TRENDS IN SOUTH AFRICA'S NUCLEAR SECURITY POLICIES AND PROGRAMS

• Document Type: FOIA
• Collection: FOIA Collection
• Document Number (FOIA) /ESDN (CREST): 05830442
• Release Decision: RIPPUB
• Original Classification: U
• Document Page Count: 35
• Document Creation Date: March 9, 2023
• Document Release Date: April 23, 2020
• Case Number: F-2018-00226
• Publication Date: October 5, 1984

pproved for Release: 2020/03/16 C05830442
Director of
Central
Intelligence
Trends in South Africa's
Nuclear Security Policies
and Programs
National Intelligence Estimate
WARNING: The material in this document is
sensitive. Distribution of this Estimate
should be strictly limited to those officials
who require access to the subject matter for
the performance of their duties.
LO
Top Secrct
Top Secret
NIE 73/5-84/CX
SC 03101-84
5 October 1984
(b)(3)
Ell
Approved for Release: 2020/03/16 C05830442
Copy 102
pproved for Release: 2020/03/16 C05830442
Warning Notice
Sensitive Intelligence Sources and Methods Involved
(WNINTEL)
NATIONAL SECURITY INFORMATION
Unauthorized Disclosure Subject to Criminal Sanctions
DISSEMINATION CONTROL ABBREVIATIONS
NOFORN� Not Releasable to Foreign Nationals
NOCONTRACT� Not Releasable to Contractors or
Contractor/Consultants
PROPIN� Caution�Proprietary Information Involved
NFIBONLY� NFIB Departments Only
ORCON� Dissemination and Extraction of Information
Controlled by Originator
REL. This Information Has Been Authorized for
Release to...
FGI� Foreign Government Information
(b)(3)
�
11
Approved for Release: 2020/03/16 C05830442
pproved for Release: 2020/03/16 C05830442
I.
-Tryp-Seraet
MOT7517177t7CV777771t1
NIE 73/5-84/CX
TRENDS IN SOUTH AFRICA'S
NUCLEAR SECURITY-POLICIES
AND PROGRAMS
SC403/01-84
MIK
Information available as of 25 September 1984 was used in
the preparation of this Estimate, which was approved by the
National Foreign Intelligence Board on 2 October 1984.
� WARNING: The material in this document is sensitive.
Distribution of this Estimate should be strictly limited to
those officials who require access to the subject matter for
the performance of their duties.
(b)(3)
(b)(3)
Approved for for Release: 2020/03/16 C05830442
pproved for Release: 2020/03/16 C05830442
S Vg.1
NOFORiv/NuLuilTrACT1012C-ON7LPRCIZEL
THIS ESTIMATE IS ISSUED BY THE DIRECTOR OF CENTRAL
INTELLIGENCE.
THE NATIONAL FOREIGN INTELLIGENCE BOARD CONCURS.
The following intelligence organizations participated in the preparation of the
Estimate:
The Central Intelligence Agency, the Defense Intelligence Agency, the National Security
Agency, and the intelligence organizations of the Departments of State and Energy.
Also Participating:
The Assistant Chief of Staff for Intelligence, Department of the Army
The Director of Naval Intelligence, Department of the Navy
The Assistant Chief of Staff, Intelligence, Department of the Air Force
The Director of Intelligence, Headquarters, Marine Corps
SC 03101-84
--T-erp�Seepet�
NM
Approved for Release: 2020/03/16 C05830442
pproved for Release: 2020/03/16 C05830442
1
�1Z5PZ57/7/FIC7Cri?7,77CC171:5
CONTENTS
(b)(3)
Page
KEY JUDGMENTS
1
DISCUSSION
5
Introduction
5
The Search for Security
5
The Nuclear Option
5
South Africa and Nonproliferation
7
South Africa's Nuclear Explosives Capability
8
Genesis and Development
8
Availability of Fissile Material
11
Foreign Suppliers
12
Taiwan and Israel
13
Prospects for South Africa's Nuclear Weapons Development
13
Size and Nature Over the Next Five Years
13
Considerations Regarding Nuclear Testing
13
Nuclear Test Plans � 1e
13
Political Considerations
15
Technical Considerations
16
Alternative Scenarios
16
Delivery System: Present Status and Future Prospects
16
Implications of South African Nuclear Weapons Development ,
17
Annex A: The Civil Nuclear Program
19
Annex B: Nuclear Test Capabilities
23
Annex C: South African Delivery System Alternatives
29
SC 03101-84
Lamm= im
CIF CC!
(b)(3)
Approved for Release: 2020/03/16 C05830442
pproved for Release: 2020/03/16 C05830442
SC 03101-84
-rcr'zeffai
NOFORN/NULc." KALT /ORCON/PRONN/IVN1NTIL
KEY JUDGMENTS
We believe that South Africa has the capability to produce nuclear
weapons on short notice. We believe�although we cannot confirm�
that South Africa has already stockpiled the components for several test
devices or first-generation nuclear weapons that use enriched uranium.
We believe that South Africa has developed a nuclear implosion
design and a gun-assembly design as
well. Given the size of its nuclear program, South Africa could stockpile
annually enough highly enriched uranium for two to four nuclear
explosives, depending on the design.-rarive-o4...
Since the furor that accompanied the discovery of the Kalahari test
site in 1977, South Africa has followed a policy of calculated ambiguity
with respect to its nuclear options by intimating that it has the
capability to produce nuclear weapons while disavowing any interest in
doing so. Such a policy allows it to avoid the intensified pressures for
safeguards and the sanctions that would inevitably follow any revelation
of a weapons capability. At the present time, Pretoria appears satisfied
with that policy and is likely to adhere to it through the five-year time
frame of this Estimate.
There is no hard evidence about the degree to which South Africa
has incorporated nuclear strategy into its military planning. The South
African defense force's clear-cut conventional military superiority in
the region makes a nuclear strike capability unnecessary. Nuclear
weaponry would have little effect against the principal military
threats�guerrillas operating from domestic or border-state rural bases
and insurrection in black urban areas. Moreover, the sense of urgency
that prevailed in the 1970s, largely because of the perceived Soviet
threat, appears to have diminished
Much of the South African capability is based on work carried out
in the mid-1970s prior to then Prime Minister Vorster's reported
direction to halt the program. However, it is reasonable to assume that
research and development has continued, with the possibility that
emphasis may have shifted more toward fabrication and delivery
systems rather than continued stockpiling of uranium and components.
Although South Africa already has several types of aircraft capable of
delivering nuclear explosives, those aircraft are aging and are becoming
1
--Top-SeeFeS�
(b)(3)
(b)(3)
(b)(1)
(b)(3)
(b)(3)
(b)(3;
Approved for Release: 2020/03/16 C05830442
pproved for Release: 2020/03/16 C05830442
NOFORN/NOcons (/ORCQNIPROPINIWNINTEL
SC 03101-84
increasingly vulnerable to the air defense systems being provided its
adversaries bv the Soviet Union.-r7F-ive-ee)---
We have no direct evidence of cooperation in nuclear weapons
development between South Africa and Israel. The two countries
maintain a dialogue on nuclear matters, however, and past or future ex-
changes of nuclear-weapons-related research, technology, or materials
cannot be ruled out. There is evidence that Israel and South Africa have
collaborated on weapon systems that have the potential to serve as
delivery platforms for nuclear weapons.
We believe that South Africa is not likely to test a nuclear explosive
device during the next five years. The increase in tensions on the
continent and with the West that would accrue from a test of a nuclear
device would be greater than the political/military gains to be derived.
However, growing tensions between the United States and the Soviet
Union, if accompanied by perceptions that the United States was losing
ground to Moscow in Africa, or was losing interest in Africa, might
provide incentive to test as a caution to the Soviets. Any attempted
intimidation by Moscow on behalf of its African clients might provoke
the South Africans to conduct a test as a warning that Pretoria will not
be bullied.*(15),
There is still considerable disagreement within the Intelligence
Community as to whether the flash in the South Atlantic detected by a
US Vela satellite in September 1979 was a nuclear test, and if so, by
South Africa. If the latter, the need for South Africa to test a device dur-
ing the time frame of this Estimate is significantly diminished.TA"-/ers).
Revelation that South Africa possessed nuclear weaponry would
immediately exacerbate, the tensions that exist in southern Africa. Black
African states, supported by the Soviet Union, would seek stricter
sanctions against Pretoria and raise the specter of South African nuclear
blackmail to achieve regional domination. They might seek some form
of protective guarantee from their patron. Although Moscow would
almost certainly not offer any explicit nuclear commitment, it might
provide more sophisticated air defense systems and step up arms and
advisory assistance.
Revelation would put the United States in an awkward position.
Black African states would hold the United States at least partially
responsible for Pretoria's nuclear status and the United States would be
criticized for not restraining South Africa's nuclear progress. Converse-
ly, strong US denunciations of South Africa's nuclear weapons capabili-
ty would be perceived as evidence to Pretoria that the United States has
taken an anti�South African position. We do not believe that a harsh US
reaction would provoke Pretoria to cut off important minerals exports
to the United States, as some observers have contended.
2
T.op L'c,rot
(b)(3)
imommostaiminl IMMOMENSEENIMIIM11111
Approved for Release: 2020/03/16 C05830442
pproved for Release: 2020/03/16 C05830442
South African testing of a nuclear device would weaken the
international nonproliferation regime and encourage the acquisition of
nuclear weapons by other countries by demonstrating that indigenous
development of nuclear weapons can be accomplished. Several states of
proliferation concern might feel fewer inhibitions about developing
nuclear weapons or openly publicizing their nuclear capabilities if
South Africa suffered no serious international repercussions
South Africa's resumption of discussions with the IAEA concerning
safeguards for the semicommercial enrichment facility at Valindaba
and its commitment to request safeguards for its own nuclear exports
may be designed to mitigate future attacks on its IAEA membership.
South Africa will not have to make major concessions to the IAEA in or-
der to continue access to nuclear technology. While foreign assistance
was crucial to South Africa's nuclear explosives development in the
1970s, Pretoria's current need for such assistance is not great-D-1:77-4.w._
3
VP VLF
1M1
Approved for Release: 2020/03/16 C05830442
Approved for Release: 2020/03/16 C05830442
TVIJ SV.6.
NOFORN/NOLUN KAt.. / PAI
DISCUSSION
Introduction
The Search for Security
1. The rapid�and largely unforeseen�escalation
of conflict in southern Africa following Portugal's
withdrawal in the mid-1970s had a traumatic impact
on South Africa. Until that time, the Portuguese
colonies of Angola and Mozambique were the furthest
extensions of the so-called "white redoubt" and, along
with the white minority regime in Rhodesia, constitut-
ed a buffer against those liberation groups seeking
black majority rule in South Africa and Namibia. By
and large, South Africans felt secure within their
borders and confident of their ability to contain black
dissidence
2. South Africa's intervention in the Angolan civil
war of 1975-76 failed, however, to turn the tide in
favor of pro-Western liberation groups, and this
sparked a major reassessment of the country's military
capabilities and policies. The intervention showed that
South Africa lacked the resources to wage a protracted
military campaign. In addition to having been over-
whelmed by the magnitude of the Soviet and Cuban
involvement in Angola, military officials also identi-
fied serious deficiencies in materiel, communications,
and logistics...
3. Those deficiencies, set against the emergence in
Angola and Mozambique of Marxist-oriented, anti-
apartheid regimes heavily dependent on Soviet and
Cuban support, generated fears in Pretoria that those
former colonies would become staging areas for a
direct Soviet-backed invasion, perhaps involving Cu-
ban as well as black African forces. A subsequent rise
in guerrilla activity in both South Africa and Namibia
by guerrillas operating out of Angola and Mozambique
and an upsurge in racial demonstrations in South
Africa added to a sense of beleaguerment
4. Thus, by the late 1970s, South Africa saw itself as
standing virtually alone against a "total onslaught"
being waged against it by black insurgents and radical
black African states supported by the Soviet Union and
its allies. In response, South African leaders adopted a
"total national strategy" aimed at creating a self-
sufficient, flexible defense force capable of deterring
SC 03101-84
conventional as well as guerrilla threats. In a crash
effort to redress the military imbalance, top priority
was given to building up the conventional capabilities
of the armed forces and to developing an indigenous
arms industry capable of reducing�and eventually
eliminating�dependence on foreign sources of supply
(see inset on page 6).-1*....
5. The transfer of power to a black majority regime
in Rhodesia in 1980 came as another blow to South
Africa's sense of security. By then, however, the South
African military had adopted a "forward defense"
strategy based on increasingly aggressive conventional
and covert military operations against guerrilla bases
and local defense positions in neighboring states, par-
ticularly Angola and Mozambique. As a result, South
Africa has reasserted itself as the dominant military
power in southern Africa. It has forced Mozambique
to accept a mutual nonaggression pact and is exerting
pressure on Angola to send home the more than 30,000
Cuban troops stationed there. In addition, it has
convinced those governments to place restrictions on
guerrillas operating from their territorieilt4s
The Nuclear Option
6. The ominous military situation that prevailed in
the mid-1970s also induced South Africa to accelerate
a nuclear explosives research and development pro-
gram that had been formally launched in 1973. That
program, like the conventional military buildup, has
shown major accomplishments. By 1977, South Africa
had constructed a nuclear test site in the Kalahari
Desert and appeared well on its way toward testing a
nuclear explosive device.TrArigre-oc.,1_
7. The international outcry that followed the dis-
covery of the Kalahari site persuaded Pretoria to
abandon any plans it may have had for a test and to.
impose tighter security over its nuclear explosives
programs. A mysterious flash in the South Atlantic
Ocean in 1979 raised fears in the international com-
munity that Pretoria finally had tested an explosive
device, although there is still strong disagreement
within the Intelligence Community as to whether a
test actually took place (see annex Brtsi...
5
up SLrt
�
Approved for Release: 2020/03/16 C05830442
Approved for Release: 2020/03/16 C05830442
op Secret
NOFORN/NOCONTRACINK. ()iv/
The Regional Military Balance
In relation to its neighbors South Africa is a super-
power, although it falls far short of that status by
Western or Soviet standards. The South African defense
force is made up of 76,500 active-duty personnel
including 28,000 regular forces. In all, South Africa
probably could muster a force of approximately
400,000 soldiers when ready reserves, home guard units,
arid older males liable for service in ease of national
emergency are added
The government-controlled Armaments Corporation
of South Africa, with assets now estimated at $1.2
billion, currently meets the bulk of the military's
requirements for ground force equipment. It produces
small arms and ammunition, artillery and rockets,
armored vehicles, tactical communications equipment,
and landmines. It also modifies and upgrades aging
equipment such as the British-supplied Centurion tank
of World War II vintage. Pretoria still faces serious
problems, however, replacing its aging fleet of fighters,
bombers, and reconnaissance aircraft-re).
In contrast, the black armies in the region are ill-
trained, inadequately supported, and poorly led. Mak-
ing the transition from bush fighter to conventional
soldier has been extremely difficult. Angola has some
30,000 to 35,000 men under arms, and Mozambique has
close to 22,000. The majority in both countries, howev-
er, are engaged in counterinsurgency operations, for
which they are inadequately prepared, or in local law
enforcement
Zimbabwe's armed forces, once second in ability only
to those of South Africa, now number about 45,000 but
have been weakened since independence in 1980 by the
loss of white professionals, desertions, political interfer-
ence, and the inevitable frictions arising from tribalism
and the largely failed integration of rival guerrilla
elements''....
While South Africa has an indigenous arms industry,
Angola and Mozambique must depend on Soviet-sup-
plied weaponry. Both countries have more battle tanks
than South Africa and together have more (and more
advanced) jet fighter aircraft. However, neither country
has the personnel trained to operate or maintain the
equipment, and the flow of spare parts is unreliable-le).-
The only weaponry in the Angolan inventory that
limits South African operations within Angola are the
sophisticated SA-8 and SA-9 surface-to-air missiles pro-
vided by the Soviet Union in mid-1983. This modern air
defense system is manned by Soviet and/or Cuban
personnel at bases 150 to 200 miles north of the Angola-
Namibia border.`(v),.
Sc 03101-84
� 8. The Soviet Union's projection of power into
southern Africa in the mid-1970s may well have been
the key factor in Pretoria's decision to step up its
nuclear explosives research and development. South
Africa's leaders long have identified the Soviet Union
as their country's major adversary. With Moscow
having once again (as it did elsewhere in Africa in the
1960s) brought the great-power rivalry to the conti-
nent�and with it the perceived threat of an invasion
of South Africa�the added protection of a credible
nuclear deterrent took on a new urgency.' 9. South South Africa's readiness to move ahead with both
conventional military buildup and nuclear explosives
development has also been motivated by the belief
that it cannot depend on the West. That belief�along
with the Soviet threat�will dominate South African
defense thinking through the 1980s and probably
beyond.77.7-Nrc-oti-
10, Prior to the 1970s, South Africa's leaders ex-
tolled the idea that their country was of strategic
importance to the West given its location and mineral
resources, Because of that importance, they argued, it
was a major Soviet target. Thus, South Africa could
count on Western assistance in the event of Soviet-
inspired aggression, Western criticism of apartheid
notwithstanding. In 1976, however, following the ill-
fated intervention in Angola, Defense Minister (now
Prime Minister) Botha, warned that realities had
changed and that South Africa no longer could count
on such assistance. That warning was based to a large
degree on the widespread belief among South African
military planners that their intervention in Angola
could have succeeded if the West had been willing to
support
11. Uneasiness over the West's reliability was rein-
forced by its outcry over the Kalahari test site and the
mandatory arms embargo against South Africa passed
by the United Nations. In addition to calling for an
embargo against conventional military assistance, the
embargo also called for a ban against UN member
states cooperating with South Africa in developing
nuclear weapons. The West's refusal to veto the
embargo deepened South African suspicions.'"*�
'The present director of the National Intelligence Service, Dr.
Barnard, wrote extensively on the strategy of nuclear deterrence
while a university professor in the 1970s. He urged that South Africa
develop a nuclear weapons capability and demonstrate it to the
world. He justified his position by observing that Moscow would
have few reservations about using nuclear weapons against South
6
-T�ep-.Seer4t�
�
Approved for Release: 2020/03/16 C05830442
Approved for Release: 2020/03/16 C05830442
Tup
NOFORN/NoL KACT/ORCON/PROPINIWNINTEL
12. Since then, South Africa's defense policy�
including nuclear policy�has been driven by a strong
desire for self-sufficiency and a fear of Western
interference. At the same time, it follows a policy of
calculated ambiguity by which it intimates that it has
a nuclear weapons capability while disavowing any
interest in testing a nuclear device,
South Africa and Nonproliferation
13. South Africa, which has been a charter member
of the International Atomic Energy Agency (IAEA)
since '1957, has been reluctant to make major interna-
tional legal commitments that would restrict its free-
dom to develop nuclear technology. It is probable that
its refusal to adhere to the Non-Proliferation Treaty
(NPT) is in large part due to the fact that the treaty re-
quires signatories to place all fissile material under
IAEA safeguards.'
14. Pretoria's unwillingness to bend to outside pres-
sure from Western supplier states concerning safe-
guards requirements was perhaps best illustrated in
the early 1980s when the South Africans had to secure
nuclear fuel for the first of the two Koeberg power
reactors. Faced with a joint US-French strategy de-
signed to force them to accept comprehensive safe-
guards in return for guaranteed fuel supply, the South
Africans canvassed the uranium market and succeeded
in obtaining a sufficient supply of low enriched urani-
um from a number of alternative sources, including
China:TN.,
15. Pretoria's position concerning IAEA safeguards
on specific South African nuclear facilities has been
more flexible. It has acceded to requests from supplier
states for such controls involving imported nuclear
materials, equipment, and technology. For example,
the country's only research reactor at the Pelindaba
nuclear research center has been under IAEA safe-
guards since 1965 because it utilizes US-origin en-
riched uranium. (US supplies were discontinued in the
mid-1970s.) Likewise, the two French-built Koeberg
power reactors are open to IAEA inspection
16. In January 1984 South Africa announced that it
would be willing to resume discussions with the IAEA
concerning safeguards for the indigenous semicom-
mercial uranium enrichment facility at Valindaba.
The motives behind this announcement are not clear.
This gesture was not the first time Pretoria has shown
a cooperative attitude. On a number of occasions
between 1973 and 1977, the South Africans demon-
strated an interest in sharing knowledge about their
unique aerodynamic enrichment process with other
countries and a willingness to place the projected
commercial facility under IAEA safeguards.2 Howev-
er, persistent and successful efforts by the black
African states and India after 1977 to restrict South
Africa's ability to participate in IAEA activities rein-
forced a South African proclivity toward a siege
mentality. As a result, Pretoria saw no incentive to
cooperate on safeguards issues with an international
organization that appeared to have become captive of
South Africa's enemies'.,,
17. The controversy with regard to South Africa's
status within the IAEA has become less intense in
recent years but the problem still remains. It is possible
that Pretoria's strategy is to head off future attacks on
its IAEA membership. There have been several UN
resolutions introduced during the past year calling on
the IAEA to terminate all contact with South Africa
until it accepts comprehensive safeguards, Several
Western nations, including the United States, have
urged that no action be taken in view of the resump-
tion in early August of South African-IAEA discus-
sions concerning the possible application of IAEA
safeguards to the semicommercial enrichment facility
at Valindaba,
18. Pretoria's public commitment in the January
1984 announcement to request IAEA safeguards for all
of its own nuclear exports can be viewed as another
positive development justifying a more restrained
approach in dealing with the South African problem.�
19. The actual technical difficulties in reaching a
safeguards arrangement for the enrichment facility
are considerable and will ensure that discussions will
The South African aerodynamic enrichment process is similar to
the West German Becker nozzle method. But South Africa asserts
that its secret process, though similar in concept, is unique'
In that announcement Dr. J. W. L. DeVilliers stated that,
"South Africa will conduct and administer its nuclear affairs in a
manner which is in line with the spirit, principles, and goals of the
Non-Proliferation Treaty (NPT) and the Nuclear Suppliers Group
Guidelines... . This would specifically include: I. South Africa will
not sell uranium to nonnuclear weapons countries without Interna-
tional Atomic Energy Agency or Euratom safeguards; H. South
Africa will not make available sensitive technology to any other
country without Agency or Euratom safeguards; III. South Africa
will not sell enriched uranium or nuclear equipment without
Agency or Euratom safeguards." Dr. DeVilliers concluded by stating
that any sales of such material or equipment must have guarantees
that it will only be used for peaceful purposes
7
(b)(3)
Sc 03101-84
Approved for Release: 2020/03/16 C05830442
ONMEMEMINIMMINIONINIONEMIMApproved for Release: 2020/03/16 C05830442
�-3
be protracted.' The enrichment process utilized by the
South Africans has unique features that Pretoria will
wish to protect as a commercial secret.
20. The South Africans are acutely aware of the
time it will take to resolve the safeguards issue and
may well have calculated that their gesture was a
relatively costless way to buy time and more protec-
tion against further attacks on their IAEA membership
status. If the black African states and others continue
their effort to expel Pretoria from the IAEA, the South
Africans will have a legitimate rationale for breaking
off negotiations at any time. We doubt that there is
some other vested interest that would persuade Pre-
toria to remain at the negotiating table under such
circumstances. The current South African need to
import nuclear fuel or technology from the United
States, or other supplier states with rigorous nonprolif-
eration policies, is not great. If necessary, Pretoria can
probably fulfill its requirements through a number of
channels without having to make major concessions to
the IAEA on safeguards issues.
South Africa's Nuclear Explosives Capability
21. Evidence that South Africa has developed a
significant nuclear explosives capability is substantial
and compelling. There are reliable reports of nuclear
explosives research and development, and there are
several facilities in South Africa, such as the Valindaba
uranium enrichment pilot plant, which have a direct
relevance to the development of nuclear explosives
(see figure 1).-Trarive-or.,1_
22. We are reasonably confident of the existence of
an active nuclear explosives program dating back to
the early 1970s. The program envisioned testing, and
efforts proceeded in that direction until the Kalahari
test site was discovered by the Soviet Union in 1977.
There are reports that South Africa in fact intended to
test a nuclear explosives device before 1977 but was
delayed by a lack of fissile materiarrVrtge-e4._
23.
1978, we believe that
South Africa has developed two nuclear explosives
designs�for an implosion device
and for a gun-assembled device.
4A serious proliferation concern is whether IAEA safeguards
coverage can be extended in some manner to the pilot enrichment
facility at Valindaba. The objective would be to prevent low
enriched uranium from the commercial plant from being used as
"feed" for the production of highly enriched uranium outside
safeguards in the pilot facility
�.11.113P11.1111M611.77b)(3)
Genesis and Development
24. According to open literature, South Africa could
have begun research on nuclear explosives devices as
early as 1968-69, when the Valindaba plant was being
designed and at least one South African scientist was in
the United States studying the application of peaceful
nuclear explosives. It was not until 1973, however, that
South Africa formally launched its nuclear explosives
program
8
the gun-assembly and
South Africa
the implosion
Research and development was turned
over to the Reactor Development Division (RDD) at
the Pelindaba nuclear research center.
25.
26.
pursued both
designs.
27. A group at Somerset West, near Cape Town, the
National Institute for Defense Research (now called
Kentron), can also be linked to the nuclear explosives
program. The institute basically is responsible for
research, development, and production of propellants
and high explosives and probably is focused primarily
(b)(1)
(b)(3)
(b)(1)
(b)(3)
(b)(1)
(b)(1)
(b)(3)
SC 03101-84
VIJ
�
Approved for Release: 2020/03/16 C05830442
IIIIMERIPEIMINEMINIPMNINIIIMENIFERIMApproved for Release: 2020/03/16 C05830442
(b)(3)
�TairSeorJaj
(b)(1)
(b)(3)
NOF R
Limmo
SC 03101-84
al
9
-1-6pleccet_
�
Approved for Release: 2020/03/16 C05830442
IIIIMPINIIMPIIIIIIIMMINIIIMApproved for Release: 2020/03/16 C05830442
NOFORN
Figure 2
Chronology Related to South Africa's
Nuclear Explosives Development
1948
1961
1968
Atomic Energy Act
Research began on
aerodynamic
enrichment process
Work on peaceful
nuclear explosives
Atomic Energy
Board formed
1970
Public announcement of
uranium enrichment capability
1973
1975
� Formal nuclear explosives
program launched
Pilot plant
began operation
1977
1978
Minister Vorster's decision
Kalahari nuclear test
site discovered
Kalahari test site mothballed,
but not decommissioned
Prime to,
halt nuclear explosives development
Optical signature detected by US
nuclear test detection satellite over
South Atlantic Ocean
� Not to scale
303781 1084
1044
SC 03101-84
10
VP CI.
Approved for Release: 2020/03/16 C05830442
pproved for Release: 2020/03/16 C05830442
on developing the conventional explosives system for a
gun-assembled and/or implosion device.lrgr-ive-4:4_
29. The Intelligence Community has not identified
full-scale nuclear-related high explosives testing any-
where in South Africa. There could have been a series
of such tests at unidentified, perhaps remote, locations
in South Africa that was not detected. Such testing is
necessary in developing implosion devices, but not for
gun-type devices, and involves the detonation of nu-
clear device mockups that employ a surrogate�usual-
ly natural uranium or depleted uranium�for the
fissile material in an actual nuclear explosives device.
These tests can provide assurance that the compression
needed for device detonation and desired yield is
attained. (See annex B.)-(rgritte-sic,1_
Availability of Fissile Material
30. The Valindaba pilot scale uranium enrichment
plant was crucial to South Africa's development of a
nuclear explosive device. The plant uses a unique
aerodynamic enrichment process developed from re-
search that began in 1961. Pretoria asserts that it is a
proprietary process and has refused to subject it to
international nuclear safeguards inspection for fear
that the design and operating data for it will be stolen.
TrAr-Pte-ar..,1_
31. Contrary to South African claims of its peaceful
use, we have reliable reports that Valindaba is de-
signed for weapons-grade uranium production, and in
significantly larger Quantities than needed for fueling
its existing research reactors. These reports also indi-
cate that technical problems initially hindered Valin-
daba from operating at its peak design capacity, which
is about 100 kilograms per year of 95-percent enriched
uranium.lrgT7r-oe44.4_
SC 03101-84
33. Valin-
daba plant has continued to operate since 1977, not-
withstanding then Prime Minister Vorster's reported
decision that year to curtail nuclear explosives devel-
opment. Thus, it seems plausible that stockpiling of
fissile material has continued since 1978. South Afri-
ca's SAFARI-1 research reactor is the only known civil
consumer of highly enriched uranium, but its design
requirement for HEU is less than one quarter of
Valindaba's designed annual production capacity. It is
difficult to assess to what extent HEU has been
stockpiled at Valindaba, but, given the size of the
facility and the excess production capacity, South
Africa could stockpile annually enough HEU for two
to four nuclear explosives, depending on the device
design. We cannot determine, however, how much
HEU has actually been produced to date. Technical
problems have limited production levels in the past
and may have continued to do so7rirrrre-e43.)--
34. South Africa currently is building a semi-com-
mercial-scale enrichment plant for producing low
enriched uranium for the two Koeberg power reactors.
The plant is theoretically capable of producing HEU,
but it could better enhance Valindaba's HEU produc-
tion capacity by providing low enriched uranium feed
to the pilot plant. The capacity of the semi-commer-
cial-scale plant will be about 30 percent higher than
the needs of the Koeberg reactors. This excess could be
used for export or making up for unscheduled plant
shutdowns.TrAT-w...
35. Although enriched uranium is the only known
source of fissile material in South Africa, a plutonium
capability could be added if South Africa developed
indigenous reprocessing and/or reactor technology (see
inset on page 12). The presence of such facilities would
not in itself constitute a clear indication of intent to
use plutonium for nuclear explosives, but it would
represent access to technology necessary for such
development-
Trivp-mc..
(b)(1)
(b)(3)
Approved for Release: 2020/03/16 C05830442
0111.11.mumiggiiimmigraWpproved for Release: 2020/03/16 C05830442
Tup Sam, v
Not-cnovirvOCONTRACT/ORCONIPROPINIWNINTFE
Heavy Water Technology and Plutonium Production
TS-177TInVe-04.__
(b)(1)
(b)(3)
(b)(1)
South Africa once had plans to develop natural
uranium power reactors. Those plans were abandoned
in the 1960s in favor of light water reactors, purported-
ly because of economics. Thus, South Africa's recent
interest in heavy water technologies could reflect
changed economic conditions that favor natural urani-
um reactors because of the relatively high cost of
indigenously produced enriched uranium, and concerns
about energy self-sufficiencyThemac)._
In comparison to the enrichment issue, we believe
that South Africa would have a less credible argument
for rejecting nuclear safeguards either on the heavy
water plant or on an indigenous plutonium production
reactor. These technologies do not have as much propri-
etary sensitivity and uniqueness as does the South
African enrichment process,1rirrive-Qc,t_
Interest in heavy-water may portend yet another
technology�spent fuel reprocessing. The South Afri-
cans have shown interest in this technology for several
years, and every country of proliferation concern that
has added heavy-water-based natural uranium reactor
technology to its program also has shown strong interest
in reprocessinernrrifs-c&
38. Thus, we believe that South Africa has the
capability to produce nuclear weapons on short notice.
We believe�although we cannot confirm�that South
Africa has already stockpiled the components for
several test devices or first-generation nuclear weapons
Ts""ITP-Tfe-Qc,,L
Foreign Suppliers
39. Foreign suppliers, predominantly West Europe-
an, have been important to South Africa's nuclear
program from its inception. The development of South
Africa's nuclear explosives program would not have
been possible without foreign technical assistance dur-
ing the 1970s,71s-x).
40. Despite embargoes and periods of political con-
troversy, foreign firms have exported equipment and
technology used in both the nuclear weapons develop-
ment program and the associated uranium enrichment
pilot plant.
These imports, consisting mostly of specialized
laboratory or industrial equipment and nuclear-quali-
ty materials, serve the interests of foreign suppliers
searching for markets and South African nuclear
program managers facing limitations on domestic ca-
pabilities.re-c&
41.
SC 03101-84
12
-Tvp-rseeret-
S Kr�Nc.142.
Approved for Release: 2020/03/16 C05830442
NIIIIIIIIIMINIMPUOINIMiumiimmosnApproved for Release: 2020/03/16 C05830442
(b)(1)
(b)(3)
(b)(1)
(b)(3)
(b)(3)
(b)(1)
Tep *Sea
Nor u7117/11t7C A904741246:4440A44441F-L-
43.
45. The South Africans have had little difficulty
acquiring materials and technology essential to their
nuclear weapons development program. There are a
number of reasons for this. First, the program was
very secret for most of its history, and nonproliferation
concerns were lower during the 1960s and early 1970s
when many important acquisitions took place. For
example, during that period the South Africans ac-
quired an IBM 370 computer for Pelindaba from the
United States. This computer could have been quite
important to nuclear weapons design calculations. The
export of an equally state-of-the-art computer today
would meet with stiff resistance
46. Second, South Africa has a large scientific and
industrial infrastructure that makes a large number of
nonmilitary purchases every year. Thus, some nuclear-
related purchases do not attract the attention they
would if made by other countries. Moreover, the South
Africans can easily purchase equipment through front
organizations.
Taiwan and Israel
47. we have no direct evidence of cooperation in
nuclear weapons development between South Africa
and Taiwan. The two countries, however, do have a
cooperation agreement on civil nuclear technology
such as uranium ore processing and nuclear medicine.
SC 03101-84
(b)(3)
There is evidence that Israel and
South Africa have collaborated on weapon systems
that have the potential to serve as delivery systems for
nuclear weapons.o.c.,1_
Prospects for South Africa's Nuclear
Weapons Development
Size and Nature Over the Next Five Years
49. There is little evidence to indicate the intentions
of the present South African Government with respect
to nuclear testing and weapons production over the
next five years. Pretoria's apparent standdown regard-
ing testing in recent years suggests that nuclear devel-
opment has met the country's military requirements,
at least to the extent warranted by current perceptions
of the strategic threat and of the risks involved in overt
actions such as testing. It also suggests that the diplo-
matic and political benefits of being seen as having
once prepared for a nuclear test are consi
as great as if a test had been carried out.�
outh Africa will not be constrained by econom-
ic considerations. During the 1970s, it made significant
progress in nuclear weapons technology
(b)(1)
(b)(3)
(b)(1)
(b)(3)
(b)(1)
(b)(3)
Given the availability of weapons-grade uranium, we
believe the additional costs for research and develop-
ment on weapons fabrication would be relatively
modest and easily absorbed in Pretoria's $16 billion
annual budget.Trn-P-wal.
Considerations Regarding Nuclear Testing
Nuclear Test Plans
52. There are strong indications that South Africa
was proceeding toward a nuclear explosives test in the
1970s, while the military conflict in southern Africa
was escalating. Along with its efforts to develop at least
two nuclear device designs, South Africa constructed
an underground nuclear test site (see figure 3) in the
'Prime Minister Botha stated in 1981 that he could not set the
minds of his enemies at rest by signing the Nuclear Non-Prolifera-
tion TreatytK
13
Tep
�
Approved for Release: 2020/03/16 C05830442
Approved for Release: 2020/03/16 C05830442
ik
Figure 3
South African Nuclear-Related Facilities
-Makgathkgach
. .
(enrichment
uranium
S 4 *enrichment facility
,,6ery Pretoria
Pelindaba nuclear 6i,6
Swaziland.
research facility *Johannesburg
Vogl Mluthane
E., Kalahari nuclear
-Yr test site
kjp,��on South'rYAfrica
0,010
South N
Atlantic �
Ocean
C-/
-"
Koeberg nuclk ear National Institute
power station \A or
,\ 1Defense Research
Cape Town�c-,. (Kentron)
�: Mosselbaai
Somerset \ltresL--L, ____s_...__,-:
�
'N..- ,/ Gourioua nuclear research
site (proposed)
""seefet]
Indian Ocean
0 190 200 Kilometers
0
100 200 Miles
(b)(3)
703343 (A04921)10-84
SC 03101-84
14
Lis
(b)(3)
mis
Approved for Release: 2020/03/16 C05830442
iMINIII11111411111111111111111111�0111111.111115151MNINApproved for Release: 2020/03/16 C05830442
(b)(1)
(b)(3)
(b)(1)
(b)(3)
�1-615-5eeret.
NOFORN/
Kalahari Desert, 80 kilometers north of Upington.
Development of the site
began as early as 1974. By the end of 1975, the
geology of the site was sufficiently understood to make
sure that no radioactivity would seep out. Logistics
structures were constructed in 1975-76. 'Pit-e�4._
53.
Trictrige-94.
54.
55. In August 1977, while the South Africans were
still awaiting an adequate supply of fissile material,
the site was identified on satellite imagery by the
Soviets, who publicized its existence. The features of
the site (emplacement area, instrumentation area,
housing, and support area), the security imposed by
the South African military,
provide a strong case that the site was
being prepared for a nuclear testirrrive-es-sAai__
Political Considerations
56. Since 1977 South Africa has followed a policy of
calculated ambiguity with respect to the nuclear op-
tion by intimating that it has the capability to produce
nuclear weapons while disavowing any interest in
doing so..),
57. Such a policy holds a number of benefits,
particularly for a pariah state such as South Africa. It
forces Pretoria's adversaries to assume that , South
Africa has a weapons capability and to factor that
assumption into their policy formulation. For example,
even though Moscow need have no fear of South
Africa's ability to launch a nuclear strike against the
Soviet Union, it must take into consideration the
damage South Africa could inflict on the Soviet
Union's African clients, as well as on Soviet and Cuban
garrisons in Africa.1317P-tte-oc,L
58. Calculated ambiguity also allows South Africa
Lesimmi to avoid the intensified pressures for nuclear safe-
SC 03101-84
guards and sanctions that would inevitably follow any
open display of a weapons capability. Under present
policy it can proceed with research and development
fairly confident that the court of world opinion lacks
the evidence to bring a case against
59. The assumption on the part of its adversaries
that South Africa has a nuclear weapons capability also
gives Pretoria a deterrent credibility while allowing it
to avoid the stigma of being the first to introduce
nuclear weapons on the African continent. That intro-
duction also might prompt neighboring black African
states to seek protection under the Soviet nuclear
umbrella or�though less likely�to allow deployment
on their territory of tactical nuclear weapons superior,
no doubt, to any South Africa might have7irwc.)...
60. There is no hard evidence about the degree to
which South Africa has incorporated nuclear strategy
into its military planning. The absence of such evi-
dence should not be taken to mean that South Africa's
military research and development is lagging, howev-
er. It may only mean that the military is more
concerned with its immediate conventional needs.
61. The South African defense force's clear-cut
superiority in southern Africa and the success of its
operations against its adversaries make a nuclear strike
capability irrelevant at the present time. In addition,
the cost and effort involved in maintaining and up-
grading the conventional capabilities of the military
and of acquiring, if possible, new and advanced
weaponry, particularly aircraft and tanks, strongly
suggest that a nuclear weapon is far down on the
military's shopping list."".}-
62. We believe that South Africa is not likely to test
a nuclear explosive device during the time frame of
this Estimate. At the present time, the sense of urgen-
cy associated with nuclear development in the 1970s
appears to have diminished. Moreover, the increase in
tensions on the continent and with the West that
would accrue from a test would be greater than the
political/military gains to be derived, particularly in
view of South Africa's military superiority in the
region and the conventional deterrent capability it
now enioysirtzr-1e-e6)--
63. There are circumstances which could induce
the South Africans to change their policy and test a
nuclear device. International considerations could lead
to a nuclear test far sooner than regional ones, at least
within the period of this Estimate. Growing tension
between the United States and the Soviet Union, if
accompanied by perceptions that the United States
15
�T-ep-Seeret�
�
Approved for Release: 2020/03/16 C05830442
_,Approved for Release: 2020/03/16 C05830442
(b)(3)
T F
NOFORN/NUCUN KAC tORCON/PROPIN/WNINTEL
(b)(1)
(b)(3)
was losing ground to Moscow in Africa, or was losing
interest in Africa, might give added weight to advo-
cates of nuclear deterrence and trigger a test.sal.
64. Stronger Soviet efforts to defend its African
allies either with conventional weapons or intimations
of stronger retaliation could force South Africa to
counter with a test and with proclamations that it will
not be bullied. A private Soviet message last Novem-
ber condemning South Africa's military policy toward
Angola, while not accompanied by threats of retalia-
tion, nonetheless was regarded by Pretoria as an
attempt at intimidation
Technical Considerations
65. Although we believe that the political consider-
ations governing whether South Africa conducts a
nuclear test are preeminent, there are, nevertheless,
significant technical considerations as well. These re-
late principally to reliability and performance of
nuclear explosives devices.7s."44.
66.
South Africa undertook development of both gun-
assembly and implosion designs. Even though gun-
assembly devices can be deployed to weapon systems
without a strong requirement that the nuclear explo-
sive be tested, there would still be some uncertainty
about the yield.' If South Africa wishes to have
confidence in an implosion design, its need to test
increases substantially. It would be possible to deploy
an implosion device without testing, but they probably
would have little confidence in its performance unless
they received outside assistance. We note that an
implosion design is much more efficient in use of
fissile material than is a gun-assembly design.
67. The relatively strong technical requirement for
testing of an implosion design provides additional
significance to the effort to resolve whether the 22
September 1979 event was, in fact, a nuclear test. If it
was, the South Africans may already have resolved
many of their basic uncertainties, and their technical
requirement for nuclear testing during the next five
years would be significantly diminished, Development
of more advanced weapons concepts such as boosting
and thermonuclear designs would require at least one
and probably more tests. For a boosted design, the
South Africans would have to test an unboosted ver-
sion to establish the necessary physics calibration
points to determine how it could be boosted. For a
thermonuclear design, tests would be required to
Even so, we note that the United States deployed its first gun-
assembly weapon in World War 11 without first testing the nuclear
explosive:1*s
SC 03101-84
16
characterize the implosion device performance and to
develop thermonuclear design conceptsis-mr*G-44._
Alternative Scenarios
68. Despite former Prime Minister Vorster's deci-
sion in 1977 to halt South Africa's nuclear explosives
program, it is possible that South Africa has leap-
frogged the testing phase and is concentrating on the
weaponizing and delivery of its nuclear explosives
device. Afrikaners are a contingency-minded people
and as such probably would prefer to have a deliver-
able nuclear weapon rather than be forced to develop
one hastily in the face of a worsening security situa-
tion.Intrr-me-Qs1_
69. The head of South Africa's National Intelli-
gence Service, Dr. Bernard, voiced that point of view
in 1979 when he publicly stated that by the time a
nuclear crisis really faces us, it will probably be too
late to make ourselves really prepared and to be able
to defend ourselves at that lever,
70. Another alternative scenario is that South Africa
is working on advanced design concepts such as
boosting or the thermonuclear device that was origi-
nally tasked in 1973. It is difficult to assess the
likelihood of this scenario because we do not know
enough about the reasoning behind South Africa's
initial decision in 1973 to develop such advanced
concepts.
Delivery System: Present Status
and Future Prospects
It is unquestionable that the Republic of
South Africa has the technological capacity to
manufacture nuclear arms as well as sophisti-
cated systems of delivery with the desired accu-
racy and penetration.
�(Commodore H. F. Nel, speech to the South
African Institute of Strategic Studies, 1980.)
(b)(1)
(b)(3)
111
Approved for Release: 2020/03/16 C05830442
11111.1ffarVimilimpiewspioned for Release: 2020/03/16 C05830442
.(b)(3)
Tu.,
NOFORN/NOLurvi mACi/ORCON/rROPIN/WNINTEI
(b)(1)
(b)(3)
72. Virtually all of South Africa's combat aircraft
can carry a nuclear payload. These include the Mirage
Fl AZ, the Mirage III EZ, the Buccaneer S Mk 50, the
Impala, and the Canberra B(I)Mk 12. We do not
believe the Impala is a sound tactical choice, however,
because of its small size and inferior performance.
This issue is discussed in detail in annex C7ts-s4_
73. In 1976, the South African Air Force (SAAF)
publicly announced that Buccaneer bombers had been
used for practicing nuclear delivery techniques. Using
conventional bombs, five Buccaneer bombers de-
stroyed a decommissioned 797-ton World War II
salvage ship off Cape Town. The bombers released
their bombs 3 to 5 miles away from the ship, then
pulled up sharply and veered away. The SAAF further
characterized the exercise as employing a -computer-
ized technique to deliver nuclear bombs and escape
the effect of the resulting explosions.".
74. The prospects for delivery systems other than
aircraft�that is, ballistic missiles or artillery systems�
are less likely during the period of this Estimate. The
extremely difficult physics problems associated with
artillery-delivered nuclear weapons militate against
successful development (see annex C)..(g-Ty<
75. However, there have been recent indications
that South Africa may be undertak-
ing ballistic missile development at Somerset West. If
this is so, it would probably take about five years to
develop an effective system. This assumes that there is
no assistance from other countries, such as France or
Israel. A missile delivery system would have a signifi-
cant advantage over aircraft in penetrating air de-
fenses. It also could serve to fill a future gap if South
Africa continues to have difficulty replacing its aging
aircraf t.17-Ril-rtr�c-e.9-3,1&4
Implications of South African Nuclear
Weapons Development
76. For the past few years at least, South Africa's
nuclear program has not been a major source of
contention in Africa nor has it had adverse implica-
tions for the United States except in relation to the
international nonproliferation regime. With the excep-
tion of the discovery of the Kalahari test site in 1977
and the mysterious flash in the South Atlantic in 1979,
South Africa's nuclear program has provoked little
condemnation. This is partly because of its low visibili-
ty and partly because Africans�who would be the
first to condemn�have been preoccupied with other
matters. The low level of open contention regarding
LSCi
03101-84
Pretoria's nuclear development has provided the Unit-
ed States with room to maneuver in international
forums
to attempt to bring
South African facilities under international nuclear
safeguards.1117-Nr-es)-
77. Revelation that South Africa possessed nuclear
weaponry would immediately exacerbate the tensions
in southern Africa. Black African states, supported by
the Soviet Union, would renew their calls for UN and
other sanctions against Pretoria. Those calls would be
more intense than previous ones in view of South
Africa's recent military operations in the region and
fears of South African nuclear blackmail to achieve
regional domination. Moscow's clients in southern
Africa might seek some form of protective guarantee
from their patron. While the Soviet Union almost
certainly would not offer any explicit nuclear commit-
ment, it might provide more sophisticated air defense
systems and step up arms and advisory assistance. In
addition, Moscow could be counted on to issue gener-
alized but ominous threats against South Africa and
those who were suspected of contributing to its nuclear
development:
78. Black African states would hold the United
States at least partly responsible for Pretoria's nuclear
status, particularly if the United States did not join in
the denunciations. In African eyes, the United States
has done little to restrain South Africa's military
operations in the region and the United States as a
nuclear superpower would be criticized for not re-
straining its nuclear progress as welrirrar)-
79. Revelation would put the United States in an
awkward position. The revived need to bring South
Africa under international nuclear safeguards and to
exert some control over its nuclear programs would be
a strong argument in favor of US-South African
nuclear cooperation. However, any such cooperation
would be viewed in Africa and much of the Third
World as contributing to South Africa's nuclear mili-
tary strength. Even the current minor level of US
assistance is rejected by many pro-Western African
states. Revelation would fire the drive for sanctions
and US disinvestment in South Africa.1s-Pi4_
80. On the other hand, strong US denunciations
would be additional evidence to Pretoria that the West
is undependable or, at the very least, that the United
States expected South Africa to meet the threats
against it with one hand tied behind its back. Many
South African officials long have believed that accept-
ing the safeguards demanded by the West puts smaller
nuclear countries at a distinct disadvantage. We doubt
that a harsh US reaction would provoke Pretoria to cut
17
Tap
Approved for Release: 2020/03/16 C05830442
11111111111161111111111Approved for Release: 2020103116C05830442
s
NOFORNmocvnir AlrerfeiReetVPRGPIN74,44=E1� .
off important minerals exports to the United States, as
some observers have contended, since economic reality
probably would prevail over South African pique. A
cutoff could become more possible, however, should
the United States actively seek to interrupt the flow of
nuclear-related equipment and technology to South
Africa."(sti.4.
� 81. South African testing of a nuclear device would
weaken the international nonproliferation regime and
encourage the acquisition of nuclear weapons by other
countries by demonstrating that indigenous develop-
ment of a nuclear weapons capability can be success-
fully accomplished. Several states (such as Pakistan,
Argentina, Brazil, and Israel) might feel fewer inhibi-
tions about developing nuclear weapons or openly
publicizing their nuclear weapons capabilities if South
Africa suffered no serious international repercussions
or technological setbacks."(r*-
18
Top `,:e,ret
Approved for Release: 2020/03/16 C05830442
I
pproved for Release: 2020/03/16 C05830442
ANNEX A
THE CIVIL NUCLEAR PROGRAM
1. Although South Africa has vast coal reserves and
exports significant amounts of coal, it saw a need for
nuclear power as early as the 1960s, This need arises
because South Africa has built its electrical power
plants where its coal is located. A number of the
developed regions in South Africa, most notably the
western Cape, are not near coal deposits, however, and
transporting coal by rail or transmitting power over
electrical lines, at distances sometimes up to 1,500
kilometers, is inefficient and prohibitively expensive.
2. Pretoria chose therefore to construct nuclear
power plants on the coast to ensure adequate cooling
water from the ocean. In fact, the recent drought in
South Africa has heightened South African interest in
additional nuclear plants since some conventional
power plants located inland have been forced to shut
down because of inadequate cooling water supply.')'.-.
3. South Africa's entry into the realm of civil
nuclear power began in 1976 when it signed an
agreement with a French-led consortium to supply
two 922-megawatt (electrical) light water reactor pow-
er plants for the Koeberg nuclear station near Cape
Town. The first of these reactors started up for initial
testing on 19 March 1984. When both reactors are
operating, they will contribute about 8 to 10 percent of
the country's electrical power needs.'
4. The problems encountered in obtaining foreign
enriched uranium fuel have Pushed Pretoria into a
program for self-sufficiency in nuclear fuel supply. It
is developing and/or constructing commercial-scale
facilities for uranium processing, enrichment, and
reactor fuel fabrication. These facilities, shown in
figure 4, should be available by the time current
foreign supply contracts are completedre-414_
5, We believe that, during the period of this Esti-
mate, South Africa will develop heavy water technol-
ogy and spent fuel reprocessing, and that it will begin
an indigenous nuclear reactor program.TM�rre-e4
Heavy Water Technology and Its Implications
6. South Africa
is negotiating for the purchase of a large (approximate-
SC 03101-84
ly 250- to 500-ton-per-year) plant, scheduled for com-
pletion in 1987-88. Such heavy water technology
presages development of natural uranium-fueled reac-
tors, which can be excellent producers of plutoniums
7. In the 1960s a South African effort to develop a
heavy-water-based power reactor was abandoned. A
competing effort on a process for uranium enrich-
ment, which would support a light-water-based nucle-
ar fuel cycle succeeded, however, and development of
the process began7nar4-....
8. South Africa's renewed interest in heavy water
technology could grow out of either civil or military
interests. A civil interest could derive from economics
and the drive for self-sufficiency. Military interests
could stem from a desire for smaller nuclear explosives
than comparable-yield uranium-based devices:TS-rev�.
9. South Africa may have determined that its en-
richment process is not economical enough to support
a domestic power program. Even though its pilot-scale
uranium enrichment plant has proved its unique
aerodynamic process on a technological level, it has
not done so on an economic (or commercial) scale.
Previous South African statements on the process have
indicated that a break-even point exists with respect to
plant size. In early 1982, a senior South African
official stated that a plant size of 3 million separative
work units (SWU) per year was needed to be competi-
tive in the world market.Trar-rae-rac,1_
10. At the present time, however, there is a large
worldwide oversupply of enriched uranium, and an
export market does not exist. Presumably for this
reason, South Africa scaled down the size of its plant to
300,000 SWU per year, or only enough to provide
refueling for the Koeberg reactors and to provide some
excess capacity. Given the lack of a world market to
support construction of a competitive enrichment
plant, we believe that the economics of self-sufficiency
for the light water fuel cycle have become less attrac-
tive than for the heavy water cycletricrr-fte-Qc.,1_
19
�
Approved for Release: 2020/03/16 C05830442
11111M1111111021.111111MIIMMIRIPMERIMmApproved for Release: 2020/03/16 C05830442
(b)(3)
Figure 4
South Africa: Nuclear Fuel Cycle Development
; Operating
-;
Under construction
--- 1 Research/development
Uranium
mining
1 Uranium ore
1i concentration
1
_J
\I/
1
1 Uranium 1
, r u, 1
),:, , 6
1 dioxide i conversion
1 1
1 conversion I 1
Heavy water
plant
303782 10.84
SC 03101-84
Uranium
enrichment
I Fuel
! fabrication
r Power
reactors
Spent fuel
reprocessing
Indigenous
reactor
development
20
(b)(3)
�
Approved for Release: 2020/03/16 C05830442
F1111�1111611.11111111111IMINEMIApproved for Release: 2020/03/16 C05830442
1
(b)(1)
(b)(3)
I
5
Spent Fuel Reprocessing
it South Africa
planned to undertake spent fuel reprocessing in the
late 1970s, At that time, scientists had studied various
reprocessing methods and had made plans for a plant
at Pelindaba for reprocessing Koeberg power reactor
f uel.-rs�grme-ec,L
12. planning
proceeded to the point where a site for the facility had
been selected and draftsmen were studying plant
specifications. In 1981
major new construction at Pelindaba would include
hot cells for Koeberg fuel reprocessing. Construction of
a building with the characteristics of a hot cell com-
plex is confirmed by satellite imagery. South African
scientists also have shown interest in computer pro-
grams that model reprocessing plant component per-
formance.-e�-..
13. There are other factors, however, that compli-
cate judgments on South African intentions regarding
reprocessing. Under the terms of the French-South
African agreement concerning the Koeberg power
reactors, all Koeberg fuel reiprocessing must be done
'outside of South Africa. The French reaffirmed that
provision in 1983. In addition, the South Africans have
announced the construction of a hot cell complex, not
SC 03101-84
11111
Approved for Release: 2020/03/16 C05830442
for reprocessing but for the examination of nuclear
fuel associated with fabrication development.
however; and South African intentions cannot be
determined at this time.7F7Trw.
14. If the South Africans decide to develop large-
scale reprocessing, we believe they are more likely to
do so at the new nuclear research center at the
Gouriqua site (near Mosselbaai) or elsewhere. The
Pelindaba site does not have the necessary space and
probably cannot be expanded.inTrwc-e6)--
Indigenous Reactors
15. South Africa has been interested in
reactor development since the 1960s.
South A rica is again
undertaking, or at least considering, such develop-
ment. These include solicitations for expertise
development of nuclear reactor components and
the existence in 1981 of a military experi-
mental uses reactor at Pelindaba. In 1983, 11 South
African scientists attended a US conference on reactor
technology (design). And more recently, a South Afri-
can official suggested the possibility of a research
reactor at the Gouriqua site..
21
Ip Ce... et
pproved for Release: 2020/03/16 C05830442
�rorSeGrast
NOFORN/NoLuN I KA CT/ORCON/fROPINIWNINTEL
ANNEX B
NUCLEAR TEST CAPABILITIES
(b)(1)
(b)(3)
(b)(3)
SC 03101-84
23
-TOTY-Sewa.t_
111111�111111111111111
Approved for Release: 2020/03/16 C05830442
pproved for Release: 2020/03/16 C05830442
NUI-U
(b)(1)
(b)(3)
SC 08101-84
7g-.<
The Kalahari Nuclear Test Site
Background
9. The Kalahari nuclear test site is located about 80
kilometers north of Upington in the Kalahari Desert.
The site's existence was discovered and publicly made
known by the Soviets in August 1977.
At the time of its discov-
ery, early preparations appeared to be under way for
an underground nuclear test. Subsequent to the inter-
national uproar created by the Soviets' announcement,
the site was mothballed, although not decommis-
sioned. A low level of activity has been observed by
satellite imagery since that time.TrWrire-es-xacbt)_
Site Description
10. The general area around the Kalahari facility is
well-suited to underground nuclear testing. The isola-
tion from important centers of activity, the relatively
low intrinsic value of the land, and the hard-rock
geology are all typical features sought in selecting a
nuclear test sitelt-rep-meri)._
11. The site itself is large, comprising a perimeter of
20 by 23 kilometers (or an area of about 450 square
kilometers). The facility consists of an operational area
and a support base, The support base generally is
upwind of the operational (test) areas as one would
expect at a test sitel'snar-saaa..
12. The operational area of the site in 1977 con-
tained two large-diameter holes (which since have
been capped) and what we believe was an instrumen-
tation area. A tower had been erected around one
emplacement hole.
24
11.11111111.11.11111
Approved for Release: 2020/03/16 C05830442
1111111111.0181111111.151111111�11111111111111191MMEMApproved for Release: 2020/03/16 C05830442
(b)(3)
C
'(b)(1)
(b)(3)
of the hole suggest the hole had a diameter of 1 meter.
The tower site was connected to the instrumentation
area by a straight trace. A large inflatable building was
erected at the instrumentation site.
The 22 September 1979 Event
The Context
16. While faced with the possibility that South
Africa conducted a nuclear test in 1979, we should
consider the following context.� In 1963, South Africa
acceded to the Limited Test Ban Treaty, which
prohibits atmospheric nuclear testing.
South Africa would be the principal though not the only
candidate for such a test.�4,,
SC 03101-84
We estimate that by 1979 Pretoria
could have produced enough of that fissile material for
a first nuclear device.rsThre-ce-ww.)_
The Event
17. Early in the morning of 22 September 1979, a
US Vela nuclear test detection satellite over the South
Atlantic recorded an optical time history characteristic
of a near-surface atmospheric nuclear detonation (see
figure 6). This Vela satellite also had a directional
sensor on board to pinpoint event location; that sensor
did not trigger on this event. The initial Vela indica-
tion suggested a nuclear test of less than 3.5 kilotons;
this was later revised to be 0.75 to 2.75 kilotons.
Between 1972 and the 1979 event, this particular
satellite had recorded 17 known nuclear events, in-
cluding some in the equivalent yield range of this
event. In all, Vela satellites had recorded 41 confirmed
nuclear events.."*�
The Search for Other Technical Evidence
18. A massive Intelligence Community effort was
launched to corroborate the occurrence of a nuclear
detonation through means other than satellite observa-
tions.
*ta),
20.
25
Top Sea..i
(b)(1)
(b)(3)
(b)(1)
(b)(3)
(b)(1)
(b)(3)
MB ins mommormiommummi moimmumna
Approved for Release: 2020/03/16 C05830442
-Approved for Release: 2020/03/16 C05830442
IUp Scuc
Figure 6
The 22 September 1979 Event
1
Equatorial Guinea
Sao Tome erld PrinCipe
\..1.406,1016.
South Atrica
(Wslris 860, �
Fa islands
.,!,(admittird (nett by U.K.. ,,-
ciaimod by Aruent loo)"
s,
South 'Pao. filo�
Ocean
South
At/an tic
Ocea r1
Circle represents area of possible nuclear event recorded
by the Vela satellite. \
703348 10-84
O'riate Edward
Islands (S. MO
26
SC 03101-84 Approved for �
for Release: 2020/03/16 C05830442
Somalia
Sum-A.111.
. , .�
Mmbique '
.4.Mad7ascar
\I nd
�--
Qcea n
Seuridery repire..1�00. I.
not nacaasatily authodtalfra.
-Seeret-
ummesummo......wmpagnimApproved for Release: 2020/03/16 C05830442
1
(b)(1)
(b)(3)
22. Given these
factors, it is not surprising that
environmental sampling in Africa, Antarctica,
aidAustralia also did not yield significant results. One
laboratory report that analysis of radioactivity in sheep
thyroids from Australia indicated a nuclear event was
investigated and found to be faulty on the basis of
incorrect instrumentation and measurement tech-
niques. Moreover, the same year (1979), that same
laboratory had published similar anomalous findings
(not correlated to any nuclear event) which the author
attributed to an unknown source of contamination.
Finally, the radioactive isotope purportedly detected
had been used in the same building that housed the
laboratory that analyzed the sheep thyroids.
Subsurface Signals
23. Analysis of data from AEDS seismic detection
stations did not reveal any signals associated with the
event. All seismic stations in the Southern Hemisphere
that had the capability to detect signals were queried
with no positive results. No evidence was found to
either prove or disprove association between hydroa-
coustic signals and the event.
SC 03'01-84
Resolution of Technical Data
27. Extensive disagreement and controversy contin-
ue to exist within the Intelligence Community about
the origin of the signals recorded by the Vela nuclear
test detection satellite on 22 September 1979. Two
outside panels have reviewed relevant data and
reached opposite conclusions. The Office of Science
and Technology Policy (OSTP) panel concluded that
the signals "probably were not from a nuclear explo-
sion" since they appear to be internally inconsistent
and could have an alternate explanation. The Nuclear
Intelligence Panel (NIP) stressed the peculiarity of the
two signals in comparison to those known from nuclear
explosions and believes that they "probably, but not
certainly" were from a nuclear explosion.
28. Barring an unforeseen breakthrough, it seems
unlikely that a definitive resolution of the 22 Septem-
27
-Ferp-Seeret�
�
Approved for Release: 2020/03/16 005830442
r1=1=IPENIMIIIMINIIIIMINIIIIIIIIIMMIMApproved for Release: 2020/03/16 C05830442
(b)(1)
(b)(3)
1.
ber event will be possible from existing technical data.
The difference in the assessments of these data flow
from different approaches to the problem and are
unlikely to change. The OSTP panel was composed
primarily of experimental physicists, who require a
rigorous standard of proof and quantitative consisten-
cy to be persuaded of nuclear origin. The NIP was
dominated by experienced senior figures in the US
nuclear weapons community who are used to dealing
qualitatively with signals from nuclear explosions and
are very conscious of potential methods for testing
clandestinely in ways designed to avoid detection
29. The Defense Intelligence Agency, stressing the
nuclear character of the signal, continues to believe
that a nuclear explosion certainly occurred.
SC 03101-84
Collateral Information
31. Despite continuing efforts since 1979 to collect
corroborative information from nontechnical sources,
there still is no information that would indicate that
the 22 September event was a nuclear test by South
Africa
32. Some reporting can be viewed as circumstan-
tially supporting such a conclusion, but it does not
provide unambiguous evidence
28
-Top-SoraeL
(b)(3)
(b)(3)
(b)(1)
(b)(3)
all
Approved for Release: 2020/03/16 C05830442
111.011MISMIIMIM=IIIIIMEIMINApproved for Release: 2020/03/16 C05830442
-ficrp-Seere4
(b)(3)
ANNEX C
SOUTH AFRICAN DELIVERY SYSTEM ALTERNATIVES '�
Aircraft
1. The aircraft in the current South African Air
Force (SAAF) inventory most likely to be selected for
the nuclear weapons delivery role are fighter-bombers
and light bombers. The fighter-bomber and light
bomber aircraft in the SAAF are the Mirage Fl AZ, the
Mirage III EZ, the Buccaneer S Mk 50, the Impala,
and the Canberra B(I)Mk 12. We do not believe the
Impala is a realistic alternative because of its small size
and inferior performance. Thus, our analysis of the
SAAF's nuclear-capable aircraft focused on the Mirage
F1, Mirage III, Buccaneer, and Canberra.-
2. Based on all performance aspects, we believe
that the best nuclear strike aircraft in the SAAF is the
Buccaneer. The Canberra has a greater combat radius
but was designed for clear-weather operations. The
Buccaneer is capable of unrestricted operations at
night and in adverse weather. Furthermore, the low-
altitude penetration speed of the Buccaneer is higher,
and the Buccaneer's superior avionics should enable it
to penetrate at a lower altitude." Although no infor-
mation was available on the electronic countermeas-
ures (ECM) of either aircraft, the Buccaneer ECM is
almost certainly superior to that of the Canberra. All
of these factors�the higher speed, lower penetration
altitude, and superior ECM�tend to make the Bucca-
neer more survivable than the Canberra. The Canber-
ra probably would not be selected in preference to the
Buccaneer unless:
� The weight and/or size of the nuclear weapon
exceeds the Buccaneer bomb bay weight or
dimension limits.
The selection criteria are such that the Canber-
ra's 200-nautical-mile combat radius advantage
over the Buccaneer is a critical factor....
'� For the purpose of this analysis we assume that South Africa can
presently produce a nuclear explosive device that uses uranium as its
fissile material, is 0.5 meter in diameter, and weighs 450 kilograms.
First-generation nuclear weapons are generally considered to weigh
several hundred kilograms, most likely on the order of 450 kilo-
grams. Such weapons should have a diameter of 0.5 to 1 meter.,),
" The Buccaneer is equipped with a Doppler radar, a search-and-
fire-control radar, and a head-up display radar. Canberra has none
of theset*�.
SC 03101-84
3. The survivability of the Mirage Fl and Mirage
III probably is comparable to that of the Buccaneer.
However, the inferior combat radius of the two Mi-
rage aircraft limits their employment to targets at
relatively short range. Even against short-range tar-
gets, the superior range of the Buccaneer would allow
it to follow a more circuitous route to the target. Such
flexibility permits an aircraft to bypass concentrations
of air defense weapons, to use terrain-masking to make
tracking and interception more difficult, or to ap-
proach the target from the most advantageous direc-
tion. Thus, we would not expect either Mirage aircraft
to be selected over the Buccaneer unless the selection
criteria are such that:
� The markedly superior combat radius of the
Buccaneer is not a critical factor.
� The SAAF believes that the superior over-target
speed of the Mirage aircraft will significantly
improve the mission's chances of success
Weapons Delivery
4. All four aircraft are capable of performing the
standard nuclear weapons delivery techniques. High-
altitude bombing, for example, with a free-fall or
parachute-retarded nuclear weapon should present no
problems, since all four aircraft have service ceilings in
excess of 10,000 meters. However, this delivery profile
may not be used in a high-threat environment since
low-level delivery makes the attacking aircraft less
vulnerable to air defenses. The four aircraft should be
capable of low-level bombing techniques�loft bomb-
ing (free-fall or parachute-retarded weapon) and level
laydown (parachute-retarded weapon)�with no un-
usual restrictions
Cruise Missile or Glide Bomb
5. South Africa
began developing a conventional warhead glide bomb
for use by the South African Air Force in the late
1970s. Its scheduled deployment was for the early
1980s. The weight and diameter of this glide bomb are
within the capabilities for a first-generation nuclear
29
op ecreT
Approved for Release: 2020/03/16 C05830442
pproved for Release: 2020/03/16 C05830442
Top 5u1
jN/FR8J'1ttjWflIHTL
The conventional glide bomb weighs 460
kilograms with a diameter of about 1.5 feet (0.5
meter). Its range reportedly is 60 kilometers and is
intended for carriage by the Buccaneer and Mirage Fl
aircraft. The bomb could not be carried by the Mirage
HI, reportedly because that aircraft's navigation sys-
tem was not sufficiently accurate for launching.
6. The South Africa Navy has Gabriel cruise mis-
siles on its Reshef patrol boats from Israel. The Gabriel
missile conventional warhead weighs 180 kilograms
and is about 0.3 meter in diameter. While the environ-
mental considerations for a cruise missile (tempera-
ture, acceleration, altitude) are similar to those for
aircraft delivery, the Gabriel is much too small to
carry the device postulated for South Africa. Modif y-
ing the design to accommodate a 60-percent reduction
in weight and a 30-percent reduction in diameter
would require an order of magnitude advance in both
design and fabrication technology over that required
to produce the original device.?')..
7. If the South Africans were to develop indigenous-
ly a cruise missile of sufficient size to carry the
postulated design, then the weaponization require-
ments would not be significantly more difficult to
master than for an air-delivered free-fall bomb. The
same conclusions apply whether the missile is surface
launched or air launched, or whether it is a glide
bomb. We believe the South Africans would have little
trouble in adapting their aircraft delivered design to a
cruise missile if no major weight and size reductions
are necessary.
Ballistic Missiles
8. There have been some indications of South Afri-
can interest in ballistic missiles. South Africa is produc-
ing an air-to-air missile similar to the Sidewinder and
short-range artillery barrage rockets. Research has
been done on antishipping missiles and on a surface-
to-surface missile. The nuclear potential of this last
project is of most concern, since it conceivably could
be under consideration as a nuclear delivery vehicle
itself or could lead to a more advanced nuclear-
capable missile. It should not be difficult to convert a
missile using a conventional warhead to a nuclear
warhead if the two warheads are matched in size,
weight, center of gravity, and moment of inertia. r*
Artillery
9. Nuclear artillery projectile design is very differ-
ent from that of aircraft or missile delivered nuclear
SC 03101-84
weapons. Artillery delivery requires that very high
acceleration be imparted directly to the projectile case
and presents formidable challenges to weapon design-
ers. Both interior and exterior ballistics considerations
must be included in the design. Center of gravity,
moment of inertia, size, weight, volume, and safety are
critical to provide the reliability and accuracy re-
quired for battlefield operaiions..
10. The South Africans manufacture their own 155-
mm field artillery guns and sophisticated extended-
range ammunition for them. This has led to specula-
tion that they may also be working on nuclear rounds
for these guns. Given the present known scope of their
program, however, we believe they would stand no
chance of succeeding in such an effort.-11"-*.
11. In general, a uranium gun-assembled device
will not fit a I55-mm shell. While it might be possible
to pack a critical (nuclear) mass of uranium in such a
small diameter, there would be no room left for other
essential nuclear components and no possibility of a
meaningful nuclear yield. It is conceivable that a
uranium implosion device could be fit into a 155-mm
shell, but the resulting device would be too long to be
ballistically stable (5 or 6 calibers). To make it work,
some sort of miniature guidance system would be
required, adding weight and reducing range. At best,
such a device might yield 100 tons or solar4.-
12. A plutonium implosion device can be built for
155-mm artillery. At present, however, South Africa
has no plutonium production capability, Even if it
were to acquire plutonium, we believe South Africa
probably would not succeed in developinz nu I
artillery.
13. It is possible the South Africans could design
and produce a larger caliber artillery piece and then
produce uranium gun-assembled devices for it. This
would be a difficult challenge, however, because South
Africa lacks some of the requisite artillery technology.
It probably does not have the conventional ordnance
for such a large caliber gun either, making it more
difficult to justify the required development expenses.
Foreign embargoes have made access to conventional
weapons technology difficult, and it would be very
expensive for the South Africans to develop a new,
very large field gun by themselves. If they did pro-
duce such a weapon, a uranium gun-assembled nu-
30
(b)(3)
in MINMEMEMMIlliii
Approved for Release: 2020/03/16 C05830442
pproved for Release: 2020/03/16 C05830442
NOI-URN/P4
Tap .1
clear round would be a fairly straightforward engi-
neering problem for them but, as noted, still a difficult
problem; and the resulting device would require very
large amounts of uranium for a very small yieldN.4_
Torpedoes
14. �The constraints placed on a nuclear explosive
device intended for use in a torpedo would be no more
severe than those for an aircraft-delivered bomb.
South Africa has a number of French-built submarines
equipped with torpedo tubes capable of firing 21.7-
inch (0.55-meter) torpedoes. We believe they would
have minimal difficulty adapting a 0.5-meter, 450-kg
SC 03101-84
device to torpedoes designed for these submarines.
The weight of the nuclear device does not exceed the
conventional warhead weight for such torpedoes, and
there is enough room for the fuzing and firing equip-
ment necessary to operate the weapon."..
15. The South Africans have a stockpile of conven-
tional torpedoes, and would not need to make their
own from scratch in order to get a nuclear-armed
stockpile. On the other hand, their torpedoes do not
have an extended range, and targets such as enemy
ports or amphibious task forces, against which a
nuclear torpedo might be used, could be more reliably
attacked from the air.11"%.4.
31
Scerct
1.111MMESMOIMM
Approved for Release: 2020/03/16 C05830442
(b)(3)
pproved for Release: 2020/03/16 C05830442
DISSEMINATION NOTICE
L This document was disseminated by the Directorate of Intelligence. Because of the
sensitive nature of some source material reflected herein, this copy is for the exclusive
information and use of the recipient only.
2. This document may be retained, or destroyed by burning in accordance with applicable
security regulations, or returned to the Directorate of Intelligence.
3. When this document is disseminated overseas, the overseas recipients may retain it for a
period not in excess of one year. At the end of this period, the document should be destroyed
or returned to the forwarding agency, or permission should be requested of the forwarding
agency to retain it in accordance with IAC-D-69/2, 22 June 1953.
4. The title of this document when used separately from the text is unclassified.
Approved for Release: 2020/03/16 C05830442