Book Review: How SA built six atom bombs


This rather repetitive and somewhat superficial book is a history of the South African nuclear weapons programme that by late 1989 had yielded six gun-type atom bombs as well as a missile programme that would shortly afterwards have allowed the country to drop a nuclear warhead anywhere in the world.

Each bomb contained 55kg of highly enriched uranium (HEU) and the country had sufficient of the metal for a seventh weapon. According to a very useful appendix to the book, a May 1994 report on the SA programme by David Albright, then president of the Washington-based Institute for Science and International Security, the weight of the completed devices was about one metric ton and the expected yield was 10 to 18 kilotons (kT) using weapons grade (90%+ enriched) uranium. Using 80% enriched uranium would have halved the yield.

The gun-type design placed the SA weapons in the same class as “Little Boy,” the device dropped on Hiroshima in August 1945 with a yield of 12.5kT. Some theoretical work was done to boost the yield to roughly 100kT using tritium, but as the “purpose of the bomb programme was to demonstrate capability, why would yield matter” asked one official.

According to Albright, the weapons programme cost the taxpayer about R10 million a year in the early 1980s, rising to between R20 and R25 million at the end of that decade.


SA commissioned its first nuclear reactor, a 20 megawatt HEU reactor acquired from the US under the latter’s “Atoms for Peace” programme in March 1965 at Pelindaba, west of Pretoria under the name SAFARI-1 or SA Fundamental Atomic Research Installation 1. Venter has it that a indigenously-designed heavy-water reactor, SAFARI-2, went critical in November 1967. It proved too expensive to operate and was shut down in 1970.

At this stage SA’s interest in nuclear technology was still peaceful, concentrated as it was in the SAFARI programme and an interest in “peaceful nuclear explosives” (PNE), perhaps for use by the mining industry. Venter notes PNE research was encouraged by the West under such schemes the American “Plowshare Programme [sic]” and that “considerable amounts of nuclear data and technology were transferred to SA under its auspices. According to a subsequently declassified Central Intelligence Agency report at least one SA scientist studied the application of PNE in the US.
“It would … be naïve to believe that the skills and knowledge South Africans acquired while studying or carrying out nuclear research abroad were not applied, where appropriate, to their country’s nuclear weapons programme. Indian, Pakistani, Israeli and other nationals did exactly that on their return … after acquiring this vital information, and there is no reason to believe the South Africans were any different.
“General HJ Martin, the SA Army Chief of Staff as quoted in December 1968 as saying that SA was ready to make its own nuclear weapons. … Martin’s statement was repudiated by the then Minister of Defence, PW Botha, but the point had been made, namely the difference between a peaceful nuclear explosion and a military one consisted costly of the fact that the latter had fins at the back.”

According to Albright, the then-Atomic Energy Board (AEB) received permission in 1971 from the Minister of Mines to do research and development work on PNE. In 1974 Prime Minister John Vorster “approved the development of a limited nuclear explosive capability and the construction of an underground test site” at Vastrap, north of Upington. The AEB produced its first nuclear device in mid-1977. A “cold test” (no HEU) was scheduled for that August but was cancelled after international intervention. A second device, codenamed “Melba” was built in 1978 and was “designed to be rapidly deployed for a fully instrumented underground nuclear test at the Kalahari site.”

Armscor took responsibility for the project in 1979 (Project Kerktoring, belfry) and built a facility known as the “Kenton Circle” within the grounds of its Gerotek vehicle testing facility between Pelindaba and Pretoria. Albright notes Armscor “considered the AEB’s November 1979 device to be an unqualified design that could not meet the rigid safety, security and reliability specifications then under development by Circle engineers.” Armscor built its device, a “pre-qualification model” in April 1982.

Then-President PW Botha decided in September 1985 to limit the scope of the programme as he recognised that costs “could escalate significantly.” Production was thus limited to seven gun-type weapons and all work was stopped on plutonium devices and tritium boosters. “But implosion development and theoretical wok on more advanced devices continued.

Armscor completed its first “qualified” device in August 1987 and this could be be delivered by a modified Blackburn Buccaneer S.50 strike aircraft, Albright avers. Shortly before President FW de Klerk cancelled the programme, a new facility, the Advena Central Laboratories, was completed at Gerotek at a cost of R36 million to build a new generation of weapons “deliverable by aircraft and most likely also by ballistic missile.”

Albright notes De Klerk halted production in November 1989 and issued “written instructions to terminate the … programme and dismantle all existing weapons” on February 26, 1990, just days after the unbanning of the African National Congress and the release of its long-imprisoned international face, Nelson Mandela. Dismantling started that July and by September the HEU had been removed from the weapons and returned to the AEB’ successor, the Atomic Energy Corporation.

Non-nuclear components, design drawing and plans were destroyed in 1992. By then SA had joined the Nuclear Nonproliferation Treaty (in July 1991) and inspections by the International Atomic Energy Agency (IAEA) were underway from November 1991. De Klerk announced the programme on March 24, 1993, by which time “most of the classified documents had been shredded and the sensitive weapons components destroyed or damaged beyond repair. Destruction of less important components continued into 1994.” The Circle and Advena were cleansed and decommissioned.

SA’s arsenal

SA nuclear expert Dr Nic von Wielligh told Venter the “first SA devices were not meant for military application but for demonstration of capability.” A “Little Boy”style uranium gun-bomb was therefore sufficient. A “Fat Boy” (the plutonium bomb dropped on Nagasaki) weapon would require the mastery of spherical geometry and the nuances of implosion, which Von Wielligh said “probably would have required actual testing”. Interestingly, while the US Manhattan Project did explode a test implosion device at Alamogordo in July 1945, this was not deemed necessary for the gun device. Little Boy was thus a “live test”.

Venter avers that the SA weapons were “regarded as clumsy, overly bulky and of outdated World War II-vintage design (although he does not say be whom). But they helped the country white male leadership achieve their “initial objective, which was to join the world’s ultra-exclusive nuclear club.”

On the specifics of the programme, Venter notes that something that “surprised IAEA inspectors … was how much of the equipment was distinctly low-tech.” This, it seems was as much to maintain secrecy as save on costs and circumvent sanctions. “They were very creative,” a US participant commented. One example mentioned was the modification of two-axis machine tools, normally used for simple manufacturing, to create complex three-dimensional shapes.

Another noteworthy aspect was the tiny workforce involved. Albright puts it at about 100 in the early 1980s of whom “only about 40 were directly involved in the weapons programme and only 20 actual built the devices. The rest were involved in administrative support and security”. By 1989 the labour pool had grown to 300 of whom “about half” were directly involved with weapons work.

Albright records that according to De Klerk “the weapons were never intended for actual use and they were never deployed militarily or integrated into the country’s military doctrine. In essence, he weapons were the last card in a political bluff intended to blackmail the US or other Western powers” in the face of a Soviet onslaught. “Whether it would have worked is impossible to determine.”

While De Klerk’s assertion may be true, many have and will refuse to accept it; believing instead the apartheid state would happily have incinerated any number of Africans to keep the Afrikaner in power.

Anyway, to work, Albright continues, this policy “required a credible nuclear weapon. And, according to Armscor officials, credibility required deliverability. If the nuclear devices were only test devices, the Western powers might not take SA’s threat seriously enough to intervene on its behalf.”

This then was the driver behind several related weapons programmes, most notably the development of a domestic ballistic missile (Project Burglar) in collaboration with Israel and – according to Venter – other powers, principally “Red” China. “What eventually emerged from the reported 1974 Israeli-SA Project Burglar initiative was a reasonably versatile family of similar solid-propelled missiles such as the two-stage Jericho-2/YA-3 IR/LRICBM [intermediate range/long range intercontinental ballistic missile]; the three/four stage Shavit/1 SLV [satellite launch vehicle]; and the subsequent four-stage Next/Shavit-2 SLV, as well as another reported ICBM variant,” Venter says. The SA versions became known as the RSA-1, RSA-2 and RSA-3. While it is not clear to this reviewer what “RSA” stood for, it seems to him to have capitalised on the patriotism of the time, taking for itself the official initials of the Republic of South Africa.

Venter gives some performance parameters but then qualifies this by saying that it “should be mentioned that these figures represent conservative and optimistic guestimates garnered from unrelated sources.” Okay…. He does quote Mark Wade, author of the Encyclopaedia Astronautica , “a comprehensive internet-based knowledge base” as calculating that a three-stage RSA-3 rocket, used as a ballistic missile, could have carried a 400kg payload to Moscow or a 340kg surprise to Washington DC. Venter argues the 340kg payload would suggest a warhead of some 200kg, “well beyond SA’s best efforts of the late 1980s.” Wade also has a four-stage RSA-4 missile that would have weighed 88mt, be 23.5m long, have a base diameter of 2.4m and would have been roughly equivalent to the US LGM-118 Peacekeeper. It would have carried an 700kg warhead to “any target on earth.”

The idea was to deploy the missiles on mobile “transport-rector-launchers” (TEL) called Beestrokke (cattle trucks, beestrok is the singular) developed under Project Wrinkle. It appears up to 39 Beestrokke were considered at one stage, divided into five flights of six TELs. The other nine TELs were apparently a “strategic reserve”. By November 1989 seven had been built and six deployed. “All these systems would probably have been based at an ultra-secure site near Hammanskraal [north of Pretoria] at a base called Rooi Lig (‘red light’), containing 22 large hangars and six even more expansive and obviously intended for maximum-sized SLVs.” This reviewer cannot say whether such a base exists or existed, although the site makes sense: there is a number of large adjacent military properties in that area, including Wallmannsthal, Boekenhoutkloof and Paardefontein. During the 1980s there was permanent missile air defence complex in that area, covering the capital from surprise air attack from the north. This air defence umbrella would also have been useful for the protection of the RSA missiles on their TELs.

Available data indicates that a RSA-1 was fired from the Overberg Test Range on June 1, 1989. A RSA-2 missile was fired on July 5, 1989. this is known to have travelled just over 1450km or 900 miles to the south east, splashing down near the Prince Edward island group. A final two-stage system was fired on November 19, 1991. Armscor apparently briefed Cabinet that the weapons could hit Nairobi (2600km distant) with a circular error probability of 275 metres.

Only one of these missiles now remain, a solitary RSA-3 on its TEL-cradle at the Air Force Museum at the former AFB Swartkop. The missiles, it appears would have been Air Force assets.


Venter’s book is not the first on the SA nuclear weapons programme, and will hopefully not be the last, as there are still many unanswered questions and taunting mysteries. It is, however, so far,the best. How SA built six atom bombs is an expansion of an earlier work, Those who had the power – South Africa: An unofficial nuclear weapons history, privately published by Nicholas Paul Badenhorst and Pierre Lowe Victor in 2006 (only a 100 copies was produced). Both are credited as associate authors with Venter, but only on the title page – not on the cover.

A second work involved former Chief of the SA Air Force Jan van Loggerenberg as a co-author along with Hannes Steyn and Richardt van der Walt. Entitled Armament and Disarmament SA’s Nuclear Weapons Experience it was published by Network Publishers in Pretoria in 2003.

Then there is also academic papers by US authors such as Helen E Purkitt and Albright in addition to more indirect titles such as Seymour Hersh’s The Samson Option: Israel’s nuclear arsenal and American foreign policy, published by Random House in New York in 1991.

How SA built six atom bombs is thus an important book. But its value is somewhat diminished by what appears to be uncorroborated and perhaps untrue claims. Some appear quite exotic to this reviewer, who, while unable to refute them is equally unable to confirm them; and leaves him wondering how much else in the book must be given a question-mark.

First and foremost are the claims that Israel “allegedly” supplied SA functional nuclear gravity bombs “around 1967”. This would not have been too long after Israel had built its first weapons. Patrick Tyler records in A World of Trouble, his account of America’s involvement in the Middle East since Eisenhower, that the Central Intelligence Agency first concluded the Dimona facility in the Negev “cannot be solely for peaceful purposes” in December 1960. It is widely believed Israel only built its first weapon after the June 1967 Six-Day War and according to the CIA from 1974 to the early 1980s the country only possessed between 10 and 20 warheads, raising the question whether they had spare weapons to lend SA even if they had the will to do so.

A more startling claim is that SA and Israel in November 1974 signed an agreement in Geneva under which the latter agreed to provide SA “eight Jericho 1 missiles. Within the framework of a project known as ‘Chalet’ it was also specifically committed to arm these weapons with ‘special warheads’. The implication here is that this might have been an actual foreign nuclear acquisition,” Venter says. As far as this reviewer can ascertain, that would have been a world first. Countries have to date been reluctant to even help each other develop weapons, let alone sell or loan them.

He claims the deal was signed by then Prime Minister John Vorster and defence minister PW Botha for SA and “premier” Shimon Peres” for Israel. Peres, currently president of Israel, and the only one of the three still alive, was in fact defence minister at the time under Yitzhak Rabin.
(The Manhattan Project was a multinational programme, but the US afterwards refused to provide the technology to even Britain, a major player in the project and its closest ally. Britain then had to reinvent the bomb for itself. Russia developed its atomic weapons partially assisted by espionage, a technique used by several other states, including Israel. France, Britain, the US certainly proliferated nuclear technology but not actual plans or weapons. North Korea may be an exception and Pakistan’s AQ Khan, apparently seeking private gain, certainly sold weapon blue-prints, know-how and technology to any interested party, including Libya.)

Venter later contradicts (or covers) himself by saying “we are still not certain whether [the missiles and nuclear warheads] were ever delivered to SA.” It is not certain who “we” are or why there is 22 pages between the two statements. He then adds it is “worth mentioning” that two British authors (Peter Hounam & Steve McQuillan, The mini-nuke conspiracy: Mandela’s nuclear nightmare, Faber and Faber, London, 1995) had noted that SA had deployed medium to long-range missiles, “some” nuclear armed, “as long ago as 1982/3”. These “were reportedly deployed at the army’s battle school at Lohatlha … in silos dug deep into the side of a mountain.” Venter then adds “the type of missile, if there were any at the battle school, remains a mystery…” Indeed. The reviewer, a frequent visitor to Lohatlha, has to this book never heard of the deployment of ballistic missiles of any type at the battle school and certainly has never seen or heard about the alleged silos, despite often flying over the base at low level. It would have been useful if Venter or his co-authors had made more effort to confirm or rebut these claims, perhaps even conducting site visits.

The same can be said about an alleged “small underground base” near the Kalahari Gemsbok national park located at the confluence of the SA, Botswana and Namibia borders as well as the claim that “more missiles were intended to b deployed at other sites in underground silos or on platforms which could be raised for firing throughout the then Transvaal (now Gauteng) province”, and the storage of the AEB 1979 device in an “abandoned coal mine” used as an ammunition depot. Perhaps it is all true, but how about some proof?

Elsewhere he quotes Hounam and McQuillan as saying SA had “the much sought-after ‘neutron shell’ … actually … available … since 1982. These would apparently have been fired from US M2 155mm “Long Tom” artillery pieces, the carriages of which Venter says had been altered to allow the fitting of “either 175mm or 203mm barrels. He then notes PW Botha’s defence minister Magnus Malan as saying SA received both 155mm and 175mm barrels, and postulates the latter could have fired nuclear shells. He then suggests Israel may have “leased or loaned … SA … several of these shells (a figure of 10 has been mentioned).”

Even if so, these weren’t the first nuclear weapons in SA either. That honour belongs to the Royal Navy that stored atomic munitions here in the 1960s. Venter has it they were stored “at a military facility called Kaalpan near the town of Warrenton” near Kimberley. The reviewer is independently aware of this but was led to believe by his source the weapons had been kept at what is now 93 Ammunition Depot in Jankempdorp, not too far from Warrenton. The reviewer is not familiar with a storage facility at Warrenton. Google Earth was not able to locate Kaalpan either. Venter has it that the weapons were likely the 900kg 20kT Red Beard tactical nuclear gravity bomb first delivered to the British navy in 1959.

Venter also adds to the speculation around the November 1987 loss of a SAA Boeing 747 (the Helderberg) by reporting the claim that the plane may have been carrying a consignment of ammonium perchlorate, an ingredient used in solid-propellant rocket fuel, apparently obtained from the People’s Republic of China. He does not provide any evidence.

Elsewhere he does make a case for missile collaboration between SA and Israel with input from China, France and the US. “As was subsequently revealed, this developed into a multi-directional flow, with subsequent missile technology developed by the Israelis and South Africans heading to China and then on to Pakistan and, more recently Iran.” For proliferation to China he cites Charles Vick of as saying that RSA-3 and RSA-4 technology “has shown up” in the four stages of the DF-31 ICBM and in the JL-1 submarine-launched ballistic missile. This Vick says proliferated to Pakistan (Shaheen-II) SLV/ICBM and Iran (Ghadr-110A) IRBM/ICBM/SLV. The latter,Vick says, “will have a striking resemblance to the RSA-4/Shavit follow-on and the Chinese DF-31/KT-1…”

If this claim is true there is some irony in Israel possibly facing an existential threat from its own technology – the reason states generally have not helped even close friends develop nuclear weapons. But again Venter does not provide much evidence, beyond Vick’s say-so, although that may be sufficient – Vick is a recognised world authority on the subject.

Al J Venter in association with Nicholas Paul Badenhorst and Pierre Lowe Victor

How SA built six atom bombs – and then abandoned its nuclear programme

Ashanti Publishing

Kyalami Estate