Over at ArmsControlWonk, Jeffrey Lewis puts to rest widely-cited but misleading estimations of the Chinese nuclear weapons arsenal being in the two thousand range. They are – apparently – just plain wrong. The culprit largely responsible for the gross overestimation is one Yang Zheng, a Singaporean student.
There are similar inconsistencies in estimations of the Indian nuclear arsenal, although not perhaps on the same scale. Among the more widely-cited figures in the West is David Albright’s estimation that India had 250-300 kgs of separated weapons-grade plutonium (WGPu) in 1998, produced by the CIRUS and Dhruva research reactors [cf. Perkovich, India's Nuclear Bomb], which would have been enough to produce about 50 bombs. Subsequent production gives Albright figures in the range of 55-115 and 65-110 weapons in 2004. Joseph Cirincione and others at the Carnegie Endowment produced a figure of 50-90 weapons in a 2002 study. The Arms Control Association in 2007 lists India as having up to 100 weapons, a nice round figure recited by journalists, academics and activists both in India and abroad. And R. Rajaraman accounts for diminished capacity factors but still produces a figure of 649 kg of WGPu by 2008, enough for 130 weapons [in Chari ed., Indo-US Nuclear Deal, pp. 133-134]. So when Praful Bidwai claims that the Indo-US nuclear deal has given India the ability to produce 500-600 nuclear weapons every year, it makes it seem as if he pulled those figures out of his—ahem—nether regions.
All of these figures appear to be overestimations for a host of reasons:
- India’s WGPu come primarily from the CIRUS and Dhruva Research Reactors, which are capable of producing about 9-10 kg and 20-25 kg of WGPu per year, respectively, according to the Nuclear Threat Initiative. Rajaraman makes a more convincing estimation – based on the work of Albright, Berkhout and Walker – of 13 kg and 32.9 kg of WGPu from the two reactors respectively at 100 per cent capacity. In addition, some WGPu may have been extracted from Kalpakkam and other research reactors or through reprocessing, but this is likely a negligible amount and offset by losses described below.
- Fuel shortages meant that CIRUS and Dhruva were operating at a fraction of their actual capacities prior to 1998 (Albright himself estimates about 50% for CIRUS and 40% for Dhruva, while Rajaraman estimates about 70% each). In fact, from 1997 to 2003 (or, according to some reports, 2005), CIRUS was completely out of operation, so for 6, perhaps 8, out of 11 years after India has had proven bomb designs it has also had but a single source of WGPu. Factoring this in, and using Rajaraman’s figure of 70% operational efficiency, that gives India a maximum stockpile of 840 kg of WGPu, enough for 120-210 weapons based on bomb design. (I use estimations of 4-7 kg of WGPu/weapon throughout, although some consider 8 kg a better figure, which should only reduce the weapon stockpile further.) Using Albright’s estimations of efficiency this would be dramatically lower, with a maximum stockpile of 536 kg of WGPu, corresponding to 77-134 weapons by 2009.
- However, India also probably lost some WGPu in testing and other processes, such as fueling India’s fast breeders and possibly even power reactors. A 2006 report coauthored by Zia Mian and others in Science and Global Security estimates a loss of 130 kgs. This brings the estimate – which is, remember, a maximum - to 710 kg (based on Rajaraman) and 406 kg (based on Albright) of WGPu. This would translate to 101-177 weapons and 58-102 weapons respectively in 2009. These still represent upper range estimates.
- Beyond the ability to produce plutonium, estimates should take into consideration the fact that India had not tested its between 1974 and 1998, so was working largely with untested bomb designs prior to 1998. It is highly unlikely that weaponization was taking place at maximum potential until that point.
- Nor was India’s weapons program operating at a steady pace, fluctuating its activity based on outside factors, such as the state of the international non-proliferation regime and the whims of the political leadership in Delhi, as documented by Perkovich among many others. Certainly, the rate of weaponization was patchy between 1963 and 1989, or so.
- If one were to believe that India had only a about ten weapons produced prior to 1998 – five of which it tested at Pokhran – and assuming (and this, admittedly, is a big assumption) that India has since produced warheads at a rate that is commensurate with its plutonium production rate, it would would mean an increase of 177.26 kg of WGPu since 1998, which is the equivalent of 25-44 weapons for a total of 30-49 warheads. That figure naturally does not account for WGPu stockpiles accumulated prior to 1998, which is a bit of a stretch, but not so much if you consider the factors listed above. If one were to go by Albright’s rather high estimate of 50 weapons in 1998, this would mean a total of 75-94 warheads today.
- Another powerful argument for a smaller arsenal than often claimed comes from Adm. Raja Menon, who notes that other “medium nuclear powers” (China, the UK, France) maintain a large percentage of their WGPu as reserves – as much as two-thirds to five-sixths. This may not be the case with India’s nascent weapons program, but remains a possibility.
- Finally, in the absence of outside verification, it is also possible that estimates of plutonium capacity provided by the Department of Atomic Energy are themselves slightly exaggerated, although it is hard to factor this in reliably.
What accounts for the frequently inflated numbers attached to India’s nuclear arsenal (ie approximately 100 weapons as opposed to the 30-49 range)? As alluded to already, estimates assume a steady (and maximum) rate of weaponization and do not always account for inefficiency or over-exaggeration of capacity. Some factor in the possibility of India having accumulated stockpiles of highly-enriched uranium (HEU) for bomb-making, but given the inefficiency of India’s power reactors, and the bomb designs tested in 1998, this is unlikely or the amounts of HEU produced are negligible. Albright also believes that if India managed to use its stockpile of approximately 600 kg of non-weapon-grade plutonium for bombs, its arsenal could be much larger. That’s a big, big if, but it also partly explains the enormous variation in subsequent estimates. That may be why many such outside estimates have, in fact, reduced over time.