Pakistan's bomb:
Out of the closet
As the Soviet Union breathed its last and Iraq's
nuclear infrastructure was destroyed under U.N. auspices, the Bush
administration redoubled its efforts to stem regional nuclear proliferation
in the Middle East, on the Korean peninsula, and on the Asian subcontinent.
At the end of 1991, with the Cold War at an end and nuclear disarmament
efforts making progress in South Africa, Brazil, and Argentina, Western
diplomats hoped that India and Pakistan also might be persuaded to
abandon their nuclear ambitions. But
this March India rejected a U.S. State Department initiative for a
regional disarmament conference. And Pakistan, which in the past has
offered to sign the Nuclear Non-Proliferation Treaty (NPT) if India
agrees to do so, signaled that if it attended a U.S.-sponsored nuclear
disarmament conference its aim would be full diplomatic recognition
as a de facto nuclear weapon state. Such an outcome is unacceptable
to Washington.
Pakistani Foreign Secretary Shahryar Khan admitted
in a February 7, 1992 Washington Post interview that his country had
the components to assemble at least one nuclear bomb. Khan claimed
he was revealing this information to bridge a "credibility gap"
created by a previous Pakistani regime. But it is more likely that
his admission was motivated by a desire to gain recognition of Pakistan's
nuclear status, and perhaps the renewal of U.S. aid to Pakistan. In
1985, Congress made aid to Pakistan contingent on the president certifying
that Pakistan did not possess a nuclear weapon; aid was halted in
1990 when President George Bush could no longer do so. Shahryar
Khan's February announcement was certainly not news to experts who
have tracked Pakistan's clandestine program. Many sources, including
previously classified U.S. government documents, indicate that Pakistan
has been pursuing nuclear weapons for more than 15 years, and that
its seasoned uranium enrichment program is capable of producing enough
weapon-grade uranium to build as many as six nuclear bombs per year.
Uranium
Pakistan began trying to enrich uranium in
earnest in 1976, secretly establishing the Engineering Research Laboratories
(ERL). According to a declassified 1983 State Department memorandum,
gas centrifuge designs "were stolen by a Pakistani national"
from the European technology holder, Urenco, the gas centrifuge enrichment
consortium.(1) Most reports identify the agent as A.Q. Khan, a Pakistani
metallurgist. The memo says that ERL had both acquired and produced
components for gas centrifuges and for nuclear weapons. (In honor
of Khan's contributions, ERL was renamed the A.Q. Khan Research Laboratories
in 1981.) The centerpiece of Pakistan's enrichment program is the
Kahuta gas centrifuge plant near Islamabad. This facility began operating
in the early 1980s, but has been plagued by chronic delays. The
1983 State Department memo says, "We believe that the Pakistanis
have experienced difficulty in making their centrifuge machines work
and that the Pakistanis have not yet produced any significant quantities
of enriched uranium." The memo
also reports that Pakistan turned to China for help. The nature and
the extent of Chinese assistance is unclear, but China may have helped
Pakistan build centrifuges or cascades. One 1989 report stated that
China had supplied enough weapon-grade uranium to make two nuclear
bombs.(2) However, one U.S. official has said that this assistance
is unconfirmed. He said that China might instead have helped by supplying
unenriched uranium hexafluoride or "hex," the chemical form
of uranium used in centrifuges. Pakistan obtained its own uranium
hexafluoride production equipment from a German company in the late
1970s, but it could still have been experiencing difficulties in producing
hex in the early 1980s. In return for
its assistance, China is believed to have received Urenco technology
from Pakistan. China has had a gas centrifuge research and development
program since 1958. It is unknown if China has built a production
facility.
In 1984, A.Q. Khan announced that Kahuta was
producing low-enriched uranium, but would not enrich uranium above
the five percent level. However, U.S. intelligence concluded by mid-1986
that Kahuta was producing highly enriched, weapon-grade uranium.(3)
According to a 1986 memo prepared for Henry Kissinger, then a member
of the president's Foreign Intelligence Advisory Board, Kahuta had
the nominal capability to produce "enough weapons-grade material
to build several nuclear devices per year."(4) But the memo did
not indicate the actual amount produced. Before her visit to Washington
in June 1989, Prime Minister Benazir Bhutto reportedly stopped weapon-grade
uranium production, a step which the U.S. government was able to verify.(5)
But when Pakistan and India clashed over Kashmir in the spring of
1990, Pakistan resumed production and continued until some time in
1991, according to Shahryar Khan's February interview with the Washington
Post. Despite a wealth of public information on the qualitative aspects
of Pakistan's program, little precise information is available about
the amount of weapon-grade uranium Kahuta can produce each year, or
about the total amount it has produced so far. We can only estimate
quantities by considering the types of centrifuges Pakistan is believed
to have obtained in Europe, and by estimating the number it has operated
at Kahuta.
Stolen centrifuges
According to a Western enrichment expert, the
first Urenco centrifuge designs Pakistan built were probably based
on two first-generation prototype centrifuges designed by Ultra-Centrifuge
Nederland (UCN), the Dutch partner in the trilateral Urenco consortium.
These machines, the CNOR and SNOR, featured aluminum rotors, connected
by bellows. The bellows act to reduce vibrations caused by resonant
frequencies at certain operating speeds. Rotors that spin faster than
the first of these frequencies are called supercritical. Bellows in
supercritical machines allow for longer centrifuges, and thus, more
separation of uranium, but they are considered difficult to master.
CNOR and SNOR machines have an estimated separative capacity of 2
to 5 separative work units (a standard measure) per year. Intelligence
reports on the activities of Pakistani agents in the Netherlands in
the 1970s concluded in 1980 that a small number of CNOR and SNOR machines
were "spinning somewhere in Pakistan."(6) Other sources
report that Pakistan had trouble getting these machines to work on
a large scale and started replacing them with more reliable machines
based on two German Urenco designs, the G-1 and G-2. First-generation
centrifuges were also being replaced by improved production models
at the UCN plant at Almelo during the early and mid-1970s. Dutch intelligence
believes that Pakistan obtained design information for the newer centrifuges
in part through Khan's efforts-in 1974 UCN asked him to translate
classified design documents for the German centrifuges. According
to the statement of a senior German official, Pakistani agents obtained
centrifuge components and design information in Germany as well.
Following the 1991 Gulf War, the International
Atomic Energy Agency (IAEA) confirmed that Urenco design information
formed the backbone of Iraq's clandestine centrifuge program. German
officials then speculated that the same unknown German sources who
provided Pakistan with information on the G-1 and G-2 might have also
passed the information to Baghdad. A G-1 centrifuge has a capacity
of between 2 and 3 separative work units per year. The G-2s have an
estimated capacity of 5 to 6 separative work units per year. The comparatively
advanced G-2 machine is a supercritical centrifuge featuring two maraging
steel rotor tubes connected by a maraging steel bellows. Because
Pakistan encountered difficulties building and operating centrifuges,
it installed considerably more machines than it has successfully operated.
In 1986, Kahuta was reported to have 14,000 centrifuges (see June
1987 Bulletin). U.S. officials confirmed that Pakistan might have
built that many, but they estimated that only about 1,000 were actually
in operation. One official added that Pakistan's centrifuge "junk
pile is sizable."
The estimate of 1,000 machines in operation
is consistent with a 1986 report in the Muslim, a daily newspaper
in Islamabad. The Muslim reported that Kahuta was "rumored to
have 1,000 centrifuges, against a planned capacity of 2,000 to 3,000
centrifuges."(7) The 1983 memo asserted that Kahuta is "eventually
to house several thousand machines." One U.S. official we interviewed
in spring 1991 said that Pakistan was operating nearly 3,000 machines
at Kahuta. Pakistan now has the manufacturing capability and know-how
to increase the number of machines. But the official said that Pakistan
was concentrating on developing more advanced machines and replacing
older centrifuges rather than increasing the number in operation.
We believe that most of Pakistan's centrifuges are based on the G-2
design, although a significant number of the machines could be based
on less capable German and Dutch designs. Assuming a mix of types,
each with a capacity of between 3 and 5 separative work units per
year, Kahuta could produce about 9,000 to 15,000 separative work units
per year. This is enough to produce about 45 to 75 kilograms of weapon-grade
uranium a year, assuming that natural uranium is fed into the plant
and that about 0.3 percent of the uranium 235 is left in the waste,
or "tails." If Pakistan had sufficient uranium hexafluoride
stocks, it could accept a higher rate of waste. With a 0.5 percent
tails assay, Kahuta's annual production could be 60 to 100 kilograms
of weapon-grade uranium. Assuming that a nuclear device requires about
15 kilograms, Kahuta has the capability to produce enough weapon-grade
uranium for 3 to 6 devices a year. Kahuta, however, has not operated
at nominal capacity for most of its history. By the end of 1991, Pakistan
had probably produced between 100 and 200 kilograms of weapon-grade
uranium, based on a variety of tails assays and separative capabilities.(8)
This is enough material for roughly 6 to 13 nuclear explosive devices.
Pakistan is building a second enrichment
plant at Golra, about six miles west of Islamabad. A 1987 British
report implied that "several thousand centrifuges" would
be installed there.(9) According to a U.S. official, Pakistan has
had problems getting equipment from abroad, and therefore progress
in completing the facility has been slow and sporadic. Western intelligence
reports suggest that Golra is being used to test a small number of
advanced supercritical centrifuges before they are installed at Kahuta.
Reprocessing
One U.S. official interviewed in 1991 said
that Pakistan had completed a small reprocessing plant called "New
Labs" at the Pinstech complex near Rawalpindi. New Labs is based
on blueprints delivered by France, with key equipment bought from
a variety of suppliers. According to the 1983 State Department memo,
New Labs would need several years to separate enough plutonium for
a nuclear weapon. The memo adds, however, that New Labs seemed to
be large enough to allow for expansion of its reprocessing capacity.
Because Pakistan lacks a supply of unsafeguarded irradiated fuel,
raw intelligence reports have claimed-and experts have speculated-
that Pakistan has been trying to build a nuclear reactor that would
generate significant amounts of plutonium. Recently, more information
about Pakistani procurement was revealed in Bundestag committee hearings,
which included the testimony of the prosecutor in the 1990 trial of
German nationals who helped Pakistan to obtain illegal materials.(10)
Rudolf Ortmayer, a German engineer convicted in 1990 of having illegally
exported nuclear goods to Pakistan, testified before his trial that
the piping he supplied in the late 1980s was "for construction
of an indigenous primitive, pool-type reactor." The aluminum
fuel-cladding material Ortmeyer supplied would not have been usable
in Pakistan's existing reactors. Based on Pakistan's procurement of
a wide array of related equipment, and the revelation of its program
to produce large amounts of nuclear-grade graphite, the West German
intelligence agency, the Bundesnachrichtendienst, reported in late
1983 that Pakistan had probably begun to develop an indigenous reactor.
It has still not been established that Pakistan is building a reactor
for plutonium production. But Pakistan has had Chinese assistance
in building a tiny research reactor that contains about one kilogram
of Chinese-supplied weapon-grade uranium fuel that is under IAEA safeguards.
Without a much larger reactor, Pakistan could separate only tiny amounts
of plutonium from the small quantities of fuel it could legally withdraw
from safeguards. A U.S. official believes that Pakistan probably did
some experimental separation of plutonium.
Weaponization
Like Iraq, Pakistan was simultaneously developing
the ability to produce fissile materials while trying to master weapon
design and production. Its weapons design efforts began over 15 years
ago, and they required substantial foreign assistance. The 1983 State
Department memo says that "nuclear explosive design and development
work began in Pakistan soon after the 1974 Indian nuclear test."
The work was assigned to an organization
within the Pakistan Atomic Energy Commission, which studied implosion
hydrodynamics, neutronics, high explosives testing, and metallurgy.
The 1983 memo says that Pakistan subsequently worked on an electronic
triggering circuit for detonating the high explosives, and experimented
with shaped charges. It concluded that "Pakistan has already
undertaken a substantial amount of the necessary design and high explosives
testing of the explosive triggering package, and we believe Pakistan
is now capable of producing a workable package of this kind."
A dummy core of natural or depleted uranium would be used to test
the non-nuclear high explosive triggering or implosion package for
a nuclear weapon. U.S. officials have
said on many occasions since the early 1980s that Pakistan received
a proven weapon design from China. It has been reported that the design
was that used in China's fourth nuclear-weapons test in 1966 at Lop
Nor. This test involved the detonation of a warhead carried by missile.
If Pakistan received a copy of this design, then its warhead is probably
much smaller than early Chinese aerial bombs and could operate under
the more exacting conditions of delivery by missile.
In a recent interview, one U.S. official cautioned
that receiving a "cookbook design doesn't mean that you can make
a cake on the first try." It took Pakistan 10 years to enrich
uranium, even with Urenco centrifuge designs and extensive knowledge
about Urenco experts and suppliers. Pakistan would probably have required
several years to build bombs after getting the design. Little is known
about the particular Chinese design. A U.S. official said in a 1991
interview that it utilized a nearly solid sphere of weapon-grade uranium
with a tiny hollow core, surrounded by a tamper and high explosives.
Such a design would require about 15 kilograms of weapon-grade uranium.
Pakistan also got bomb components and test equipment in Europe and
the United States. According to the 1983 State Department memo, in
late 1981 through 1982 European companies sold Pakistani procurement
agents metal components that were "unambiguously identified as
those of a nuclear device." Pakistani agents also tried to get
precision lathes and associated equipment, specifically for the manufacture
of bomb components. Following U.S. government protests, Pakistan shifted
"from procurement of weapons components themselves to procurement
of machinery necessary for their manufacture."
One of the more difficult problems in building
a nuclear bomb is the neutron initiator, which starts the fission
chain reaction at the right moment by injecting a spurt of neutrons.
Recording a sharp neutron signal while testing an implosion package
would provide confidence that the bomb would work. Pakistan would
have several initiator options. All, however, require either an alpha-particle
source-most likely polonium 210-or tritium. In the first type, polonium
is mixed with beryllium at the right instant, and the alpha particles
from the polonium 210 interact with beryllium to produce neutrons.
A tritium initiator requires fusion of tiny amounts of tritium and
deuterium, resulting in high-energy neutrons. Tritium initiators are
generally considered more difficult to build than polonium-beryllium
initiators. (Iraq was pursuing a polonium-beryllium initiator, probably
for this reason.) But polonium-beryllium initiators have a great disadvantage.
The half-life of polonium is only about 140 days; tritium has a half-life
of 12 years. To keep weapons with polonium initiators functional would
require a constant resupply of polonium. Pakistan's most probable
source of polonium is safeguarded reactors, but producing polonium
for nuclear weapons in civilian reactors would violate Pakistan's
safeguards agreements with Canada, the IAEA, and the United States.
And if Pakistan were caught, its ability to field usable weapons would
be jeopardized.
One U.S. official said that the United States
is satisfied that Pakistan has not created polonium in its safeguarded
reactors. If not, then Pakistan has probably developed tritium initiators,
even though they are considered more difficult to develop. China may
have given Pakistan the design for a small tritum initiator. The initiator,
located at the center of the weapon-grade uranium core, would require
a design that could produce a more symmetric converging shock wave.
In 1987, Pakistani agents smuggled 0.8 grams of pure tritium gas they
had obtained from German parties who were convicted of illegally exporting
tritium in 1990.(11) This would be enough tritium for a number of
neutron initiators. Pakistan is probably looking at the development
of more sophisticated fission bombs, boosted fission bombs, and perhaps
hydrogen bombs. One of the Germans convicted in 1990 testified that
his Pakistani clients were trying to obtain equipment and materials
for an H-bomb program.(12) Declassified documents confirm that Pakistan
has engaged in a long and increasingly sophisticated nuclear weapons
program, and Pakistan's attempts to build a tritium production facility
during the late 1980s may also have been motivated by a program to
develop tritium-boosted fission weapons. Information
from the trial of Ortmayer and others in 1990 suggested that Pakistan
did not get all the equipment it needed to begin operating a tritium
purification plant. But some equipment was supplied from Germany and
was tested in Pakistan. Moreover, the testimony recorded that "plans
and the know-how for the tritium facility were supplied" in 1987.
However, without a proven design for
a significantly more sophisticated fission weapon, a boosted fission
weapon, or a thermonuclear device, it is doubtful that Islamabad could
develop one without full-scale testing.
