactor core.

This estimate of enrichment capacity also has implications for the size of the nuclear submarine fleet that can be sustained by India. At 200 kgSWU/kg-U-235 and 90 kg U-235/reactor core, a 4,800 SWU/year facility could produce a reactor core in about four years. Nuclear strategists have argued that India requires a fleet of three to five submarines.^[26] Since each submarine would need a new core in ten years, the estimated current capacity may be just insufficient to produce the enriched uranium requirements for a three submarine fleet. But since the enrichment capacity can be increased, this may not be a major bottleneck.

Literature:

• Albright, David, Frans Berkhout and William Walker, 1997, Plutonium and Highly Enriched Uranium 1996, Oxford University Press, New York.
• Alvarez, Robert and Debra Sherman, 1985, US to resume uranium production for weapons, Bulletin of the Atomic Scientists, April, pp.28-30.
• Anonymous, 1999, Reactor for N-submarine by next year, The Hindu, August 17.
• Bukharin, Oleg, 1996, Analysis of the Size and Quality of Uranium Inventories in Russia, Science and Global Security, Vol. 6 No. 1.
• Chellaney, Brahma, 1987, Indian Scientists Exploring U Enrichment, Advanced Technologies, Nucleonics Week, March 5.
• DAE and DRDO (Department of Atomic Energy/Defence Research and Development Organisation), 1998, Joint Statement by Department of Atomic Energy and Defence Research and Development Organisation; www.indianembassy.org/pic/PR_1998/May98/prmay1798.htm, accessed on October 1, 2003.
• FAS (Federation of American Scientists), Undated, Project 670 Skat / Charlie I & Project 670M Skat-M / Charlie II; www.fas.org/nuke/guide/russia/theater/670.htm, accessed on October 1, 2003.
• Fera, Ivan and Kannan Srinivasan, 1986, Keeping the Nuclear Option Open: What It Really Means, Economic and Political Weekly, December 6, 11(49), pp. 2119-2120.
• Gopalakrishnan, A., 2000, Undermining nuclear safety, Frontline, June 24.
• Hansen, Chuck, 1995, The Swords of Armageddon: US Nuclear Weapons Development Since 1945, CD-ROM.
• Hibbs, Mark, 1992a, India and Pakistan Fail to Include New SWU Plants on Exchanged Lists, Nuclear Fuel, March 30.
• Hibbs, Mark, 1992b, Second Indian Enrichment Facility Using Centrifuges Is Operational, Nuclear Fuel, March 26.
• Hibbs, Mark, 1997, India to Equip Centrifuge Plant with Improved Rotor Assemblies, Nuclear Fuel, December 1.
• Kumar, Dinesh, 1998, India Inching Towards Indigenously Build N-powered Submarines, The Times of India, October 3.
• Mærli, Morten Bremer, 1998, Criticality Considerations on Russian Ship Reactors and Spent Nuclear Fuel, Norwegian Radiation Protection Authority, Oslo.
• Nilsen, Thomas, Igor Kudrik and Alexandr Nikitin, 1996, The Russian Northern Fleet: Sources of Radioactive Contamination, Bellona Report, Bellona; www.spb.org.ru/bellona/ehome/russia/nfl.
• Nuclear Engineering International, 1998, World Nuclear Industry Handbook 1998, Business Press International, Sutton, England.
• Ølgaard, Povl L., Undated, The Potential Risks from Russian Nuclear Ships, NKS-SBA Status Report, Risø National Laboratory, Roskilde, Denmark; www.svanhovd.no/nrpa/nks/paper_povl.pdf.
• Podvig, Pavel (ed.), 2001, Russian Strategic Nuclear Forces, The MIT Press, Cambridge.
• Raghuvanshi, Vivek, 2001, Indian Navy Reaches Nuclear Power Milestone, Defense News, November 5.
• Ramana, M.V., 2004, An Estimate of India’s Uranium Enrichment Capacity, Science and Global Security, Vol 12 No 1-2, pp. 115-124.
• Reddy, C. Rammanohar, 2003, Nuclear Weapons Versus Schools for Children: An Estimate of the Cost of the Indian Nuclear Weapons Programme, in: M.V. Ramana and C. Rammanohar Reddy (eds.), Prisoners of the Nuclear Dream, Orient Longman, New Delhi, pp. 360-408.
• Rethinaraj, T. S. Gopi, 1998, ATV: All at sea before it hits the water, Jane’s Intelligence Review, June 1, pp. 31-35.
• von Hippel, Frank, David H. Albright and Barbara G. Levi, 1986, Quantities of Fissile Materials in US and Soviet Nuclear Weapons Arsenals, Princeton University/Center for Energy and Environmental Studies, Princeton.
   

---

For example, the location of the uranium enrichment plant was not included in the list of nuclear sites exchanged by the governments of India and Pakistan as a confidence building measure, Hibbs, 1992a. Albright et al., 1997, pp. 269-270. Chellaney, 1987. Fera and Srinivasan, 1986. Hibbs, 1992b. Fera and Srinivasan, 1986. Albright et al., 1997, pp. 270. Though the yield of the nuclear test and therefore the success of the design have been questioned, there is no reason to doubt official Indian assertions that a two-stage thermonuclear bomb was tested. Hibbs, 1992b. Hibbs, 1997. For a more detailed analysis of these requirements see Ramana, 2004. For a useful history of the program see Rethinaraj, 1998. Raghuvanshi, 2001; Kumar, 1998; Gopalakrishnan, 2000. One report claims that these tests are of a “scaled down reactor” but this has not been corroborated elsewhere. Even if true, it is still likely that the Rattehalli complex would have produced sufficient enriched uranium to fabricate the entire reactor core. It is of course possible that the necessary enriched uranium was produced much earlier. But the reports that the Rattehalli facility was not working well suggests that the necessary enriched uranium would have likely been produced only close to the time the reactor began to be tested. See for example Raghuvanshi, 2001. Anonymous, 1999. von Hippel et al., 1986. shp = shaft horsepower; 1 shp = 0,746 kilowatts. My calculations based on information from Podvig, 2001, pp. 273-278, Bukharin, 1996, and Nilsen et al., 1996. FAS, Undated.

M.V. Ramana is Fellow at the Centre for Interdisciplinary Studies in Environment and Development, ISEC Campus, Nagarabhavi, Bangalore 560072, India; tel. +91-80 23 21 70 13 (X24); ramana@isec.ac.in; http://www.cised.org. CISED aims to promote environmentally sound and socially just development by contributing critically and constructively to public and academic debates. Its mandate is to work on issues at the interface of environment and development. My calculations based on information from Mærli, 1998 and Ølgaard, Undated. Hansen, 1995, pp. I-94. The spark plug is a fissile mass placed in the center of the fusion fuel. When the fusion fuel is compressed by the shock wave, the fissile mass is compressed to supercritical densities setting off a chain reaction, which increases the yield and produces additional neutrons that could induce ignition of the secondary. Alvarez and Sherman, 1985. DAE and DRDO, 1998. This would presumably increase the chances of igniting the secondary, a likely goal for a first thermonuclear test. Nuclear Engineering International, 1998, pp. 60. VVER = Voda-Vodyanoi Energiticheskiy Reaktor, the Russian pressurized water reactor. Reddy, 2003, pp. 380-381.