Notes and references
- Spector , Leonard S. and Smith , Jaqueline R. 1990 . Nuclear Ambitions , Boulder, Colorado : Westview Press . For more thorough treatments of the history of South Africa's nuclear program, see
- Spector , Leonard S. 1984 . Nuclear Proliferation Today , New York : Vintage Books, A Division of Random House .
- Albright , David , Berkhout , Frans and Walker , William . 1993 . World Inventory of Plutonium and Highly Enriched Uranium 1992 , Oxford : SIPRI and Oxford University Press .
- Stumpf . “South Africa's Nuclear Weapons Programme,” . Atomic Energy Cooperation of South Africa Ltd. . 1993
- Spector . Nuclear Proliferation Today , 280
- Spector and Smith . Nuclear Ambitions 270 President Eisenhower had launched the “Atoms for Peace” program in his speech before the United Nations on 8 December 1953.
- Stumpf , Waldo E. 1993 . “South Africa's Limited Nuclear Deterrent Programme and the Dismantling Thereof Prior to South Africa's Accession to the Nuclear Non‐Proliferation Treaty,” . a presentation at the South African Embassy Annex . July 23 1993 , Washington DC. Dr. Stumpf is Chief Executive Officer of the Atomic Energy Corporation of South Africa. A transcript of his talk has been released by the South African Embassy.
- Spector . Nuclear Proliferation Today 281
- Spector . Nuclear Proliferation Today 283
- Spector . Nuclear Proliferation Today 284
- 1993 . The Arms Control Reporter , May : 455.B.81
- 1993 . The Arms Control Reporter , May : 455.B.77 Stumpf, South Africa's Limited Nuclear Deterrent Programme.” The program was code‐named “Kraal,” an Afrikaans word for the stone walls used to fence in cattle,
- 1993 . The Arms Control Reporter , May : 455.B.82
- Benedict , Manson , Pigford , Thomas and Levi , Hans . 1981 . Nuclear Chemical Engineering , 876 – 895 . New York : McGraw‐Hill Book Co. . Stumpf, “South Africa's Limited Nuclear Deterrent Programme.” The Y‐plant uses an aerodynamic separation process similar to that developed by Becker in West Germany. It is described more fully in
- Z‐plant . 1993 . “Valindaba” is a Zulu word meaning “we don't talk about this at all,” . The Arms Control Reporter , May : 455.B.77 Contrary to some reports, it does not use the Helikon cascade technique that was incorporated into the
- Spector . 1984 . (Nuclear Proliferation Today , 290 New York : Vintage Books .
- Jaster , Robert S. 1984 . "Politics and the ‘Afrikaner’ Bomb,” . Orbis , winter : 28 claims Roux “announced that his country would build a commercial‐scale plant capable of producing 5,000 tons of low‐enriched uranium per year.” To produce 5,000 tonnes of 3.25 percent‐enriched LEU, at a tails assay of 0.3 percent, would require 20 million SWUs, on the order of the total U.S. enrichment capacity at its peak. If, as seems more likely, the intention was to enrich annually the uranium in 5,000 tons of U3O8, then only two million SWUs annually would be required to produce 540 tonnes of 3.25 percent‐enriched LEU annually, at 0.3 percent tails.
- Spector and Smith . Nuclear Ambitions 277 Ibid., and
- 1993 . The Arms Control Reporter , May : 455.B.80
- Stumpf, “South Africa's Nuclear Weapons Programme.”
- Benedict , Pigford and Levi . Nuclear Chemical Engineering 893 – 895 . Stumpf, “South Africa's Limited Nuclear Deterrent Programme.” The Z‐plant also uses the aerodynamic, or jet nozzle process, and incorporates the Helikon cascade technique permitting several separation stages to be incorporated in a single module. See
- Berkhout , Albright, and Walker . World Inventory of Plutonium and Highly Enriched Uranium 1992 187
- Following this point in the paper, one must differentiate between (a) the declared amount of HEU on hand—presumably reported by South Africa to the IAEA, and subject to conformation by direct measurement; and (b) South Africa's estimate of the amount of HEU produced, when calculated using material balance equations and measured values of product and tails inventories and U‐235 assays. Alternatively, the declaration (a) and estimate (b) can be given in terms of the amount of U‐235 in the HEU, labeled (c) and (d), respectively. Since South Africa has not revealed (a) through (d) publicly, except for the uncertainty in (d), we will be making our own estimates, (e) through (h), of the South Africa's measurements and best estimates, (a) through (d). Each of our best estimates has a corresponding uncertainty.
- Stumpf, “South Africa's Limited Nuclear Deterrent Programme.” ō[XHH] is calculated from measured product, other than HEU, and tails assays and amounts. Here it is not measured directly.
- Bevington , Philip R. 1969 . Data Reduction and Error Analysis for the Physical Sciences , 60 New York : McGraw‐Hill Book Co. . See, for example,
- Stumpf, “South Africa's Limited Nuclear Deterrent Programme.” We infer that “actual discrepancy” is the difference between the amount of U‐235 in the HEU, as calculated from the measured inventories and assays of tails, scrap, and enrichment products, other than the weapons HEU, and the amount of U‐235 in the HEU on hand as measured directly.
- We infer this, since the 526‐kilogram figure was not reported as being more than six times the “actual discrepancy.” The “actual discrepancy,” of course, could be even less than 88 kilograms.
- Benedict , Pigford and Levi . Nuclear Chemical Engineering 678 – 679 .
- Ibid., equation (12.204), p. 681, and equation (12.209), p. 682.
- Ibid., p. 895.
- The production of 953 kilograms of 90 percent‐enriched product from natural uranium feed, leaves 333,643 kilograms of 0.456 percent‐enriched tails = 370,643 kilograms total—37,000 kilograms associated with start‐ups.
- 1990 . “One‐Stop Irradiation Services from the Safari Material Test Reactor, Pelindaba, South Africa,” . Atomic Energy Corporation of South Africa, Limited . The core is in the form of a 9 x 8 grid. In 1963 its design was reported as having 22 fuel elements, 5 control rods, 22 beryllium reflectors and 23 aluminum filler pieces; “Research Reactors,” International Atomic Energy Agency, information on Safari‐I is from 1963. In 1990 it was reported as having 28 fuel elements and 6 control rods;
- We assume a capacity factor of 0.65, a fuel burnup of 40 percent, and 1.23 grams U‐235 consumed per megawatt‐day (MWd‐1), thus, (365 days per year) . (0.65) . (5 megawatts) . (1.23 gm MWd‐1) / [(3,604 grams per core) . (0.4)] = 1.0 cores per year = 3.6 kilograms per year.
- We assume a lower capacity factor of 0.5 due to the additional refuelings, thus, (365 days per year) . (0.5) . (20 megawatts) . (1.23 gm MWd‐1) / [(3,357 grams per core) . (0.4)] = 3.34 cores per year = 11.2 kilograms per year.
- If Safari‐I had operated with 90 percent‐enriched fuel, only one half the amount of product, 40 kilograms, would have been required. This would have required approximately the same SWUs, feed, and tails—6,428 kilograms SWU, 14,046 kilograms of feed, and 14,006 kilograms of 0.456 percent‐enriched tails; and therefore does not affect our calculations. Some of the Y‐plant's 90 percent‐enriched product may have been produced to supply future Safari‐I fuel requirements. Since it would be fungible with the HEU allocated for weapons, we draw no distinction. Some of the HEU from the now dismantled weapons also may be reserved for Safari fuel.
- Hibbs , Mark . 1988 . NuclearFuel , 25 July
- Albright , Berkhout and Walker . 1993 . World Inventory of Plutonium and Highly Enriched Uranium 1992 , 189 Oxford : SIPRI and Oxford University Press . reported that a West German middleman arranged for the export from China to South Africa of 30 tonnes of three percent‐enriched uranium and 30 tonnes of 2.7 percent‐enriched uranium in the form of UF6;
- For XP = 0.325, XF = 0.00711, and XT = 0.00456, the ratio of feed to product (F / P) = 10.96, and the ratio of separative work to product (SWU / P) = 2.96.
- The uncertainty takes into account the ± 16,000 kilogram uncertainty in tails associated with start‐up, the ± 4,200 kilogram uncertainty in tails associated with production of Safari‐I fuel, and the ± 7,400 kilogram uncertainty in the tails associated with production of Koeberg fuel.
- 0.00071 / 0.00456 = 0.1557, where 0.00071 is the one standard deviation uncertainty (the square root of the variance), and 0.00456 is the best estimate of the tails assay.
- Found by plugging data from table 1, into equation (7).
- Cochran , Thomas B. , Arkin , William M. and Hoenig , Milton H. 1984 . Nuclear Weapons Databook, Volume I: U.S. Forces and Capabilities , 32 Cambridge, Massachusetts : Ball‐inger Publishing Co. .
- 1993 . The Arms Control Reporter , May : 455.B.78
- Paine , Christopher E. and Cochran , Thomas B. “Strengthening International Controls on Military Applications of Nuclear Energy,” chapter 9 ” . In Controlling the Atom in the 21st Century , Edited by: O'Very , David P. , Paine , Christopher E. and Reicher , Dan W. Boulder, Colorado : Westview Press . in press).