Order Code RS21391
Updated July 29, 2003
CRS Report for Congress
Received through the CRS Web
North Korea’s Nuclear Weapons:
How Soon an Arsenal?
Sharon A. Squassoni
Specialist in National Defense
Foreign Affairs, Defense and Trade Division
Summary
In December 2002, North Korea ended the 8-year-old freeze on its nuclear program
by expelling inspectors and reopening its plutonium production facilities. The CIA
assessed that North Korea could produce 5-6 weapons by mid-2003, in addition to 1 or
possibly 2 weapons it might already have. In April 2003, North Korean officials
claimed they had completed reprocessing all 8000 spent fuel rods (containing enough
plutonium for 5-6 weapons), a claim which few believed. Since then, however, there
are indicators that some reprocessing has occurred. This report, which will be updated
as warranted, describes North Korea’s steps in producing plutonium nuclear weapons.
Background
In the early1980s, U.S. satellites tracked a growing indigenous nuclear program in
North Korea. North Korea’s small reactor at Yongbyon (5MWe), capable of producing
about 6kg of plutonium per year, began operating in 1986.1 Later that year, U.S. satellites
detected high explosives testing and a new plant to separate plutonium (a necessary step
before turning the plutonium into metal for a warhead). In addition, the construction of
two larger reactors (50MWe at Yongbyon and 200MWe at Taechon) added to the
mounting evidence of a serious, clandestine effort. Although North Korea had joined the
Nuclear Nonproliferation Treaty in 1985, nuclear safeguards inspections began first in
1992. Those inspections raised questions about how much plutonium North Korea had
produced covertly that still have not been resolved. In 1994, North Korea signed the
Agreed Framework with the United States, agreeing to freeze its plutonium programs and
eventually dismantle them in exchange for several kinds of assistance.2 Western
intelligence agencies at that time estimated that North Korea had produced an amount of
1 5MWe is a power rating for the reactor, indicating that it produces 5 million watts of electricity
per day (very small). Reactors are also described in terms of million watts of heat (MW thermal).
Although the reactor was optimized to produce plutonium, and is not connected to any electricity
grid, North Korea has always referred to it by the 5MWe rating.
2 See CRS Issue Brief IB91141, North Korea’s Nuclear Weapons Program.
Congressional Research Service ˜ The Library of Congress

CRS-2
plutonium equal to that needed for one to two bombs. Other sources suggested North
Korea had enough material to produce 4-5 bombs.
This report describes the steps in producing plutonium-based nuclear weapons and
suggests potential signs of North Korean activities. Although the current crisis was
ignited by October 2002 revelations of a North Korean clandestine uranium enrichment
program, North Korea apparently is still in the construction phase.3 Estimates of a 5-6-
weapon nuclear stockpile by mid-2003 are based on plutonium production.
Weapons Production Milestones
Most proliferation analysts agree that the key obstacle in nuclear weapons
development is acquiring special nuclear material – plutonium-239 or highly enriched
uranium (HEU).4 Producing these two materials is technically challenging; in
comparison, many experts believe weaponization to be a relatively easy process.5 This
may help explain intelligence estimates that countries like Iraq could assemble a device
within months if they acquired fissile material from elsewhere. In general, outside
assistance often helps tip-off clandestine programs. North Korea is, however, largely self-
sufficient, with industrial-scale uranium mining, and plants for milling, refining,
conversion, fuel fabrication, reactors, and reprocessing. Key steps include fuel
fabrication, irradiation in reactors, and reprocessing of fuel. North Korea makes magnox
fuel -- natural uranium (>99%U-238) metal, wrapped in magnesium-alloy cladding. A
fuel core for the 5MWe reactor contains 8000 small fuel rods. The fuel fabrication plant
may have fallen into disrepair, according to one source, but North Korea has a supply of
fresh (unirradiated) fuel for the 5MWe reactor.6
When irradiated in a reactor, natural uranium fuel absorbs a neutron and then decays
into plutonium (Pu-239). The longer the fuel remains in the reactor, the more it becomes
contaminated by the isotope Pu-240, which can “poison” the functioning of a nuclear
weapon.7 Thus, a key consideration is how long the fuel must remain in the reactor to
produce the right kind of plutonium. According to North Korea, the 5MWe reactor
operated from January 1986 to April 1994 but many analysts believe some (or possibly
all) of the fuel core was removed in 1989. Spent or irradiated fuel, which poses
radiological hazards, must cool after removal from the reactor. The cooling phase,
estimated by some at 5 months, is proportional to the fuel burn-up.
Reprocessing – or separating the plutonium from waste products and uranium – is
the next step. North Korea uses a separation process similar to that used in the United
States. After shearing off the fuel cladding, the fuel is dissolved in nitric acid.
3 Unclassified CIA point paper distributed to Congressional staff on November 19, 2002.
4 Highly enriched uranium (HEU) is uranium comprised of 20% or more U-235 and is commonly
used as fuel in research reactors. Weapons-grade uranium is 90% or more U-235.
5 The physical principles of weaponization are well-known, but producing a weapon with high
reliability, effectiveness and efficiency without testing may provide other challenges.
6 David Albright, President of the Institute for Science & International Security, January 10, 2003
7 Plutonium that stays in a reactor for a long time (reactor-grade, with high “burn-up”) contains
about 20% Pu-240; weapons-grade plutonium contains less than 7% Pu-240.

CRS-3
Components (plutonium, uranium, waste) of the fuel are separated into different streams
using organic solvents. In small quantities, separation can be done in hot cells, but larger
quantities require significant shielding to prevent deadly exposure to radiation.8
Many experts agree that North Korea has mastered the engineering requirements of
plutonium production. Its 5MWe nuclear reactor operated from 1986 to 1994, and North
Korea separated plutonium in hot cells and tested its reprocessing plant. On the other
hand, some analysts have reported that the 5MWe reactor, which was thought to be a test
reactor rather than a plutonium production reactor, operated at low efficiencies and one
IAEA official called the reprocessing plant “extremely primitive” when he toured it in
1992. A key consideration for a crash program to build nuclear weapons is how quickly
the reprocessing plant can become operational; for building a larger arsenal, the
completion of the two larger reactors is key.
There is little information on whether North Korea has a workable nuclear weapons
design. The simplest nuclear weapon design, a gun-type assembly, cannot use plutonium.
Many believe North Korea to be capable of manufacturing implosion-type devices, which
require sophisticated lenses of high explosives to compress plutonium in the core. As
long ago as 1986, U.S. satellites detected high explosive testing with the kind of
compression patterns associated with implosion devices, although North Korea claimed
the tests were for civilian purposes.9 There have been reports of Soviet scientists aiding
North Korea, although CIA officials in the mid-1990s reportedly said that North Korean
scientists did not receive training in nuclear weapon technologies from Russia or China.10
Although states that have developed nuclear weapons typically first have used relatively
crude delivery methods, North Korea has concurrently produced ballistic missiles with
sufficient range and payload to carry nuclear warheads. Nonetheless, such a warhead
would need to be small and light enough to fit on a missile, and robust and sophisticated
enough to tolerate the variety of conditions encountered through a ballistic trajectory.
Estimating Nuclear Material Production
Estimating nuclear stockpiles is difficult, but in North Korea’s case, it may be more
important to know when it crosses the threshold between a demonstration capability (1-2
bombs) and a modest deterrent (10-20 bombs). Without empirical data, estimates rely on
assumptions of how much plutonium is produced and how much plutonium is needed per
bomb. (The IAEA has stated it cannot determine how much plutonium North Korea has
produced already without physical samples of the reactor fuel rods, knowing their exact
position in the reactor core and inspecting a suspected waste site.) There is unlikely to
be any data at all on weaponization. To determine how much plutonium is produced, key
estimates include: the average power level of the reactor; days of operation; how much
of the fuel loading is reprocessed and how quickly, and how much plutonium is lost to
production processes. According to North Korea, the 5MWe reactor performed poorly
8 Hot cells are heavily shielded rooms with remote handling equipment. They can be used for
examining irradiated targets or reactor fuel.
9 Don Oberdorfer, The Two Koreas, (MA: Addison-Wesley), 1997, p. 250.
10 David Albright, Frans Berkhout ,William Walker, Plutonium and Highly Enriched Uranium
1996: World Inventories, Capabilities and Policies
, Oxford University Press, 1997, p. 307.

CRS-4
in the early years, unevenly irradiating the rods. Maintenance reportedly was performed
at the reprocessing plant in 2002, but it is not clear that the plant has run reprocessing
campaigns beyond the “hot test” in 1990. North Korea told the IAEA that during the
1990 test, it recovered 62 grams of plutonium, losing almost 30% in the waste streams.11
Although this is not surprising for a plant starting up, it is difficult to assess when the
plant would improve its efficiency. Another consideration is whether or not the
reprocessing plant is operated continuously. The capacity of the plant is estimated at 220-
250 tons of spent fuel annually, about 4-5 times the amount of fuel in the 5MWe reactor.
Estimates of how much plutonium is needed per bomb vary according to judgements
about North Korea’s technical sophistication. Although the international standard is 8kg
of Pu per weapon(and 25kg for HEU), technical experts agree that it is possible to make
nuclear weapons with less than half that amount. Many of the higher estimates of North
Korea’s nuclear weapons production assume between 4 and 6 kg of plutonium per
weapon.
What Does North Korea Have Now?
In 1994, Western intelligence agencies estimated that North Korea had produced an
amount of plutonium equal to that needed for one to two bombs. One analyst estimated
that North Korea had separated between 6 and 10kg of plutonium in the late 1980s, and
that at least 8 kg of plutonium was needed for a first bomb, because of losses in the
production process, but the scraps could be saved and used potentially in a second bomb.12
Other sources ranged as high as North Korea being able to produce 4-5 bombs.
Recent assessments seem to emphasize that North Korea has assembled weapons.
Secretary of State Powell stated in December 2002 that “We now believe they [North
Koreans] have a couple of nuclear weapons and have had them for years.”13 An
unclassified CIA paper in November 2002 stated that the “North has one or possibly two
weapons using plutonium it produced prior to 1992.”14 However, the CIA paper stated
that this was an assessment that has not changed since the 1990s. In that time, the CIA
has consistently reported that North Korea “has probably produced enough plutonium for
at least, one, and possibly two, nuclear weapons.”15 Evidence connected to
weaponization, if it exists, has not been revealed publicly.
11 David Albright and Kevin O’Neill, editors, Solving the North Korean Nuclear Puzzle, ISIS
Report, ISIS Press, 2000, p. 88.
12 Ibid, pp. 111-126.
1 3 T r a n s c r i p t o f D e c e m b e r 2 9 , 2 0 0 2 “ M e e t t h e P r e s s ” ; s e e
[http://www.msnbc.com/news/852714.asp]
14 CIA unclassified point paper distributed to Congressional staff on November 19, 2002.
15 See Acquisition of Technology Relating to Weapons of Mass Destruction and Advanced
Conventional Munitions
,” [http://www.cia.gov/cia/publications/bian/bian_jan_2003.htm]

CRS-5
Adding to the Arsenal
Reprocess Existing Fuel. The likely quickest source of additional fissile
material is to reprocess the spent fuel from the 5MWe reactor, now in sealed canisters,
which contains an estimated 25 to 30kg of weapons-grade plutonium (5 to 6 weapons).
The sealed canisters do not create a technical barrier to reprocessing; contrary to North
Korean assertions about corroding fuel rods, the canisters simply keep track of the fuel
rods. Given the need to remove the fuel cladding on individual rods mechanically, the
canisters would need to be opened first, either in situ, or in a shielding cask.
The fuel rods are trucked to the reprocessing plant within the same complex. In
1992, IAEA inspectors estimated the reprocessing plant to be 80% complete (with one
reprocessing line) with 40% of equipment installed. In 1994, the second reprocessing line
was nearly complete, but still lacking some instrumentation. On April 17, 2003, North
Korean officials announced they were successfully reprocessing plutonium but a week
later, officials softened that statement to “successfully going forward to reprocess work.”16
On July 13, 2003, North Korean officials told U.S. officials in New York that they had
completed reprocessing the 8000 fuel rods on June 30.17 Apparently, preliminary tests to
detect Krypton-85 (a noble gas that is a by-product of reprocessing) were inconclusive but
later suggested that some reprocessing had taken place.
According to one report, in December 2002, China shipped North Korea 20 tons of
tributyl phosphate, which is a key ingredient in reprocessing.18 However, the reprocessing
plant, which conducted small-scale separations in the early 1990s, could encounter
difficulties in large-scale production. According to Robert Alvarez, who helped secure the
reactor fuel in the mid-1990s, “Because of its 1950s design, running the antiquated
reprocessing plant will involve a significant amount of ‘hands on’ operation...[making it}
more prone to spills, leaks and chemical failures, which could lead to extended shutdowns
and the loss of plutonium into the waste stream.”19 Assuming that one reprocessing line
was operating continuously, separations could take between 3.5 and 5 months.20 If North
Korea reprocessed about 11 tons/month, it might produce enough plutonium for 1 bomb
per month. The shortest estimate would include: 1 month to ready the reprocessing plant
and prepare fuel for reprocessing, 3 months to reprocess and 1-2 months to convert the
material into metal and shape a weapon. A longer estimate would include: 3 months to
ready the plant and fuel, 5 months to reprocess and 1-2 months to convert, for 10 months
total. In contrast, IAEA officials have estimated that North Korea could take two years
to produce 25 kg of separated plutonium.21 North Korea may also be able to reprocess
some material in existing hot cells, perhaps in the isotope production laboratory
associated with the IRT reactor.
16 “North Korea Shifts Tone on Nuclear Plan,” International Herald Tribune, April 22, 2003.
17 “North Korea Says It Has Made Fuel For Atom Bombs,” New York Times, July 15, 2003.
18 “China Ships North Korea Ingredient for Nuclear Arms,” Washington Times, December 17,
2002.
19 Robert Alvarez, “North Korea’s Weapons Challenge,” San Diego Union-Tribune, May 9, 2003.
20 Albright, Plutonium and Highly Enriched Uranium 1996, p. 305.
21 “Pyongyang May Operate Separation Plant,” Nuclear Fuel, December 23, 2002.

CRS-6
Make New Plutonium. Bush administration and other officials estimated that it
would take one year to 18 months before the 5 MW reactor began operating, based on
resolving engineering, safety, and personnel readiness issues stemming from the reactor’s
being closed for six years.22 However, North Korean officials announced on February 6
that the reactor was operating, and commercial satellite photography confirms that there
was activity in March. Although a common estimate is that the reactor generates 6kg per
year, the reactor would likely be operated for several years before fuel is withdrawn. In
3 years, it could generate about 14-18kg of plutonium, enough for 2 to 3 weapons.
Shorter cycles are possible, but would waste considerable fuel. Assuming a 6-month
cooling period for plutonium, North Korea would be ready to reprocess by August 2006,
and ready to convert into metal by February 2007.
Bring New Reactors On-Line. North Korea may also finish construction on the
reactors at Yongbyon (50MWe) and Taechon (200MWe), which are both several years
from completion. At the Yongbyon reactor, the main building appears to be completed,
while other buildings are still in skeletal form. Commercial satellite photography shows
no presence of cranes or trucks.23 The CIA estimates that together, the two reactors could
generate about 275kg of plutonium per year.24 If necessary, the Soviet-supplied IRT
reactor could produce 2-4 kg of plutonium per year by irradiating additional targets, but
this assumes North Korea acquires HEU fuel rods.
Likely Indications of North Korean Activity
The outside world now relies largely on remote monitoring and defector information
to detect North Korea’s nuclear weapons production. Some activities, like construction
(reactors at Yongbyon, Taechon), and reactor operations and shutdowns, have distinct
signatures. Satellites detected truck movements at Yongbyon in late January, but there
was no further information that confirmed the trucks were moving spent fuel to the
reprocessing plant.25 Although some activity was detected at the reprocessing plant in
April using satellite imagery, U.S. officials could not confirm as of May that large-scale
reprocessing was taking place.26 Reprocessing apparently has been detected remotely by
tracking the release of Krypton-85 (a noble gas that is a byproduct of reprocessing), but
there appears to be some uncertainty about the location of the emissions – whether they
are from Yongbyon or another, second unknown site.27 In addition, there were some
recent reports of a new high explosives test site; it is likely that actual nuclear tests – even
underground – would be detected. It probably will not be possible to detect when North
Korea mates warheads with missiles. Additional information could be obtained from
defectors or human intelligence sources, but these are typically rare.
22 ibid.
23 Briefing by Corey Hinderstein, Assistant Director, Institute for Science and International
Security, on April 3, 2003.
24 CIA unclassified point paper distributed to Congressional staff on November 19, 2002.
25 “Reactor Restarted, North Korea Says,” Washington Post, February 6, 2003.
26 “US Suspects North Korea Moved Ahead on Weapons,” New York Times, May 6, 2003.
27 “2nd N. Korean Nuclear Site Not Likely,” Washington Times, July 22, 2003.