Blogger Templates by Blogcrowds

North Korea’s Nuclear Weapons Programme

সোমবার, ১৩ জানুয়ারী, ২০১৪

N
orth Korea’s nuclear efforts can be divided into four distinct phases. During the first phase
(1959–80), the country’s nuclear programme was primarily focused on basic training and research. North Korea relied on assistance from the Soviet Union, which trained North Korean scientists and engineers and helped to construct basic research facilities – including a small research reactor and a radioisotope production laboratory. These facilities were placed under International Atomic Energy Agency (IAEA) safeguards in 1977.
The second phase (1980–94) covers the growth and eventual suspension of North Korea’s indigenous plutonium production programme. Around 1980, Pyongyang embarked on a major campaign to build a series of industrial-scale nuclear facilities that could produce substantial amounts of nuclear energy and weapons-grade plutonium. Although North Korea acceded to the 1968 nuclear Non-Proliferation Treaty (NPT) in December 1985, it denied the existence of these facilities until 1992, when it finally concluded a safeguards agreement with the IAEA permitting full inspections. Subsequently, North Korea’s refusal to cooperate with the IAEA and to account for possible plutonium production prior to 1992 led to the 1993–94 nuclear crisis and, eventually, to the 1994 US–North Korea Agreed Framework. Although this accord froze North Korea’s plutonium production facilities and placed them under IAEA monitoring, the US estimated that Pyongyang could have recovered enough plutonium for one or two nuclear weapons before this agreement came into force. The actual amount of plutonium produced by North Korea before 1992 is unknown.
The third phase (1994–2002) covers the period of the nuclear freeze. During these eight years, North Korea’s indigenous plutonium production facilities remained in a state of suspended animation, and its known plutonium stocks – some 25–30 kilograms (kg) in spentfuel rods – were subject to IAEA monitoring. With the plutonium route blocked by the Agreed Framework, which also required North Korea to eventually declare the plutonium produced prior to 1992, Pyongyang instigated a secret programme in the late 1990s to develop the means to produce weapons-grade enriched uranium utilising gas centrifuge technology.
The final phase examines the period from the end of the freeze in late 2002, following the exposure of Pyongyang’s secret uranium enrichment programme, to the present day. North Korea has taken some initial steps to revive its plutonium production facilities after nearly a decade of dormancy, and claims it has extracted all the plutonium it has on hand in spent fuel rods (though this cannot be confirmed by independent means). Presumably, North Korea is also continuing with its efforts to complete development of a significant uranium enrichment capability. However, very little is known about this project.
Phase 1 – origins of North Korea’s nuclear programme North Korea’s nuclear programme was born with assistance from the Soviet Union. The two countries signed a nuclear cooperation agreement in 1959 and, over the next 30 years, Moscow provided Pyongyang with training and technology useful in the development of basic nuclear technology. The type of aid granted to North Korea was typical of that on offer during the Cold War, when both the Soviet Union and the US supplied some of their allies and client states with basic nuclear technology and training. The 1959 agreement enabled a variety of technical and scientific exchanges and projects, including construction of the Yongbyon Nuclear Research Centre, training of North Korean scientific and technical personnel, and geological surveys for nuclear applications. Soviet assistance was not specifically intended to assist the development of nuclear weapons, but it allowed Pyongyang to master the basic technologies needed to produce and separate plutonium, which North Korea later employed in its nuclear-weapons programme.
In the early 1960s, with Soviet assistance, North Korea began construction of the Yongbyon Nuclear Research Centre, which became the centrepiece of its nuclear programme. Initially, the principal facilities housed at Yongbyon comprised a small research reactor, the IRT-2000, designed to conduct basic nuclear research and to produce small quantities of medical and industrial isotopes, and an adjacent radiochemical laboratory for extracting isotopes from ‘targets’ irradiated in the IRT-2000. The IRT-2000 is a ‘pool-type’ research reactor fuelled by a mixture of fuel elements of 10%, 36% and 80% enriched uranium, moderated and cooled by ‘light’ (i.e. ordinary) water.1 Construction of the IRT-2000 began in 1963. It became operational in 1965 at a power rating of 2MW(th), which was upgraded to 4MW(th) in 1974, and to 8MW(th) in 1987.
The radiochemical laboratory, which became operational in 1977, was originally fitted with 20 shielded hot cells and glove boxes, used to process radioactive isotopes for medical and industrial


‘North Korea’s nuclear programme was born with assistance from the Soviet Union’ … ‘Soviet assistance was not specifically intended to assist the development of nuclear weapons, but it allowed Pyongyang to
master the basic technologies needed to produce and separate plutonium’



Major North Korean nuclear sites
1. Pakchon                        Location of a uranium mine and milling facility (known as the April Industrial Enterprise), declared to the IAEA in 1992.The uranium milling facility reportedly processes ore from mines in the Sunchon area. Current status is unknown.
2. Pyongsan                      Location of uranium mining and a uranium milling facility, declared to the IAEA in 1992. The milling facility in Pyongsan reportedly processes ore from the Pyongsan and Kumchon uranium mines. Current status is unknown.
3. Pyongyang          Laboratory-scale hot cells, provided by the Soviet Union in the 1960s.
4. Sinpo                             Location of two 1,000 MW(e) light water reactors being constructed by the Korean Energy Development Organization (KEDO) under the terms of the Agreed Framework, which set a ‘target date’of 2003 for completion of the project.Various events have delayed the project. Construction began in mid-1997. The major non-nuclear element for the first reactor, defined in the Agreed Framework as a ‘significant portion’of the LWR project, was scheduled for completion by mid-2005. KEDO ‘suspended’the project for one year in December 2003.
5.Taechon                Location of an incomplete 200MW(e) graphite-moderated nuclear power reactor.
Construction began in 1989 and was frozen in 1994 (under the 1994 Agreed Framework). Current status is unknown, but there are no reports of major construction resuming after North Korea renounced the nuclear freeze in December 2002.
6.Yongbyon                       Location of a Nuclear Research Centre, comprising a 5MW(e) graphite moderated research nuclear power reactor, an unfinished 50MW(e) graphite moderated prototype power reactor, reprocessing facility, uranium conversion plant, fuel fabrication plant, and spent fuel and waste storage facilities. Also location of a Soviet-supplied IRT research reactor and radioisotope laboratory. Operation of the 5MW(e) reactor, the uranium conversion plant, the fuel fabrication facility and the reprocessing plant were frozen in 1994, along with construction of the 50MW(e) reactor. Since 2002, North Korea has restarted the 5MW(e) reactor and reportedly reprocessed some or all of the 8,000 spent fuel rods at the site.
No resumption of work on the 50MW(e) reactor has been reported.
7.Youngdoktong                Reported location of site (active in the 1990s) for nuclear weapons-related high- explosive testing.
8. Sunchon                        Location of an important uranium mine. Other mines reportedly located in Kumchon, Pyongsan, and Hwangsan.
Note:There are assumed to be several additional nuclear facilities associated with North Korea’s enrichment programme, including a possible production-scale centrifuge plant.The location of these facilities is unknown.
Sources: Carnegie Endowment for International Peace, www.ceip.org; Federation of American Scientists, www.fas.org;
Nuclear Threat Initiative, www.nti.org; and David Albright and Kevin O’Neill, Solving the North Korea Nuclear Puzzle, (Washington, DC:The Institute for Science and International Security, 2000).






purposes from irradiated targets. Typically, target materials were irradiated in the research reactor and then transferred to the hot cells and glove boxes where the desired isotope was chemically separated from waste products. Soviet experts also assisted in the construction of an underground facility at the site for storing radioactive waste from isotope production.
Although the IRT-2000 reactor and the radiochemical laboratory were intended for basic nuclear research and isotope production, the materials and equipment also provided North Korea with the means to experiment with the production and extraction of small amounts of plutonium, which Pyongyang acknowledged to the IAEA in 1993. These Sovietsupplied facilities were placed under IAEA inspections in 1977. However, due to IAEA procedures for monitoring facilities of this type, they were not subject to close scrutiny. The total amount of plutonium produced is, therefore, uncertain.
Although Soviet training of North Koreans is reported to have begun a few years before 1959, under the 1959 agreement and subsequent arrangements, the Soviet Union trained more than 300 (an exact figure regarding the number of experts who received training and in what fields is not publically available) North Korean engineers and physicists at Soviet institutes, including the Joint Institute for Nuclear Research at Dubna, the Moscow Engineering Physics Institute, and the Bauman Higher Technical School. (North Korean nuclear experts also received training in Canada, Japan and the former East Germany.)
Meanwhile, geological surveys conducted by the Soviet Union determined that North Korea possessed large deposits of uranium ore and graphite – Pyongyang subsequently developed these to form the building blocks of its plutonium production programme.
Phase 2 – North Korea’s plutonium production programme
Around 1980, North Korea launched a concerted national programme to build a series of industrial-scale facilities capable of producing significant amounts of plutonium for a nuclear-weapons programme, as well as for the country’s nuclear-power industry. The core of this programme were three gas-cooled, graphitemoderated, natural-uranium-fuelled reactors:
        a small 5MW(e) (25MW(th)) research reactor at Yongbyon;
        a larger 50MW(e) (200MW(th)) prototype power reactor at Yongbyon; and
North Korean plutonium fuel cycle
Mining
        a fullscale 200MW(e) (800MW(th)) power reactor at Taechon.
The power levels of nuclear reactors can be measured in terms of megawatts of heat produced (MW(th)) or megawatts of electricity generated ( MW (e)). Plutonium production capacity is a function of thermal power, but North Korea always designated its reactors according to their electrical output, presumably to emphasise their civil intent. As a result, North Korea’s reactors are usually identified by their nominal electrical output capacity, a method followed in this document.
Around this trio of reactors, North Korea also constructed facilities for the full plutonium fuel cycle. At the ‘front end’ of the fuel cycle were uranium mines, factories to process and refine uranium ore, as well as plants to purify natural uranium, to convert it to metal, and to fabricate fuel. At the ‘back end’ of the fuel cycle was an industrial-scale reprocessing plant at the Yongbyon Nuclear Research Centre designed to extract plutonium from spent reactor fuel along with facilities to treat and store nuclear waste.
‘Around 1980, North Korea launched a concerted national programme to build a series of industrial-scale facilities capable of producing significant amounts of plutonium for a
nuclear-weapons programme, as well as for the country’s nuclear-power industry’
The technology that North Korea chose was attractive for several reasons. Based on 1950s technology originally developed by France and the UK (to produce plutonium for their nuclear-weapons programmes), the basic reactor designs were available in the public domain and relatively straightforward to build and operate. Since the raw materials for these reactors – large quantities of natural uranium and graphite – could be found locally, Pyongyang was able to pursue an indigenous nuclear programme with minimal dependence on foreign assistance.
In addition, these reactor designs are well suited to producing weapons-grade plutonium.2 In contrast to light-water reactors (LWRs), which use low enriched uranium, graphite-moderated reactors fuelled by natural uranium are relatively more efficient in producing plutonium. In addition, the fuel is designed for moderately low levels of irradiation or ‘burn up’, minimising the accumulation of certain plutonium isotopes that are undesirable for weapons purposes. Moreover, the type of fuel used in the graphitemoderated reactors developed by North Korea is generally unsuitable for long-term storage because the metal cladding surrounding the uranium fuel rods tends to corrode rapidly in water and presents a fire hazard if exposed to air. Consequently, the irradiated fuel is usually reprocessed – that is, the plutonium and uranium are chemically separated from radioactive waste products, which can be safely stored in specialised tanks and containers.
Finally, just as graphite-moderated reactor technology was initially developed for military production but later adapted for energy production in western countries, this choice of reactor provided North Korea with a ‘dual-use’ capability – technology that could be utilised in nuclear-weapons production as well as nuclearpower production.
The main facilities that comprise North Korea’s plutonium production complex are detailed below.3
Uranium mining and milling
On the basis of geological surveys conducted by the Soviet Union, North Korea began large-scale uranium mining operations at various locations near Sunchon and Pyongsan in the late 1970s or early 1980s. The raw uranium-bearing ore was shipped to uranium milling factories at Pakchon and Pyongsan, where it was crushed and chemically processed to produce U3O8 or ‘yellow cake’, which was then transported to the Yongbyon nuclear centre for further processing and fabrication into nuclear reactor fuel. Typically, one tonne of North Korean uranium ore contains about one kilogram of uranium, meaning that some 50,000 tonnes of ore had to be mined and processed in order to obtain the 50 tonnes of natural uranium needed for the initial fuel load for the 5MW(e) reactor. It has been estimated that, at its peak in the early 1990s, North Korea was able to produce about 300 tonnes of yellow cake annually, equal to approximately 30,000 tonnes of uranium ore. Actual production of yellow cake in the
Uranium-conversion and fuel fabrication facilit
decade before the nuclear freeze is unknown. North Korea’s current mining and milling capacity is also unknown, but it appears unlikely that yellow cake production is a significant constraint on its immediate nuclear requirements.
Uranium conversion and fuel fabrication
Between 1980 and 1985, North Korea built a substantial factory at Yongbyon to refine yellow cake and to produce uranium metal fuel elements for its graphite-moderated reactors. The yellow cake was chemically refined in a series of buildings using a process known as ‘conversion’ into more purified forms of uranium, from U3O8 to UO3 ( uranium trioxide) to UO2 (uranium dioxide). At a high temperature, the UO2 was mixed with highly caustic gaseous hydrofluoric acid to produce a more purified uranium compound – uranium tetrafluoride (UF4) – which was converted into uranium metal ingots in vacuum furnaces. In the final stage of fuel fabrication, the uranium metal ingots were melted and alloyed with small amounts of aluminum, with the resulting alloy extruded and machined into fuel rods of around three centimetres (cm) in diameter and 50cm in length. Finally, each uranium fuel rod was inserted into a magnesiumzirconium alloy cladding tube (known as ‘magnox’) to produce the final fuel assembly. Approximately 8,000 such fuel assemblies, containing about 50 tonnes of natural uranium, were necessary for the 5MW(e) reactor core. If completed, the larger 50MW(e) reactor would require about 400 tonnes of uranium fuel, equal to some 64,000 fuel assemblies, while the 200MW(e) reactor would require about 1,400 tonnes of uranium fuel, equal to about 224,000 fuel assemblies.
The uranium-conversion and fuel-fabrication facility at Yongbyon was designed to produce fuel for the entire line of graphite-moderated reactors under construction in North Korea during the 1980s. However, the level of production before the nuclear freeze is unknown. According to North Korean officials, in 1992, the plant was producing roughly 100 tonnes of uranium fuel per annum, equal to approximately 16,000 fuel assemblies, although its nominal annual capacity was larger, perhaps 200–300 tonnes of uranium fuel (32,000–48,000 fuel assemblies). At a minimum, North Korea in known to have produced enough fuel prior to the freeze for the initial core load for the 5MW(e) reactor and at least one fresh core load. It is also known to have produced slightly more than one-half of the fuel required for the 50MW(e) reactor under construction. North Korea could have produced a significant amount of additional fuel before 1992 that it failed to declare to the IAEA.
The 5MW(e) ((25MW(th)) experimental power reactor
5MW(e) [(25MW(th)] experimental power reactor4
During the 1980s, the most important facility in North Korea’s plutonium production programme was a small research reactor located at the Yongbyon Nuclear Research Centre, which was designated an
‘experimental power reactor’. Based on the same design concept as the UK’s 50MW(e) Calder Hall plutonium production reactor, which became operational in 1956, the North Korean reactor is fuelled with magnox-clad, natural-uranium fuel elements cooled with carbondioxide gas (CO2) and moderated and reflected with high purity graphite. In design, the reactor core consists of some 300 tonnes of graphite blocks into which 812 fuel channels have been drilled vertically. Each fuel channel is designed to hold ten fuel assemblies stacked vertically on top of one another, giving a total of about 8,000 fuel rods (50 tonnes of natural uranium) in a full core load. In addition to the fuel channels, some 40 control channels have been drilled into the graphite blocks to allow for the insertion of barium carbon control rods to maintain control over the nuclear reaction. The graphite blocks holding the fuel and the control-rod channels are surrounded by an additional 300 tonnes of graphite reflector (which serves to reflect neutrons back into the core) and the entire core is encased in a steel pressure vessel to contain the cooling gas. Cooling takes place using pressurised CO2, which is blown through the core by electric motors. A large loading and unloading machine refuels the reactor from the top of the core.
Reactor construction began in 1980, and the reactor went critical in August 1985. It operated intermittently from 1986 until 1994 when it was shut down under the Agreed Framework. According to North Korea, the reactor was designed to have a power rating of 25MW(th), which North Korea expressed in terms of its nominal 5MW(e) electrical output, although the reactor was not actually used to produce electricity. In theory, operating at full power for 300 days per year, this reactor could produce approximately 7.5kg of weapons-grade plutonium annually in the discharged spent fuel.5 Of course, actual annual plutonium production would depend on the fuel’s irradiation level, which is a function of the reactor’s power level and the number of days per year that it was operational, usually expressed as megawatt thermal days per tonne (MW(th-d/t)).
‘By the time the reactor was completely de-fuelled in June 1994, the spent fuel probably contained about 25–30kg of plutonium’
The operational history of the reactor between 1986 and 1994 (and hence how much plutonium was produced) is shrouded in mystery. In 1992, when IAEA inspectors were first allowed access to the reactor, Pyongyang claimed that the reactor had experienced serious start-up and control difficulties, which prevented full-power operations and resulted in frequent shut downs during the first several years of operations. According to the North Koreans, reactor operation caused distortions in the neutron flux (the total number and speed of neutrons within a specific volume), which caused fuel in some parts of the core to overheat and to fail. In April–May 1989, the reactor was reportedly shut in order to remove a few hundred damaged fuel rods, after which North Korea claimed that it was able to operate more regularly at 20MW(th) – close to full power. At the time of the initial IAEA inspections in April 1992, North Korea claimed that about 17kg of plutonium had been produced in the reactor fuel. By the time the reactor was completely defuelled in June 1994, the spent fuel probably contained about 25–30kg of plutonium.
The accuracy of North Korea’s operating record prior to 1992 has never been verified, and independent evidence is mixed. Through satellite reconnaissance, the US detected the construction of the 5MW(e) reactor at an early stage, and was able to confirm initial operation of the reactor in 1986 by noting the emission of steam plumes from its cooling tower, which indicated the reactor was venting excess heat. However, satellite monitoring of the facility was not frequent enough during the 1980s to compile a complete record of operations and cloud cover sometimes prevented plumes from being detected. In any event, ‘plumology’ is an inexact science. If operated at low power levels or under certain climatic conditions, the reactor might not produce a clearly visible steam plume. Nonetheless, US observation of the reactor over the years did reveal some discrepancies in North Korea’s declared operating
The 50MW(e) (200MW(th)) prototype reactor
history, suggesting that the reactor may have functioned more frequently than Pyongyang had claimed.
The IAEA was also never able to complete its investigations into the reactor’s operating history. Some of the evidence collected by the Agency supported North Korea’s assertions, while some of it did not. Furthermore, Pyongyang obstructed the IAEA‘s efforts to take further measurements that might have provided answers to these questions. In fact, North Korea deliberately destroyed evidence which would have been helpful in reconstructing and verifying the reactor’s operating history. In June 1994, when North Korea completely de-fuelled the reactor, it refused to allow the IAEA to record the exact location of each of the 8,000 fuel rods in the core. Armed with this information, and subsequent measurements of the radioactivity level of each rod, the IAEA had hoped to ‘map’ the reactor’s operational history, especially to determine how long the fuel was in the reactor. To foil this effort, North Korea not only prevented the IAEA measurements, but also mixed the rods together in
The 200MW(e) (800MW(th)) power reactor
different storage baskets in the cooling pond so that the evidence was destroyed.
The 50MW(e) (200MW(th)) prototype power reactor In 1984, North Korea began construction of a larger reactor at Yongbyon, using the same basic materials and technology as utilised in the 5MW(e) reactor – magnox-clad natural uranium fuel, graphite moderation, and CO2 gas cooling – although the core design concept in this case resembles the Frenchdesigned G2 reactor commissioned at Marcoule in 1956 with fuel rods placed horizontally rather than vertically, as in the Calder Hall-type design. Nominally rated at 50MW(e) and with a core load of 400 tonnes of natural uranium fuel this reactor is, theoretically, capable of producing about 55kg of weapons-grade plutonium per year, if operated at full power for 300 days and assuming a discharge of 100 tonnes of the most heavily irradiated spent fuel. According to North Korean officials, the reactor was within a year of initial service at the time of the nuclear freeze, although this claim was never verified. Because it was never completed, it is unknown whether the reactor is capable of full-power operations, and there are no clear signs that contruction on the reactor has resumed since the nuclear freeze ended in 2002.
The 200MW(e) (800MW(th)) power reactor
In the late 1980s, North Korea began construction at Taechon of a fullscale version of the 50MW(e) reactor, based on the same technology – magnox-clad naturaluranium fuel, graphite moderation, and CO2 gas cooling – and the same core design as the French G2. Nominally rated at 200MW(e), and with a core load of 1,400 tonnes of uranium fuel this reactor would, theoretically, be capable of producing up to 220kg of weapons-grade plutonium annually, if operated at full power for 300 days per year. This reactor was in the early stages of construction when the nuclear freeze came into effect in 1994.
Radiochemical laboratory/reprocessing plant
In 1984, North Korea began construction of an industrial-scale reprocessing plant to separate plutonium from spent nuclear fuel at the Yongbyon Nuclear Research Centre. During construction, the exact purpose of the facility was debated within the US intelligence community. Some analysts believed that it was most likely a reprocessing facility, while others argued that it could be engaged in non-nuclear industrial activities. It was not until the IAEA conducted inspections in 1992 that it was confirmed as a reprocessing plant, which Pyongyang euphemistically called a ‘radiochemical laboratory’. The operation of the plant is based on the Purex (plutonium uranium extraction) process, a standard procedure widely used in the nuclear industry, in which uranium and plutonium are separated from nuclear waste and then from each other through a series of chemical processes. Lorries transport spent fuel from the 5MW(e) reactor in buckets stored inside heavy casks to the front end of the reprocessing facility, where the nuclear-fuel assemblies are mechanically dis-assembled, and the uranium fuel is dissolved in nitric acid. The liquid mixture is treated with various chemicals and passed through a series of stainless-steel mixer settler tanks in which the plutonium and uranium is selectively precipitated from the spent fuel’s highly radioactive fission products. The aqueous plutonium-uranium  mixture is then passed through another set of mixer settler tanks to separate out the plutonium. Because of high radiation, all of the operations are carried out remotely behind heavy shielding. Finally, at the back end of the plant, the concentrated plutonium is further purified and
Radiochemical laboratory/reprocessing plant
collected as plutonium oxide (PuO4) in a series of glove boxes. The oxide powder can then be converted into plutonium metal ingots, which could be melted and cast into components for nuclear weapons. Adjacent to the main building are a series of tanks and vaults intended to concentrate and store the large volumes of liquid and solid radioactive waste produced during reprocessing.
In 1992, IAEA inspectors discovered that one reprocessing line had been completed at the plant and that a second was under construction. According to North Korean officials, at this time, the reprocessing facility was designed to process spent fuel containing 0.7kg of plutonium per tonne of spent fuel, and each line was designed to process one tonne of spent fuel over the course of about three days of continuous operations. During reprocessing, plants typically operate around-the-clock, although it is possible (if less efficient) to process individual batches of spent fuel one at a time. Theoretically, the facility’s one completed line is capable of processing the 5MW(e) reactor’s entire 50 tonne core load in a single campaign, lasting approximately 150 days. If both lines were operating continuously for 300 days per year, the plant would have a total nominal capacity to process annually some 200 tonnes of magnox spent fuel, more than sufficient to handle the spent fuel that would typically be discharged each year by the 5MW(e) and the 50MW(e) reactors. In 1994, when IAEA inspectors returned to monitor the nuclear freeze, they found that North Korea had made considerable progress in installing equipment for the second reprocessing line, which was scheduled for completion in 1996.
The 1992 plutonium mystery
North Korea’s accession to the NPT in December 1985 necessitated that it place all its nuclear facilities and materials under international inspection and that it pursue nuclear technology solely for peaceful purposes. Although North Korea had 18 months under the treaty to negotiate a comprehensive safeguard agreement with the IAEA, it did not sign this agreement until January 1992. In all, six official inspection missions took place in North Korea in 1992, before Pyongyang denied inspectors access to suspect nuclear waste storage facilities and threatened to withdraw from the NPT.
‘In 1984, North Korea began construction of an industrial-scale reprocessing plant to separate plutonium from spent nuclear fuel at the Yongbyon Nuclear Research Centre’
During the first inspection in 1992, North Korea told the IAEA that it had test run the reprocessing plant between March–May 1990, during which 86 damaged fuel rods (which had been removed from the 5MW(e) reactor in 1989), as well as 172 fresh fuel rods, were reprocessed in a single campaign of three batches. According to North Korea, this resulted in the recovery of 62 grams of plutonium oxide from the nearly 90 grams of plutonium contained in the spent fuel, a claimed high loss rate of approximately 33%.6 To verify the accuracy of North Korea’s declaration, the IAEA took samples from the extracted plutonium, as well as from the waste tanks and work areas at the reprocessing plant, including ‘swipe samples’ from the glove boxes in which the plutonium was processed. Such samples contained individual dust particles that could be analysed to deduce the fractional content and ratio of different isotopes produced in the irradiated fuel.
Analysis of these samples highlighted several discrepancies in North Korea’s initial declaration to the IAEA. The Plutonium-240 (Pu-240) content of the declared plutonium was uniform, but the dust particles contained a range of Pu-240 content that tended to cluster in three distinct groups, suggesting that additional batches of fuel (containing slightly different levels of Pu-240) had been reprocessed. Contrary to North Korea’s declaration that it had only conducted reprocessing in 1990, the ratio of Americium-241 to Pu241 found on these dust particles at the site suggested that at least three reprocessing campaigns had occurred, in 1989, 1990 and 1991. Finally, the IAEA discovered that the percentage of Pu-240 in the declared plutonium was different to that in the waste tanks, suggesting that there was a quantity of ‘missing’ nuclear waste. To explain these discrepancies, North Korean officials said that a a few hundred milligrams of plutonium had been separated in 1975 from targets irradiated in the IRT-2000 reactor, and that the waste from these experiments had been mixed with the waste from the 1990 test run, contaminating the IAEA’s findings. This explanation did not seem plausible to IAEA analysts, because the additional waste declared by North Korea seemed too old and too small to explain the discrepancies.
As the IAEA began to discover these discrepancies, the US reviewed the record of satellite imagery of the 5MW(e) reactor. Based on the absence of steam plumes from the cooling tower, the reactor appeared to have been shut down for about two months in early 1989, as North Korea claimed. During this time, North Korea contended that it had replaced some 300 fuel rods damaged by overheating due to flaws in the reactor’s design. Of these damaged fuel rods, North Korea said that 86 were subsequently reprocessed, and that the others were stored in a dry storage pit at the reactor site. US analysts, though, estimated that, in the worst case, North Korea could have unloaded a much larger portion or even the entire core load during the twomonth shut down. This conclusion was based on a calculation of how quickly one or two fuel machines could unload spent fuel and load fresh fuel, assuming the North Koreans worked around-the-clock.7 Based on Pyongyang’s account of the reactor’s operating history prior to the 1989 fuel discharge, the entire core would have contained about 9.5kg of plutonium, which, assuming a likely range of potential 10–30% reprocessing losses, could have yielded as much as 6.5–8.5kg of plutonium. If the most heavily irradiated half of the core fuel-rods were discharged in 1989, they would have contained about 7kg of plutonium, which could have yielded about 5–6kg of separated plutonium if reprocessing losses ranged between 10–30%. In this scenario, North Korea could have loaded a new core or a substantial portion of a new core in 1989 and then falsified the operating records from 1989–92 to make it appear to be the original one.
To add credence to this scenario, US analysts accumulated satellite evidence that North Korea had built, operated and concealed two underground waste storage sites at Yongbyon. One site, originally constructed in the late 1970s, was associated with the IRT-2000 reactor and the radiochemical laboratory. In 1992, as IAEA inspections were underway, satellite imagery indicated that North Korea had covered this old radioactive waste site with earth and had planted vegetation on top of it in order to conceal the location, while building a new waste site nearby. North Korea told the IAEA that this was the original waste site. IAEA inspectors were denied access to the original waste site, which could contain radioactive waste from undeclared plutonium production in the IRT-2000 reactor. Although North Korea admitted in 1993 that it had produced a few hundred milligrams of plutonium in the IRT-2000 reactor in 1975, and claimed that it had mixed waste from these experiments with waste from the 1990 test run at the reprocessing plant, its efforts to conceal the old waste site suggested that additional plutonium had been produced. IAEA analysts estimated that the IRT-2000 reactor could have produced between two and 4kg of plutonium between 1965, when the reactor first became operational, and 1992, when regular IAEA monitoring was authorised.8 Actual plutonium production is, of course, unknown.
An even more significant waste storage site, known as Building 500, was constructed near the reprocessing plant in 1991. This structure contained a basement that appeared to be divided into compartments for storing liquid and solid radioactive waste. Concrete slabs were laid over the basement and a single-story building was erected on its foundation. In the winter of 1991–92, trenches – presumably to hold piping – were dug between Building 500 and the reprocessing facility, which could have allowed North Korea to pump waste from the reprocessing facility to Building 500 prior to the arrival of IAEA inspectors. If North Korea did carry out substantial undeclared reprocessing and diverted the waste to Building 500, tonnes of radioactive waste would be stored in the basement, which could not be concealed from a close inspection. Measurement of the volume and content of any such waste would provide evidence with which to calculate the extent of undeclared reprocessing and thus the amount of additional plutonium that North Korea could have produced before 1992.
Tipped off by US information, the IAEA sought access to the two suspect waste sites to determine whether radioactive waste, produced by undeclared reprocessing activity, was stored at the sites. In September 1992, IAEA inspectors were allowed to visit Building 500 (which was being used as a militaryvehicle repair workshop) but they were told it did not have a basement. After North Korea refused repeated demands for greater access, including for the extraction of samples from underneath the building, the IAEA requested a ‘special inspection’ of the two suspect waste sites in February 1993. Maintaining that the sites were non-nuclear military facilities beyond the scope of the IAEA inspection mandate, Pyongyang responded, in March 1993, by invoking its right to withdraw from the NPT, setting in motion the 1993–94 nuclear crisis.
Washington’s assessment that North Korea might have produced ‘enough plutonium for one or possibly two nuclear weapons’ or between 8–12kg of separated plutonium before 1992 was based on five factors.
      Firstly, analysis by the IAEA, based on samples from the reprocessing facility, strongly indicated that North Korea had not fully declared its plutonium production prior to 1992, although these sampling ‘discrepancies’ could not determine how much plutonium North Korea was hiding.
      Secondly, based on satellite surveillance of past operations of the 5MW(e) reactor and estimates of North Korea’s de-fuelling capability, the US estimated that most or all of the core could have been unloaded in April–May 1989, containing some 6.5–8.5kg of plutonium after reprocessing. Pyongyang could have falsified the reactor’s operating records in order to disguise the insertion of a second fresh core in 1989. North Korea’s decision to unload the reactor in June 1994 in a manner that deliberately made it impossible to reconstruct its operating history reinforced suspicion that it wanted to conceal such information.
      Thirdly, North Korea could have employed the small IRT-2000 research reactor to produce small amounts of plutonium every year, perhaps generating between 2–4kg in total.
      Fourthly, two suspect nuclear waste sites provided plausible locations for North Korea to divert and hide substantial quantities of nuclear waste produced as a result of undeclared reprocessing of spent fuel from the 5MW(e) and the IRT-2000 reactors.
      Finally, North Korea’s unwillingness to cooperate with the IAEA to resolve discrepancies pertaining to its plutonium declaration and, particularly, its refusal to allow access to the suspect nuclear waste sites – which precipitated a major international crisis – convinced many in Washington that Pyongyang must be hiding something significant.
In short, there was substantial evidence to support the conclusion that North Korea was concealing some plutonium, and plausible scenarios could be constructed to account for enough plutonium for ‘one or possibly two’ nuclear weapons. The actual amount of plutonium acquired by North Korea before 1992, though, is unknown.9
Phase 3 – plutonium programme frozen (1994–2002)
Under the terms of the 1994 Agreed Framework, North Korea agreed to freeze and eventually to dismantle the key facilities associated with its plutonium production programme, including the uranium-conversion and fuel-fabrication plant, the 5MW(e), 50MW(e) and 200MW(e) reactors, and the reprocessing facility. The IRT-2000 reactor (and its related radiochemical laboratory) was exempt from the freeze, on the grounds that it could be used to produce radioisotopes for medical and industrial purposes – subject to IAEA inspections. To verify that the frozen plutonium production facilities were not operating and that all construction had been terminated, the IAEA placed seals on the main access points, installed monitoring devices, and stationed a small team of resident inspectors at Yongbyon, who were allowed to conduct short-notice inspections of different parts of the facilities subject to the freeze.
‘Under the terms of the 1994 Agreed Framework, North Korea agreed to freeze and eventually to dismantle the key facilities associated with its plutonium production programme’
During the course of the nuclear freeze, the IAEA and North Korea held a series of ‘technical discussions’, during which Pyongyang agreed to a number of additional containment and surveillance measures to verify the freeze. But Pyongyang resisted any Agency activities that it believed could shed light on its past plutonium production efforts, such as the installation of IAEA monitoring equipment at nuclearwaste tanks, additional sampling at the reprocessing facility or the taking of measurements to determine the plutonium content of the 5MW(e) reactor’s spent fuel rods.10 Under IAEA supervision, North Korea also undertook various initiatives to maintain the readiness of its plutonium production facilities. At some establishments, such as the 5MW(e) reactor and the reprocessing plant, the maintenance procedures appeared regular and thorough, while other facilities, such as the 200MW(e) reactor and the fuel-fabrication plant, received little attention.
In addition, the Agreed Framework called on the US to help North Korea to ‘stabilise’ the 8,000 spent-fuel rods discharged from the 5MW(e) reactor in May–June 1994, pending their eventual removal from North Korea when the first unit of the Light Water Reactor (LWR) project (see below) was completed. After their removal from the reactor, these rods had been stored in a spentfuel pond next to the reactor building for over two years, during which time a considerable amount of corrosion had occurred. As a result, much of the magnesium cladding and some of the uranium metal had broken loose from the fuel rods themselves, posing a safety risk. Over the course of the next few years, under IAEA monitoring, the spent fuel rods were placed in 400 stainless-steel canisters, each containing approximately 20 rods or fragments. These canisters were filled with inert gas and sealed by US contractors on-site in North Korea, and then placed in underwater racks under IAEA seal. From Washington’s perspective, the canning operation helped to prepare the fuel for eventual shipment out of the country because the canisters were designed to be fitted inside shielded shipping casks. From Pyongyang’s standpoint, however, canning prevented further corrosion and helped to preserve a reprocessing option if the Agreed Framework failed.
Under the terms of the Agreed Framework, the US, along with Japan and South Korea, formed an international consortium, the Korean Peninsula Energy Development Organization (KEDO), which undertook to provide North Korea with a LWR project, consisting of twin 1,000 MW(e) reactor units, by a ‘target date’ of 2003. Although construction was delayed for a variety of reasons, site preparation at Sinpo on North Korea’s east coast commenced in July 1997, and the concrete foundations of the first reactor unit were laid in August 2002. In accordance with the Agreed Framework, no significant nuclear components were delivered to the LWR project because North Korea was first required to cooperate with the IAEA in accounting for its plutonium production prior to 1992. In any event, the project was formally suspended for one year by KEDO on 4 November 2003, following the collapse of the Agreed Framework. If they were ever to be completed, the LWRs would produce substantial quantities of plutonium in spent fuel, which, in theory, North Korea could extract and divert to weapons production. In practice, North Korea could not acquire spent fuel from the LWR project without detection by the IAEA, and it has no known facility to separate plutonium from the type of fuel used in the LWRs.
During the nuclear freeze, the US and its allies remained vigilant to the possibility that North Korea might seek to evade the restrictions imposed under the Agreed Framework and continue with its nuclear efforts at clandestine facilities. In early 1998, the US intelligence community concluded, based on data from a variety of sources, including information supplied by defectors and satellite imagery, that North Korea was constructing a large underground facility near Kumchang-ri, about 40km northeast of Yongbyon, which potentially could house a secret reactor and reprocessing facility to produce plutonium. Adding to suspicions, the construction company involved in the Kumchang-ri project was the same one that built the Yongbyon Nuclear Research Centre. Washington’s original inclination was to keep the site under observation in order to determine whether North Korea had begun to transport nuclear reactor components to the site, but suspicions about the site were leaked to the press in August 1998.11
The suggestion that North Korea might be about to violate the Agreed Framework, combined with the subsequent test of a Taepo-dong-1 missile in late August, put enormous political pressure on Washington to address concerns about the construction at Kumchangri. After a series of negotiations, North Korea agreed, in March 1999, to allow a US team to visit the Kumchangri site in exchange for US food assistance. In May 1999, the US team inspected the site, which consisted of numerous underground tunnels and storage rooms. The inspection revealed that the site was not configured to house an underground reactor and reprocessing facility, much to Washington’s embarrassment.
The secret uranium enrichment programme
Under the Agreed Framework, North Korea’s capacity to produce additional plutonium at the Yongbyon complex was effectively frozen. As a consequence, its presumed nuclear arsenal was limited to one or two nuclear weapons. Moreover, under the terms of the Agreed Framework, North Korea would eventually be required to account for its undeclared plutonium holdings and to dismantle its plutonium production facilities as a condition for receiving the LWR project. Unless Pyongyang decided to pay the political costs of openly reneging on the provisions of the agreement, it would be forced to give up whatever nuclear weapons capability it had required before 1992. To the extent that maintaining a nuclear hedge was perceived as essential to the survival and defence of the regime, Pyongyang had a strong incentive to develop an alternative means of producing nuclear material, which would allow it ostensibly to comply with the Agreed Framework, while preserving a secret nuclear-weapons programme.
Centrifuge machine
Although neither country has acknowledged the arrangement, it is widely reported that North Korea provided Pakistan with No-dong missiles and production technology in exchange for gas centrifuge technology and perhaps other assistance for North Korea’s nuclear weapons programme, probably around 1997.12 For Islamabad, the arrangement was a short cut to acquiring a longer-range missile capable of delivering a nuclear payload, at a time when China was cutting back on its assistance for Pakistan’s missile programme. For Pyongyang, the deal provided an attractive alternative technology with which to produce weapons-grade fissile material, in contravention of the terms of the Agreed Framework.
There are many different kinds of centrifuge technology, but the basic principle involves rapidly spinning uranium in gaseous form (uranium hexafloride) in tubes called rotors. Depending on the design, rotors can be made of high-strength metals (such as certain alloy types of aluminium or steel) or carbon fibre. The centrifugal forces inside the rotor cause a slight separation of lighter U-235 and heavier U-238  atoms and the two ‘streams’ of uranium hexafloride are siphoned off under separate withdrawal systems. Each centrifuge machine is capable of a small amount of separation – measured in Separative Work Units (SWUs) – but by passing the slightly enriched stream through an interconnected series of hundreds or thousands of machines, known as a cascade, it is possible to increase the percentage of U235 from the low level (0.7%) found in nature to higher percentages necessary for nuclear-reactor fuel or nuclear weapons. Generally, enrichment levels of 3–5% U-235 are required to fabricate fuel for the LWRs, while higher levels of U-235 (around 90%) are most desirable for nuclear weapons. A number of countries have built centrifuge plants to produce reactor fuel for civilian nuclear power requirements, including China, Japan and Russia, as well as Uranium Enrichment Services Worldwide (URENCO), a European consortium comprising British, Dutch and German firms.
It is not known exactly what kind of nuclear assistance North Korea received from Pakistan, but it is generally assumed that it could have included technical specifications, sample centrifuge machines, and training that would allow North Korea to duplicate the technology and to assemble a productionscale centrifuge facility. Based on information regarding Pyongyang’s procurement efforts, it appears that North Korea is still seeking to obtain components for a type of centrifuge that matches the specifications of one that Pakistan is known to possess.
While working as a metallurgist for a Dutch contractor employed by URENCO in the early 1970s, a Pakistani scientist, A.Q. Khan, obtained the designs for at least three types of centrifuges. The most advanced was the G-2, a German-designed, supercritical centrifuge consisting of two maraging steel rotors connected by flexible bellows, which is capable of rotating at around 500 metres per second. Each machine is about one-metre long with a diameter of 148mm, and is capable of achieving five Separative Work Units (SWUs) per year.
The G-2 eventually became the workhorse of Pakistan’s nuclear-weapons programme after Khan returned home to lead its centrifuge programme ( as head of Khan Research Laboratories). By the 1990s, Khan was also in charge of Pakistan’s efforts to develop a nuclear-capable, long-range, liquid-fuelled missile. Consequently, he became the principal point of contact with regard to nuclear and missile cooperation with North Korea, reportedly making numerous trips to North Korea, beginning in the late 1990s. Because of his role in the development of nuclear weapons, missile warheads and centrifuge technology, it is thought that Pakistani and North Korean technicians may have collaborated on nuclear weapon designs in addition to enrichment and missile technology.
By 2000, US intelligence had begun to detect North Korean attempts to procure equipment and materials that could be used in a centrifuge programme. However, the quantities were small, suggesting a research and development effort, and technical opinion was divided on whether the items were intended for use in centrifuges or for other purposes, such as in missiles. Nonetheless, there was sufficient concern that, in March 2000, US President Clinton waived a legislative certification that required him to certify that ‘North Korea is not seeking to develop or acquire the capability to enrich uranium’. In 2001, a source – said to be a North Korean defector – reported that North Korea had been pursuing a centrifuge enrichment programme for several years, although the location of the production plant and related facilities were apparently not identified.13 Moreover, North Korea reportedly began seeking large quantities of materials and components that were uniquely associated with centrifuge production, such as high-strength aluminium tubes of specific dimensions and equipment suitable for uranium feed-and-withdrawal systems.14
Based on this information, the US intelligence community concluded, in June 2002, that North Korea had embarked on an effort to build a clandestine production-scale centrifuge facility to produce weapons-grade uranium. In October 2002, during his first visit to Pyongyang to open high-level talks between the US and North Korea, US Assistant Secretary of State James Kelly accused North Korean officials of pursing an enrichment programme. According to subsequent US accounts, Vice Foreign Minister Kang Sok Ju, the most senior North Korean official responsible for dealing with the US, ‘acknowledged’ that the US accusation was accurate. More recently, though, North Korean diplomats claim that Kelly ‘misunderstood’ Kang, who was merely saying that North Korea had a ‘right’ to an enrichment programme in view of US ‘violations’ of the Agreed Framework. In November 2002, the US Central Intelligence Agency (CIA) stated publicly that North Korea is ‘constructing a plant that could produce enough weapons-grade uranium for two or more nuclear weapons per year when fully operational – which could be as soon as mid-decade’.15
From the available evidence, it appears that North Korea obtained substantial centrifuge assistance from Pakistan and that it is seeking to build a clandestine production-scale centrifuge plant to produce highly enriched uranium for nuclear weapons.
However, the status of North Korea’s centrifuge programme, and particularly, how close it is to completion, is very difficult to ascertain, given various uncertainties and Pyongyang’s attempts to conceal its activities.
‘It appears that North Korea obtained substantial centrifuge assistance from Pakistan and that it is seeking to build a clandestine production-scale centrifuge plant to produce
highly  enriched uranium for nuclear weapons’
Firstly, it is unclear exactly how much and what type of assistance Pakistan provided. Even if it supplied a full set of machines for a pilot-scale cascade, normally consisting of a few hundred machines, North Korea would need to manufacture or assemble additional machines for a production-scale facility, which typically requires a few thousand centrifuges of the type in Pakistan’s possession in order to produce annually enough weapons-grade uranium for a few nuclear weapons. Pakistan denies that it has provided any nuclear assistance to North Korea, although Islamabad is investigating several nuclear scientists who apparently sold centrifuge technology to Iran in 1987.
Secondly, and more important, it is not known whether North Korea has been able to purchase all of the components, materials and equipment necessary to assemble a production-scale centrifuge plant or to what extent it is able to manufacture such items indigenously. There is at least some fragmentary evidence to suggest that North Korea is still shopping around. In April 2003, French and German authorities cooperated to halt a shipment of 22 tonnes of highstrength aluminium tubes, the first instalment of a larger order for 200 tonnes of such tubes. The particular type of aluminium and the dimensions of the tubes closely match the requirements of the rotor casings for the G-2 centrifuge.16 Allowing for some loss in processing, the 200 tonnes of tubes could be used to manufacture around 3,500 G-2 centrifuges, enough to produce about 75kg of weapons-grade uranium a year or about enough for three nuclear weapons of a first generation uranium-based implosion design, assuming 20–25kg per weapon.17 Also in April 2003, Japanese authorities, working with officials in Hong Kong, thwarted a North Korean effort to obtain three inverters.18 These are electronic devices used to generate the direct-current power necessary for the operation of centrifuges, although they also have an application in missile guidance systems. Hundreds of inverters would be required for a production-scale centrifuge plant because each centrifuge machine is run by its own motor.
The interception of the aluminium tubes shipment in April 2003 reinforces the conclusion that North Korea is seeking to build a production-scale centrifuge facility, but these failures in Pyongyang’s procurement effort suggest that North Korea may still lack key components. Moreover, aluminium casing tubes are only the ‘tip of the iceberg’ in relation to the necessary components, materials, and equipment needed to complete a production-scale centrifuge plant. Other critical components, even more difficult to manufacture, include nuclear-grade maraging steel rotors and caps, rotor bearings, and electrical systems. Of course, it is possible that North Korea has been able to establish a completely undetected procurement apparatus to obtain such items. It seems more likely, though, that North Korea is still in the process of acquiring at least some of the necessary items. As a result, the pace of its enrichment programme could be slowed by stronger interdiction efforts, as envisaged under the new Proliferation Security Initiative (PSI).
Thirdly, the locations of North Korea’s centrifuge facility and key ancillary facilities are unknown. Intelligence and military experts believe it is likely that North Korea would choose to build a centrifuge plant underground to guard against detection and military attack. However, sites identified in press reports identifying possible locations for an underground centrifuge plant are mostly based on conjecture, unconfirmed defector reports and analysis of satellite imagery of underground facilities with an unknown purpose.19 North Korea has myriad military-related underground sites that, in theory, could house a centrifuge plant large enough to produce annually enough weapons-grade uranium for a few nuclear weapons.
In addition to the actual centrifuge facility itself, a North Korean uranium enrichment programme would also require the production of large quantities of uranium hexaflouride (UF6) feed material, the gaseous form of uranium required for centrifuges. For example, a centrifuge plant consisting of 3,500 G-2-style centrifuge machines would require around 13.5 tonnes of UF6 feed material per year. North Korea’s fuelfabrication facility at Yongbyong was able to process large amounts of natural uranium yellowcake (U3O8) into uranium dioxide (UO2) and then into uranium tetrafluoride (UF4), the immediate precursor to the production of UF6. The fluoride processing lines at the facility are badly corroded, though, and would need to be rebuilt and refitted to resume UF4 production, perhaps taking a year or so to complete. Of course, North Korea could decide to build a UF6 feed plant at another location. In the end, production of sufficient UF6 feed material should not be a major technical hurdle for North Korea’s enrichment programme, whether or not such a plant already exists is unknown.
Finally, assuming that North Korea is able to complete a production-scale centrifuge plant, a lengthy period of testing is normally necessary before fullscale sustained production can commence. Centrifuge machines are notoriously temperamental. Operating at high speeds, they can suffer catastrophic failure (known as ‘crashing’) due to manufacturing for operational errors, requiring that the entire line be shut down in order to replace or bypass the damaged machine. For example, any fluctuation in, or interruption to, the electrical current can prove fatal for centrifuge machines, and North Korea’s electrical system is known to be highly unreliable. To overcome this potential problem, North Korea would need to deploy independent generators to produce an uninterrupted power supply in the event of a failure of the national system.
In conclusion, Washington’s assessment that a production-scale centrifuge facility could be completed by ‘mid decade’ is a ‘worst case’ estimate based on analytical judgements and assumptions, rather than on a wealth of factual information. Fundamentally, the US estimate draws from a sense of how long it would take a country of North Korea’s perceived industrial, scientific and engineering potential to complete a production-scale centrifuge facility, assuming that it possessed the necessary technology and that it made a political decision to devote the necessary resources to it. It is possible that North Korea’s enrichment programme is even more advanced than US assessments suggest, especially if it has been able to obtain undetected significant quantities of materials and equipment from foreign sources. However, North Korea’s enrichment programme probably still faces a number of technical obstacles, which would put the estimated date of completion at the far end of ‘mid decade’ or even later. In the absence of additional information, it is impossible to make a decisive judgement either way.

Phase 4 – the plutonium programme unfrozen (2002–present)

‘North Korea’s enrichment programme probably still faces a number of technical obstacles, which would put the estimated date of completion at the far end of ‘mid decade’ or even later’
Following the revelation of North Korea’s clandestine enrichment programme and the collapse of the Agreed Framework in late 2002, North Korea disabled IAEA monitoring equipment at the 5MW(e) reactor, the spent-fuel storage pond and the reprocessing facility, expelled IAEA inspectors from the Yongbyon Nuclear Research Centre and took steps to revive its plutonium production programme, which had been suspended since 1994. In the absence of inspectors, it is very difficult to determine the exact status of North Korea’s plutonium production programme, although satellite photography and other forms of international monitoring do give some clues. To complicate matters, North Korean officials have made a variety of public and private statements regarding their country’s nuclear activities since inspections ended, but these may be for political effect and they cannot be taken at face value. Whatever the uncertainty about the exact status of North Korea’s plutonium production facilities, enough is known about the technical capabilities of these facilities to produce an informed assessment of North Korea’s ability to manufacture additional plutonium in the short term.
Of greatest immediate importance is the fate of the nearly 8,000 irradiated fuel rods discharged from the 5MW(e) reactor in 1994 and stored in an adjacent pond, near to the reprocessing facility. Although the IAEA was not allowed to measure the irradiation levels of the rods, it believes that they contain some 25–30kg of plutonium. Notionally, this is enough for between two and five nuclear weapons, depending on the amount of plutonium lost in the reprocessing process and on the quantity required for each nuclear weapon of North Korean design. If the spent fuel contains 25–30kg of plutonium, the amount actually recovered from reprocessing could be 17.5–27kg, assuming a reprocessing loss of 10–30%. Furthermore, assuming that a first generation implosion design requires 5–8kg of plutonium for each weapon, the separated plutonium would be enough to produce as few as two or as many as five nuclear weapons.
The reprocessing facility was mothballed for nearly eight years, but some maintenance work occurred during the freeze. Most experts believe that North Korea could have restored the facility to operational status within a few months and has likely done so since the inspectors were expelled in December 2002. In theory, the single completed line in the reprocessing plant is capable of reprocessing the 8,000 fuel rods (50 tonnes of uranium) if operations ran continuously for approximately five months, assuming no technical difficulties.
Whether North Korea has reprocessed some or all of the fuel is not known for certain. In early 2003, satellite imagery detected the presence of lorries at the storage site, suggesting that the fuel rods were being removed from the area. In theory, the fuel rods could be moved to the reprocessing facility, or potentially to some other location for processing or protection from military attack. In April 2003, North Korean diplomats told US officials in private that the country had begun reprocessing, and in July they said that reprocessing was completed. In October 2003, North Korea announced publicly that reprocessing of the rods had been completed successfully by the end of June and that the state was using the resulting plutonium to increase its ‘nuclear deterrent force’.20
The forensic evidence is ambiguous. In June, US monitoring devices located near North Korea reportedly detected slightly elevated levels of Krypton85 (Kr-85), a radioactive gas released during reprocessing. However, it is extremely difficult to quantify how much spent fuel might have been reprocessed based on these emissions. Although the US has reportedly improved its detection capabilities over the years, evaluation of levels of Kr-85 is complicated because of the presence of background Kr-85 from reprocessing operations in nearby countries, such as China, Japan and Russia, and variations in wind patterns, especially if the amount of fuel reprocessed is relatively small. It is also unknown whether North Korea might attempt to employ technical measures to reduce Kr-85 emissions at the Yongbyon facility, or, indeed, whether there is a second reprocessing facility hidden elsewhere in the country.21 As of the end of 2003, elevated Kr-85 levels have not been detected since June 2003, and analysts monitoring the Yongbyon reprocessing facility have not observed evidence of continuous operations that would indicate a fullscale reprocessing campaign.
From the available evidence, most government analysts believe that North Korea probably carried out limited reprocessing at the Yongbyon facility in June – perhaps enough to produce one or two nuclear weapons but that Pyongyang stopped short of completing a full campaign. This may have been a test run to appraise the reprocessing facility, or a political tactic to press the US to agree to negotiations. It is possible that North Korea began reprocessing and experienced technical difficulties, or that it exercised caution in response to strong private warnings from Washington that reprocessing would scuttle negotiations. In this case, Pyongyang’s public declarations in October that it had completed reprocessing could be intended to strengthen its bargaining position and provide political cover if Pyongyang decides to finish reprocessing at some point in future. An alternative view, is that North Korea’s public statements are accurate, and that, in fact, it has completed reprocessing of the available spent fuel, which could not be detected by intelligence means. Once the plutonium is separated, it would be virtually impossible to track and monitor within North Korea using available intelligence resources.
‘Once the plutonium is separated, it would be virtually impossible to track and monitor within North Korea using available intelligence resources’
North Korea’s ability to produce fresh plutonium in the near term is limited. Since the end of the freeze, it is believed to have refuelled the 5MW(e) reactor at Yongbyon, and restarted it in March 2003 – a view based on the observation of steam plumes from the reactor’s cooling tower. The reactor has apparently experienced some start-up problems, which is not surprising after eight years of inactivity, but such difficulties are not likely to be insurmountable. Assuming maximum power for 300 days, the reactor is capable of producing up to 7.5kg of plutonium per year, perhaps enough for one nuclear weapon, depending on assumptions concerning reprocessing losses and the amount of plutonium required for a nuclear weapon of North Korean design.22 In its October statement, Pyongyang vowed to reprocess spent fuel from the 5MW(e) reactor as it became available. Plutonium-bearing fuel could be discharged from the 5MW(e) reactor in spring 2004, and assuming some storage time for cooling, reprocessing of the fuel to extract plutonium could be complete by summer or autumn 2004.
In the longer term, North Korea’s ability to produce larger amounts of plutonium depends on how quickly it can complete the two larger nuclear reactors that were under construction when the nuclear freeze came into effect in 1994. Estimating the completion time for these larger reactors is difficult. It depends on how far construction had proceded by the time of the freeze, the amount of maintenance work that North Korea has performed in the interim, and the degree of effort and resources that Pyongyang is prepared to invest to finish the projects. It is also unknown whether North Korea has secretly built components or stockpiled materials for the reactors before or during the freeze.
Of the two larger reactors, the 50MW(e) was closest to completion in 1994. At that time, North Korean officials told the IAEA and the US that the reactor was 9–12 months from initial service. However, Pyongyang had an incentive to exaggerate the status of construction because the amount of heavy fuel oil (HFO) delivered to the country under the Agreed Framework was calculated according to the time when the 50MW(e) reactor was expected to be completed. Thus, after one year, the amount of HFO delivered to North Korea under the Agreed Framework was increased from 50,000 tonnes to 500,000 tonnes, the rise roughly representing the expected energy output that North Korea was sacrificing by halting construction of the 50MW(e) reactor.
No technical assessment of the reactor’s status in 1994 was conducted. According to IAEA inspectors who visited the reactor throughout the period of the freeze, external work on the main reactor building was complete and the reactor pressure vessel was installed. IAEA inspectors also accounted for slightly more than half of the graphite blocks needed for the reactor core, which were stored in a nearby warehouse, and about one-third of the fuel pieces required for the initial fuel load. These are tagged and sealed under the terms of the Agreed Framework. However, the IAEA was not
Basic implosion design
able to determine the status of several major pieces of equipment and components essential for completing the reactor, such as the fuel-loading machine and the blowers for circulating CO2 coolant.
Since the end of the freeze in late 2002, there have been no reports of increased activity at the 50MW(e) reactor suggesting that construction has resumed, although some work on essential components could be occurring undetected off-site. Even assuming that key components such as the fuel loading machine and blowers are ready for installation, it would still likely take a few years to complete the reactor. Fuel fabrication, in particular, could pose a delay. Some of the metal parts and equipment used on the process line at the Yongbyon fuel fabrication plant have been badly corroded by fluoride residues because they were not properly cleaned and maintained during the 1994–2002 freeze. Consequently, new equipment would need to be installed before fuel fabrication could resume. Once production recommenced, it would take approximately one year to produce the remaining fuel for the initial core load, unless North Korea has undeclared stocks of fresh fuel or even an undeclared fuel fabrication plant. Finally, once the reactor is complete and fully fuelled, one would normally expect a period of testing before any attempt is made to sustain full-power operations.
‘Assuming no major technical problems, and that Pyongyang takes the political decision to complete the
50MW(e) reactor, full-power operations could be underway in a few years at the earliest’
Thus, assuming no major technical problems, and that Pyongyang takes the political decision to complete the 50MW(e) reactor, full-power operations could be underway in a few years at the earliest. Resumption of large-scale construction at the site and installation of critical equipment and components would probably be detected by satellite, giving advance warning that North Korea was seeking to finish the reactor. In the same timescale that it would take to finalise the 50MW(e) reactor, North Korea could also probably expand its reprocessing capability by completing the second line at its reprocessing facility at Yongbyon. This would provide surplus redundant capacity to process the normal 100-tonne annual spent-fuel discharge from the 50MW(e) reactor. In theory, operating at full power for 300 days a year, the 50MW(e) reactor could produce about 55kg of plutonium per annum, enough for about five to ten nuclear weapons, depending on reprocessing losses and the amount of plutonium required for each weapon of North Korean design.23 Operating at lower power levels for shorter periods would generate correspondingly less plutonium. Of course, there is no way to know the actual performance capability of the 50MW(e) reactor.
Compared to the 50MW(e) reactor, the much larger 200MW(e) power reactor at Taechon was at an early stage of development in 1994, and has suffered from poor maintenance and exposure to the elements during the period of the Agreed Framework. Indeed, some experts believe that it may be a complete write-off. As far as is known, none of the key components, graphite blocks, or fuel for the reactor has been fabricated. Furthermore, North Korea would need to construct a new reprocessing line to separate plutonium from the reactor’s spent fuel. Notionally, the 200MW(e) reactor could produce hundreds of kilograms of plutonium annually, enough for tens of nuclear weapons, but there seems little prospect that it could be completed for many years.
Nuclear weapon design and fabrication
There is virtually no substantial information on North Korean efforts to design and manufacture nuclear weapons, although certain assumptions can be derived from the basic principles that apply to all countries. The common assumption is that North Korea’s nuclear weapon design is based on a first generation implosion device, the logical choice for states in the initial stage of nuclear weapon development.
In a first generation implosion device, such as the atomic bomb dropped on Nagasaki in 1945, a solid ball or core of fissile metal (either plutonium or high enriched uranium (HEU)) surrounded by a metal tamper/reflector (usually natural uranium) is compressed by a spherical system of shaped highexplosives, known as a lens. To produce super criticality, a burst of neutrons is introduced at a key instant of compression. The main technical challenge lies in creating a spherical implosion of highexplosives, which requires precise fabrication of the high-explosive lens and exact timing. Failure to achieve this could result in significant loss of nuclear yield or even a dud. Another technical challenge concerns the design of the neutron generator needed to release a burst of neutrons to trigger a chain reaction in the compressed fissile core. The Nagasaki weapon had about 6kg of plutonium in its core, and produced a yield of slightly more than 22 kilotons. Overall, the weapon was some 1.5 metres in diameter, 3.6 metres long, and weighed approximately 4.9 tonnes. Over the years, the advanced nuclear-weapon states have developed a number of different techniques to reduce the amount of plutonium or HEU needed to achieve a desired yield, and to decrease significantly the size and weight of implosion weapons.
An alternative to an implosion design is a gun type device in which a smaller piece of weapons-grade uranium is fired into a larger piece of weapons-grade uranium in order to create a supercritical mass. A chain reaction is initiated with the introduction of a burst of neutrons at a key moment. Unlike an implosion device, which can have either plutonium or uranium as its fissile core, a gun-type device can only be built with HEU because the spontaneous neutrons emitted by plutonium are likely to cause premature criticality (that is, before the nuclear core is fully assembled), significantly reducing the overall explosive yield. Guntype devices are generally simpler to design and construct than implosion devices, mainly because the high-explosive system used to assemble the critical mass is less complex, but they require considerably more nuclear material to achieve the same yield as is produced by an implosion device. Consequently, implosion designs are generally more attractive to countries with limited amounts of nuclear material. The first gun-type weapon was that dropped on Hiroshima in 1945, which contained about 60 kilograms of HEU and produced a yield of between 12 and 15 kilotons. The overall bomb was 71cm in diameter, 3.4m long, and weighed around 4.3 tonnes.
In estimating the number of nuclear weapons that North Korea might be able to produce, this study assumes that it would require between 5–8kg of weapons-grade plutonium and 20–25kg of HEU for each implosion device, which roughly corresponds to the range of fissile material used by the nuclear-weapon states in their early designs. Using more advanced techniques or aiming to achieve lower yields, nuclear weapons can be built with smaller amounts of plutonium or weapons grade uranium. This study, though, presupposes that North Korea may not have access to such advanced techniques and, therefore, is more likely to pursue simpler and more reliable designs in the range of 10–20 kilotons. However, without knowing the details of North Korea’s nuclear-weapon design the actual amount of fissile material used in such a device cannot be determined. The ranges posited in this study cover the most likely possibilities.
Since at least the mid-1980s, North Korea has conducted a series of high-explosive tests, which appear to be related to the development of an implosion system for a nuclear weapon. Prior to 1992, North Korea carried out high-explosive nuclear-related development tests at the Yongbyon Nuclear Research Centre in a nearby stream bed. According to a KGB report of 22 February 1990, leaked to the Russian press in March 1992, the Soviet intelligence agency had already concluded that North Korea had succeeded in developing a ‘nuclear explosive device’ at the Yongbyon Nuclear Research Centre.24 The IAEA visited this test site during its various inspections of the Yongbyon establishment in 1992, but it found no evidence of nuclear materials. Later, high-explosive tests were conducted at a nearby site with more sophisticated facilities, known as Youngdoktong. According to a South Korean intelligence report leaked to the country’s National Assembly, satellites detected some 70 high-explosive tests at Youngdoktong.25
It is difficult to evaluate these North Korean tests with the information available. High-explosive testing for nuclear weapons development has certain distinct characteristics that help distinguish it from highexplosive testing for the development of conventional military ordnance – although any work with shaped charges has some similarities. In theory, using surrogate material for the fissile core, such as natural or depleted uranium metal, such tests can be used to develop an effective nuclear weapon design without the need for a full nuclear test. Whether these tests indicate that North Korea is having difficulty establishing a reliable system, or is seeking to improve on an existing design or to develop new design types, cannot be determined from the available evidence. Given the length of time over which North Korea has apparently conducted nuclear-related high-explosive tests, its ability to manufacture shaped high-explosive charges for conventional munitions, and availability in the public domain of basic information on early implosion designs, the US has believed – since the mid1990s – that North Korea is capable of designing and building a simple implosion-type nuclear weapon, assuming that it has sufficient stocks of plutonium or highly enriched uranium for such a device. Since North Korea has continued high-explosive testing over the last decade, the current US assessment, that North Korea has built ‘simple fission-type’ nuclear weapons without nuclear testing, has become more confident.26
If this assessment is correct, a key uncertainty concerns the size and weight of the nuclear weapons, which determine the means of delivery. Clearly, from Pyongyang’s standpoint, it would be highly desirable to develop a nuclear weapon small enough and light enough to be delivered by the missiles in its inventory, such as the No-dong, which is likely to be more survivable and effective than military aircraft. North Korea’s ability to threaten targets beyond the Korean Peninsula, such as Japan or, eventually, the US, would be much more credible if it was able to deliver a nuclear warhead using missiles in its inventory. In this regard, the question of Pakistani assistance is critical. According to press accounts, US intelligence believes that Pakistan may have provided North Korea with nuclear weapon design information and even supplies of high enriched uranium (HEU) under the missile-for-nuclear barter agreement of the late 1990s.27 With North Korean and Pakistani nuclear and missile personnel apparently working closely together for several years, it is plausible that some discussion of weaponisation would take place. Pakistan’s nuclear weapon design, an implosion system utilising HEU instead of plutonium, is based on an early Chinese design, and it is small and light enough to be delivered using the No-dong missile – one of the reasons why Pakistan wanted to acquire the No-dong. In the worst case, if Islamabad provided North Korea with nuclear weapons design information, it would substantially assist North Korean efforts to develop nuclear weapons that could be delivered by No-dong missiles. Even without Pakistani assistance, North Korea may have been able to develop a warhead over the past decade that is small and light enough to be delivered using a No-dong missile, but there is insufficient information to make a confident assessment.
‘The US has believed – since the mid-1990s – that North Korea is capable of designing and building a simple implosion-type nuclear weapon’
In recent years, a few North Korean defectors have come forward claiming that the country possesses nuclear weapons, but much of their information is second-hand and cannot be confirmed. An example  of this type of information comes from former Secretary of the North Korean Workers’ Party, Hwang Chang Yop, and his assistant, Kim Duk Hong, who defected in 1996. Speaking in 1999, Kim said that Jong Pyong-Ho, a senior party official in charge of military matters, had told Hwang in 1996 that North Korea had five plutonium-based nuclear weapons.28 Even assuming that Jong actually told Hwang that North Korea possessed nuclear weapons, the possibility remains that the regime would want to create the impression among the party elite that North Korea had a nuclear deterrent in order to maintain internal morale. Other defectors have made more extravagant statements, such as a North Korean who said that he was a general in the Korean People’s Army and claimed that North Korea had ‘dozens’ of nuclear weapons.29 Since the collapse of the freeze, Pyongyang has issued a variety of public and private statements intended to reinforce the impression that North Korea has nuclear weapons, although the regime has not issued a formal public declaration to that effect. The current public formulation is that North Korea possesses a ‘nuclear deterrent force’. Of course, Pyongyang has a strong interest in creating the impression that it has nuclear weapons, so its private and public statements are not definitive either way.
In conclusion, the current US assessment that North Korea ‘has one or possibly two’ nuclear weapons is based on analytical judgements that North Korea has sufficient fissile material and is technically capable of building a simple implosion device, without requiring a full nuclear test, and that Pyongyang has made the political decision to exercise its nuclear option. The original basis for these judgements were developed during the 1993–94 nuclear crisis, and the judgements have become more confident over time. If analysts judged that North Korean scientists and technicians could probably build a first generation device in the early 1990s, it makes even more sense that they could do so a decade later. High-explosive testing has continued during that period, and Pakistani experts may have provided assistance to help North Korea develop a nuclear warhead deliverable by the No-dong missile. Whether North Korea actually has nuclear weapons, of course, is not known, but it is impossible to be confident that it does not.
Conclusion
North Korea’s current and projected nuclear-weapons capability depends on several key factors. How much nuclear weapons usable material (either separated plutonium or highly enriched uranium) does it possess? How much additional plutonium and HEU will it be able to produce in the future and in what timeframe? What is its capacity to design and fabricate nuclear weapons from its current and projected stocks of nuclear material? In particular, how ‘deliverable’ would such weapons be?
As with most issues concerning North Korea, there is no definitive answer to any of these questions. The information needed to answer these questions either cannot be obtained or is ambiguous and fragmentary because North Korea has gone to great lengths to conceal its nuclear activities. Moreover, Pyongyang has actively tried to shape the perceptions of the outside world in one direction or another. In the early 1990s, Pyongyang tried to emphasise the civilian intent of its nuclear programme and sought to downplay the extent of its nuclear capabilities. Since the end of the nuclear freeze in 2002, North Korea has seemingly tried to broadcast its nuclear strength to reassure the party faithful and to deter and intimidate perceived enemies. Whether these statements constitute a boast or a bluff cannot be determined.
During the 1993–94 nuclear crisis, the US assessed that North Korea could have produced enough plutonium prior to 1992 for ‘one or possibly two nuclear weapons’. This judgement was based on informed analysis rather than direct empirical evidence of how much plutonium North Korea possessed. That North Korea was pursuing a nuclear-weapons programme throughout the 1980s was clear, despite Pyongyang’s belated efforts to justify its programme in terms of nuclear energy production. Moreover, the ‘discrepancies’ in North Korea’s declarations to the IAEA and its refusal to allow access to suspect nuclear waste sites supported the conclusion that it was hiding some plutonium. US experts were able to construct possible scenarios in which North Korea could have manipulated the operations of the Soviet-supplied IRT-2000 reactor and the 5MW(e) reactor to produce something in the range of 8–12kg of plutonium, enough, in theory, for one or possibly two nuclear weapons of the type that North Korea was assumed to be able to manufacture. Given North Korea’s long history of nuclear-related high-explosive testing, which began in the mid-1980s, it seemed plausible that North Korea could produce first generation nuclear weapons, assuming that enough plutonium was available.
‘Whether North Korea actually has nuclear weapons, of course, is not known, but it is impossible to be confident that it does not’
Throughout the confrontation with the IAEA and the subsequent negotiations with the US over the Agreed Framework, it appeared probable that Pyongyang would not go to such lengths to avoid revealing its plutonium holdings unless it was protecting a quantity of strategic significance – that is, enough for its first nuclear weapon. At the same time, one cannot rule out the possibility that Pyongyang’s primary objective was to sustain strategic ambiguity. Whatever its actual plutonium holdings, as long as Pyongyang maintained the perception that it could have enough for one or two nuclear weapons, which was the essence of Washington’s public statements during the 1993–94 crisis, it would enjoy some degree of nuclear protection. Over time, the US view that North Korea had ‘one or possibly two nuclear weapons’ has grown more confident, although the US does not have proof that North Korea has nuclear weapons.
Whatever its current nuclear inventory, North Korea clearly has the capacity to produce enough fissile material for nuclear weapons in the future – only a few in the near term, but a larger quantity over a number of years. The most immediate threat is from the roughly 8,000 spent fuel rods from the 5MW(e) reactor, which the IAEA estimates contain about 25–30kg of plutonium, enough for between two and five nuclear weapons, assuming a reprocessing loss range of 10–30% and that 5–8kg of plutonium is required for each weapon. The status of this plutonium is unknown. North Korea reportedly began reprocessing this fuel in June 2002, but may not have completed a full campaign, for technical or political factors. North Korea’s repeated assertions – first made in private and now publicly – that it has completed reprocessing all of the fuel rods are open to interpretation. Pyongyang may be telling the truth or may be seeking to bolster its perceived nuclear deterrent or it may be trying to create the conditions to finish the job when it believes that the political circumstances are right.
In addition to the plutonium in North Korea’s existing stockpile of spent fuel, North Korea has also restarted the 5MW(e) reactor, which, theoretically, can produce enough plutonium for approximately one nuclear weapon per year. However, the reactor may be experiencing some technical difficulties in achieving full power for sustained periods.
Assuming, therefore, one or two nuclear weapons from plutonium separated before 1992, between two and five nuclear weapons from plutonium in North Korea’s existing spent fuel, and approximately one additional bomb’s worth annually from plutonium produced by the 5MW(e) reactor, North Korea’s maximum nuclear arsenal is likely to be limited to 6–12 nuclear weapons over the next several years, if no new facilities to produce plutonium or HEU are completed. This assessment does not include the possibility that North Korea acquired additional nuclear weapons useable material from foreign sources, such as weapons-grade uranium from Pakistan.
‘Whatever its current nuclear inventory, North Korea clearly has the capacity to produce enough fissile material for nuclear weapons in the future – only a few in the near term,
but a larger quantity over a number of years’
In the longer term, North Korea’s ability to produce significantly larger quantities of fissile material depends on whether it can complete the 50MW(e) reactor and its presumed centrifuge enrichment plant. It is impossible to predict accurately when these facilities might be completed. From available information, it appears most likely that the 50MW(e) reactor is probably at least a few years from completion and full operation. However, the status of key pieces of equipment are unknown, as is whether the reactor will be able to operate as designed. Presuming that it is completed and operates as designed, the 50MW(e) reactor could produce up to 55kg of plutonium annually, enough for five to ten nuclear weapons, assuming a 10–30% reprocessing loss rate and that between 5–8kg of plutonium is required for each weapon.
Even less is known about the enrichment project. Publicly, the US claims that a production-scale centrifuge facility that is able to produce enough weapons-grade uranium for ‘two or more nuclear weapons per year’ could be operational as soon as ‘mid-decade’. It is not known, though, whether North Korea has been able to obtain the equipment and materials necessary to complete such a facility or the extent to which it can produce such items indigenously. A more conservative estimate is that completion of the plant could be delayed until the end of the decade, especially if interdiction efforts (several of which took place in 2003)  can slow the acquisition by North Korea of foreign equipment and materials.
In short, it is impossible to reach a firm conclusion about North Korea’s current nuclear weapons capability. On the one hand, a plausible case can be made that North Korea has enough plutonium for a very small number of nuclear weapons, including plutonium that it may have separated before 1992 and plutonium that it may have separated since 2002, and that it is technically able to manufacture a deliverable nuclear weapon from this plutonium. On the other hand, we cannot confirm how much plutonium North Korea has and whether it is able to fabricate a deliverable nuclear weapon from this material. From a public policy standpoint, and given the stakes involved, the case is strong enough that it would be imprudent to conclude that North Korea does not have nuclear weapons.

0 মন্তব্য(গুলি):

একটি মন্তব্য পোস্ট করুন