|Fig. 1: Pelindaba nuclear research center (Source: Wikimedia Commons)|
During the latter half of the Cold War (1970-1990), the republic of South Africa pursued a covert research and development program that ultimately resulted in the production of seven nuclear weapons.  This effort reflected both the significant technological capabilities of the South African industry and the grave security concerns of the political leadership. With the end of the apartheid era, all nuclear weapons, along with their related production infrastructure, were voluntarily dismantled. South Africa thus became the first (and so far only) country to renounce its indigenously developed nuclear potential, providing an interesting case in the study of proliferation.
Almost immediately after declaring its independence in 1961, South Africa was faced with a growing degree of isolation in the international scene. This situation was the direct result of controversial policies pursued by the ruling white minority, such as the practice of strict racial segregation and discrimination and the occupation of South-west Africa (modern day Namibia). During the 1970s and 80s, South Africa was also actively involved in the Angolan civil war, fighting against a large number of Soviet-backed Cuban troops and their local allies. Heightened regional instability, coupled with an inability to modernize or replace its conventional arms due to sanctions and embargoes, made exploring the nuclear option an attractive alternative for South Africa. 
The involvement of South Africa in the field of nuclear energy was primarily motivated by that country's large reserves of natural uranium.  Starting from the 1950s, South Africa actively sought and secured foreign assistance in obtaining the basics of a nuclear industry, particularly in the areas of training qualified personnel and equipment procurement. A major aspect of this early collaboration was the supply by the US of the Safari-1 research reactor, which was commissioned in 1965 at the Pelindaba nuclear research center near Pretoria.  Subsequent indigenous developments followed two distinct paths. One was the construction of a sodium-cooled, heavy water-moderated reactor fueled by natural uranium for energy and plutonium production. However, this reactor was found to be uneconomical and was shut down in 1969, after only 2 years of operation.  The second, more successful, component of the indigenous effort focused on the production of enriched uranium fuel, using the helikon vortex separation process.  The latter is a type of aerodynamic enrichment method, where a highly compressed mixture of uranium hexafluoride gas and hydrogen is injected tangentially into a tube that tapers to a small exit aperture. The isotopic separation is provided by the centrifugal force due to the spiral motion of the gaseous mixture. A-publicly acknowledged-enrichment facility based on this process, called the Y-plant, was commissioned in 1974 and continued operations until 1990. This plant produced fuel for the Safari-1 reactor and the Koeberg nuclear power station (built by France and commissioned in 1984) but also served as the main source for the highly enriched uranium (HEU) used in South Africa's nuclear weapons program.
The origins of the nuclear weapons program can be traced back to studies conducted in the 1960s by the South African Atomic Energy Board (AEB) on the possibility of developing peaceful nuclear explosives for use in the mining industry (along the lines of the American Plowshare program).  In the early 1970s, in light of the perceived deterioration of South Africa's security situation, the decision was made to embark on a covert program for the production of deliverable nuclear weapons. The design chosen by AEB was a gun-type device, similar in concept to the Little Boy bomb that was dropped on Hiroshima. In this type of nuclear weapon, a projectile of HEU is shot against a target which is also made of weapon-grade uranium, thereby creating a supercritical mass of fissile material that starts an uncontrolled chain reaction (nuclear explosion). Gun-type devices have the advantage of being simpler to design and build but are also less safe than implosion weapons. After some research into gun-type ballistics, AEB produced two such test devices in 1977 and 1978. The HEU core consisted of about 55 kg of uranium from the Y-plant with more than 80% enrichment while the explosive yield was in the 10-18 kT range.  A cold test (without the fissile core) in the Kalahari desert was planned for 1977 but had to be aborted when the discovery of the test site and preparations caused an international uproar. To gain confidence that their weapons were workable, AEB scientists conducted a successful criticality test at a Pelindaba facility in 1979. 
After that point, weapons development was transferred to Armscor, the country's state-owned armaments industry. Armscor was tasked with turning the largely experimental AEB design into a device that could meet the stringent safety requirements of a deployable weapon. At the time of the program's cancelation in 1990, four such qualified weapons had been assembled and a seventh was under construction as a test device. Weighing approximately 1000 kg, they had a diameter of 65 cm and a length of 1.8 m and could be delivered by bomber aircraft. Interestingly, they did not require a neutron initiator but relied instead on background neutrons to start the chain reaction. Advanced concepts such as implosion or boosted-fission weapons for use by aircraft or ballistic missiles were also considered but never progressed beyond the theoretical study phase.  It should be emphasized that the gun-type devices were always kept under tight security in an Armscor storage facility and were never placed under the direct control of the South African Defence Forces (SADF). In fact, their existence was to be made public (and possibly demonstrated by an underground test) only as a means of forcing the Western powers to come to the aid of South Africa, in case she ever faced an overwhelming external threat. 
By the late 1980s, the Cuban/Soviet withdrawal from the civil war in Angola and the granting of independence to Namibia removed most of the security issues that had prompted the development of nuclear weapons. Thus, in anticipation of the transition to a democratic (majority rule) system of government, president F.W. de Klerk ordered the dismantling of the stored weapons, a process which was complete by time he publicly acknowledged the existence of the (now terminated) program in 1993.  The HEU inventory (about 400 kg in total) was recycled for use as reactor fuel while South Africa became a signatory to the Non-Proliferation Treaty in 1991. The total cost of the weapons program is estimated at $400 million (in 1994 values). 
At the height of the apartheid era, South Africa acquired a small number of indigenously developed gun-type nuclear devices. This covert nuclear weapons program stemmed from a larger involvement of South Africa with nuclear energy but also highlighted the leadership's acute fears due to their country's isolation in the world. As soon as South Africa was poised to become a respected member of the international community again, she abolished this small nuclear deterrent and became an active proponent of non-proliferation.
© Panos Baxevanis. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.
 D. Albright, "South Africa and the Affordable Bomb", Bull. Atomic Scientists 50, No. 4, 37 (Jul/Aug 1994).
 H. Steyn, J. Van Loggerenberg, and R. Van der Walt, Nuclear Armament and Disarmament, South Africa's Nuclear Experience (iUniverse, 2007).
 Z. Masiza, "A Chronology of South Africa's Nuclear Program", Nonproliferation Review 1, No. 1, 35, (Fall 1993).
 H. E. Purkitt and S. F. Burgess, South Africa's Weapons of Mass Destruction (Indiana U. Press, 2005).
 M. D. Laughter, "Profile of World Uranium Enrichment Programs", Oak Ridge National Laboratory, ORNL/TM-2009/110, April 2009.