Portal de Periódicos da CAPES

China's HEU and Plutonium Production and Stocks

2011; Taylor & Francis; Volume: 19; Issue: 1 Linguagem: Inglês

10.1080/08929882.2011.566469

1547-7800

Hui Zhang,

Nuclear and radioactivity studies

Abstract This article discusses the history of China's production of highly enriched uranium and plutonium for nuclear weapons and uses new public information to estimate the amount of highly enriched uranium and plutonium China produced at its two gaseous diffusion plants and two plutonium production complexes. The new estimates in this article are that China produced 20 ± 4 tons of HEU, 2 ± 0.5 tons of plutonium, and currently has stockpiles of about 16 ± 4 tons of HEU and 1.8 ± 0.5 tons of plutonium available for weapons. The values for China's fissile material production are at the low end of most previous independent estimates, which range from 17–26 tons of highly enriched uranium and 2.1–6.6 tons of plutonium. These new estimates would be significant to assess China's willingness to join a fissile material cutoff treaty and a multilateral nuclear disarmament. Acknowledgments The author would like to thank Harold Feiveson, Zia Mian, and Frank von Hippel for helpful discussions. Notes 1. China's highly enriched uranium is assumed to be 90 percent uranium-235 and weapon–grade plutonium is taken to be 94 percent plutonium-239. 2. For previous estimates of China's fissile material stocks, see, e.g., Robert S. Norris, A. S. Burrows, and R. W. Fieldhouse, Nuclear Weapons Databook, Volume V: British, French, and Chinese Nuclear Weapons (Washington, DC: Westview Press, 1994); David Albright, Frans Berkout, and William Walker, Plutonium and Highly Enriched Uranium 1996 (New York: Oxford University Press, 1997); David Wright and Lisbeth Gronlund, “Estimating China's Production of Plutonium for Weapons,” Science and Global Security 11 (2003): 61; David Albright and Corey Hinderstein, “Chinese Military Plutonium and Highly Enriched Uranium Inventories,” ISIS, 30 June 2005. These estimates have drawn mainly on translations of China's official nuclear history: Li Jue, Lei Rongtian, Li Yi, and Li Yingxiang, eds., China Today: Nuclear Industry (Beijing: China Social Science Press, 1987) (in Chinese). Selections were translated into English and published by the U.S. Foreign Broadcast Information Service, JPRS-CST-88–002, 15 January 1988; and JPRS-CST-88–008, Washington, DC, 26 April 1988; and John W. Lewis and Litai Xue, China Builds the Bomb (Stanford, CA: Stanford University Press, 1988). 3. The U.S. Department of Energy (DOE) estimate was first reported in Bill Gertz and Rowan Scarborough, “A Nation Inside the Ring,” The Washington Times, 9 July 1999. A more detailed report appeared in Robert S. Norris and William M. Arkin, “World Plutonium Inventories,” Bulletin of the Atomic Scientists, (September/October 1999): p. 71. The U.S. Department of Energy estimate also assigned China a stockpile of 1.2 tons of civil plutonium. 4. It is reported that China is also operating its own demonstration CEP near Lanzhou. See, e.g. “China's Indigenous Centrifuge Enrichment Plant,” Uranium Intelligence Weekly, Vol. IV, No. 43, October 25, 2010. 5. China Today: Nuclear Industry, op. cit. (in Chinese), p. 172. Subsequent references are to this Chinese version. 6. Director of Central Intelligence, Special National Intelligence Estimate (SNIE) 13-2-63, “Communist China's Advanced Weapons Program,” 24 July 24 1963, National Security Archive. . 7. “Summary and Appraisal of Latest Evidence on Chinese Communist Advanced Weapon Capabilities,” U.S. Arms Control and Disarmament Agency, ACDA-957, 10 July 1963. As of August 1964, U.S. intelligence agencies still believed the Lanzhou uranium enrichment plant was not complete. The U.S. Central Intelligence Agency (CIA) assumed that China would not have enough fissile material for a test until after the end of 1964. Director of Special Intelligence, “The Chances of an Imminent Communist Chinese Nuclear Explosion,” Special National Intelligence Estimate, 13-4-64, August 26, 1964, declassified version in Kevin Conley Ruffner (ed.), CIA Cold War Records Series, Corona: America's First Satellite Program (Washington, D.C.: Center for the Study of Intelligence, 1995). 8. Joel Ullom, “Enriched Uranium versus Plutonium: Proliferant Preferences in the Choice of Fissile Material,” Nonproliferation Review 2 (1994): 1, 1–5. A reactor was detected at Baotou nuclear complex by a March 1963 U-2 flight. U.S. intelligence mistakenly believed that it was a production reactor with a thermal power around 30 MWt, able to produce plutonium for one or two bombs a year. They estimated that a likely date for China to test its first plutonium-based device would be late 1964 or early 1965. U.S. Arms Control and Disarmament Agency, op. cit. 9. William Burr and Jeffrey Richelson, “Whether to Strangle the Baby in the Cradle,” International Security 25 (2001/2001): 3, 91. 10. See, e.g. R.E. Lawrence and Harry W. Woo, “Infrared Imagery in Overhead Reconnaissance,” Studies in Intelligence 11 (1967): 2, 17–40. 11. U.S. Defense Intelligence Agency, “People's Republic of China Nuclear Weapons Employment Policy and Strategy,” Report No. TCS-654775-72, Washington, D.C., March 1972. The Manhattan Project physicist Ralph Lapp used the early experience at the Oak Ridge Tenn. gaseous diffusion plant to estimate in 1971 that, at start-up, the Lanzhou plant may have produced about 130 kg per year of weapon-grade HEU and been able to double its annual production by 1966 as operators gained experience with the enrichment process; Lapp projected that the production capacity could increase to over 350 kg per year of HEU by the 1970s, Charles Murphy, “Mainland China's Evolving Nuclear Deterrence,” Bulletin of the Atomic Scientists, January 1972, pp. 28–35. 12. Albright and Hinderstein, op. cit. They assumed China's weapon-grade HEU was 93 percent uranium-235. If it is assumed to be 90 percent uranium-235, this production rate is equivalent to 23,000–51,000 SWU per year at a tails assay of 0.5 per cent, or 29,000–64,000 SWU per year for 0.3 per cent tails. 13. China Today: Nuclear Industry, op. cit., p. 179. 14. Xie Wuzhan, “504 Chang: Gongheguo Nongsuoyou Shiyue de Lingpaozhe,” Gansu Ribao (in Chinese) (“Plant 504: the leading runner of the cause of China's uranium enrichment,” Gansu Daily) 31 May 2008. 15. “Mainland China Talking to French, Germans, about Nuclear Power,” Nucleonics Week, 12 January 1978. 16. “Zhongguo Younongsuo ji Ranliao Yuanjian Zhizao (“China's Uranium Enrichment and Fuel Fabrication”) 28 March 2009, , and discussions with Chinese experts. 17. China Today: Nuclear Industry, op. cit., p. 180. 18. Ann MacLachlan and Mark Hibbs, “China Stops Production of Military Fuel: All SWU Capacity Now for Civil Use,” Nuclear Fuel, 13 November 1989. 19. Mark Hibbs, “China Said to be Preparing for Decommissioning Defense Plants,” Nuclear Fuel, 17 May 1999. 20. It is not clear whether or not the contract also allows the production of LEU for naval-reactor fuel. 21. By leaving more of the uranium-235 in the tails (i.e., a tails assay of 0.5 percent rather than 0.3 percent), China could achieve a given HEU production rate with a lower enrichment capacity. This would require about twice the amount of natural uranium feed, however 22. Albright and Hinderstein, op. cit. 23. This assumed capacity is lower than the 1972 DIA estimate. In its 1972 estimate, the DIA estimated that this plant could produce 750–2950 kg of weapon-grade uranium per year. This would correspond to about 145,000–569,000 SWU per year at a tails assay of 0.3 per cent. Since this estimate was made several years before the plant was put into operation, it is not clear whether it was based on the existing building or on an assumption that the building would be expanded, U.S. Defense Intelligence Agency, op. cit. 24. It is noted below that the plutonium production reactor at Guangyuan, which was built as a third line facility, has the same design power as the original Jiuquan reactor that it was backing up. 25. See, e.g., “Zhongguo Younongsuo ji Ranliao Yuanjian Zhizao,” op. cit. 26. China began in 1970 enriching uranium recovered from the irradiated fuel discharged from its plutonium production reactors. China Today: Nuclear Industry, op. cit., p. 186. 27. Producing about 2 tons of plutonium (as discussed in the following section) would have resulted in about 4000 tons of reprocessed uranium. To enrich all of this uranium to 90 percent HEU, with a tails assay of 0.3 percent, would require about 3.2 million SWU. This is 154,000 SWU more than would be required to produce the same amount of HEU from natural uranium. 28. China also has about 4 Miniature Neutron Source Reactors (MNSR). Each requires a long-lived core containing about 1 kg of 90 percent HEU. One of them shut down in 2007 and China has decided to shut down the other three MNSRs and replace them with LEU-fueled neutron sources. In addition, China sold one MNSR each to Ghana, Iran, Nigeria, Pakistan, and Syria. China has a project to convert those reactors to LEU cores . 29. Assuming that the 125 MWt HFETR had an average burnup of 40 percent of the urainuim-235 in its fuel and operated 12 weeks per year (IAEA research reactor database), and assuming 1.26 g of uranium-235 would be consumed per MWd (thermal) (Alexander Glaser, Neutronics Calculations Relevant to the Conversion of Research Reactors to Low-Enriched Fuel, Ph.D. Dissertation, Darmstadt, 2005) the HFETR would have used about 994 kg 90 percent HEU before conversion. The 5 MWt MJTR, with an average burnup of 45 percent and an operation of 14 weeks per year (IAEA research reactor database), would have used about 25 kg before conversion. 30. T. Dedik, I. Bolshinsky, and A. Krass, “Russian Research Reactor Fuel Return Program Starts Shipping Fuel to Russia,” paper presented at 2003 International RERTR Meeting, Chicago, Illinois, USA, 5–10 October 2003 . 31. David Albright and Kimberly Kramer, “Civil HEU Watch: Tracking Inventories of Civil Highly Enriched Uranium,” ISIS, February 2005, revised August 2005 . 32. “Entering a New Era,” Nuclear Engineering International, 8 January 2010 . 33. China Today: Nuclear Industry, op. cit. p. 239; The Magical Sword Literary and Art Society of Nuclear Industry, eds., The Secret Course (the second version) (Beijing: Atomic Energy Press, 1993), p. 301 (in Chinese). 34. China Today, Nuclear Industry, op. cit. pp. 305–306. 35. It is expected that China's first third-generation attack submarine (Type 095) will come into service around 2015, HansKristensen, “China'sNoisyNuclearSubmarines,” . 36. China reportedly has been building two (Type 094 Jin-class) ballistic-missile submarines since around 2003–2004. It is expected that about five such SSBNs will be built. HansKristensen, “TwoMoreChineseSSBNsSpotted,” . 37. See, e.g., Type 091 (Han Class) Nuclear-Powered Attack Submarine . 38. It is assumed the reactor operates with an average output of one-sixth of full power and the spent fuel has a uranium-235 burnup of 50 percent, and one kg of uranium-235 fission generates about 940 Megawatt-days of energy. Chunyan Ma and Frank von Hippel, “Ending the Production of Highly Enriched Uranium for Naval Reactors,” Nonproliferation Review 8, (Spring 2001); 95. 39. If a submarine was scheduled for launch before 1985, LEU may have had to be produced before 1980 to allow time for fuel fabrication, loading fuel into the reactor, and possible delays. The launch and initial operational capacity (IOC) of Han-class nuclear attack submarines are: Changzheng 1 (launched December 1970; IOC August 1974); Changzheng 2 (launched December 1977; IOC December 1980); Changzheng 3 (launched October 1983; IOC December 1984); Changzheng 4 (launched December 1985; IOC December 1987); Changzheng 5 (launched April 1990; IOC December 1990). Assuming the refueling interval is 10 years, then a total five cores were used for SSNs launched before 1985: Changzheng 1 (2 cores), Changzheng 2 (1 core), Changzheng 3 (1 core), Changzheng 4 (1 core), Changzheng 5 (0 core). “Type 091 (Han Class) Nuclear-Powered Attack Submarine,” . 40. For natural uranium feed, producing 1 kg of 5 percent LEU with a tails assay of 0.3 percent requires 7.2 SWU. 41. “China's Nuclear Tests: Dates, Yields, Types, Methods, and Comments,” . 42. This assumes 20 kg of HEU was used in each of the fission weapon tests. In the June 1967 3–3.3 MT thermonuclear weapon test, it is assumed that about 20 kT of the total yield came from the fission primary and about one-quarter of the yield in the thermonuclear secondary came from the fission of HEU, with about half of the HEU having fissioned. Frank von Hippel, Princeton University, personal communication, September 2010. This test would have consumed about 100 kg of HEU in the secondary. After plutonium became available in 1968, China may have shifted to using plutonium or composite uranium-plutonium pits since they allow the primaries to be more compact. 43. There were 18 tests after 1968 with yields above 20 kT that are assumed to have been thermonuclear weapon tests. The total yield of these 18 tests was about 19 MT. 44. In the U.S. enrichment program, “normal operating losses” were about 5 tons out of total production of about 1000 tons from gaseous diffusion plants, i.e., 0.5 percent losses. 45. Simon Henderson, “Nuclear Scandal: Dr. Abdul Qadeer Khan,” The Sunday Times, 20 September 2009. 46. One contribution to the ± 25 percent uncertainty assumed for the estimated HEU production is due to the range of possible tails. For natural uranium feed producing 90 percent HEU, at a given separative work capacity, a tails assay of 0.5 percent would produce about 25 percent more HEU than a tails assay of 0.3 percent. There is no official information about tails assays in China's gaseous diffusion enrichment program. 47. This is significantly less than the 21.5 ± 4.5 tons of HEU estimated in Albright and Hinderstein, op. cit. 48. China Today: Nuclear Industry, op. cit., p. 205. 49. Ibid. pp. 210–211. 50. W. Lewis and L. Xue, “Chinese Strategic Weapons and the Plutonium Option,” Critical Technologies Newsletter, U.S. Department of Energy, Washington, D.C., April–May 1988, p. 12. 51. China Today: Nuclear Industry, op. cit., p. 211. 52. Ibid., p. 212. 53. Ibid., p. 214. The reactor's power was increased by 10–15 percent through improvements in the cooling system. Fuel burn-up also was increased, and the number of full-power days went from the original design value of 288 days to 324 days per year. 54. See, e.g., Zhou Zhi, “Hegongyue 404 Jidi Chuangyue Huiyi,” (“Recollections of the Pioneering Work of Plant 404”), in Chinese, 19 August 2007. The author was vice-minister of the former Ministry of Nuclear Industry, . 55. During the early 1980s, China planned to convert the reactor to the dual mission of producing electric power as well as plutonium. Work started in September 1984 and was planned to be completed in 1987, but the modification seems to remain unfinished. No electricity substation or transmission lines connected to the site have been seen on satellite images. China Today: Nuclear Industry, op. cit., p. 91. 56. China Today, Nuclear Industry, op. cit., p. 227. 57. For a burn up of 800 MWt-days/ton, each ton of spent fuel would contain about 0.7 kg weapon-grade plutonium. See, International Panel on Fissile Materials, “Appendix B, Production of Highly Enriched Uranium and Plutonium for Weapons,” Global Fissile Material Report 2010, 2010 . 58. China's Nuclear Tests: Dates, Yields, Types, Methods, and Comments, . 59. Hui Zhang and Frank von Hippel, “Using Commercial Imaging Satellites to Detect the Operation of Plutonium-Production Reactors and Gaseous-Diffusion Plants,” Science and Global Security 8 (2000): 219. Figure A-2 shows that for a seasonal average temperature of 10°C and a typical temperature increase between 5°C and 15°C, the amount of heat discharged by the cooling towers would range from 0.02 MWt/m2 to 0.2 MWt/m2. For a top diameter of 30 meters, this corresponds to 14–140 MWt for each tower. 60. Operating at a capacity factor of 80 percent, a reactor of power 70–660 MWt could produce about 20–200 kg per year of weapon plutonium. This is a large range but it excludes a 1972 U.S. intelligence estimate that the Jiuquan reactor produced 300–400 kilograms of plutonium per year. U.S. Defense Intelligence Agency, op. cit. 61. International Panel on Fissile Materials, “Russia: Plutonium,” Global Fissile Material Report 2010, op. cit., p. 46. 62. See, e.g., News from Chongqing Cable TV, 26 April 2010 (in Chinese); Peng Yining, “Nuclear Reaction to Tourist Attraction,” China Daily, 22 June 2010 ; and “Former Nuclear Plant Opening as Tourist Attraction,” China Daily, 13 April 2010 . 63. The amount of plutonium produced per year by a reactor is estimated by: R(kg/yr) = CPth (MW)α(kg/MWd)365(d/yr), where R is plutonium production rate (in kilograms per year); C is the capacity factor; Pth is the thermal power of the reactor (in megawatts); and α is the amount of plutonium produced per megawatt-day of operation. 64. It also assumes the amount of plutonium produced per MWt-day by the Jiuquan reactor is the same as for the U.S. Hanford graphite-moderated, water-cooled reactors (see International Panel on Fissile Materials, op. cit.). Between 1967–69, the reactor operated at an average burn-up of about 400 MWt-days/ton and produced 0.9 grams of plutonium per MWt-day. From 1970 until 1984, the reactor operated at an average burn-up of 800 MWd/t and produced 0.85 g/MWt-day. 65. Zheng Jingdong, retired senior engineer from Plant 821, blog, “Zai Qiangjian 821 Chang de Rizi Li,” (“The Days of Racing to Complete the Plant 821”) (in Chinese) 26 September 2009. . 66. “The days of racing to complete the Plant 821,” op. cit.; and “Liangwei Hedian Gongchen de Zhihui Rensheng, Jiangsu Gonggong Kejiwang Wendang: Ren Wu,” (“The intelligent life of two heroes of nuclear power,” document web of Jiangsu public science and technology: People) (in Chinese) 14 May 2007 . 67. “The Intelligent Life of Two Heroes of Nuclear Power,” op. cit. 68. The new enterprise was called the Sichuan Wuzhou Industry Company, a subsidiary of the China National Nuclear Company. The company declared bankruptcy in 2009 and the residents in the complex will move to new living areas . 69. Norris et al., op. cit., p. 350. 70. See, e.g., “The Days of Racing to Complete the Plant 821,” op. cit. 71. It also assumes the amount of plutonium produced per MWt-day by the Guangyuan reactor is the same as for the U.S. Hanford graphite-moderated, water-cooled reactors (International Panel on Fissile Materials, op. cit.). For the whole operating period, the reactor is assumed to have operated at an average burn-up of 800 MWd/t and produced 0.85 g/MWt-day. 72. This is halfway between the 6 kg used in the first U.S. plutonium nuclear weapons and the 4 kg assumed for the pits of modern nuclear weapons. 73. The uncertainty of ± 25 percent stems primarily from the uncertainty of the initial design powers of the two reactors. 74. Reported by Bill Gertz and Rowan Scarborough in “A Nation Inside the Ring,” The Washington Times, 9 July 1999; and Norris and Arkin, op. cit. 75. See, e.g. Wright and Gronlund, op. cit. For the Jiuquan reactor, they assume a design power of 250 MWt that was later increased to 500 MWt. For the Guangyuan reactor, they assume a design power of 500 MWt that increased to 1000 MWt. 76. Communication received from China Concerning its Policies Regarding the Management of Plutonium, IAEA, INFCIRC/549/Add.7/8, 1 April 2008. 77. Deng Guoqing, China National Nuclear Corporation, “Overview of spent fuel management in China,” International Conference on Management of Spent Fuel from Nuclear Power Reactors,” Vienna, 31 May–4 June 2010 .