【May 27, 2019】Methods for Refining Estimates of Cumulative DRPK Uranium Production (PSNA-WP-8)

Category : PSNA Activities

Methods for Refining Estimates of Cumulative DRPK Uranium Production

David von Hippel
Nautilus Institute Senior Associate

PSNA Working Paper Series (PSNA-WP-8)1

May 27, 2019

[PDF version]


This paper summarizes the history of what is known about uranium mining in the DPRK; describes the major uncertainties regarding DPRK uranium production; notes some of the key techniques, as used in preparing estimates of nuclear sector activity in other nations, that might be available to assist in narrowing the range of estimates of DPRK uranium and processed fissile material production; summarizes estimates of enriched uranium and plutonium production prepared by other authors; describes existing estimates of the amounts of fissile materials used in nuclear weapons tests and exported; provides a demonstration of the potential impact of remote sensing methods and testing in the DPRK in reducing uncertainties in cumulative historical uranium production, and thus in fissile materials inventories; and, offers conclusions, resulting from the potential impacts of uncertainty reduction approaches, as to which verification procedures should be key targets during negotiations with the DPRK.

1 Introduction and Statement of Problem

Recent weeks and month have seen what appears to be a series of improvements in relationships between the Democratic People’s Republic of Korea (DPRK) and the international community. In particular, summit meetings between Republic of Korea (ROK) President Moon and North Korean leader Kim Jong Un, a meeting between Kim Jong Un and United States President Donald Trump, and advancing planning for a second Trump-Kim summit have raised hopes for progress in talks to “denuclearize” the DPRK. Reaching agreement on, and then implementing a process of dismantling the DPRK’s nuclear weapons production facilities, and of removing from the Korean Peninsula (or otherwise isolating under international oversight) the DPRK’s nuclear weapons and fissile material, is likely to take years, perhaps more than a decade. In the meantime, it is important that the international community be able estimate as accurately as possible the amount of uranium—enriched and natural—that has been produced and used in the DPRK over the decades since its first foray into nuclear technologies, starting in the 1960s or before,2 as well as the amount of plutonium that has been produced in irradiated fuel and separated via reprocessing. Determination and verification of the potential quantity of fissile material existing in DPRK as a result of mining and processing activities in advance of the DPRK’s own declaration of its inventory of fissile materials will require a combination of remote sensing analysis and on-the-ground tests at DPRK uranium mines and other facilities to narrow the range of estimates of historical uranium production, and of the resulting inventory of fissile materials likely to exist in the DPRK.


2 Description of history (or what is known about it) of U mining in the DPRK

2.1 Mining sites: location, extent, dates of known mining

Different researchers and sources offer different opinions on the number of mines from which the DPRK has extracted uranium. It has been reported that uranium has been mined to supply the DPRK’s domestic nuclear industry from more than a dozen mines located in various areas around the country, including Pyongsan, Pakchon, Hongnam, Jusong, Ungki, Sunchon, Cholsan, Kujang, Hamheung, Hekumkang, Wolbisan (possibly same as Wolbingsan, or Kŭmch’ŏn), Hwangsan, Cheonmasan, Wiwon, Musan, and Najin.3 Another source refers to a uranium mine near Hungnam (probably the same as “Hongnam”), where the Japanese built a cyclotron in 1943-44.4

In a report prepared for Nautilus by Edward Yoon,5 two major uranium ore mines in the DPRK: the Pyong-San mine, and the Woong-Gi mine, are described as below.

•    “Pyong-San Uranium Mine: This mine is located in Pyong-San–Gun, Hwang-Hae province and has been operating for 30 years under the control of the People’s Army Department. The deposit in this mine area was estimated at 1.5 million tonnes (as Uranium ore), and the mine’s annual production capacity is 10,000 tonnes. The mine has own separator for concentration of ore. All products are sent to the Nyung-Byun (Yongbyon) Nuclear power station under armed guards. Recently, a new facility for Uranium extraction has been built in the Pyong-Won area.

•    Woong-gi Uranium Mine: This mine is located in Woong-gi, Ham-Kyung province and has been operating for 35 years under the control of the People’s Army Department. The deposit in this mine area was estimated at 10 million tonnes (as Uranium ore), and its annual production capacity is 19,000 tonnes. The mine has its own separator for concentration of ore. All products are sent to the Nyung-Byun [Yongbyon] Nuclear power station under armed guards. The mine’s operation has been kept secret from outsiders, and even from North Koreans, due to the fact that output from the mine is known to have been used for nuclear weapon development purposes. As a result, the workers and engineers in the mine have been reportedly restricted to the area within the mine facilities even if they suffered from nuclear radiation-related disease.”

2.2 Processing sites and complexes

A recent report by 38 North includes references to a uranium concentration plant located alongside the Pyongsan uranium mine (and nearby mines), about 100 km southeast of Pyongyang, as well as another facility, referred to as “the Pakchon Uranium Concentration Pilot Plant”, located in North Pyongan province.6 The Pyongson uranium concentration plant, of which satellite images over time suggest has processed uranium in a number of periods over recent years, is shown in Figure 1. The Nuclear Threat Initiative describes this plant as “having [a]bout 1,500 workers; as of 1999, [and] a capacity to process 200,000 tons of uranium ore into of 290 tons of yellow cake per year”.7 Figure 2 shows an image of the large building (about 100 by 120 meters) located at the latitude/longitude coordinates identified by Jeffrey Lewis in a 2015 38 North publication as those of the Pakchon Uranium Concentration Pilot Plant.8 This location appears to show limited signs of recent activity, though the lighter circles of earth in the upper left of the photo could indicate that structures were recently removed.

Figure 1: Uranium Concentration Plant at Pyongsan, DPRK9

Figure 2: Uranium Concentration Pilot Plant at Pakchon, DPRK10

The Nuclear Threat Initiative (2018, ibid) lists a third uranium processing site, the Cheonmasan Uranium Milling Facility (also listed as a “Suspected Uranium Enrichment Facility”), which it suggests (variously) has 535 or 670 employees. NTI describes the facility as “located underneath Mt. Cheonma”. Joseph Bermudez lists four milling sites, those listed above at Pyongson (the “Namchon Joint Chemical Industrial Company”) and Pakchon (the “April Industrial Enterprise”), one at Sunchon, where other sources have indicated mining has taken place, and one at Kusong used for both mining and milling.11 Bermudez also lists mines at both Sinpo and Unggi, although these may be covered in the list of mines above under different names.

2.3 Ore quantity and quality

Figures on the DPRK’s reserves of uranium and the quality of its ore are difficult to obtain, and their accuracy is unknown.

Two sources suggest that the DPRK’s uranium deposits “are estimated at 26 million tons”12,13. One of the sources describes these deposits as “high grade ore”, so it seems virtually certain that the references are to tonnes of ore, not tonnes of uranium metal (or uranium oxides). Another source states:

“It has been estimated that, at its peak in the early 1990s, North Korea was able to produce about 300 tonnes of yellow cake [U3O8] annually, equal to approximately 30,000 tonnes of uranium ore.”14

The information from this source implies that the DPRK’s uranium ore has (or had) an average U content of about 0.9 percent, which is quite high—most uranium ore in the Northeast Asia region has an average U content of closer to 0.2 percent or less15—so this estimate may be in error. Other analysts of the subject have reported estimates of 3 and 4 million tonnes of “reasonably assured resources”, based on older OECD and ROK estimates, respectively. Still another source cites a figure of 4.5 million tonnes of uranium ore, and quotes “Russian scientists who have visited North Korea” as saying that the DPRK’s “mining and milling capabilities produce 2000 tons of natural uranium, per year”.16 A 2004 Nautilus report described DPRK deposits as “uraniferous black shale occurrences (perhaps similar to that at Ok´chon in South Korea) occurring at a depth about 200 meters. The ore grades are about 0.2%”.17

The Nuclear Threat Initiative lists the uranium ore extracted from the “Kŭmch’ŏn Uranium Mine” as being reported by a defector as having “a uranium content of 0.8 percent, a vanadium content of 1.4%, and other rare metals such as nickel, molybdenum and radium”. The same uranium content is listed by NTI for the Pyongsan mine, although the figures in the quote above from the NTI reference, 290 tons yellowcake from 200,000 tonnes ore, would seem to imply a U content of more like 0.2 percent, assuming about 20 percent losses in processing. The Nuclear Threat Initiative lists the Hamhung Uranium Deposit as having “approximately 4 million tons of uranium deposits”. In his 2015 (ibid) 38 North article on NK U mining and milling facilities, Jeffrey Lewis includes a 1979 quote from a telegram from the Hungarian ambassador to the DPRK, in turn quoting a Soviet source, and giving the average ore quality at the two main NK U mines at the time as “0.26 percent, while in the other it is 0.086 percent”.18 Lewis goes on to say “[b]ased on other information released by the Soviet Union, it appears these mines are near Pakchon and Pyongsan, with Pyongsan likely having the higher quality ore.” These values are considerably lower than the 0.8 to 0.9 percent estimates described above, but closer to international averages.

Reports on a pair of missions (1987 and 1990) undertaken by International Atomic Energy Agency (IAEA) experts at the request of the DPRK government,19 indicate that at the time DPRK engineers were working with fairly rudimentary uranium prospecting equipment, often of Russian origin and decades old, although DPRK government news releases in the last few years have praised progress in uranium processing technology.20

2.4 Trade in uranium and related ores with other nations (Russia, China)

The DPRK is reported to have exported significant amounts of uranium ore over the years, starting in (at least) the 1947-1950 period, with the export of “over 9,000 tons of uranium [presumably ore] and an unknown amount of monazite to the USSR”, and continuing with a reported “$6 billion worth of uranium ore” to the USSR in 1985, “1,500 tons of monazite21 annually” in the 1990s to “China, Japan, Spain, and Hong Kong”22. More recently, an advertisement by the DPRK’s International Chemical Joint Venture Corporation was published in an English-language DPRK trade journal in 2001 and 2002 advertised ammonium diuranate (ADU), a processed form of yellowcake (U3O8), for sale on the international market23. A report in late 2006 that the DPRK and Russia had been negotiating, apparently since 2002, a deal that would give Russia “exclusive rights” to the DPRK’s uranium deposits “in exchange for Moscow’s support at six-party talks aimed at denuclearizing Pyongyang” suggested that Russia would enrich DPRK uranium for re-export to Vietnam and China as nuclear fuel. The report was dismissed as “rumors” by Russian authorities24.

Exports from the DPRK to China of 90.54 and 34.9 tonnes of “Thorium Ore and Concentrate” were listed in China Customs statistics (and the United Nations Comtrade database) for the years 2004 and 2007, respectively. The listed values for these shipments, about $22,000 and $7,400 USD, suggests that the exports were of ore, not refined metal, or yellowcake, which would have been 200 to 1000 times as costly. Uranium and thorium exports from the DPRK to China are not listed for other years between 1996 and 2017, but trades with China in the same category (thorium) with a total value of about $95,000 were listed in 1994 and 1995.25 No quantities are listed for the latter trades, but assuming thorium ore at costs per tonne similar to the trades in 2004 and 2007, the 1994 and 1995 trades would total about 400 tonnes of ore. Note that this review of Comtrade data certainly does not rule out other, off-the-books trades in uranium and/or related ores between the DPRK and other parties.

2.5 Existing estimates of total uranium production as prepared by others

We have thus far been unable to identify existing estimates of the cumulative production to date of uranium in the DPRK, either as refined yellowcake or as mined ore. This is likely because most analysts have focused on the estimation of the DPRK’s cumulative production and inventories of fissile material, namely highly enriched uranium and plutonium (see below). Inventories of these fissile materials imply minimum historical production levels of uranium ore and of the refining of ore into uranium oxide and/or metal, but cannot encompass uranium or ore that might be in storage, and available for processing into feed for enrichment facilities or reactors.

3 Major uncertainties in preparing estimates of historical DPRK uranium production

Barring past or future imports of uranium or plutonium, the total uranium production over time (meaning since the beginning of significant mining) in the DPRK serves to set an upper limit on the DPRK’s potential production of nuclear materials for weapons. Calculating cumulative DPRK uranium production, however, requires the knowledge of a number of factors, many of which, as indicated above, are uncertain. The key uncertainties involved are summarized below.

3.1 Volume of ore removed

Once a uranium ore deposit has been identified and mine development has taken place, the quantity of ore produced by the mine is a determinant (but not the only one) of the amount of uranium extracted. In the case of the DPRK, as indicated above, there are uncertainties as to how many mining sites have seen significant production, what the capacity of many of the sites are (although we have some estimates for major mines), and how much, on average, the mines have operated relative to capacity (average capacity factor) over the years. All of these factors are required for an accurate total estimate of the volume of ore removed, and thus should be the focus of various on-the-ground and remote methods of reducing uncertainty.

3.2 Ore quality

As noted above, there are a wide range of reported ore quality (fraction of uranium per unit weight of ore) in the DPRK. If the average is indeed closer to 0.8 to 0.9 percent, as opposed to 0.2 percent or lower, the amount of uranium mined when a given volume of ore is removed could be a factor of four or more higher. As such, obtaining a better understanding of the ore quality in the DPRK, probably through sampling of ore bodies, mined ore samples, and/or ore processing wastes, is likely to be required to reduce this uncertainty.

3.3 Uranium remaining in wastes after processing

Also uncertain is what fraction of the uranium in the ore mined in the DPRK remains in the product yellowcake. Uranium ore can be processed in a number of different ways, with the choice of technology depending on factors including the concentration of uranium minerals in the ore, the composition of the other minerals in the ore, and, of course, the technologies and other inputs (such as chemicals for processing) available. Different processing approaches, and the way that processes are operated, yield different levels of uranium losses, that is, different amounts of uranium end up in tailings and other wastes, as opposed to the product yellowcake. Uranium losses during processing can range from 10 to 30 percent,26 and reducing the uncertainty as to the loss rate is helpful in identifying, how much ore has been processed and how much concentrated uranium has been produced, particularly if measurements of the uranium content of tailings and other wastes are an input to the determination.

3.4 Unreported exports to other nations

Also unknown is how much of the DPRK’s uranium, whether as ore, yellowcake, or uranium hexafluoride (UF6) might have been exported. It has been reported that the DPRK shipped 1.6 tons of (UF6) to Pakistan, which was later shipped on to Libya,27 and likely provided at least technology for the Syrian plutonium production reactor destroyed by Israel in 2007, though whether the DPRK provided uranium for that reactor is not clear. Various analysts suggest that these known instances of nuclear cooperation with other nations are only a part of the DPRK’s trade in nuclear technologies. As such, unreported trades of uranium, beyond the few recorded “on books” trades and the several sales of ore to Russia, China and others, as described above, would affect the possible inventories of fissile material that remain in the DPRK

3.5 Uncertainties as to whether DPRK has uranium in storage

The DPRK may have uranium in storage as ore, or, more likely, as yellowcake or other forms of uranium concentrates, which requires a much smaller storage volume, and is easily concealed. To the extent that these stores exist, they could complicate efforts to understand the DPRK’s holdings of uranium. Although uranium in storage would presumably be accounted for if an accurate understanding of ore produced and processed could be obtained, in practice it is likely that the uncertainties in ore volumes produced and processes will accommodate a large range of potential volumes of concentrated uranium in storage.

4 Techniques and tests to try to narrow uncertainties

Given that the uncertainties identified above are multiplied with each other when one attempts to use data from the uranium production and use chain to help calculate the uranium that the DPRK (or another nation) has on hand, it is desirable to reduce uncertainties as much as possible at each level. Some of the methods for doing so are discussed below. Some uncertainty reduction methods can be undertaken through the use of satellite images, but others will require access to uranium-related sites.

4.1 Satellite imagery measurements of ore slag piles, and other features near mines and mills

In a recent summary of satellite and other remote methods of assessing uranium mining and milling, Jeffrey Lewis and co-authors include the use of changes of geographical features seen in series of satellite images over time, using Synthetic Aperture Radar (SAR) to look at small changes in elevation over time, and in some cases to see inside structure, and using “hyperspectral imagery” to possibly identify specific minerals.28

•    Mining activity, and in particular underground mining, typically results in the accretion of piles of spoil—the rock in which ore is found—in the vicinity of the mine. Satellite images of an area over time can be used to identify areas where mining is active, and, in some cases, estimate the volume of material removed, though sometimes such estimation is complicated by complex terrain and other factors. Additionally, uranium milling releases tailings, the minerals remaining as waste after uranium (and in some cases, other valuable metals, such as vanadium) are removed. These tailings may end up in piles or ponds near the mill, and can be tracked from space. Lewis used these techniques in a 2015 paper looking at recent DPRK uranium mining and milling activity.29 Others at the Center for Nonproliferation Studies have used these techniques to estimate the capacity of the DPRK’s Pyongsan uranium milling facility, obtaining estimates of between 273 and 885 tonnes of uranium per year, depending on assumptions about ore grade and on the number of counter current decantation units assumed to be contained in buildings (and therefore not visible in images) at the Pyongsan site.30 An extension of human analysis of satellite images is to use detection algorithms to screen series of images for changes indicative of mining activity. To use detection algorithms, a researcher would identify a physical area of interest within a satellite image, then use one (or hybrids) of a range of techniques to analyze the digital data in the satellite image to detect changes in that area over a temporal series of images. Some of the analytical techniques that can be employed to look for differences occurring in an area over time include image differencing, ratio of means, change vector analysis, inner product and spectral correlation analysis, and multi-variate correlation detection.31

•    Synthetic Aperture Radar is capable of imaging areas with resolution to the tens of centimeters, and is sometimes capable of seeing through thin roofing material to detect and size equipment within buildings. It can also be used to detect changes in piles of material, such as mine spoils or tailings, or change in roads or other infrastructure.

•    Hyperspectral sensor capture light reflecting from objects over a wide range of spectra, in both the visible and non-visible ranges. Hyperspectral images, together with processing software, can be used to look at “dozens or hundreds” of spectral bands, and thus to discern between different types of ground cover and minerals. A recent review of the use of hyperspectral remote sensing for mineral exploration includes the following description:32 “Hyperspectral remote sensing combines two sensing modalities: imaging and spectroscopy. An imaging system captures a picture of a remote scene related to the spatial distribution of the power of reflected and/or emitted electromagnetic radiation integrated over some spectral band… [at the same time] spectroscopy measures the variation in power with the wavelength or frequency of light, capturing information related to the chemical composition of the materials measured. The instrumentation used to capture such spectral information is called an imaging spectrometer or a hyperspectral sensor…”. The spectral information from surface features is compared with the known spectral properties of minerals—for example, uranium or the minerals in which uranium is found in or with—and the spatial distribution of the minerals of interest are mapped through the analysis of the acquired hyperspectral data. An example of a hyperspectral image processed for minerals detection is provided in Figure 3. analysis has been used to evaluate uranium deposits in Australia on a test basis.33 Case studies of the large Ranger uranium mine matched spectra of known minerals with those found in hyperspectral images of known mines, and a case study of uranium/rare earth element deposits in Queensland used techniques such as Spectral Angle Mapper, Mixture Tuned Matched Filtering. Additional details on these investigations can be found in the source references. At present, the drawback of this technology is that adequate hyperspectral coverage is not currently available by satellite, as NASA’s Hyperion orbital sensor, the most readily available source for such data, functioned from 2000-2017. New images thus require flyovers by planes or drones fitted with suitable detectors, which is only possible in some circumstances (not, currently, in the DPRK). A technical issue is moving the large amounts of spectral data from space to earth-based receivers. Several ventures, are working on the data transfer problem for a variety of potential applications of hyperspectral data.

All three of these approaches—evaluation of conventional satellite images, of SAR data, and of hyperspectral data—can be combined with machine learning (including “deep convolutional neural networks”) and ground truthing at known facilities to develop and refine algorithms that allow computers to process image data to identify potential sites and changes in sites over time. Research in this area is ongoing. In addition, satellite imagery can be used to preliminarily identify sites that should be visited on the ground to confirm or rule out the presence of uranium-related activities such as mining or refining (and possibly enrichment. For those sites determined to be related to uranium fuel cycle activities, satellite imagery can be used to help guide direct physical measurements, for example of mine spoil piles or of tailing dumps from uranium refining, as well as sampling, for example, of uranium contents and for concentrations of related decay-chain isotopes. Satellite imagery can also be used to help identify places that should be asked about in interviews of DPRK scientists, and about which information should be sought in documentation provided by DPRK officials as part of a denuclearization agreement.

Figure 3: Example of Hyperspectral Image Processing for Minerals Detection34

4.2 Satellite imagery measurements of uranium mills

In a 2015 study, three Indian authors researched methods of using satellite images to identify the capacity of known uranium mills.35 Measuring the diameter of the Counter Current Decantation (CCD) units that are common to uranium mills (these are also sometimes called thickeners in common parlance) from satellite images of mills around the world where the mill capacity was fairly well known, these authors developed a “methodology for estimation of an appropriate production function”. The methodology estimates annual mill production capacity as a function of ore grade (effectively, % U in ore), the number of CCD units present, and the area of CCD units. CCD units are among the common features of a uranium mill, and are among the features most likely to be visible from space. They then used the methodology to check the individual prediction for the throughput capacity of the known mills sites included in their database, and to estimate the capacity of a known uranium mill in Pakistan. The authors note that the accuracy of their methods depend greatly on having accurate data on the actual quality of ore used both in the mills that form their database and in the mills to which the methodology is applied. They also note that an estimate of the capacity of a mill is only one input to understanding its output, as the mill may not operate at full capacity.

A first step to applying the type of methodology these authors developed is to identify a site as definitively being a uranium mill, as opposed to a mill used to process some other mineral, such as copper. Mineral processing mills designed to concentrate different metals often use equipment that, at least from space, looks similar. In a separate paper, the same three authors described means of determining whether minerals refining mills are or are not uranium mills, based in part on, for example, the presence of equipment, such as smelters, that would not typically be a part of uranium milling facility, and of equipment that would be present at a uranium facility but not elsewhere.36

4.3 On-site measurements of ore slag piles near mines

Measurements of ore slag piles located near mines can help to identify the volume of material removed from the mine, including both overburden/slag and ore, which can in turn, along with information about ore uranium content, help to determine the amount of ore removed. The use of core sampling techniques can be used to indicate the variation in ore content of slag over the period in which the slag was deposited, providing indications of how the ore quality in the deposit may have varied over time. In addition, there are many potential DPRK uranium mines where little if anything is known about the production capacity, history of production, or indeed whether the mine has ever operated. Site visits and on-site measurements at these locations will help to reduce the spread of uncertainty as to overall DPRK ore production capacity and output.

4.4 On-site evaluation of mines to estimate fraction of U-bearing ore expected to be extracted versus overburden

Expert evaluation of the structure of a mine itself can provide information about the expected ratio between ore and overburden that would have been (or is being) extracted, and can be used together with on-site measurements of slag piles to help to estimate the amount of ore taken from a mine over time.

4.5 Isotopic examination of ore slag piles

Isotopic examination of samples from ore slag piles for long-lived uranium decay products such as protactinium-231, thorium-232, and radium-230 could provide clues as to the concentration of uranium in the original ore.

4.6 Evaluation of uranium content of ore samples by mine

Evaluation of the uranium content of ore samples by mine can help to determine at least the current average uranium content of ore being extracted, and in combination with information about mine output (tonnes of ore), can help to indicate the amount of uranium being produced in a given year. To the extent that older ore samples that are both dated and representative of average ore mined can be tested, additional information about the history of mine output may be available. It is possible that testing ore present in older, no longer mined, sections of a mine can also help to determine the history of uranium production at a given location.37

4.7 Evaluation of uranium content and isotopic composition of waste ponds from uranium ore processing to yellowcake

When ore is processed to remove and concentrate uranium the resulting uranium-depleted slurry is typically dumped in waste ponds.38 Measuring the volume of these “tailings” can help to establish the volume of ore processed over time in a given mill. Samples, including core samples, are taken from tailings, and evaluating those samples for uranium content, which would typically be related to the concentration of uranium in the ore, depending on the way that the process is/has been operated, may provide clues as to the original concentration of uranium in the incoming ore. Core samples of tailings piles and waste pond sediments provide a way of going “back in time”, as the samples farther down in the core represent material deposited earlier in the life of the uranium processing facility. Slices of a given core sample are tested for content of uranium, and the resulting concentrations are evaluated based on where in the core a given sample is taken. In addition, using techniques such as gamma and alpha spectrometry can help to determine the relative concentration of uranium isotopes and other radionuclides within sediments in waste ponds, which can help to determine when the ore was processed and its original U concentration, as well as which mines the ore came from.39 In this way, isotopic examination of tailings samples for long-lived uranium decay products should also provide clues as to the concentration of uranium in the original ore. Other types of analyses, including raman spectroscopy, can track minerals associated with uranium in ores, and thus also help to determine which mines or ore bodies the uranium processed at the facility originally came from (see below).

4.8 Establishing Nuclear Forensic Signatures for mines

The uranium ore and uranium concentrate produced from a given mine has characteristic non-uranium impurities and isotopic composition that can be traced from mine to yellowcake, and thus help to determine the origin of uranium used in the nuclear fuel cycle, and by extension, help to determine the output of a given mine. Examples of the types of measurements that can be made of ore and uranium concentrate include measurement of the abundance of rare earth minerals as a reflection of the geochemical conditions in the mine, other trace element measurements of the impurities in processed uranium and in mines, and measurements of uranium/lead (Pb) and lead/lead isotope ratios, both as an indicator of mine identity and to date the uranium deposit, itself also an indicator of mine identity.40

4.9 Information for confirmation of past U ore imports by trade partners

As noted above, only a very few trades of ores containing uranium and thorium appear in at least aggregate trade statistics in recent years. Piecing together more of a history of the DPRK’s trade in uranium ores (and/or metals) will therefore depend first on a thorough search of literature including, for example, available historical Soviet and Eastern Bloc diplomatic communications, as well as detailed trade statistics showing firms involved, quantities, values, and related information from, for example, Russia, China, and other nations thought to be or have been importers of DPRK ore. Next might come requests to possible importer nations for information on historical trades with the DPRK, although one imagines that in perhaps the majority of cases, such information will not be forthcoming. Finally, information on past exports could be sought from the DPRK itself as a part of a denuclearization deal. Requests for information on uranium exports should be accompanied by requests for information from at least the DPRK, and possibly other nations, on imports of uranium by the DPRK, though we assume that such trades, apart from, for example, imports of enriched uranium in small quantities from Russia for the DPRK’s small research reactor at Yongbyon, are unlikely to have taken place in significant volume

5 Estimates of enriched uranium produced

Uranium used in nuclear weapons is enriched from natural levels (about 0.7% U235) to about 90 percent U235, referred to as highly enriched uranium, or HEU. For the last several decades, the primary enrichment technology employed has been to use large arrays of gas centrifuges that separate uranium isotopes fed into the centrifuge system as uranium hexafluoride gas. Although the international community had long suspected that the DPRK had a clandestine uranium enrichment program, or at least was pursuing research into enrichment technologies and application, the first concrete proof of the DPRK enrichment program came in 2010 when Professor Siegfried Hecker was shown the centrifuge hall at the Yongbyon Nuclear Complex by his North Korean hosts. Since then, analysts have suspected that the DPRK may have, or may have been developing, at least one other uranium enrichment facility, and possibly two. Another large facility, estimated to be larger than the enrichment plant at Yongbyon, has been tentatively identified at a site called Kangsong, in the town of Chollima, on the outskirts of Pyongyang.41 Although the team that identified the Kangsong site, led by Professor Jeffrey Lewis, stops short of definitively identifying it as a centrifuge plant, Lewis was quoted as saying “this [the Kangsong building] is a suitable building that has a number of signatures consistent with that [being an enrichment plant] and no obvious inconsistencies”, and that it is “clearly a sensitive national defense site.” Figure 4 shows a Google Earth image of the plant. The large building in the middle of the image, measuring roughly 50 m by 100 m, is the possible enrichment hall.

The estimated number and size of the centrifuge cascades within the Yongbyon facility, the possible Kangsong facility, and a possible third enrichment plant is a major uncertainty in all of the estimates of the DPRK’s enriched uranium inventory summarized below.

Figure 4: Possible Uranium Enrichment Facility at Kangsong, near Chollima, DPRK42

•    Hecker, Braun, and Lawrence (2016) estimated that the DPRK had a stockpile of 300 to 450 kg of HEU by 2016, and was producing additional HEU at a rate of 150 kg/yr.43 Key assumptions (and thus uncertainties) in this estimate are that the estimate is based in part on an assessment of the availability of key, possibly limiting components and materials that the centrifuge program would require, and that total centrifuge capacity was 35,000 kg SWU (separative work units) per year, or the output of about 8750 P-2 type centrifuges. Implied in this estimate is a “tails assay” (fraction of U235 in the depleted uranium gas exiting the centrifuge cascade) of about 0.2%, and an annual input feed of natural uranium of about 27 tonnes of uranium.44 Use of a higher value for the tails assay implies greater production of HEU, and greater natural uranium requirements. For example, raising the tails assay to 0.5% means that about 230 kg/yr HEU can be produced, with an input feed rate of about 100 tonnes of natural uranium per year.

•    David Albright of the Institute for Science and International Security (ISIS) estimates that the DPRK HEU stocks as of the end of 2017 were between 230 and 760 kg, “where 230 kilograms corresponds to a median estimate for the case of one centrifuge plant and 760 kilograms corresponds to the median estimate for the case of two centrifuge plants”.45 Here the major uncertainty seems to be the existence of the second centrifuge plant, as well as how long it has been operating (and at what capacity factor).

•    Signs that the Experimental Light Water Reactor (ELWR) built at Yongbyon may have started up or be in a testing mode as of 201846 imply that the DPRK has produced at least enough low-enriched uranium (LEU) for one fueling of that unit (thought to be 4 tonnes of LEU), if not more.47

The estimates above yield a possible overall range on the order of 300 to 900 kg of HEU stocks produced by the DPRK (including HEU used in weapons tests) by late 2018. At these levels, the DPRK’s estimated HEU stocks represent about 60 to 400 tonnes of natural uranium, assuming 5 percent losses in converting yellowcake to UF6,48 and similar losses in converting enriched UF6 to HEU metal, and using a range of enrichment tails assays from 0.25% to 0.5% U235. These values, in turn, assuming a range of ore uranium contents of 0.2 to 0.9 percent (see above), and losses of about 20 percent in ore processing and uranium conversion to yellowcake, imply uranium ore requirements of about 9,000 to 300,000 tonnes. If an additional 4 tonnes of LEU has been produced to fuel the ELWR, an additional 30 to 60 tonnes of natural uranium would have been required, which translates to about 4,000 to 40,000 tonnes of ore.

6 Estimates of plutonium (Pu) produced

The DPRK has produced plutonium in its (nominally) 5 megawatt electric (equivalent) reactor (5 MWe) at the Yongbyon nuclear complex starting in approximately 1989, and continuing, with several interruptions, to the present day. The 5 MWe reactor uses fuel made of natural uranium (0.7% U235) in its graphite-moderated core. A number of authors have prepared or cited estimates of total plutonium production over time, including the following:

•    Hecker et al (2016, ibid) estimated Pu production by the DPRK of 42 to 63 kg by 2016, less 10 percent losses in reprocessing. This estimate includes Pu used in nuclear weapons tests (see below). Key uncertainties in these estimates, as described by the authors, include the average power levels in the plutonium production reactors, the efficiency of reprocessing, and the amount of Pu in the waste streams from reprocessing. With the 5 MWe reactor capable of producing an additional 6 kg of Pu annually, these estimates would presumably be in the 54 kg to 75 kg range by 2018.

•    Albright (2018, ibid) provides an estimate of 30 kg of separated Pu by the end of 2017.49

•    An assessment that the DPRK possessed about 50 kilograms of Pu was published by the Korea Times in 2017, citing a 2016 ROK Defense White Paper.50

The estimates above, ranging from about 36 to about 75 kg of Pu production by late 2018, can be used to calculate implied input uranium requirements for the Yongbyon graphite-moderated plutonium production reactor (5 MWe). Assuming plutonium conversion in spent fuel from that reactor at about 0.6 kg/t U,51 the DPRK’s estimated plutonium stocks represent about 70 to 140 tonnes of natural uranium, assuming 10 percent losses in fuel fabrication. These values, in turn, assuming a range of ore uranium contents of 0.2 to 0.9 percent (see above), and losses of about 30 percent in ore processing and uranium conversion to uranium metal, imply uranium ore requirements of about 10,000 to 100,000 tonnes.

7 Estimates of uranium and plutonium used in DPRK weapons tests

Some of the production of Pu and HEU described above has been used by the DPRK in its six nuclear weapons tests to date (through late 2018). Estimates of the amount of fissile material used in those tests include the following:

•    Hecker et al (2016, ibid) assume that three of the first five tests undertaken by the DPRK used Pu, and that on the order of 15 kg of Pu was used in those tests, plus an approximate 10 percent “production losses in plutonium purification and metal fabrication”. Hecker et al note that the amount of Pu used during these tests is hard to determine due to lack of knowledge about the designs of the explosive devices tested, and how much Pu was used in each device.

•    Jeffrey Lewis, quoting, respectively, a DPRK defector and DPRK state sources, suggests that the first nuclear weapons test carried out by the DPRK, which was judged to have failed by the international community, used only 4 or 2 kg of Pu.52 Lewis suggests that the DPRK has purposely pursued smaller bomb designs as to learn more from their tests, and thus save fissile material.

•    David Albright offers a “median estimate” of 7 kg of Pu used in the two weapons tests in 2006.

•    We have yet to find direct estimates of the amount of HEU used in the (assumedly) three, or perhaps four, tests where HEU was likely to have been used, but if we use the range suggested by Albright (2017, ibid), of 15 to 25 kg HEU per detonation, we get a range of about 45 to 100 kg HEU used in the tests.

The above imply that perhaps 7 to 15 kg of Pu were consumed in the DPRK’s nuclear tests, along with about 45 to 100 kg of HEU. These estimates lie in the range of about 5 to 20 percent of the DPRK’s estimated fissile material production, as described above.

8 Quantitative demonstration of impact of narrowing uncertainties in uranium production on estimate of overall current DPRK processed uranium and fissile material inventories

An exhaustive estimate of the total uranium produced by the DPRK, and thus of an upper bound on the amount of fissile material available to the DPRK (assuming no or insignificant imports of same) would in theory require determined values or estimates for a number of different parameters. These parameters would include:

•    The average capacity of DPRK uranium mines, ideally by historical period

•    The average capacity factor of DPRK uranium mines, also by historical period

•    The capacity of uranium mills, and/or the fraction of uranium ore milled to produce natural uranium

•    The fraction of uranium ore (or other forms of uranium) exported to other nations, also ideally by historical period

•    The average fraction of uranium in uranium ore (which may have changed over the history of DPRK uranium production)

•    The efficiency with which uranium mills convert uranium to yellowcake (U3O8)

Estimates of the amount of fissile materials produced from the DPRK’s uranium output would further require:

•    Estimates of the efficiency of conversion of uranium in yellowcake to uranium hexafluoride and back to uranium oxide for reactor fuel

•    Estimates of the fraction of purified forms of natural uranium (UF6, UO3, uranium metal) retained in storage, versus the fraction undergoing enrichment or conversion to natural uranium reactor fuel (for example, for use in Yongbyon “5 MWe” reactor)

•    Estimates of the fraction of uranium remaining in uranium “tails” (depleted uranium) from enrichment to HEU and LEU

•    Estimates of the fraction of Pu in reprocessed fuel prior to reprocessing

•    Estimates of the efficiency of Pu recovery during reprocessing

We have undertaken what must be considered initial estimates of the amount of uranium that may have been mined and milled over time in the DPRK using two different procedures. The first, which might be called “top-down” starts with the estimates of the amount of fissile materials thought by others (as described above) to be held by the DPRK and/or have been used in weapons tests, and calculates the volumes of uranium and uranium ore implied to have been produced. The second approach starts with estimates of the DPRK uranium mining capacity and estimates overall uranium ore and refined uranium production. Comparison of the differences between the two estimates, including sensitivity analysis with the second approach, is used to try and determine what types of information should be sought, either on the ground or by remote means, to try and narrow down the estimates of uranium available to the DPRK as a part of nuclear weapons/nuclear materials verification in support of a denuclearization agreement. Below, and in the annexes that follow, we present the methods and results of our estimates.

8.1 Presentation of central estimate and range of estimates at present

“Top-down” Estimates of Uranium and Uranium Ore Produced by the DPRK

Starting with the ranges of estimates of enriched uranium and plutonium production prepared by others, and summarized above, and selecting “central” estimates from those ranges, we estimate that on the order of 200 to 800 tonnes of natural uranium (as U, not uranium oxides) would be required to produce the ranges of enriched uranium—as HEU for weapons (or possibly some for the Yongbyon research reactor) and as LEU for the DPRK’s ELWR—and plutonium described above. A central estimate is 400 tonnes uranium. These calculations are shown in Annex 1. Please note that in both Annex 1 and 2, although results are shown for convenience to many digits, the figures shown should probably not be considered accurate to more than one significant figure.

Using a range of average uranium concentrations in DPRK ore of 0.15 percent to 0.9 percent, with a central estimate of 0.25 percent, we get uranium ore requirements implied by the production estimates above ranging from about 20 thousand to 600 thousand tonnes, with a central estimate of somewhat under 200,000 tonnes.

Note that these are cumulative, not annual figures. Note also that these figures do not include any estimates of the amount of uranium in processed or ore form that the DPRK may have in storage. We would consider the possibility of significant amounts of ore being held in storage less likely than storage of processed uranium, as to the bulk of ore that would have to be stored would be on the order of one hundred to (more likely) over one thousand times the volume of processed yellowcake with the same uranium content.

“Bottom-up” Estimates of Uranium and Uranium Ore Produced by the DPRK

Starting with estimates of DPRK uranium mining capacity, and implicitly assuming that most, if not all, mined uranium is processed to yellowcake (and that therefore the processing capacity is sufficient to more or less keep up with ore production), we estimate that between about 200 and 7,000 tonnes of natural uranium (tonnes U) has been produced by the DPRK, with a central estimate of about 800 tonnes U. Note that this estimate excludes uranium that may have been exported (mostly as ore, but possibly in small part in refined form) and thus remains in the country in some form—as yellowcake or other uranium oxides, uranium metal, UF6, HEU, LEU, depleted uranium or as Pu produced from uranium.

The lower end of this estimate starts with mine capacity of 29,000 tonnes of ore per year, which is the sum of the estimates for two major mines described by Yoon (ibid). The higher end of the estimate is capacity doubles that estimate, and a central value was chosen as 38,000 tonnes/yr, which assumes that the other mines (that is, apart from Pyongsan and Woong-gi) that have been noted in various articles sum to a capacity of about 10,000 tonnes of ore annually. Using the same range of ore concentrations—from 0,15% U to 0.9% uranium—implies total cumulative uranium ore output (net of exports) of about 200,000 to 1 million tonnes of ore, with a central estimate of about 400,000 tonnes. See Annex 2 for details of these calculations.

Figure 5 shows the differences, in percentage terms, between uranium production (net of exports), as the differences between the minimum and central and central and maximum estimates of production capacity are reduced. Figure 6 shows how the difference between these pairs of estimates is reduced as the range of estimates of ore quality are reduced.

Figure 5:

Figure 6:

A key input assumption in these calculations are the annual average capacity factors for different periods (each a decade or more) in which the DPRK produced uranium. These annual capacity factors are our rough estimates and range from 5 percent—of what are assumed to have been the highest capacities achieved by the DPRK over the years—in the early years of production, to up to 50 percent in the 1990s.

Comparison of Top-down and Bottom-up Estimates

The ranges of the top-down and bottom-up estimates of natural uranium and uranium ore produced above do not match up entirely, but do overlap. Particularly at the low end of the range, for total uranium production as expressed as tonnes of elemental U, the two estimates are relatively close, at near 200 tonnes uranium. The estimates diverge somewhat in going from the “minimum” end of the range through central estimates to the “maximum” values, which could imply uranium in storage as natural (probably refined) uranium. We prepared a rough calculation of the amount of space that might be required to store all of the refined (for example, as yellowcake) uranium produced by the DPRK (net of exports, but including all forms of uranium, and uranium and plutonium that has been used in weapons tests. The space required for uranium storage ranges from 1000 square meters (smaller than the buildings on site at the known uranium refining facilities in the DPRK) to about 40,000 square meters (several such buildings), but in either case is not so large that it would be difficult for the DPRK to conceal the stored uranium.

8.2 Using direct sampling and satellite analysis to reduce the range of uncertainty in key parameters

Direct sampling and satellite image analysis can be used together to reduce the range of uncertainty in a number of key parameters. Perhaps the most important among these is the average ore content of uranium mined in the DPRK. Our guess is that the average value for ore quality is much more likely to be near 0.2 percent than 0.9 percent, as the latter would be quite high and seems to be contradicted by the typical ore qualities in the nations in the area, as well as by Russian reports of ore quality (see above). Although multi-band analysis of satellite imagery may provide some way of assessing the ore quality in uranium mine spoil or in piles of ore waiting to be processed, it is likely that direct access and sampling (of ore, mine spoil, and mill tailings, for example) will be needed to reduce uncertainty in the range of historical uranium output. Satellite techniques for assessing the growth in spoil piles and tailings dumps would ideally be combined with on-the-ground measurements, including cores, to attempt to determine the rate of ore production and uranium refining over time.

8.3 Sensitivity analysis of degree sampling/satellite analysis would reduce range of estimates of existing uranium

Annex 3 provides calculations of how reducing the difference between the minimum and central estimates of key parameters, and between the central and high estimates, can reduce the difference in results. Differences in each of the parameters was reduced by 5 to 50 percent. A reduction of 50 percent in the difference between the minimum and central estimates of uranium mine capacity decreases the difference in uranium production net of export (as U) by about 10 percent, and in uranium ore by about 30 percent. (A reduction in the difference between minimum and central annual capacity factors for mines would have had a similar impact). A reduction of 50 percent between the central and maximum estimates for mine capacity would reduce the difference in tonnes of uranium and uranium ore available for domestic use by about 60 and 40 percent, respectively. Reduction in the differences between the estimates for ore quality by 50 percent reduce the volume of ore needed for non-exported production by about 20 percent (for minimum to central estimates) and by 80 percent for central to maximum estimates. The reduction in the latter is particularly pronounced because the difference between the central and maximum values we assumed for ore quality is large.

9 Conclusions

9.1 Overall current understanding of potential range of historical U production in the DPRK, and of potential net inventories of U

Based on the above, total historical uranium production (net of export) in the DPRK is estimated to be in the range from about 200 to 1000 tonnes of uranium as U, with ore production in the range from tens of thousands of tonnes to up to a million tonnes, with hundreds of thousands of tonnes being more likely, in our opinion.

9.2 Implications for verification tests that should be a part of a negotiated settlement with the DPRK

It will be crucial to obtain a better understanding of ore quality in the DPRK in particular, so sampling of ore bodies, waste piles, and tailings ponds/piles at identified major mines and at uranium processing facilities will be crucial, and should be a part of any negotiated agreement. Satellite image analysis will provide a useful addition to data acquired on the ground, but likely cannot, particularly in the short term, substitute for direct sampling.

9.3 Practical issues associated with verification testing

There are, of course, a wide range of issues associated with reducing the uncertainty of uranium production estimates. These include (but are not limited to):

•    Site access—that is, whether the DPRK will allow access to all identified and suspected sites involved in uranium production and processing.

•    Hidden facilities, including underground facilities, are widely expected to exist for a variety of military installations, with uranium processing and enrichment among them.

•    Sampling error or bias, caused by not being able, through physical or access impediments, or to lack of equipment, manpower, and/or time, to sample all of the relevant parts of a given uranium-bearing stratum, spoils pile, tailing pile, or other input to our output of the uranium industry.

•    Access to mining and processing records at all the sites and facilities involved in uranium mining, processing, enrichment, and fabrication.

Maximizing access to sites and records and making arrangements that will allow the minimizing of sampling errors and of missed hidden facilities, as well as reducing, to the extent possible, any other issues likely to limit the effectiveness of verification testing, will be a goal of negotiations on verification protocols.


Annex 1:53

Annex 2:

Annex 3:

1 This working paper is published under a 4.0 International Creative Commons License the terms of which are found here. It is published by Nautilus Institute here; by the Asia-Pacific Leadership Network for Nuclear Non-Proliferation and Disarmament here; and by the Research Center for the Abolition of Nuclear Weapons, Nagasaki University, here

2 See, for example, Jonathan McLaughlin (2017), “North Korea Nuclear Milestones – 1962-2017“, The Wisconsin Project, dated September 29, 2017, and available as Other sources refer to a nuclear cooperation agreement signed by the DPRK and Russia in 1959, predated by nuclear prospecting by both the Japanese during World War II and by the Russians as early as 1946.

3 Document in the authors files, referencing a number of Korean and international literature sources [ELE-96]; Andrea Berger (2014), “What lies beneath: North Korea’s uranium deposits: All signs point to Pyongyang regime’s ongoing nuclear program development”, NK News, dated August 28, 2014, and available as; and Nuclear Threat Initiative (NTI, 2018), “North Korea, Nuclear”, Last Updated: October, 2018, and available as Some of the areas listed may be identified deposits of uranium, but not active mines.

4 Federation of American Scientists, “Hungnam N39°49 E127°37′ Hungnam Chemical Engineering College Hungnam Fertilizer Complex”, available as

5 Edward Yoon (2011), Status and Future and Future of the North Korean Minerals Sector, Nautilus Institute Special Report dated January 6, 2011, and available as

6 Frank V. Pabian, Peter Makowsky, and Irv Buck (2018), “North Korea’s Uranium Mining and Milling Operations Continue at Pyongsan”, 38 North, dated November 2, 2018, and available as

7 NTI (2018), ibid.

8 Jeffrey Lewis (2015), “Recent Imagery Suggests Increased Uranium Production in North Korea, Probably for Expanding Nuclear Weapons Stockpile and Reactor Fuel”, 38 North, dated August 12, 2015, and available as

9 Source, Google Earth, probably mid-2018, available as,126.4333268,594m/data=!3m1!1e3.

10 Source, Google Earth, probably early-2018, available as,125.5688776,409m/data=!3m1!1e3.

11 Joseph S. Bermudez, Jr. (2017), Overview of North Korea’s NBC Infrastructure, the US-Korea Institute at Johns Hopkins SAIS, dated June 2017 and available as

12 Larry A. Niksch, United States Congressional Research Service (CRS), CRS Issue Brief for Congress: North Korea’s Nuclear Weapons Program, updated January 17, 2006. The same figure is also quoted in Yo-Taik Song, “IN OUR TIMES SERIES, PART 6, The North Korean Nuclear Program: Technical and Policy Issues”, available as

13 The DPRK has been highly reluctant to reveal the extent of its deposits of uranium ore and its annual production capacity to the outside world. This same estimate of reserves (26 million tonnes of ore), however, was provided in information from private sources in China and DPRK business contacts compiled by E.Yoon in the Nautilus report referenced above. Much of the discussion of uranium reserves and mining presented here is adapted from Mr. Yoon’s work, which should be consulted by the reader wishing further details and reference sources.

14 North Korea’s Nuclear Weapons Programme, by the International Institute for Strategic Studies, 2006, available as

15 As just two anecdotal example, an undated (but probably late 1990s) article available on the International Atomic Energy Agency (IAEA) website describing uranium mining in Benxi, in the Northeast province of Liaoning, China (about 150 km from the DPRK border) lists an ore U content of 0.34% U (Zhang Rong, “New Development Stage of China’s Uranium Industry”, available as; and the World Nuclear Association lists Russian mines with ore U contents ranging from 0.05 % to 0.2% uranium, albeit the mines listed are not particularly close to the DPRK (World Nuclear Association (2018), “Russia’s Nuclear Fuel Cycle”, updated May 2018, and available as

16 “North Korean Hullabaloo”, by Paul Vos Benkowski, 6 – Nukewatch Pathfinder, Winter, 2006-2007, page 6.

17 Peter Hayes (2004), “North Korea’s Uranium Exports: Much Ado About Something”, dated May 25, 2004, and available as

18 These same values are also quoted in Balazs Szalontai and Sergey Radchenko (2006), North Korea’s Efforts to Acquire Nuclear Technology and Nuclear Weapons: Evidence from Russian and Hungarian Archives, COLD WA R International History Project, Working Paper #53, dated August 2006, and available as

19 Milan Matolin and Mohamad Tauchid (1987), Report to the Government of the Democratic People’s Republic of Korea: Uranium Prospecting DRK/3/003 Evaluation Mission, dated 15 May, 1987, available as; and Milan Matolin (1990), Report to the Government of the Democratic People’s Republic of Korea: Uranium Prospecting DRK/3/003-04 Laboratory Gamma Ray Spectrometry, available as In personal communication, Professor Matolin indicated that his mission did not have access to data on ore quality, but he did measure the uranium content of “0.1 percent and up” in uranium ores, though these measurements were made for the purposes of training as part of the mission, and thus do not reflect a DPRK average or even the average for the deposit where the sample was taken.

20 See, for example, Andrea Berger (2014), “What lies beneath: North Korea’s uranium deposits: All signs point to Pyongyang regime’s ongoing nuclear program development”, NK News, dated August 28th, 2014, and available as

21 Monazite is a name for a group of rare earth phosphate minerals, the most common form of which (Monazite-(Ce)) contains Cerium, Lanthanum, Thorium, Neodymium, and Yttrium. Monazite is radioactive, and it seems likely to have been exported in this instance primarily as a source of Thorium, though that is just the authors’ conjecture. A description of Monazite can be found at Amethyst Galleries “THE MINERAL MONAZITE”,

22 “North Korea Profile, Nuclear Exports”, prepared for the Nuclear Threat Initiative by the by the Center for Nonproliferation Studies at the Monterey Institute of International Studies, 2003, and previously available as (but that link is now inactive). In retrospect, the reference to “$6 billion worth of uranium ore” seems improbable, as it would imply on the order of tens of millions of tonnes of ore, so perhaps it could be a units error in reporting.

23 Foreign Trade of the DPRK, 1 Jul 2001, and 1 Oct 2002.

24 NUKEWARS, “Moscow Dismisses Rumors on Uranium Deal with Pyongyang” by Staff Writers
Moscow (AFP—Agence France-Presse), Dec 04, 2006; and “NKorea, Russia in secret deal over nuclear talks: report”, Tokyo (AFP) Dec 03, 2006. Available as

25 China Customs statistics as compiled by N. Aden for N. Aden (2006), North Korean Trade with China as Reported in Chinese Customs Statistics: Recent Energy Trends and Implications, as prepared for the DPRK Energy Experts Working Group Meeting, June 26th and 27th, 2006, Palo Alto, CA, USA). Dr. Aden’s paper is available as Additional data from (used for 2007 and 1994/1995 values above) show no other reports of trades between the DPRK and any country in HS 2612, uranium and thorium ores and concentrates, from 1990 on.

26 See, for example, chapters 4 through 6 in IAEA (1993), Uranium Extraction Technology, Technical Reports Series No. 359, available as

27 See, for example, Joshua Pollack (2010), North Korea’s Nuclear Exports: On What Terms?, 38 North Special Report 9, dated October 14, 2010, and available as

28 Jeffrey Lewis, Melissa Hanham, Joshua Pollack, Catherine Dill, Raymond Wang *2017), Open-Source Monitoring of Uranium Mining and Milling for Nuclear Nonproliferation Applications, James Martin Center for Nonproliferation Studies Occasional Paper #34, Middlebury Institute for International Studies at Monterey, dated December, 2017, and available as

29 Jeffrey Lewis (2015), “Recent Imagery Suggests Increased Uranium Production in North Korea, Probably for Expanding Nuclear Weapons Stockpile and Reactor Fuel”, 38 North, dated August 12, 2015, and available as

30 Melissa Hanham, Grace Liu, Joseph Rodgers, Mackenzie Best, Scott Milne, and Octave Lepinard (2018), Monitoring Uranium Mining and Milling in China and North Korea through Remote Sensing Imagery, James Martin Center for Nonproliferation Studies Occasional Paper #40, Middlebury Institute for International Studies at Monterey, dated October, 2018, and available as

31 A full description of these methods is beyond the scope of this paper, but a description of their applications to a similar problem, the analysis of human settlement dynamics, can be found in Ranga R. Vatsavaia and Jordan Graessera (2012), “Probabilistic Change Detection Framework for Analyzing Settlement Dynamics Using Very High-resolution Satellite Imagery”, prepared 7 International Conference on Computational Science, ICCS 2012, published as Procedia Computer Science 9 (2012) 907-916, Elsevier Ltd, and available as

32 See, for example, Enton Bedini (2017), “The use of hyperspectral remote sensing for mineral exploration: a review”, Journal of Hyperspectral Remote Sensing 7 (2017) 189-211, available from

33 See, for example, Jose Lim, G. A. Borstad, Leslie N. Brown and Q.S. Bob Truong (2006), A Systematic Approach to Hyperspectral Interpretation of Uranium Mines”, available as, and Rodrigo dos Reis, Salles, Carlos Roberto de Souza Filho, Thomas Cudahy, Luiz Eduar do Vicente, and Lena Virgínia Soares Monteiro (2017), “Hyperspectral remote sensing applied to uranium exploration: A case study at the Mary Kathleen metamorphic-hydrothermal U-REE deposit, NW, Queensland, Australia”, Journal of Geochemical Exploration, Volume 179, August 2017, Pages 36-50, available from

34 Source: Enton Bedini, 2017 (ibid). Original source as cited by Bedini is “Clark R.N., Swayze G.A., Livo, K.E, Kokaly, R.F, Sutley, S.J, Dalton, J.B, McDougal, R.R, Gent, C.A., 2003. “Imaging spectroscopy: Earth and planetary remote sensing with the USGS Tetracorder and expert systems”. Journal Geophysical Research-Planets 108, 44.”

35 S.Chandrashekar, Lalitha Sundaresan, and Bhupendra Jasani (2015), Estimating Uranium Mill Capacity Using Satellite Pictures, dated December 2015, and available as

36 S.Chandrashekar, Lalitha Sundaresan, and Bhupendra Jasani (2015), “Identification of Uranium Mill Sites from Open Source Satellite Images”, dated December 2015, and available as

37 See, for example, IAEA (1990), Manual on Laboratory Testing for Uranium Ore Processing, Technical Report Series # 313, dated June, 1990, and available as, for general procedures for uranium ore sampling and analysis.

38 An exception here would be for mines that use a process of in situ leaching (ISL) where an aqueous solution is pumped through ore-bearing rock in a uranium deposit and a uranium-enriched leachate solution is collected. ISL does not appear to have been widely used in DPRK mines to date.

39 See, for example, A.C. Keatley, P.G. Martin, K.R. Hallam, O.D. Payton, R. Awbery, F. Carvalho, J.M. Oliveira, L. Silva, M. Malta, and T.B. Scott (2018), “Source identification of uranium-containing materials at mine legacy sites in Portugal”, Journal of Environmental Radioactivity 183 (2018) 102-111, available from, and Zsolt Varga, Judit Krajko, Maxim Penkin, Marton Novak, Zsuzsanna Eke, Maria Wallenius, and Klaus Mayer (2017), “Identification of uranium signatures relevant for nuclear safeguards and forensics”, Journal of Radioanalytical and Nuclear Chemistry (2017) 312:639-654, available as

40 See, for example, Manny Mathuthu and Ntokozo Khumalo (2017), “Developing Nuclear Forensics Signatures and National Nuclear Forensics Libraries for the African Continent: A Case Review for South Africa”, International Journal of Applied Science – Research and Review, 2017, Vol. 4, No 1:1, available as This study lists the types of data that forensic signatures for uranium mines can be sourced from as including rare earth elements concentrations (as “these exhibit consistent patterns under varying geochemical conditions”), trace element compositions (which “show the impurities for each uranium processing”), the U-Pb isotopic composition, and Concordia age dating of samples. See also Michael J. Kristo, Amy M. Gaffney, Naomi Marks, Kim Knight, William S. Cassata, and Ian D. Hutcheon (2016), “Nuclear Forensic Science: Analysis of Nuclear Material Out of Regulatory Control”, Annu. Rev. Earth Planet. Sci. 2016. 44:555-79, available as

41 Ankit Panda (2018) Exclusive: Revealing Kangson, North Korea’s First Covert Uranium Enrichment Site”, The Diplomat”, dated July 13, 2018, and available as See also David Albright and Sarah Burkhard (2018), Revisiting Kangsong: A Suspect Uranium Enrichment Plant, dated October 2, 2018, and available as

42 Image from Google Earth, 2018,,+Pyongyang,+North+Korea/@38.9572239,125.6120409,290m/data=!3m1!1e3!4m5!3m4!1s0x357e17cb26380e57:0x624eae2b358f3f13!8m2!3d38.9333992!4d125.5821994.

43 Siegfried S. Hecker, Chaim Braun, and Chris Lawrence (2016), “North Korea’s Stockpiles of Fissile Material”, KOREA OBSERVER, Vol. 47, No. 4, Winter 2016, pp. 721-749, available as

44 Tails assay and implied uranium inputs estimated by the authors of this report using the online “Uranium Enrichment Calculator” by the WISE Uranium Project, last updated 23 Nov 2009, and available as

45 David Albright (2018), “Understanding North Korea’s Nuclear Weapon Capabilities”, presentation dated May 9, 2018, and available as

46 See, for example, K.K. Rebecca Lai, William J. Broad, and David E. Sanger (2018), “North Korea Is Firing Up a Reactor. That Could Upset Trump’s Talks With Kim”, The New York Times, dated March 27, 2018, and available as This article references work by Jane’s Intelligence Review and the Center for International Security and Cooperation at Stanford University, as well as ISIS.

47 The estimate of the size of the core of the ELWR at Yongbyon is from “Redefining denuclearization in North Korea” by Siegfried S. Hecker, 20 December 2010, Bulletin of the Atomic Scientists, available as

48 See, for example, Oak Ridge National Laboratory (ORNL, 2009), Model of a Generic Natural Uranium Conversion Plant-Suggested Measures to Strengthen International Safeguards, Report # ORNL/TM-2008/195, dated November 2009, and available as

49 David Albright (2017), “North Korea’s Nuclear Capabilities: A Fresh Look”, presentation dated April 22, 2017, and available as

50 Jun Ji-hye (2017), “N. Korea can make 10 nuclear bombs”, Korea Times, updated 2017-01-11, and available as

51 This figure is based on figures estimated from the late 1980s/early 1990s operating history of the reactor as declared by the DPRK, and presented by ISIS (undated, but after 2006), in “ISIS Course, Introduction to Reactors and Fuel Cycle: Small Yongbyon Nuclear Reactor”, available as

52 Jeffrey Lewis (2017), “The Game Is Over, and North Korea Has Won”, Foreign Policy, dated August 9, 2017, and available as

53 The calculations performed by the “WISE Enrichment Calculator” follow the equations provided in the Wikipedia article “Separative work units”, last edited 28 August 2017, and available as The “Implied tonnes natural U in feed for HEU production” in the table above is calculated for each of the three cases as (Tonnes Natural U feed per kg HEU, as obtained from the WISE Enrichment Calculator for the U-235 tails percentages shown) * (Estimated 2018 HEU inventories (kg HEU))/(1 – Conversion Losses, Enriched UF6 to HEU metal).


【May 20, 2019】Verification of DPRK Nuclear Disarmament (PSNA-WP-7)

Category : PSNA Activities

Verification of DPRK Nuclear Disarmament: The Pros and Cons of Non-Nuclear-Weapon States (Specifically, the ROK) Participation in This Verification Program

John Carlson1, A Member of Asia Pacific Leadership Network for Nuclear non-proliferation and Disarmament (APLN)

PSNA Working Paper Series (PSNA-WP-7)2

May 20, 2019

[PDF version]


In the expert and diplomatic communities, it is generally considered that disarmament verification should be undertaken as far as possible on a multilateral basis. Partly this reflects experience with the International Atomic Energy Agency’s safeguards system, and partly it reflects the view of non-nuclear-weapon states that international participation is required to ensure transparency and credibility in the disarmament process. The main argument against this is proliferation risk from the diffusion of proliferation-sensitive information. However, a number of aspects of disarmament verification will not involve sensitive information, and where sensitive information is involved there are ways of enabling effective verification while protecting such information.

As yet no specific details have been negotiated on how nuclear disarmament in the Democratic People’s Republic of Korea (DPRK) will proceed, and how this will be verified. Whatever is negotiated, the international community will certainly want assurance of the integrity of the verification process. In particular, the ROK has a very direct interest in what is happening across the DMZ and has every reason to be involved in the disarmament effort. This paper discusses how this can be possible consistent with non-proliferation principles.

1. Introduction

As yet no specific details have been negotiated on how nuclear disarmament in the Democratic People’s Republic of Korea (DPRK) will proceed, and how this will be verified. Internationally, there is no established model for conducting and verifying nuclear disarmament. There have been bilateral arms control agreements between the United States and Russia (or the Soviet Union), but these are of limited scope compared with what would be required for complete disarmament.

To date the only precedent for a state that had produced nuclear weapons disarming completely is South Africa, which dismantled its warheads secretly, and submitted the recovered fissile material (highly enriched uranium – HEU) to International Atomic Energy Agency (IAEA) safeguards as part of joining the Nuclear Non-Proliferation Treaty (NPT). Other precedents are:

(a) Ukraine, Belarus and Kazakhstan, which at the dissolution of the Soviet Union had Soviet nuclear weapons on their territories and agreed to transfer these weapons to the Russian Federation; and

(b) Iraq, Iran and Libya which were found to have nuclear weapon programs at varying stages of development.3

None of these precedents is comparable to the situation of the DPRK. Accordingly, whatever process is developed for the DPRK will be a pioneering effort, important in itself and also in helping to set a precedent for eventual disarmament in other nuclear-armed states.

In 2015 the International Partnership for Nuclear Disarmament Verification (IPNDV) was established to facilitate international collaboration on verification approaches and methods in support of nuclear disarmament. So far, the IPNDV has focused its studies on a specific aspect – monitoring and inspection of a notional nuclear weapon dismantlement process, what it calls the Basic Dismantlement Scenario. The IPNDV has not yet placed this dismantlement process into a broader disarmament framework. However, there has been substantial discussion of this subject within the verification expert community. Drawing on these discussions, this paper outlines a model framework for disarmament verification, discusses how this might apply to the DPRK, and discusses who might be given responsibility for the various verification tasks.

In the expert and diplomatic communities, the general view is that disarmament verification should be undertaken as far as possible on a multilateral basis. The establishment of the IPNDV reflects this view – from the outset the IPNDV has been focused not just on developing disarmament verification, but specifically how non-nuclear-weapon states can be involved in such verification.

Partly this reflects the experience gained with the IAEA safeguards system, which has a multilateral inspectorate, and partly it reflects the view of non-nuclear-weapon states that international (that is, multilateral) participation in nuclear verification is required to ensure transparency and credibility in the disarmament process. It is a matter of trust – non-nuclear-weapon states are not prepared to leave it to the nuclear-weapon states to inspect each other. The main argument in favor of non-nuclear-weapon state participation in nuclear disarmament verification, therefore, is to ensure international confidence in the integrity of the process.

The main argument against non-nuclear-weapon state involvement is the risk of proliferation arising from the diffusion of sensitive information. Some states, notably Russia, have taken a firm position against non-nuclear-weapon state involvement, maintaining disarmament verification can be undertaken only by personnel from nuclear-weapon states. However, this position fails to consider two key factors:

(a) a number of aspects of disarmament verification will not involve classified or proliferation-sensitive information, and in this case, there should be no objection to a multilateral process. For example, once ex-weapons nuclear material is in non-classified form and composition, it is no different to other comparable nuclear material and can be safeguarded accordingly – see the discussion in section 3 below; and

(b) where classified information is involved, it may well be possible to develop approaches and methods that enable effective verification while ensuring such information is fully protected.

In developing multilateral verification, therefore, the key issue to address is protection of classified information – how to ensure that involvement of non-nuclear-weapon state personnel in disarmament verification does not result in them acquiring nuclear weapon designs and know-how, which would be a violation of the NPT (discussed further in section 6). This is a particular focus of the IPNDV’s current work. It is absolutely crucial to both nuclear-weapon states and non-nuclear-weapon states to ensure effective protection of classified information – but states should be prepared to consider on their merits internationally-developed approaches to meet this objective.

In the case of the DPRK, the international community as whole (which predominantly comprises non-nuclear-weapon states) certainly wants assurance of the integrity of the verification process: apart from anything else because this is an important precedent for future disarmament efforts in the nuclear-weapon states. In particular, the Republic of Korea (ROK) has a very direct interest in what is happening across the DMZ and has every reason to be involved in the disarmament effort. This paper will discuss how this can be possible consistent with the NPT’s non-proliferation principles.

2. A model approach to nuclear disarmament

A generic approach to nuclear disarmament in a state would look something like this:

Stage 1 Cease production of fissile materials (HEU, separated plutonium – Pu)

(a) Declaration of all fissile material production facilities (enrichment and reprocessing facilities).

(b) Monitoring of these facilities to ensure production has ceased.

(c) In addition, tests of nuclear weapons and nuclear-capable missiles are to be terminated

– these tests are not covered by this paper.

Stage 2 Declaration of all nuclear material and all nuclear facilities

(a) Nuclear material – (i) total quantities per material category for all nuclear material in the state, including in warheads; and (ii) inventories at each nuclear facility

– total material per category (HEU, Pu) in warheads or military custody would be black boxed – the overall quantity within the military program would be declared, but without any breakdown by forms and locations;

■ this is because such information is sensitive and the state is unlikely to be prepared to declare it – of course if the state is prepared to give any details these would be extremely useful for verification purposes;

■ materials in warheads would not be available for verification until the warheads are dismantled (stage 5).

(b) Nuclear facilities – enrichment and reprocessing facilities should be declared in stage 1. Here all related facilities would be declared: reactors, fuel fabrication, conversion, mines/mills, storage, radwaste, etc.

(c) Historical nuclear material flows (production, consumption, losses)

– declarations, and supporting documentation, will be required in due course, but are not essential at the outset.

(d) Nuclear-related locations – declarations including:

– centrifuge R&D;

– centrifuge manufacturing;

– activities, items and materials covered by the IAEA Additional Protocol;

■ Annex I – items and materials specially prepared for nuclear use;

■ Annex II – dual-use items and materials;

– dual-use activities with potential nuclear weaponization application (based on the Iran JCPOA4).

(e) Tritium – declaration of relevant facilities (reactors, separation plant) and material flows also required in due course

– by stage 6 – earlier if production is proscribed at outset.

Stage 3 Inspections of declared facilities and related nuclear materials

(a) Where facilities are shut down/decommissioned – status to be verified.

(b) Where facilities remain in operation – inspections to verify they are operated as agreed (all nuclear material under safeguards; quantity and quality limits if applicable).

(c) Nuclear materials – safeguards to verify that materials remain in peaceful use and are transferred only to safeguarded locations and activities.

(d) Nuclear-related locations – activities at these locations should be terminated if the related nuclear activity (e.g. enrichment) is shut down. Inspections are required to verify shutdown, or that continuing activities are as agreed.

(e) This stage would also include establishment of a procurement channel where required for agreed nuclear-related activities and potential weaponization activities.

Stage 4 Excess nuclear materials in military program to be declared and removed from the state or transferred irreversibly to the safeguarded nuclear program

(a) There should be no valid reason for the state to retain separated plutonium. This would be removed from the state.

(b) Likewise, there is no valid reason for the state to retain HEU. This would be removed from the state. If the state is operating reactors requiring low-enriched uranium (LEU) fuel, the state could be supplied with LEU fuel corresponding to the quantity of HEU removed.

Stage 5 Progressive reduction in warheads (and missiles)

(a) Declaration of types and numbers of warheads and missiles will be required at an appropriate time (arrangements regarding missiles are not covered in this paper).

(b) Warheads are to be dismantled, and fissile materials are to be converted to unclassified forms and treated as excess materials (see stage 4 – materials to be verified and removed from the state).

(c) An issue to be negotiated is how dismantlement would be monitored/verified

– the usual concept is for warheads to be dismantled by the possessor state under black box/chain of custody arrangements, so the verifying entity can confirm that a warhead entered dismantlement and a corresponding quantity of fissile material exited.

Stage 6 Verification activities to provide assurance against existence of undeclared nuclear facilities and nuclear materials

(a) This is likely a contentious area as it requires intrusive activities including challenge inspections. The state needs to understand this is a necessary aspect of verification, without which confidence is not possible.

The state can be assured that a mandate to look for undeclared facilities and materials is not carte blanche for access anywhere for any purpose. In the verification context undeclared means something that should have been declared in accordance with the agreements applicable at the time in question.

Obviously until the state is required to give up all its warheads it will have some nuclear material it is not yet obliged to submit for inspection – verification activities will not be seeking to locate items and materials unless the state is required to declare them and has not done so. The purpose of verification against undeclared facilities and materials is to detect possible violations of applicable agreements.

(b) This stage will include establishing a historic nuclear materials balance, drawing on declared material flows (stage 2 (c)), facility operating records, sampling and analysis of materials, interviews of personnel and related activities.

(c) Activities to provide assurance against undeclared warheads and missiles will also be required but are not covered in this paper.

(d) Also required, but not covered in this paper, are programs, and appropriate verification/monitoring, to, inter alia:

– convert nuclear weapons-related labs, workshops and factories to peaceful purposes;

– redeploy specialists from the nuclear weapon program to peaceful purposes;

– track key specialists to ensure they don’t become involved with nuclear weapon programs elsewhere.

Stage 7 End of the disarmament process – the state is shown to meet the requirements for a non-nuclear-weapon state

At the end of the disarmament process the state would become a non-nuclear-weapon state. In the case of a non-NPT party the state should join the NPT. In either case – whether a former non-NPT party or a former NPT nuclear-weapon state – the state would be a non-nuclear-weapon state, subject to the most rigorous form of IAEA safeguards.

Recognising that the state had nuclear weapon capabilities (therefore the capability to rebuild its nuclear weapon program – and even the possibility that it has successfully hidden parts of its former program), it will also be subject to additional verification, confidence-building and transparency measures, including those referred to in 6 (d) above.


3. Applying this model to the DPRK

As yet it is not known if agreement can be reached with the DPRK for applying this model. It would be possible to apply a more limited version initially, reflecting more limited goals (for example, cessation of fissile production, dismantling of a specified number of warheads). However, as discussed in a complementary paper, Denuclearizing North Korea: The Case for a Pragmatic Approach to Nuclear Safeguards and Verification (see References), achievement of complete disarmament will require all of the elements outlined in the model.

Who should be responsible for undertaking the various monitoring and verification tasks? Most of these tasks are the same as or very similar to activities conducted by the IAEA in the implementation of safeguards around the world. While these tasks would not necessarily be undertaken by the IAEA, there seems no in-principle reason why they should not be. For example:

Stage 1 – cease production of fissile materials

This requires declaration of all enrichment and reprocessing facilities, and monitoring of these to ensure they are no longer operating. Monitoring the status of nuclear facilities is a standard part of IAEA safeguards procedures. The IAEA has previously undertaken monitoring of the reprocessing plant and 5 MWe reactor at Yongbyon.

Stage 2 – declaration of all nuclear material and all nuclear facilities, and nuclear-related activities, items and materials

Receipt and analysis of declarations of nuclear facilities, and nuclear material inventories and flows, are a standard part of IAEA safeguards procedures.
While the IAEA does not usually verify inventories and flows of non-nuclear materials such as tritium (stage 2 (e)), it could do so, INFCIRC/66 safeguards agreements5 allow for this possibility.

Stage 3 – Inspections of declared facilities and related nuclear materials

Inspections to verify the operational status of nuclear facilities, and to verify nuclear material inventories and movements, are a standard part of IAEA safeguards procedures. Where proliferation-sensitive activities are involved (such manufacturing of centrifuge components) it may be necessary to use inspectors drawn from technology-holder states.

Stage 4 – Excess military nuclear materials to be declared and transferred from the DPRK or transferred irreversibly to safeguarded program

This involves verifying materials that are declared excess, and tracking them to ensure they are transferred from the DPRK or are placed under safeguards in the DPRK and remain under safeguards. These activities are similar to standard IAEA safeguards procedures.

Stage 6 – Verification for assurance against possible undeclared nuclear facilities and materials

This involves a range of activities, such as:

– information collection and analysis (including open-source information, satellite imagery, possibly wide-area environmental sampling, information from states) looking for possible indicators undeclared nuclear activities and materials;


– establishing a historic nuclear materials balance, looking for discrepancies and inconsistencies in declared information;

– investigation of possible indicators, including through inspector access to suspect locations, using mechanisms such as complementary access, technical visits, or special inspections.

All of these activities are part of standard IAEA safeguards procedures. Special arrangements may be required if the IAEA has to investigate possible weaponization activities (this may require specially cleared inspectors from nuclear-weapon states). Such arrangements have applied during IAEA investigations in Iraq, Iran and Libya, and in South Africa.

Stage 7 – The DPRK qualifies as non-nuclear-weapon state

At this point standard IAEA safeguards arrangements will apply, as in any other non-nuclear-weapon state. As noted above, additional confidence-building and transparency measures will also be required.

Monitoring and verification that would not be undertaken by the ILEA

Stage 1 (c) – no tests of nuclear weapons and nuclear-capable missiles

Activities for detection of any nuclear tests would be undertaken by the Comprehensive Nuclear-Test-Ban Treaty Organization (CTBTO). There is no international inspectorate for detection of missile tests, this is a matter for national intelligence and technical means.

Stage 2 (d) – monitoring at nuclear-related locations

In appropriate cases might be undertaken by supplier states or through cooperation among the authorities of relevant states – or through establishment of a Joint Commission along the lines of the JCPOA.

Stage 3 (e) – procurement channel

Clearance and monitoring of procurement might be undertaken by supplier states or through cooperation among the authorities of relevant states – or through a Joint Commission.

Stage 5 – Reduction and dismantlement of nuclear warheads

This is the main area where new verification arrangements need to be developed. This is the current focus of IPNDV studies. (Stage 5 would also include destruction of missiles, but this is not covered by this paper).


4. Dismantlement of nuclear warheads

This is an area of high secrecy from two perspectives:

(a) National security – while the state continues to hold nuclear weapons, it does not want others to learn specifics of its capabilities (including questions of warhead yield and reliability);

(b) Non-proliferation – there is an over-riding international interest to ensure that information potentially helpful to a proliferator is totally protected.

These considerations, as well as the requirement for verification effectiveness, will influence the specifics of the verification arrangements on which agreement can be reached.

A threshold question is whether the DPRK is prepared to simply hand over warheads (for example, to a team of specialists from the nuclear-weapon states). If so, monitored dismantlement would not be necessary. However, it is likely the DPRK will be concerned to protect national security information, so for this paper it is assumed the DPRK will not hand over intact warheads.

Safety must be paramount   Apart from DPRK sensitivities, a compelling argument for warheads to be dismantled by DPRK personnel is for reasons of safety. First, transporting the warheads elsewhere could be dangerous. Second, those who made the warheads know their design and characteristics and are in the best position to dismantlement them safely. Particular care will be required to build a dismantlement facility that provides adequate protection for surrounding populations in case of accidental explosion (it will also be essential to warn neighbouring states when dismantlement operations are proposed).

A further threshold question is whether verification of warhead dismantlement is essential. The immediate reaction is, of course it is. However, this is not a straightforward issue, it depends on the objective sought. If the objective is immediate elimination of an agreed number of warheads, then monitored dismantlement will be required. On the other hand, if (as is likely) it is assessed that the DPRK has limited holdings of fissile material (so has limited ability to replace dismantled warheads if it sought to do so), it might be considered acceptable to have dismantlement without monitoring, with the DPRK simply handing over the quantities of HEU and plutonium estimated for the agreed number of warheads.

There may be some concern that dismantlement by the DPRK without monitoring would leave the possibility that warheads declared to be dismantled have really been concealed – but this is an issue anyway, because the number of warheads actually produced by the DPRK is not known (so calling for the elimination of a specific number of warheads may be of uncertain utility). It will probably not be until the end of the disarmament process that there is sufficient information to conclude that all warheads and nuclear materials are satisfactorily accounted for.

This paper is not recommending against requiring monitored dismantlement, but simply noting that if there are difficulties in establishing monitored dismantlement, the pros and cons could be further considered. In support of monitored dismantlement, it can be pointed out to the DPRK this would have a substantial confidence-building benefit.

The following diagram shows IPNDV’s visualization of key steps in the process of dismantling nuclear weapons.6


The concept of monitored warhead dismantlement

The basic approach is that the state would be responsible for dismantling its own warheads, thereby ensuring it maintains secrecy over warhead characteristics (design, fissile material quantity, quality and shape, and so on). Dismantlement would take place in a black box – this black box would comprise a specially constructed facility together with appropriate procedures. Movement of warheads into the facility and objects and materials out of the facility would be monitored by inspectors of the verifying entity (see section 5 below). All fissile material exiting the facility would be transferred to monitored storage and disposition.

The generic concept is illustrated in the following diagram from IPNDV documentation.7


A description of monitored warhead dismantlement is as follows:

(a) Dismantlement facility

This facility would be specially designed and constructed for safe and secure dismantlement of warheads in circumstances that enable confidence that all movements of objects and materials into and out of the facility can be monitored effectively. Inspectors would be given the facility design information and would have access during construction to verify there are no hidden exit pathways (doorways, pipework) or places where objects and materials could be hidden for subsequent removal. Inspectors would have regular access to the facility to check there have been no alterations and that objects and materials are not being retained within the facility.

In the (unlikely) event that warhead reductions proceed at a faster pace than the construction of the dismantlement facility, warheads could be held in monitored storage until the dismantlement facility is ready.

(b) Confirmation that an object entering the facility is a warhead

It is assumed the DPRK will wish to conceal the specific characteristics of its warheads. There are two possible situations:

(i) the DPRK presents a warhead to inspectors to check prior to dismantlement; or

(ii) the DPRK presents a container declared to contain a warhead. In the latter case standardized containers would be used, approved by inspectors for the purpose.

In either case inspectors would perform a range of measurements designed to confirm, without revealing classified information, that (i) the object presented is a warhead or (ii) the container holds a warhead. This approach, described as attribute measurement, is discussed below.

(c) The dismantlement process

DPRK personnel would dismantle each warhead, and re-form the fissile components (weapon cores, or pits) into unclassified shapes and mass (for example, 1 kg or 2 kg buttons), and possibly convert the materials into other forms (for example, from metal to oxide).

Because re-forming or converting the fissile material involves very different processes to dismantlement (for example, melting or chemical reactions), it is possible the DPRK may wish to undertake these processes in a separate facility. In this case it would be necessary to establish a system for verifying transfers of materials from one facility to the other and maintaining a chain of custody over these materials.

(d) Transfer of fissile material from the dismantlement facility to storage and disposition

When plutonium or HEU is ready to be transferred from the dismantlement facility, inspectors would measure the material to confirm its mass and isotopic composition. Inspectors would also check the cumulative mass for outgoing transfers in a given period to ensure this is at least equal to the cumulative threshold values for the warheads that entered the dismantlement facility during the period.

There will be some uncertainties in deriving a material balance between fissile materials entering and exiting the dismantlement facility, because material inputs will be calculated on minimum threshold values for mass and isotopic composition, while material outputs will be precisely measured. Because the threshold values are minimums, total material outputs can be expected to exceed total inputs. Total output may be reduced by conversion losses, but these would be very small (and it should be possible to confirm losses by measurement of wastes and discards).

(e) Rigorous monitoring of all movements into and out of the dismantlement facility

In addition to declared transfers of warheads into the facility, and declared transfers of nuclear materials out of the facility, all other movements of objects and personnel will require rigorous monitoring to ensure there are no undeclared movements of nuclear materials.

(f) Regular inspections of the dismantlement facility

Inspectors will need to check for undeclared alterations to the building, and for possible concealment of nuclear materials. As inspectors should not have the possibility of access to classified information, these inspections would be conducted between dismantlement campaigns, when there are no warheads or intact pits in the facility.

Attribute measurement

Attribute measurement is an approach by which inspectors can take measurements to confirm whether an object is a warhead, or a container holds a warhead, without accessing classified information. The approach is based on information barriers, enabling instruments to be used to measure for expected attributes without revealing classified details to the inspector.

A series of attributes would be defined for particular warhead types. The attributes would be described as threshold numeric values, for example:

(i)   a mass of plutonium above a specified threshold;

(ii)   a Pu-240/Pu-239 ratio below a specified threshold;

(iii)  a mass of U-235 above a specified threshold;

(iv)  a U-235/U-238 ratio above a specified threshold;

(v)   presence of high explosives.

Modified instruments, that would give a go/no go (or green light/red light) indication but not specific readings, would be used for these measurements. The result is that inspectors would be confident that a warhead containing “x” kilograms or more of weapon grade plutonium, or “y” kilograms or more of weapon grade HEU, has entered the dismantlement facility.

One form of attribute measurement involves the use of templates. Where there are a number of identical warheads, inspectors would take readings from a randomly selected warhead, to create a template against which the other warheads could be compared. It is not clear whether the characteristics of the DPRK’s nuclear arsenal are such that templating would be useful.

The idea of an attribute measurement system with information barriers was developed and demonstrated as part of the Trilateral Initiative undertaken by the United States, Russia and the IAEA in the period 1996 to 2002.8 The concept is proven, but further development may be required before it is ready for practical application. One area requiring further research is cyber-security aspects, ensuring that information barriers and authentication measures are not defeated. Attribute measurement was one of the techniques trialled in the United Kingdom-Norway Initiative on the Verification of Nuclear Warhead Dismantlement9, discussed below.

The dismantlement concept outlined above has been developed with a large weapon program in mind, and it may be possible to simplify it for the relatively small DPRK program. For instance, for a small number of warheads being dismantled in relatively short campaigns, attribute measurement might not be considered essential. If inspectors witness the transfer of say five warheads, each declared to contain at least “x” kilograms of HEU, into the dismantlement facility for a campaign expected to take say “z” weeks, then the DPRK would be expected to hand over to inspectors at least 5x kilograms of weapon grade HEU at the end of this period.

Cheating scenarios can be envisaged, for example:

(a) if the threshold value is set too low, the DPRK could submit four real warheads and a dummy (thus retaining one real warhead), knowing that the total recovered material will meet the expected threshold value:

– say the threshold value is 15 kg HEU/warhead, but each warhead actually contains 20 kg. The DPRK could submit four real warheads and one dummy. The inspectors would expect an output of 75 kg HEU (5 x 15), and would be presented with 80 kg, so all would appear to be in order, when actually the DPRK has withheld one warhead;

– this example suggests it is preferable to have attribute measurement of all warheads submitted for dismantlement;

(b) the DPRK could submit five dummy warheads each containing the threshold mass (say 15 kg HEU), while retaining the real warheads that contain a larger mass (say 20 kg HEU):

– on this scenario the DPRK appears to dismantle five warheads – in reality it has given the inspectors 75 kg of HEU, but still has the warheads.

Attribute measurement is more important if there are large numbers of warheads and there could be an extended period (maybe years) before the recovered fissile material could be correlated with the warheads submitted. With a small program the risk of cheating is reduced, but cannot be excluded. Ultimately confidence in disarmament depends on availability of complementary, mutually reinforcing information, such as nuclear archaeology (historical nuclear material balance substantiated by contemporary documentation and sampling at facilities and waste storage) and verification activities for providing assurance against undeclared missiles.10

5. The verifying entity

As discussed in section 3, most of the verification activities that would be involved in denuclearization in the DPRK are the same as or very similar to activities conducted by the IAEA in safeguards implementation. It follows that these activities could be undertaken by the IAEA, pursuant to a mandate given under a safeguards agreement concluded between the DPRK and the IAEA, or a mandate given by Security Council resolution. In due course a new safeguards agreement will be required between the DPRK and the IAEA. While some of these verification activities do not correspond exactly to a standard IAEA safeguards agreement, the IAEA Statute provides flexibility to conclude an agreement as requested by the parties.11

Other possibilities for the verifying entity, touched on below, include:

● nuclear-weapon states, or P5 (the Permanent Members of the Security Council) – either all the P5 or those most engaged with the DPRK (the United States, China and Russia);

● the parties to agreements with the DPRK pursuant to the denuclearization process – at this point it is not clear which states might be directly involved, the Six Parties again (the DPRK, United States, China, Russia, the ROK and Japan) or some other grouping. Possibly the parties might decide to establish a Joint Commission along the lines of the Iran JCPOA;

● bilateral arrangements between the DPRK and the United States;

● bilateral arrangements between the DPRK and the ROK, along the lines of ABACC (the Argentine-Brazilian Agency for Accounting and Control of Nuclear Materials);

● a regional safeguards inspectorate, along the lines of Euratom.

IAEA inspections can involve staff from non-nuclear-weapon states or nuclear-weapon states, commonly a mix of the two. Usually no distinction is made between the two groups of states. However, if the subject of an inspection is proliferation-sensitive, it is established practice to form a team of inspectors from nuclear-weapon states, comprising individuals having appropriate security clearances from the relevant national authorities.

For example, where the IAEA has been responsible for establishing that a nuclear weapon program had been terminated (South Africa) or investigating suspected nuclear weapon programs (Iran, Iraq, Libya, Syria and the DPRK), much of the verification work was undertaken by normal safeguards inspectors but, where necessary to protect classified information, tasks were assigned to inspectors who were appropriately cleared nationals from nuclear-weapon states, as just discussed. In some cases, teams were established that included non-staff specialists provided by nuclear-weapon states. Thus, the IAEA has developed substantial expertise in dealing with and appropriately protecting classified information.

As regards monitoring and verification of warhead dismantlement, the attribute measurement approach was developed in the context of bilateral arms control inspections between the United States and Russia – the objective was to enable an inspector from one state to confirm that an object presented by the other state is a warhead, without the inspector gaining classified information. Clearly this approach could also be valid for an inspector from a third state, or an international inspector, which is why the IAEA participated in the Trilateral Initiative. In other words, application of attribute measurement could be undertaken by inspectors from non-nuclear-weapon states.

The possibility of warhead dismantlement being verified by inspectors from non-nuclear-weapon states has been trialled in the United Kingdom-Norway Initiative, which has successfully conducted several practical exercises. The Initiative has involved three areas of work:

● managed access – how inspections can be carried out in practice;

● information barriers – procedural and technical measures to enable unclassified measurements to be made of a classified object;

● confidence in verification processes – including multinational participation in verification research.

The work of the United Kingdom-Norway Initiative has been an important input to the work of IPNDV. IPNDV has stated that “… actual dismantlement is the most important, complex, and technically challenging task of nuclear disarmament verification”, and has expressed the judgment that:

… while tough challenges remain, potentially applicable technologies, information barriers, and inspection procedures provide a path forward that should make possible multilaterally monitored nuclear warhead dismantlement while successfully managing safety, security, non-proliferation, and classification concerns in a future nuclear disarmament agreement.12 (underlining added)

In line with this judgment, there seems no reason why dismantlement of warheads in the DPRK could not be monitored by IAEA inspectors, which could include ROK nationals, and/or also by ROK government personnel. There is one caveat – because attribute measurement, and also the concept of monitored warhead dismantlement, are still in the development stage, there will likely be a need for specially qualified and cleared personnel from one or more nuclear-weapon states to oversee the operation to ensure there is no inadvertent transfer of classified information.

Non-IAEA monitoring and verification

It is possible there may be some resistance to early involvement by the IAEA in monitoring and verification in the DPRK. If this is delayed for any reason, monitoring and some other verification tasks could be undertaken by suitably qualified personnel from states involved in the denuclearization process (for example, the Six Parties, or a Joint Commission?) and from other states willing to support the process and acceptable to the parties.

There is some speculation that the DPRK may prefer bilateral verification arrangements, that is, inspections by United States personnel. This would present two difficulties. First is the question of credibility and integrity – will the international community have full confidence in inspections undertaken by the nationals of only one state, especially if there might be political pressures to reach favourable results? For this reason, multilateral inspections are the well-established international practice. Second, it should be recognized that the IAEA must be involved as soon as possible, having regard to the Agency’s nuclear verification mandate, its specialized expertise and equipment, and its international standing. The objective should be to develop DPRK-IAEA cooperation as soon as possible.

Summary of verification options

The following table summarizes the above discussion. In this table, Joint Commission is used to encompass either a formally constituted Joint Commission along the lines of the Iran JCPOA or a less formal grouping of parties to the denuclearization agreement(s) with the DPRK.


6. Possible ROK participation in denuclearization verification activities

If the ROK wished to participate in inspections in the DPRK the possibilities seem to be as follows:

(a) If initially, prior to agreement on IAEA involvement, monitoring and verification activities are carried out by Six Party or Joint Commission personnel, this is an opportunity for ROK participation. There are obvious advantages in having Korean inspectors in the team.

(b) Once IAEA activities start, ROK safeguards inspectors on the IAEA staff could join the team that the IAEA is likely to establish to carry out inspections in the DPRK. Here too there are obvious advantages in having Korean inspectors. However, it must be kept in mind that an inspected state can reject inspectors of specific nationalities, so it will be essential to ensure that the DPRK has no objection to ROK inspectors (this is also a possible issue under (a)).

(c) Another possibility is either bilateral safeguards arrangements between the ROK and the DPRK, or a wider regional safeguards inspectorate.

On a bilateral arrangement, the ROK and the DPRK might consider concluding arrangements similar to ABACC, under which safeguards inspections would be undertaken jointly by the IAEA and an ROK/DPRK bilateral inspectorate. It should be noted that although ABACC is generally thought of as a bilateral arrangement, actually it is more complex – it is a quadripartite arrangement, between Argentina, Brazil, ABACC and the IAEA.

It is for the ROK and the DPRK to consider whether a bilateral safeguards arrangement would be useful, for example, for transparency and confidence-building. It is important to note that the ABACC arrangements are reciprocal, so following this model would result in DPRK inspectors participating in inspections in the ROK as well as vice versa.

On a regional arrangement, the precedent is Euratom. Euratom was established a decade before the NPT, and it can be questioned whether a regional safeguards entity is warranted in today’s circumstances. Nonetheless, this is something states in the region, or states in the immediate neighbourhood, might consider – for example, whether an entity comprising ROK, DPRK, China, Japan, and maybe Russia and the United States (that is, the Six Parties) would serve a useful purpose. One way to look at this, quite different to the Euratom precedent, would be in support of the creation of a North Asia nuclear-weapon-free zone. If a regional safeguards entity were to proceed, the responsibilities of the IAEA would have to be accommodated, for example through a partnership approach as established between Euratom and the IAEA.

Treaty issues relating to ROK participation in denuclearization verification

NPT   The key issue, in terms of the NPT, is whether the ROK’s participation in denuclearization verification activities could result in it acquiring information that could materially assist in the design or manufacture of a nuclear weapon. As a non-nuclear-weapon state Party to the NPT, the ROK has undertaken

“… not to receive the transfer from any transferor whatsoever of nuclear weapons … or … control over such weapons … directly, or indirectly; not to … acquire nuclear weapons …; and not to seek or receive any assistance in the manufacture of nuclear weapons …”13

Although the language of the NPT is not explicit, there is no doubt that acquisition by ROK nationals of data that could materially assist in the design or manufacture of a nuclear weapon would be considered a violation of the NPT.14 Also acquisition of data that could assist in the production of fissile material would raise difficult issues because of international concerns about any spread of proliferation-sensitive data.

The NPT places a corresponding obligation on nuclear-weapon states not in any way to assist any non-nuclear-weapon state to manufacture or otherwise acquire nuclear weapons.15 Thus the ROK must be scrupulously careful not to acquire, even inadvertently, any classified or proliferation-sensitive data, and nuclear-weapon states having control of such data through verification and monitoring in the DPRK must be scrupulously careful to prevent access to the data by a non-nuclear-weapon state.

Accordingly, the ROK must not be involved in any activity where it could acquire proliferation-sensitive data, and nuclear-weapon states in a position to do so must ensure that the ROK and other non-nuclear-weapon states do not acquire such data in the DPRK. As discussed in this paper, this does not mean a blanket exclusion from denuclearization verification in the DPRK. Many of the stages involved in denuclearization do not involve sensitive technology or information, or fissile material in sensitive forms or composition. There should be no objection to ROK personnel being involved in these stages.

Areas where ROK and other non-nuclear-weapon state personnel would have to be excluded include facilities where sensitive technology and information could be accessible (including weaponization activities, manufacturing of centrifuge components, and so on), and areas where nuclear weapon design and know how could be revealed. This is especially the case with warhead dismantlement (described as stage 5 in this paper), unless a black box approach with rigorous protective measures is established.

United States-ROK agreement concerning the peaceful uses of nuclear energy

The current agreement was concluded in 2015. The agreement reaffirms the Parties’

… strong partnership on strengthening the global nonproliferation regime … and close cooperation on advancing their shared objective to address the security and proliferation threat posed by North Korea’s nuclear program.

There are no provisions in the agreement that have a direct bearing on the issue of ROK participation in denuclearization verification in the DPRK. The Parties may agree on cooperation in research, development and demonstration, including safeguards and physical protection, and other areas as mutually agreed16 , but this language does not readily apply to verification implementation in the DPRK and there seems no reason why the Parties would seek to bring this under the terms of the agreement. Likewise, the Parties may agree to include under the High Level Bilateral Commission established pursuant to the agreement any topics related to peaceful nuclear cooperation mutually agreed to … by the Parties17, but there seems no reason why the Parties would seek to apply this to denuclearization verification.

The agreement could apply in the case of nuclear supply to the DPRK (for example, if nuclear supply is part of a denuclearization agreement), but this is beyond the scope of this paper.

Pros and cons for the ROK in participating in monitoring and verification in the DPRK


● It would be a major plus for the ROK government to achieve DPRK acceptance of such a role; and also, recognition of the ROK’s co-equal status with the nuclear-weapon states.

● Most importantly, it could be a confidence-building measure between the DPRK and the ROK, smoothing the way for extending monitoring and verification arrangements to non-nuclear arms control measures in support of reducing tensions on the Korean peninsula.

● Taking a long-term perspective, in-depth involvement in dismantling the DPRK’s military program would reinforce the ROK’s understanding and capacity to deal with the DPRK’s nuclear weapon capabilities in a unified Korea.

● The ROK’s involvement might be implemented as part of a bilateral or a multilateral nuclear-weapon-free zone inspectorate that would also create a binding legal framework for the monitoring and verification activity between the disarmament process and the DPRK’s re-entry into the NPT, and give the three proximate nuclear-weapon states a formal role in DPRK denuclearization.

● ROK inspectors are the most likely of all to pick up cultural and other signals of deception and/or misunderstandings related to safety, security, and other limits imposed by the DPRK on monitoring and verification of its disarmament. Typically, the DPRK provides access and transparency in precise calibration to a mutually agreed rationale for such, and no more than minimally required. Being able to understand and negotiate that boundary is a critical conflict-avoidance issue in a monitoring and verification activity, to defuse such situations before they escalate into wars of words and then actions.

● The ROK may provide considerable logistical, technical, and financial support that could be hard to mobilize in the nuclear-weapon states.


● The DPRK reaction may be strongly negative, adhering to the past view that this is a matter for the United States only (because the DPRK treats compliance with monitoring and verification as a way to get the United States’ attention, not because it wants monitoring and verification per se, let alone the involvement of the IAEA or other parties).

● It may complicate the negotiations over monitoring and verification in general, for example, by providing an argument for Japan that it too deserves to be confident that the DPRK has disarmed and to be treated co-equally.

● It could complicate the IAEA’s role if the DPRK objected to the ROK’s involvement.

● In the short to medium term some may suspect that the ROK wants to be involved as a way of gaining knowledge of how to produce nuclear weapons.

● It might be read as validating somehow that in the long run, a reunified Korea will combine ROK technological prowess with DPRK nuclear weapons knowledge.


7. Conclusions

There is a general view in the international community that nuclear disarmament verification should be undertaken as far as possible on a multilateral basis, in order to establish confidence in the integrity and credibility of the disarmament process. The main argument against a multilateral process is the possibility of classified and proliferation-sensitive information being compromised. However, a number of aspects of disarmament verification will not involve such information – for example, once fissile materials have lost classified form and composition, they are no different to comparable materials that are covered by IAEA safeguards. Further, IAEA safeguards demonstrate that a multilateral approach, incorporating special arrangements where necessary, can ensure the protection of sensitive information.

While it is absolutely crucial to both nuclear-weapon states and non-nuclear-weapon states to ensure effective protection of classified information, states should be prepared to consider on their merits internationally-developed approaches to meet this objective – a major focus of the IPNDV is to develop verification appropriate arrangements for non-nuclear-weapon state participation.

In the case of the DPRK denuclearization effort, the ROK has an obvious interest and every reason to be involved. This paper discusses a number of approaches to enable ROK participation consistent with the NPT’s non-proliferation principles. There are some challenges, but the parties involved in the denuclearization effort should be prepared to work collaboratively to address these.


A Verifiable Path to Nuclear Weapon Dismantlement, Dismantlement Walkthrough, IPNDV (accessed 14 November 2018),

Denuclearizing North Korea: The Case for a Pragmatic Approach to Nuclear Safeguards and Verification, John Carlson, 38 North Special Report, 24 January 2019,

Phase I Summary Report: Creating the Verification Building Blocks for Future Nuclear Disarmament, IPNDV November 2017,

IPNDV Working Group 2 – 2016-17 Output Report: Inspection Activities and Techniques, November 2017,

Innovating Verification: New Tools and New Actors to Reduce Nuclear Risks, NTI, July 2014,

Nuclear Disarmament: The Legacy of the Trilateral Initiative, Thomas E. Shea and Laura Rockwood, Deep Cuts Working Paper 4, March 2015,

Nuclear disarmament verification: the case for multilateralism, David Cliff, Hassan Elbahtimy, David Keir and Andreas Persbo, VERTIC Brief 19, April 2013,

Trilateral Initiative: IAEA Authentication and National Certification of Verification Equipment for Facilities with Classified Forms of Fissile Material, Eckhard Haas, Alexander Sukhanov, John Murphy, IAEA Safeguards Symposium 2001,

UK-Norway Initiative on the Verification of Nuclear Warhead Dismantlement,;

1 John Carlson was director general of the Australian Safeguards and Nonproliferation Office. He was appointed as chairman of the IAEA’s Standing Advisory Group on Safeguards Implementation by former IAEA Director General Mohammed ElBaradei and served from 2001 to 2006. He also served as Alternate Governor for Australia on the IAEA Board of Governors. He is an Australian member of the Asia Pacific Leadership Network.

2 This report is published under a 4.0 International Creative Commons License the terms of which are found here. It is published by Nautilus Institute here; by the Asia-Pacific Leadership Network for Nuclear Non-Proliferation and Disarmament here; and by the Research Center for the Abolition of Nuclear Weapons, Nagasaki University, here.

3 Mention should also be made of Syria, where the IAEA’s investigation of a suspected nuclear weapon program has not progressed due to civil war. Of course, the other case of safeguards non-compliance was the DPRK itself, but the IAEA’s investigations were thwarted by the DPRK’s withdrawal from the NPT.

4 Joint Comprehensive Plan of Action.

5 INFCIRC/66 safeguards agreements are described as “item-specific” agreements. They are used for non-NPT states, specifically India, Israel and Pakistan. Prior to joining the NPT in 1985, the DPRK had an INFCIRC/66 agreement covering the Soviet-supplied IRT reactor.

6 From IPNDV Working Group 2 Report of November 2017, page 89.

7 From IPNDV Working Group 2 Report of November 2017, page 36.

8 See Nuclear Disarmament: The Legacy of the Trilateral Initiative (References).

9 See References.

10 Missiles are considerably larger than warheads, hence are harder to conceal.

11 IAEA Statute Article III.A.5.

12 IPNDV, Phase I Summary Report, page 6.

13 NPT Article II.

14 Though probably not relevant in the context of DPRK denuclearization, it might be argued that acquisition of data in the public domain would not constitute a violation. However, there would be international concerns about a state’s motives in acquiring such data, and the NPT’s prohibition on seeking to manufacture nuclear weapons applies regardless of the status of the data involved.

15 NPT Article I.

16 See Article 3.

17 Article 18.


【Apr. 9, 2019】Reducing Nuclear Dangers on the Korean Peninsula: Bilateral versus Multilateral Approaches (PSNA-WP-6)

Category : PSNA Activities

Reducing Nuclear Dangers on the Korean Peninsula:
Bilateral versus Multilateral Approaches

Thomas Graham, former US Ambassador,
Executive Chairman, Lightbridge Corporation

PSNA Working Paper Series (PSNA-WP-6)1

April 9, 2019

[PDF version]


This paper addresses the important issue of nuclear weapons on the Korean Peninsula. It reviews alternate solutions: an essentially bilateral solution with the United States as an associated party and a multilateral regime establishing a nuclear weapon free zone in a designated part of Northeast Asia which would include the militarily significant states in the region along with the NPT nuclear weapon states as associated parties. The effectiveness of a non-binding pledge versus a legally binding agreement and the possible availability of a nuclear assurance commitment itself a non-binding declaration or a legally binding obligation is analyzed. The verification requirements of a legally binding arrangement are outlined and associated issues such as transit through the zone established by an agreed arrangement are considered. The political salience of the two types of solutions, bilateral and multilateral is commented upon: for example what has the Democratic People’s Republic of Korea [DPRK] indicated it would accept and the likelihood that the DPRK is now prepared to be, or can be, persuaded to make the hard decision to eliminate weapons and accede to the vast verification requirements of a legally binding regime; and whether the United States would be willing to provide a negative nuclear assurance of any sort in a bilateral non-binding agreement situation as well as the level of verification it might demand in a legally binding agreement.

1 This working paper is published under a 4.0 International Creative Commons License the terms of which are found here ( It is published also by the Nautilus Institute as a NAPSNet Special Report ( and Asia-Pacific Leadership Network for Nuclear Non-Proliferation and Disarmament (here), and by the Research Center for the Abolition of Nuclear Weapons, Nagasaki University (RECNA), as a Working Paper ( of Panel on Peace and Security of Northeast Asia (PSNA). PSNA acknowledges its kind contribution from the Nautilus Institute on publication of this paper. The views expressed here is of author’s own and do not necessarily reflect views of PSNA or RECNA.



The Korean Denuclearization Declaration (signed January 20, 1992, into force February 19, 1992) committed the two Koreas to agree not to test, manufacture, produce, receive, possess, store, deploy, or use nuclear weapons; to use nuclear energy solely for peaceful purposes; and not to possess facilities for nuclear reprocessing and uranium enrichment. The two Koreas also committed to conducting inspections of locations in the other Korea and established the South-North Joint Nuclear Control Commission (JNCC) to implement the inspections. However, since 1993, the JNCC has been unable to reach agreement on reciprocal inspections and none have taken place under its purview. Article 4 commits the two Koreas to verify the nuclear-free status of the Korean Peninsula within about two months from 19 March with inspections to begin in less than a month.[1]

However, the two sides were unable to agree on the frequency of inspections, and their demands on what would be inspected were also a-symmetric, with the ROK demanding inspections of any suspect locations in the DPRK; and the DPRK demanding that it also be able to inspect any US military facilities in the ROK. In June, the issue came to a head over the DPRK’s demands and the ROK’s rejection of North Korean inspections of US military facilities in the ROK, and by October 1992, multiple disagreements surfaced in the JNCC talks, and the implementation of the inspection regime ground to a halt in December 1992 as suspicions increased about North Korean reprocessing, and the US and the ROK began to prepare for the next Team Spirit exercises.

Not only was the agreement not legally binding (because it was struck between the two Koreas on the basis of their special interim relationship stemming from the process towards unification as stated in the preamble of the South-North Basic Agreement and in the recently legislated ‘Act for development of the south-north relationship’),[2] but it also lacked a number of the key attributes of a meaningful NWFZ as defined by UN Disarmament Commission in its 1999 report, including:

1. Total absence of nuclear weapons: any states should not develop, test, manufacture, produce, acquire, possess, store, transport and deploy nuclear weapons within a nuclear weapons-free zone;
2. Effective verification of compliance;
3. Clearly defined boundaries;
4. Negative Security Assurance: legally binding commitments to the zone by the nuclear weapon states not to use or threaten to use nuclear weapons against the zone parties

Ironically, the 1992 Joint Declaration went beyond a standard NWFZ in that it not only banned nuclear weapons in Korea (although it did not define the boundaries of the zone) by banning the possession of “nuclear reprocessing and uranium enrichment facilities,” it failed to establish an effective mechanism for verification of compliance (Article IV, V). Also, the Joint Declaration did not address the transit or transportation of nuclear weapons in or over its territorial waters, straits and international water, or airspace by a nuclear armed state (although the DPRK raised this issue at several JNCC meetings, it is a generic issue that relates equally to Chinese and Russian nuclear weapons in or around the Korean peninsula).[3]

Of great importance, the Joint Declaration imposed no obligation on the nuclear weapon states not to use or threat to use nuclear weapons against the two Koreas, and lacked any protocols that would have bound the nuclear weapon states to the Joint Declaration in this regard. Political statements by Russia and the United States welcoming the Joint Declaration and calling for the full implementation were no substitute for negative security assurances in a standard NWFZ treaty.

As noted above, the Joint Declaration foundered on the inspection mechanism, and is now moribund. However, moribund is not dead, in spite of the DPRK’s nuclear testing and declared armament, and the Denuclearization Declaration is an important common reference point referred to in the 1994 US-DPRK Agreed Framework, and in the September 19, 2005 Six Party Talks Principles and Joint Statement.

It also contains no provision for termination and therefore is arguably still in force with respect to the two Korea’s initial commitments, no matter how much (or little) each of them is in breach of the Joint Declaration.

As the two Koreas and the great powers, especially the United States, consider their options of the best political and legal framework in which to realize a nuclear-free Korean peninsula, it is prudent to note the strengths and weaknesses of the 1992 Joint Declaration, and to ascertain if a new but more comprehensive bilateral agreement might be used today, or whether an alternative, more robust multilateral framework exists such as a nuclear weapons-free zone that might subsume the 1992 Joint Denuclearization commitments, and overcome its evident shortfalls.

Need for Robust Legal Framework Today

As noted above, the bilateral framework has inherent limitations. It is conceivable that a new bilateral declaration might address some of the deficits of the 1992 Joint Declaration. For example, it is possible that the United States might consider issuing a nonbinding executive statement containing a security assurance directly to the DPRK. More than this, however, in the non-binding area does not appear possible. In general, on this and other critically important dimensions, it is necessary to search for a more robust legal framework in which to realize denuclearization of the Korean peninsula.

Following the Singapore Summit another attempt is being made to create such a framework. The DPRK and the ROK may prefer a solution that in its eyes would represent a broader commitment than a bi-lateral treaty. A multilateral treaty is not stronger than a bi-lateral treaty in a technical legal sense. But by creating legal obligations to more than one other state, violating such a treaty breaches treaty relations with more than one state thereby creating a somewhat greater penalty for the violating state. In this sense a multi-lateral treaty perhaps can create somewhat broader legal obligations and potentially greater penalties than a bi-lateral treaty. The best-known form to address nuclear disarmament in a broader regional setting is the nuclear weapon free zone treaty. Both sides say that the agreed objective is a denuclearization Agreement which covers the Korean Peninsula. However even though the objective of the two negotiating parties is the same in writing, substantively they are far different. For the United States the term means North Korea relinquishing all of its nuclear weapons. But to the DPRK the term means reducing nuclear weapons in a “balanced and synchronous way” as Kim Jong-un said to China President XI prior to the Singapore Summit. This means to North Korea that weapons are reduced only if certain conditions are met. These likely include ending the United States nuclear umbrella in East Asia and terminating the American presence in South Korea. The DPRK may want commitment to reductions by the U.S. as well. So, the two negotiating parties at this point are not even on the same page as to the desired end result. The DPRK appears unlikely to even consider Washington’s demand for “complete, verifiable, and universal” denuclearization—it would challenge the fundamental structure of North Korea’s political system.

What North Korea appears to mean by a security guarantee as a prerequisite for a commitment to denuclearization is a regime guarantee—an undertaking to keep the Kim hereditary political system intact and the absolute authority of the leader in place. However, it should be noted that in the 1994 Agreed Framework between the U.S. and the DPRK the North settled for the following in the text. “The U.S. will provide formal assurances to the DPRK, against the threat or use of nuclear weapons by the U.S.” This Agreement was not a Treaty but it was an international agreement with legal force.

Nuclear Weapons Free Zone (NWFZ) Treaty

The nuclear weapon free zone treaty form is quite familiar to the world community. There exist five functioning treaty-based nuclear weapon free zones: Latin America (the Treaty of Tlatelolco—1967); South Pacific (the Treaty of Rarotonga—1985); Africa (the Treaty of Pelindaba—1996); South-east Asia (the Treaty of Bangkok—1997); and Central Asia (the Treaty of Semipalatinsk—2006); when the Nuclear Non-Proliferation Treaty (NPT) was negotiated in the late 1960s it explicitly codified—in Article VII—the right of the parties to the NPT to establish nuclear weapons free zones in their regions.

Responding to a proposal submitted by Finland, the United Nations General Assembly (UNGA) commissioned a comprehensive expert study (which was carried out at the Conference of the Committee on Disarmament in Geneva) of NWFZs in 1974. It was completed in 1975. And it established multiple criteria for such a treaty:

“1 – Obligations related to the establishment of a nuclear weapon-free zone may be assumed not only by groups of states, including entire continents or large geographical regions, but also for smaller groups of states and even individual countries.

2 – Zonal arrangements must ensure the complete absence for the present and future nuclear weapons in the region covered by the treaty.

3 – The initiative for the zonal treaty must come from within the region concerned.

4 – If the zone is intended to embrace a region, the participation of all militarily significant states, and preferably all states, would be important.

5 – The zonal treaty must have an effective system of verification (the experts were of the view that the viability of the nuclear weapon free zone will largely depend on this).

6 – The treaty established must be of indefinite duration.”

In 1975 at the 30th session of the UNGA the United Nations on the initiative of Mexico defined the concept of a nuclear weapon free zone.

“A nuclear-weapon-free zone shall, as a general rule, be deemed to be any zone recognized as such by the United Nations General Assembly, which any group of states in the free exercise of their sovereignty, has established by virtue of a treaty or convention, whereby: (a) The statute of the total absence of nuclear weapons to which the zone shall be subject, including for the procedure of the delineation of the zone, is defined; (b) An international system of verification and control is established to guarantee compliance with the obligations deriving from that statute.”[4]

At the same time the General Assembly also provided that in every case of a nuclear-weapon-free zone treaty recognized by the UNGA the NPT nuclear weapon states should conclude an internationally legally binding instrument in which they undertake the following obligations:

“(a) To respect in all its parts the statute of total absence of nuclear weapons defined in the treaty or convention, which serves as the constitutive instrument of the zone;

(b) To refrain from contributing in any way to the performance in the territories forming part of the zone of acts which involve a violation of the aforesaid treaty or convention;

(c) To refrain from using or threatening to use nuclear weapons against a state included in the zone.”[5]

All nuclear weapon free zone arrangements are different and the negotiations upon occasion have had to develop novel solutions to different problems. In Latin America while the Treaty of Tlatelolco was being negotiated there were two potential nuclear weapon programs within the zonal limits. The lead negotiator, Ambassador Garcia Robles of Mexico, developed a flexible structure whereby a state would become subject to the obligations of the zonal treaty only when that state had deposited with the nation serving as the depository—Mexico in this case—a certificate of waiver of the Treaty’s Article 28 entry into force requirements as well as the instrument of ratification of the Treaty. Thus, Brazil could and did ratify the Treaty in 1968 but was not subject to its obligations until 1994 when it had shut down its program entirely and formally delivered a waiver of the Article 28 requirements. Argentina signed the Treaty in 1968 but deposited its instrument of ratification and as well as the waiver document when it joined the Treaty in 1994.

Some important issues for other NWFZ treaties were:

(a) The dumping of radioactive substances in the high seas by the French (banned by the Rarotonga Treaty, approximately 10 years before France became a protocol party and could make its own commitment to this). Of course, the Rarotonga parties (the Pelindaba Treaty also has this provision) do not have jurisdiction over the high seas, no state does, but the 1975 United Nations London Dumping Convention parties decided in 1993 at the consultative meeting of contracting parties that the disposal of radioactive waste into the high seas was prohibited.[6] Thus the unrestricted dumping of radioactive substances into the high seas is internationally prohibited.

(b) The status, under the Treaty establishing the African Nuclear Weapon Free Zone, of islands off shore of Africa claimed by the Organization of African Unity to be part of Africa as well as states outside the Treaty zone placing territories they possess within the zone under the Treaty obligations (the Latin American and South Pacific free zone treaties had variants of this issue as well), for example, Réunion Island possessed by France. The most difficult such problem was the Chagos Archipelago, which includes the Island of Diego Garcia, a major U.S. naval base leased from Britain. The Archipelago comprises the British Indian Ocean Territory, located approximately 1,000 miles south of India, and 2,000 miles east of the African mainland. It is also claimed by Mauritius, an island state near Réunion Island. This claim is supported by the Organization of African Unity;

(c) The continuing security relationship with Russia, a nuclear weapon state, by four of the five parties to the Treaty of Semipalatinsk;

(d) The insistence by the parties to the Treaty of Bangkok that the non-nuclear weapon statute of the Treaty apply to the high seas out to the 200-mile Exclusive Economic Zone (EEZ) limit of each party. This has the effect of blocking China from basing nuclear weapons on islands and other partially submerged reefs claimed by China to be islands in the Southeast Asian high seas area.

The Korean Peninsula likely will present comparable problems for negotiations to establish a NWFZ in Northeast Asia.

Treaty of Tlatelolco Phased Implementation

The Treaty of Tlatelolco was negotiated over the period 1964-67 in Mexico City. The impetus for the Treaty was the then recent Cuban Missile Crisis. Latin America had had enough of nuclear crises. Garcia Robles and his fellow negotiators spent much time and effort developing the provision for the entry into force of the Treaty. As said, at the time of the negotiation there were two potential nuclear weapon programs in Latin America, in Argentina and Brazil. Garcia Robles wanted all Latin American states to have the opportunity to be involved with the Treaty at the beginning, regardless of their situation. There were two groups of states among the negotiating parties. The first group proposed that the Treaty would come into force among those states that had ratified it once eleven Latin American states-representing a majority of the participants in the Preparatory Commission which managed the negotiations—had ratified the Treaty. The second group took the position that the Treaty would only come into force after the following four things had happened.

1 – The signature and ratification of the Treaty by all states to which it was opened.

2 – The signature and ratification of Protocol I (for outside states—the United States, Britain, France, and the Netherlands—which had jurisdiction over territory in Latin America) by all states to which it was opened.

3 – The signature and ratification of Protocol II (for the five NPT recognized nuclear weapon states) by all states to which it was opened; and

4 – The conclusion of safeguards agreements with the International Atomic Energy Agency (IAEA) by all contracting parties of the Treaty and of its Protocols.

Article 28 of the Treaty provides that the Treaty will come into full force and effect only when the four requirements are met. Article 28 of the Treaty however permits a ratifying state to submit a formal waiver along with its instrument of ratification and the treaty will apply to that party’s land area, including its internal waters and territorial seas—but not the adjacent expanded sea areas. This then bridges the views of the two groups.

In practice each party to the Treaty of Tlatelolco has brought the Treaty into force for itself by depositing its instrument of ratification and the formal waiver document with the government of Mexico, the Depository. The fourth condition may never be met—and the broad adjacent sea areas brought under the Treaty which only happens upon full force and effect—as expensive IAEA safeguard arrangements likely will never be in the economic interest of small states, like the Bahamas. Such states probably will never have nuclear facilities for which safeguards would be applied.

For many years after the conclusion of the negotiations in 1967, Argentina while a signatory had not ratified, while Brazil and Chile had signed and ratified but had not waived. Cuba did nothing for a long time. Argentina and Brazil had active national programs, while Chile just followed Brazil and Cuba alone, marched to its own drummer.

By 1994 both Argentina and Brazil no longer wished to keep the nuclear weapon option open. Argentina ratified the Treaty of Tlatelolco and waved the four requirements in 1994. Brazil and Chile deposited their declarations of waiver with the government of Mexico the same year. Argentina and Chile became parties to the NPT in 1995, Brazil in 1998. With the full accession to the Treaty by Cuba in 2002, all Latin American states had signed, ratified and submitted their waiver documents pursuant to Article 28 and thus the Treaty applied to the land area of all Latin American states. Only three of the four requirements have been met. But for all practical terms the Treaty is in full force. The Treaty’s de facto zone of application is the land area territory of the parties. If the fourth requirement is ever met, the area of the application of the Treaty will expand to large ocean regions surrounding Central and South America, in the west touching the border of the South Pacific sea area which surrounds the states of the Treaty of Rarotonga.[7] Should this ever happen it could negatively affect the navigational rights of nuclear ships including warships.

Inclusion of Japan?

In addressing the specific possibility of a nuclear weapon free zone for the Korean peninsula some variation of the Garcia Robles formula might be workable. But first the following should be noted. United Nations rules provide that in order for a NWFZ treaty to be recognized by the UN, it must include, as mentioned before, “all military significant states in the region”—that may make it desirable to include Japan if the object is a treaty with a broad non-nuclear weapon commitment—and the Protocol must include all five NPT recognized nuclear weapon states as Protocol parties which among other things binds them to a negative security assurance (NSA) for treaty parties. A more detailed treatment is provided in Attachment 1.

The NPT Conundrum

The DPRK cannot rejoin the NPT until the IAEA pronounces it nuclear weapon free, a process likely to require a great many years to complete. Probably it will prove to be the case that a verification system is required that would be more stringent than that of the so-called Iran Agreement, the Joint Comprehensive Plan of Action or JCPOA—IAEA anytime, anywhere inspection rights (some facilities could be exempted by specific agreed provision) 24 hours a day, 365 days a year to last many, many years, then perhaps the U.S. would be comfortable bringing the Treaty into force. And as with South Africa, after the special verification system comes to an end, the DPRK would be subject to ongoing IAEA inspections to verify that it remains in a non-nuclear weapon status. The principal obligation that the DPRK would defer, since the DPRK would not be bound by the Treaty during verification and the U.S. would not ratify the Protocol during this period either would be the requirement to reduce and eliminate nuclear weapons. That would only come when verification is complete and the Treaty is brought into force. And the U.S. would defer its Protocol obligations of no U.S. nuclear weapon facilities in South Korea and a legally binding nuclear negative security assurance (NSA) for the DPRK.

However, during the years that the DPRK was being verified, the international law rule (Article 18 of The Vienna Convention of the Law of Treaties) that a signatory to a Treaty prior to ratification (which in this case would happen after verification) would not take acts that would “defeat the object and purposes of the Treaty” would apply. This presumably would mean no testing, transfer, or fabrication of new weapons or components for the DPRK while entry into force is pending. These constraints would impose serious limitations on the DPRK and it might want reciprocity from the U.S.

Once the DPRK is declared nuclear weapon free and it submits its instrument of ratification it would be bound by the Treaty and no longer by the international law rule. The U.S. would then ratify the Protocol where as indicated the principal obligations the U.S. would implement would be not to have nuclear weapons or related facilities in the Treaty Zone and not to use or threaten to use nuclear weapons against any Treaty party in good standing. There could be additional U.S. obligations in its Protocol. There could be important obligations applicable to the DPRK that the U.S. will want to see included in the Treaty and to happen promptly—likely non-transfer, no weapon testing and no fabrication of weapons or components. As stated above the International law rule could be sufficient but this also could be explicitly agreed in a separate agreement to reinforce the international law rule referred to above.

Again, the DPRK cannot rejoin the NPT until the IAEA formally declares the country to be nuclear weapon free and also that any sensitive nuclear technology or material it possesses are under full scope IAEA safeguards. The DPRK has no relationship to the NPT regime until it returns to the Treaty as a party.

Interim Measures

As indicated specific provisions additional to those in the NPT and the NWFZ Treaty for the interim period while verification is ongoing—arguably covered by the International Law rule—could for additional confidence be established by a separate agreement. For example, a separate protocol designed to come into force immediately signed by the DPRK and the U.S. could provide such interim arrangements. Or it could be signed by all parties—including protocol parties and provide for an interim or anticipatory regime, while the Treaty itself was proceeding towards entry into force. This has been done before—or something somewhat similar—in the Conventional Armed Forces in Europe Treaty. It was called the Provisional Application Protocol, designed to prepare for the treaty regime, pending entry into force. It could mirror the rule of international law mentioned above or it could provide specific interim obligations to preserve the treaty regime until entry into force. As said, to make such a concept palatable to the DPRK there might be a need to have a United States commitment as well, perhaps some easily reversible undertaking such as suspension of military maneuvers in South Korea, or at least those involving nuclear weapon delivery platforms.

Then there is the question of reciprocity. One might assume that, among other things, part of the final settlement would be that all U.S. nuclear weapons and nuclear weapon related facilities and components on the land territory of any of the parties within the Zone, presumably including the Republic of Korea (ROK) and Japan (as explained above Japan’s inclusion as a party may be a requirement for this Treaty to be supported by the United Nations) would be prohibited. This obligation would be also verified by the IAEA. There also would be the NSA in the Protocol.

Inspections of any U.S. facilities located in the Treaty Zone (presumably the Korean Peninsula and Japan, but not parts of Russia and China—should they be Protocol parties) would also have to be with the consent of the host party for the facilities, the ROK or Japan. For the United Nations facilities along with the western islands to be inspected it would seem sufficient to include a provision calling for this in the Treaty with an acceptance letter by the Commander of UN Peacekeeping Forces. If the Depository is to be the Secretary General as with many such treaties like this one today the Commander again could make a declaration accepting the inspections. At first glance there does not appear to be any reason for this Treaty to in any way affect the work of the Neutral Nations Supervisory Commission with respect to the Armistice. If part of the agreed package is a Peace Treaty the Commission likely would go away in any case.

The NWFZ Treaty being discussed herein-pursuant to existing U.N. rules would have a permanent duration like the other NWFZ Treaties. It would be a legally binding international agreement forever barring nuclear weapons from the area of applicability of the Treaty. The reintroduction of nuclear weapons would be prohibited by the Treaty absent a material breach.


Transit has been an important issue in the NWFZ process. It was not addressed in the Tlatelolco treaty but the United States as part of its ratification of Protocol I to the Treaty submitted several understandings, which are formal communications to other parties as to the Treaty’s interpretation. If an understanding is not challenged by another party it is included in the legal structure of the Treaty regime. If it is challenged renegotiation would be necessary. None of the U.S. understandings were challenged so they became parts to the Tlatelolco Treaty regime. One of them explicitly stated that the transit of nuclear weapons through the zone by ship or airplane is not affected by the Treaty. This provision was made explicit in all subsequent NWFZ treaties. The Rarotonga Treaty set the form this provision would take. This provision was drafted to be acceptable to Australia which permitted port visit by warships carrying nuclear weapons and New Zealand which did not. The relevant provision states that in the exercise of its national sovereignty “each party is free to decide for itself” whether to permit port visits and transit through its territorial waters by ships and visits and overflights by aircraft “in a manner not covered by the right of innocent passage”. Territorial waters are under the sovereign control of the littoral state. Innocent passage through such waters is guaranteed to all states but subject to certain rules. Therefore, whether nuclear powered or nuclear weapon capable ships or nuclear capable aircraft can make port calls or traverse the territorial sea and make visits as well as overflights in the case of aircraft depends on the policy of the states whose waters or airspace it is. If that state chooses to prohibit transit by such ships and aircraft then traversing the territorial waters or airspace would not be “innocent”.[8]

The Treaty of Pelindaba utilizes the same language as Rarotonga to the same effect that each party in the exercise of its sovereignty can decide whether to allow port visits and transit of territorial waters by ships and overflights by aircraft in a manner not subject to innocent passage. All aircraft in order to overfly or utilize the airport of a state must only do so pursuant to permission of that state which for normal commercial aircraft is part of the world-wide commercial air traffic control system regulated by international agreement. This is of course different from innocent passage through territorial waters but considered under NWFZ treaties in a similar way for certain aircraft. Port visits, transit of territorial waters by nuclear powered and nuclear capable ships and nuclear capable aircraft overflights and landing at airfields are treated together. And the program applies in the NWFZ treaties to ships and planes that are nuclear weapon capable as well as to nuclear powered in the case of ships. Overflights by aircraft of the territorial sea designed to be threatening if carried out without the permission of littoral state—as they wouldn’t be—are a breach of sovereignty and not permitted under international law.[9] Such overflights along the land border—such as along the DMZ might not be contrary to international law but they certainly would be contrary to the spirit of the Treaty regime.


Now turning to verification provisions, it seems likely that these provisions have to be at least as severe—and probably more so than the JCPOA regime, applicable to Iran. The following is an outline of what this regime possibly might look like in general.

A preambular paragraph could contain a similar no nuclear weapons ever pledge that Iran undertakes in the JCPOA making the DPRK the second country in the world to make this pledge: “The Democratic People’s Republic of Korea affirms that in the future under no circumstances will the Democratic People’s Republic of Korea again seek, develop, or acquire nuclear weapons.” (It is good to have this general principle stated.).

Second, the plutonium production reactor at Yongbyon should be verifiably destroyed according to agreed procedures. (The only purpose that this reactor has is to make plutonium for nuclear weapons therefore it should be destroyed.)

Third, the enrichment plant, likely built with the design and components supplied to the DPRK by A.Q. Khan should be dismantled according to agreed procedures. This facility was built with designs and components illegally acquired and there is no reason to permit the DPRK to enrich uranium on the large scale that this plant permits. Almost certainly other enrichment plants exist. Enrichment is permitted to NPT parties in good standing under the Treaty. But under the most positive scenario the DPRK is many years away from such a status. Indeed, it may be that a decision will be taken years from now to readmit the DPRK to the NPT with less than complete certainty that the country is in fact free of nuclear weapons because complete certainty is impossible to achieve. The DPRK may contain too many caves, tunnels, etc., to enable certainty to be achieved no matter how many highly intrusive inspections are carried out. Former U.S. Defense Secretary William Perry, a veteran of years of negotiations with the DPRK, believes this and has said so in print. As a result, enrichment is simply too dangerous a technology for the Nuclear Supplier’s Group to permit to be exported to the DPRK or for it to continue to possess or to develop such technology for the foreseeable future.

Fourth, there must be 24 hours a day, 365 days a year highly intrusive inspections on-site by the IAEA—and, for political reasons, the inspection teams must include some American experts—looking for nuclear weapons and related technology. This process will continue for many years, likely years after the DPRK has returned to the NPT and this Treaty has entered into force. At least ten years beyond the entry into force of this Treaty should be the duration of the enhanced inspection regime. After some agreed point in time perhaps a more normal verification regime could be agreed upon—such as other NPT parties have agreed to work with the IAEA—but this would be many, many years in the future. Verification that the DPRK is nuclear weapon free will be excruciatingly different.

Fifth, DPRK borders should be placed under tight control 24 hours a day, 365 days a year, with respect to any nuclear weapon related technology. The duration of these controls should be the same as for the IAEA inspections.

Sixth, likewise for the same duration, the uranium mines and any thorium production should be closely monitored under an identical regime.

Seventh, research and development with respect to any nuclear weapon related technologies should be closely monitored continuously and for the same duration.

Eighth, as specific identifiable progress is confirmed by the IAEA sanctions could slowly be lifted over a number of years. Specific goals should be set and periodic reports by the DPRK to the IAEA Board of Directors required.

Ninth, any violations of the inspection process reported by the IAEA will result in the return of all sanctions that have been lifted and new ones added.

A nuclear weapon free zone treaty for Northeast Asia certainly should be the longer-range objective for the region. It was Garcia Robles’ vision that “We should attempt to achieve a gradual broadening of the zones of the world from which nuclear weapons are prohibited to the point where the Territories of Powers, which possess these terrible tools of mass destruction, will become ‘something like contaminated islets subjected to quarantine.’”[10] This is the philosophy we all should follow. But in Northeast Asia we are far away from this now. The DPRK currently has said that all it will accept in terms of a denuclearization agreement is a non-binding Resolution akin to the 1992 Resolution. This could be seen by the world community as worthless but that isn’t necessarily so. It depends on what it’s linked to and accompanied by. The Helsinki Final Act of 1975 and The Stockholm Document on Confidence Building Measures in Europe, 1986 are examples of non-legally binding agreements which proved very valuable. They led to the Conventional Armed Forces in Europe Treaty in a few years, and other developments in Eastern Europe and the end of the Cold War in 1991.

It would seem that the primary objective should be a Peace Treaty ending the Korean War. Hostilities associated with a state of war on the Korean Peninsula ended 65 years ago. That would appear to be long enough to wait. A Peace Treaty—actually probably two peace treaties one between the two Koreas—but that would involve recognition—so perhaps the inter-Korean Peace Treaty could be with the United States as an intermediary third party—and there could be a separate Peace Treaty between the DPRK and the U.S. Much would follow from this: U.S. diplomatic recognition of the DPRK, an exchange of Ambassadors and the end of threatening overflights and close border flights. Such flights except arguably flights along the DMZ border, would be breaches of sovereignty, and inconsistent with peaceful relations. And the land border flights would certainly be questionable. There could also be a protocol to the U.S.-DPRK Peace Treaty or it could be an associated standalone agreement. This agreement could put in legally binding obligations the DPRK decision to do no more nuclear weapon or missile tests and no exports of such technology. The U.S. would have to give something for this, as suggested above, an ending of U.S. military exercises in the South, at least those involving strategic platforms.

These would be very positive steps which would add considerably to stability. But into this will be thrust by the media the non-legally binding and unverifiable denuclearization declaration. There would be little confidence anywhere that the DPRK would actually eliminate any of its nuclear weapons. The U.S. would not be able to give the DPRK a legally binding NSA. Almost certainly the U.S. Senate would never consider abandoning the U.S. nuclear deterrent absent complete confidence that the DPRK stockpile had been or was about to be eliminated. A two-thirds vote in the Senate would be required to approve a legally binding NSA and such a vote is tough to get. The U.S. of course could give to the DPRK a non-binding NSA in the form of a national statement, which all of the P-5 gave to all NPT parties in 1995. The DPRK also was the recipient of such assurances while it was an NPT party. It is essential that the denuclearization declaration not be presented as an end in itself but as the key to the Peace Treaty and all of its benefits.

The issue of NSAs for non-nuclear weapon states from the NPT nuclear weapon states goes back to the beginning of the NPT.

In the 1960s, as the number of states possessing nuclear weapons rose to five, there were projections that 20-30 additional states would acquire nuclear weapons over the next two decades, and if such a scenario had occurred, there would have likely been many more in the following decades. As part of an effort to stem the trend toward the widespread proliferation of nuclear weapons, 62 states signed the NPT on July 1, 1968, the first day the Treaty was open for signature. During the NPT negotiations, the Non-Aligned Movement (NAM) states sought negative security assurances from the nuclear-weapon states, arguing that after all, if the non-nuclear-weapon states were to foreswear nuclear weapons, the least the nuclear weapon states could agree to was not to threaten or use nuclear weapons against non-nuclear-weapon states.[11] In 1965, the United Arab Republic (UAR) rejected the idea of bilateral security guarantees, claiming that it would result in “a situation where vast areas were divided under a nuclear trusteeship of this or that Power.”[12] Several non-nuclear-weapon states requested that assurances or guarantees from the nuclear-weapon states accompany or be included in the emerging non-proliferation treaty. Soviet Premier Kosygin proposed (on February 1, 1966 in the Soviet draft of the Non-Proliferation Treaty) “a clause on the prohibition of the use of nuclear weapons against non-nuclear States parties to the treaty which have no nuclear weapons on their territory.”[13] The UAR, Mexico, Nigeria, and India (ultimately not a signatory of the NPT) supported this initiative. UAR Ambassador Khallaf submitted treaty language that incorporated Kosygin’s proposal, specifying that “each nuclear-weapon state undertakes not to use, or threaten to use, nuclear weapons against any non-nuclear-weapon state Party to this Treaty which has no nuclear weapons on its territory.”[14] Romania and Switzerland made similar proposals.

US President Lyndon Johnson had assured nations that did not seek nuclear weapons that they would, if the need arose, enjoy strong US support “against nuclear blackmail threat,”[15] but the United States refused to accept the Soviet proposal. Canada also refused such a proposal arguing that reaching a consensus to include it in the Treaty would prove difficult, and attempting to do so would unacceptably prolong negotiations. Canadian representative Burns suggested instead that the nuclear-weapon states make parallel declarations that could include negative security assurances. More specifically he proposed that the nuclear-weapon states pledge in these declarations “not to use nuclear weapons against non-aligned non-nuclear parties.”[16]

In the beginning of 1968, the revised draft treaty still did not include any security assurances for non-nuclear-weapon states.[17] At the Eighteen Nation Disarmament Commission (ENDC), certain non-nuclear-weapon states voiced their regret regarding the absence of any such assurances and the Federal Republic of Germany stated that the treaty should ban nuclear blackmail against the non-nuclear weapon states. Romanian Ambassador Ecobesco again requested that the nuclear-weapon states include an undertaking not to use or threaten to use nuclear weapons.[18] In March 1968, the United States, the Soviet Union, and the United Kingdom agreed to offer some positive security assurances.[19] Such assurances generally refer to action that would be taken by the Security Council or by its permanent members to assist an NPT non-nuclear-weapon state if it was attacked or threatened with nuclear weapons. However, US Ambassador de Palma stated that the draft treaty did not include security assurances because the issue proved “too difficult and complicated to be reduced to a treaty provision.”[20] Thus, NATO concerns about the conventional superiority of the Warsaw Pact and the credibility of the Western Alliance’s “flexible response” policy, as well as the Soviet Union’s reluctance to give negative security assurances to non-nuclear-weapon states members of NATO, precluded any agreement among the nuclear-weapon states on negative security guarantees at that time.[21] Only China (not a NPT party until 1992) unilaterally pledged a no first use policy.

During the Cold War, mutual fear on both sides of the Iron Curtain prevented further progress in this area as the nuclear-weapon states denied repeated requests by non-nuclear-weapon states for the NSAs to be made legally-binding. The primary reason lay in distrust across the East-West divide. It was contended that non-nuclear-weapon states in the Warsaw Pact countries, as an alliance, possessed conventional superiority over NATO, and were closely allied to the Soviet Union. For its part, the Soviet Union argued that NATO stationed nuclear weapons on the national territories of its non-nuclear-weapon state members.

In 1995 at the NPT Review and Extension Conference, the five nuclear-weapon states reaffirmed, and to a degree harmonized, their political commitments not to threaten NPT non-nuclear-weapon states parties with nuclear weapons in the context of the NPT extension in 1995. The NPT nuclear weapon states agreed to legally binding NSAs when signing the relevant protocols to the African, South Pacific, Central Asian, and Latin American Nuclear Weapon Free Zone (NWFZ) Treaties. The Southeast Asian Nuclear Weapon Free Zone Treaty remains unresolved on this point because of the extension of the treaty limits to the high sea areas in Southeast Asia by extension to the EEZ limits. The protocols to these NWFZ Treaties strengthen the NSAs as they require the nuclear weapon states (they have all signed the relevant protocols) not only to refrain from the use of nuclear weapons against the states parties to the NWFZ Treaties, but also from the threat of use of nuclear weapons. The Latin American Nuclear Weapon Free Zone signed in 1967, a year before the NPT was the first to require a legally binding NSA for the nuclear weapon states—which non-nuclear weapon states not in a NWFZ treaty ask for to this day. All of the other Zonal treaties have followed the Latin American lead on this.

One other comment, the P-5 countries have only given legally binding NSAs to parties to NWFZ treaties. This is recommended by the UN rules and the P-5 members have judged that in general NWFZ treaties arguably one exception to this practice, the NSA given to the DPRK by the United States in the 1994 Agreed Framework is part of a legally binding international Agreement—not a treaty but an international agreement—as a result of being one of the provisions of the Agreement. This NSA could have been considered legally binding as well.[22] However, the language used in the text is not such as to indicate legal obligation: “The U.S. will provide formal assurances to the DPRK, against the threat or use of nuclear weapons by the U.S.”[23] Thus, the impact of the provision is not clear. The same issue exists with respect to the relevant provision of the 2005 joint statement of principles developed in the Six Party Talks involving the DPRK, the ROK, the U.S., China, Russia, and Japan. The relevant provision there reads, “The United States affirmed that it has no nuclear weapons on the Korean Peninsula and has no intention to attack or invade the DPRK with nuclear or conventional weapons.”[24] Unlike the provision in the Agreed Framework this language almost implies that it is only present intention that is referred to. However, the joint statement does contain a general provision of assurance in which the two parties agree to respect each other’s sovereignty, co-exist peacefully and take steps to normalize their relations. This did not appear to satisfy the DPRK.

Two treaties signed with the Soviet Union in 1961 provided the basic security assurance for the DPRK for many years, however with the collapse of the Soviet Union the DPRK realized that it would have to look for its security assurances elsewhere. The DPRK therefore undertook a series of negotiations with the United States. In general, the DPRK seemed satisfied or at least found acceptable to the NSA in the 1994 Agreed Framework. With respect to the Six Party Talks neither the security assurance nor the general reassurance clause in the Joint Framework appeared to be sufficient. The only regime assurance undertaking with the United States which has been valued by the DPRK is the 2000 Clinton-Myong-Rok “No Hostile Intent” communiqué signed by the two parties in October, 2000. More precisely it said that neither government would have “hostile intent” against the other and both sides were intent on building a “new relationship free from past enmity.”[25] This document was of great importance to the DPRK, it was regarded by North Korea in much the same way that the Chinese regarded the Shanghai communiqué of 1978, as the foundation of a new relationship.

But in all this the DPRK was seeking a better relationship with the United States and the survival of the regime. The 1994 Agreed Framework and the 2000 “no hostile intent” communiqué appeared to be important steps on the road to achieving that, the Joint Statement of Principles of 2005 did not. In the Singapore process the DPRK is seeking an NSA, perhaps a non-legally binding assurance will suffice when associated with a Peace Treaty.


An arrangement between the DPRK and the United States and other countries on the elimination of the DPRK nuclear weapon stockpile and related equipment and technologies could of course take many forms and can be shaped to the desires and requirements of the parties. But if it ever becomes possible there is merit in the NWFZ form. It is well understood around the world. If done according to United Nation rules it would virtually guarantee the full support of the United Nations which could be very valuable. It would also help the DPRK to become a full-fledged member of the world community. The UN played a central role in the negotiation of the African Nuclear Weapon Free Zone Treaty and the Central Asia Nuclear Weapon Free Zone Treaty providing staff, expert and logistical support. This of course would not be lacking in any negotiation involving the United States and the DPRK but the UN would bring a special perspective. The CTBT Verification Center in Vienna with its unrivaled capability in seismic and other technologies of monitoring would be part of the package. There is no reason that such an agreement would have to precisely mirror the five NWFZ treaties that have gone before. They all differ among themselves. But their provisions provide past solutions which can be guides and precedents for this most important project. Let us hope one day this can be done.

So as a final note what to make of the two approaches.

The bilateral approach:


It appears to be negotiable, the DPRK has so indicated.

It can unlock the door to the Peace Treaty which can bring a number of good things.

• diplomatic recognition
• exchange of Ambassadors
• commerce and trade
• a possible side agreement halting the weapon and missile testing and export of nuclear weapon related technology permanent on a legally binding basis.
• an end to threatening overflights
• an agreed policy promulgated to eventually eliminate nuclear weapons and stop threatening their use.


The Denuclearization Declaration will be seen by countries around the world and the political class in the U.S. as worthless, non-binding, unverifiable and unreliable. This may poison the Peace Treaty process and the good things that can come from it.

Only a non-binding NSA for the DPRK will be possible, likely nothing for the ROK. The ROK already has a non-binding NSA through the NPT.

This agreement will not significantly bring the DPRK into the world community.

The multilateral approach:


It would solve the problem of nuclear weapons on the Korean Peninsula. It would have the full support of the world community and all six powers. The six power talks participants all would be centrally involved. DPRK security would be protected.

The DPRK would receive legally binding NSAs from the P-5 as would the ROK and Japan.

It would bring the DPRK into the world community as a full-fledged member. This would be very positive for the DPRK economy. It could even bring a close relationship with the U.S.

The verification process would significantly change DPRK society (likely seen as a con in some quarters).


This is not negotiable at this time.

The verification process would be exceedingly long.

In conclusion, the multilateral approach is where we should eventually be. Garcia Robles should remain our guide. In the short term we should make the best of the bilateral approach and keep our focus on the Peace Treaty.



This attachment expands on the suggestion on page 7 of this essay that: “In addressing the specific possibility of a nuclear weapon free zone for the Korean Peninsula some variation of the Garcia Robles formula might be workable.” It provides more specific explanation of how a variation of the Garcia Robles formula might be applied to the Korean Peninsula.

In developing his concept of a nuclear weapon free zone, Garcia Robles was undoubtedly influenced by the thinking about the subject that existed already at the time of the negotiation of the Treaty of Tlatelolco that established a nuclear weapon free zone for Latin America.[26]

The central thought was that any such negotiated zonal treaty should be based on four basic principles, viz:

1. Participating countries must undertake a legally binding international commitment not to produce or deploy nuclear weapons anywhere in their territories and not to permit other countries to do so. This obligation remains valid in times of war and times of peace.
2. Nuclear weapon states recognized under the NPT must agree to respect such a zone and not to use or threaten to use nuclear weapons on participating states.
3. An international verification system must be established.
4. A zone must be established in such a way that it enhances international peace and security, which suggests that all states in the region affected directly by the zonal treaty should concur with it.

To these four fundamental principles, Garcia Robles made three important innovations.

First, he held that all significant states in the region to which the zonal treaty applies should be parties; second, that the motivation for the zonal treaty should come from countries within the zone; and third, that the right of the parties to use nuclear energy for peaceful purposes must be preserved.

Garcia Robles also established an implementing body for the treaty, OPANAL, with an implementing staff and periodic high-level meetings of the parties to provide oversight. Verification would be by the International Atomic Energy Agency with parties signing Safeguards Agreements with the Agency. Also pursuant to an Understanding submitted by the United States but accepted by all parties, the transit of nuclear weapons through the treaty zone by sea and air is not covered by the constraints of the zonal treaty.

Garcia Robles produced two Protocols for relevant outside states to attach themselves to the Treaty: One Protocol invites outside states responsible for administrating territory in the zone, such as the United States with Puerto Rico and the Virgin Islands, to place these areas under the zonal treaty’s constraints. Another Protocol is open only to the five recognized nuclear weapon states and invites them to assume the obligations never to use or threaten to use nuclear weapons against treaty parties.

All subsequent nuclear weapon free zones are based on the above principles: South Pacific, Africa, Southeast Asia and Central Asia. All of the regional treaty regimes work closely with the United Nations through their implementing bodies. Each of these zones was also custom-tailored to regional circumstances. Northeast Asia is no different.

Applying these principles to the Korean Peninsula would involve the following:

First, it would require that the Republic of Korea, the Democratic People’s Republic of Korea, and Japan all become parties to the Treaty and undertake the fully meet all the non-nuclear weapon constraints referred to in the first pillar above.

Second, a verification system would have to be established. Given the diplomatic situation that exists today, it would have to be at least as intrusive as the system established for the Iran Agreement with the Security Council, the JCPOA, which also is administered by the IAEA.

Third, an implementing body should be established by the three parties with a Secretary General and international staff.

Fourth, the recognized nuclear weapon states, the United States, the United Kingdom, France, Russian and China, would sign a Protocol in which they would undertake never to use or threaten to use nuclear weapons against the parties to the “North-East Asian Nuclear Weapon Free Zone Treaty (NEANWFZ).” Pursuant to the diplomatic situation of today, the United States might wish to delay its ratification of this Protocol until the nuclear weapon stockpile of the DPRK has been fully verified by the IAEA and process of reductions has at least begun.

In the course of actually negotiating the Treaty of Tlatelolco, Garcia Robles wove a set of linkages to enable the different parties to commit to the treaty from the outset, thereby keeping all the relevant states “inside the tent” (or in this case, the zone) while coming into compliance and/or ratifying the treaty. In Article 4, he set up four requirements for entry into force, viz:

1-signature and entry into force of the Treaty by all eligible parties (all Latin American states).
2-signature and ratification of Protocol I by all eligible states( United States, United Kingdom, France, the Netherlands).
3-signature and ratification of Protocol II by all eligible states (United States, United Kingdom, France, Russia, China).
4-Completion of IAEA Safeguards Agreements with the IAEA by all eligible states (all Latin American states).

Conditions 1-3 are fully met and have been since 2002 (Cuba was the last to join). As noted above, condition 4 will probably never be met as small Latin American states don’t want to spend money on Safeguards when they are never going to have safe-guardable nuclear facilities. Thus, it is likely the Treaty will never fully come into force, yet it provides effective constraints on all the parties to the treaty.

Article 28, paragraph 2 permits each individual Latin American state to sign, ratify and then formally waive the requirements and thereby apply the obligations to its national territory and territorial sea (twelve-mile zone). A party therefore could sign, ratify and waive and apply the treaty to its national territory only and join the treaty individually (no broad ocean areas).

By exploiting this clause, Garcia Robles permitted the two states in the region with potential nuclear weapons programs in 1967 to associate themselves with the Treaty but not have the Treaty obligations apply until they were ready: by signing and ratifying but not waiving. Garcia Robles wanted to involve all Latin American states with the Treaty from the outset if he could. Brazil chose to take advantage of this option (as did Chile which followed Brazil in this area) but Argentina did not.

This is a different situation from the Korean Peninsula but a variant of Garcia Robles’ formulae could be pursued in Northeast Asia.

To describe this approach in Northeast Asia in Garcia Robles terms: There would be only three requirements for full Treaty entry into force, the first three of Garcia Robles Tlatelolco requirements cast in Korean Peninsula terms, that is:

1-signature and ratification of the Treaty establishing a Northeast Asia Nuclear Weapon Free Zone.

2-signature and ratification of the Protocol placing territory in the Nuclear Weapon Free Zone controlled by an outside state (probably not required in the case of a Northeast Asia Nuclear Weapon Free Zone Treaty).

3-signature and ratification of the Protocol adhered to by the NPT recognized nuclear weapon states (the P-5).

In the Korean case probably only requirements 1 and 3 apply, so in practice there would be only two requirements.

These requirements are only met when the Treaty and the necessary Protocol or Protocols are signed and ratified by all parties to which the Treaty or Protocol or Protocols are opened for signature-or in other words by all relevant parties.

The relevant states for the NWFZ for NE Asia Treaty would be the ROK, the DPRK and Japan. The relevant states for the Protocol would be the P-5.

Thus, at the outset, in a Robles approach, Japan would sign, ratify and waive; four of the P-5 might do the same with the Protocol; but probably the United States for the Protocol and South Korea and North Korea for the Treaty would sign, ratify but not waive (like Brazil) or do nothing (like Argentina) until the complete DPRK nuclear weapons production program has been satisfactorily verified by the IAEA and DPRK nuclear weapon reduction is at least well underway and perhaps completed.

An additional possibility, suggested by Morton Halperin, is that ROK and Japanese commitment may be contingent upon specified denuclearization of the DPRK over timelines and that if sufficient progress in this regard is not achieved, for example, within three or five years, these states reserve the right to withdraw from the treaty.[27]

A further important precedent was set in the Treaty of Tlatelolco. This was the inclusion of the US territory of Puerto Rico and US Virgin Islands in the territories covered by Protocol 1 of the zone. Originally, the United States opposed this inclusion, as was noted in the UN 1976 Comprehensive Report.[28] Ultimately, however, Protocol I of the Treaty of Tlatelolco was signed by President Carter in 1977, it was approved by the Senate in 1980 pursuant to three Understandings making clear that transit is unaffected-not objected to by other nations- and it was ratified for the US in 1981 by President Reagan. President Nixon dealt with Protocol II.

US adherence to Protocol I provided that all US possessions in the Treaty Zone are placed under the nuclear weapon free zone provisions of the Treaty of Tlatelolco. Thus, nuclear weapons may not ever be stationed, deployed, manufactured or tested in Puerto Rico or the Virgin Islands.

This precedent has obvious application to the possible inclusion of Guam in a regional nuclear weapons free-zone, given the fixation of the DPRK’s leadership on this territory as a site from which nuclear attack might be launched against it; which in turn raises the issue of symmetry, and whether areas of Northeast China and the Russian Far East might also be included—noting that every addition introduces complicated asymmetries of interest and negotiation costs.



[1] Nuclear Threat Initiative, “Joint Declaration of South and North Korea on the Denuclearization of the Korean Peninsula,” Last Updated: October 26, 2011, at:

[2] Ji Hyun Lee, “Assessing the idea of South Korea being a virtual NWFZ since the 1992 Joint Declaration for the denuclearization of the Korean peninsula,” paper presented to the “Strong connections: Australia-Korea strategic relations – past, present and future” workshop, Nautilus Institute, Seoul, June 15, 2010, pp. 3-4, at:

[3] Ji Hyun Lee, op cit, page 5.

[4] Shaker, Nuclear Non-proliferation Treaty, p. 920.

[5] Ibid., pp. 923-924.

[6] Convention on the Prevention of Marine Pollution by Dumping of Wastes and Other Matter (1975); Consultative Meeting of Contracting Parties (1993). It should be noted that the 1975 Convention is still a treaty, therefore, as in the Rarotonga Treaty, the states parties cannot create obligations for non-parties in areas beyond their national jurisdiction.

[7] Garcia Robles, Latin American Nuclear-Weapon-Free Zone, pp. 18-19.

[8] Article 5, Treaty of Rarotonga.

[9] Article 4, Treaty of Pelindaba.

[10] Alfonso Garcia Robles, quoted in UN General Assembly, First Committee Provisional Verbatim Record, 13 and 32; Michael Hamel-Green, “Peeling the Orange: Regional Paths to a Nuclear-Weapon-Free World,” accessed February 16, 2015,

[11] On March 9, 1967, the Nigerian representative argued before the ENDC that it would be unfair to ask any responsible government to adhere to an NPT without guarantees. The Brazilian delegate added that non-nuclear-weapon states signatories to the NPT would be surrendering “the most important means they might otherwise have at their disposal to counter possible aggression.” See International Negotiations of the NPT, ACDA (US Arms Control and Disarmament Agency), p73.

[12] Statement by the UAR representative in the UN First Committee, see International Negotiations of the Treaty on Nonproliferation of Nuclear Weapons, The Arms Control and Disarmament Agency, publication 48, January 1969, p26.

[13] International Negotiations on the NPT, ACDA, p42.

[14] Ibid., p89.

[15] Ibid., p73.

[16] Ibid., pp89-90.

[17] Ibid., p112.

[18] Ibid., p112.

[19] For more detail, see Tripartite Proposal on Security Assurances, March 7, 1968, ibid., p112.

[20] Ibid., p112.

[21] George Bunn, “Security Assurances to Non-Nuclear-weapon states,” The Non-proliferation Review, The Monterey Institute of International Studies, Fall 1994, Volume 2, Number 1.

[22] Agreed Framework between the United States of America and the Democratic People’s Republic of Korea.

[23] Ibid.

[24] Joint Statement of the Fourth Round of the Six Party Talks, September 19, 2005.

[25] US-DPRK Joint Communiqué, Washington, D.C., October 12, 2000.

[26] Garcia Robles, The Latin American Nuclear Weapon-Free Zone (1979) is available for download here:

[27] “The treaty could be structured so that it goes into effect when the three nuclear weapons states (the US, Russia and China) and the two non-nuclear states (Japan and South Korea) ratify it. However, South Korea and Japan would have the right to withdraw from the treaty after three or five years if the provisions were not being enforced effectively throughout the Korean Peninsula. Effective enforcement would occur if either North Korea ratified and implemented the treaty, or if it collapsed and the Peninsula were unified under South Korea.” Morton H. Halperin, “A Proposal for a Nuclear Weapons-Free Zone in Northeast Asia”, NAPSNet Special Reports, January 03, 2012,

[28] “In 1965, the United States declared that neither the United States Virgin Islands nor Puerto Rico could be included in the nuclear-weapon-free zone because the Virgin Islands were part of the territory of the United States and Puerto Rico had a special relationship with the United States. The Canal Zone, the United States added, could be included, provided that the rights of transit through the Panama Canal were not affected, as well as the Guantanamo base, if Cuba joined the Treaty. In 1974, the representative of the United States declared at the twenty-ninth session of the General Assembly that the position of his Government with respect to Additional Protocol I remained unchanged (A/C.l/PV.2023, p. 12).” In COMPREHENSIVE STUDY OF THE QUESTION OF NUCLEAR-WEAPON-FREE ZONES IN ALL ITS ASPECTS, SPECIAL REPORT OF THE CONFERENCE OF THE COMMITTEE ON DISARMAMENT, A/10027/Add.1, New York, 1976, page 14.


Citizen’s Watch on the Implementation of Korean Denuclearization Agreements

Category : PSNA Activities

Peace Depot launched a new project “Toward a Northeast Asia Nuclear Weapon Free Zone: Citizen’s Watch for a Fair Implementation of Korean Peninsula Denuclearization Agreements” (in short: “Citizen’s Watch on the Implementation of Korean Denuclearization Agreements”) on Nov. 14, 2018. The project is meant to help successful implementation of the historic agreements regarding the peace on the Korean Peninsula and beyond.

This project publishes the “Watch Report” roughly once every three weeks, first in Japanese, then later in Korean and English. The “Watch Report” is published on the following free-access blog website.

Watch Report [English] [Korean] [Japanese]

They welcome your comments and advices, so please write to “office■”(Please change ■ to @).

*PSNA is an advisor of this project.
*Dr. Hiromichi UMEBAYASHI, Director of this Project, is Member of PSNA, Former Director of RECNA and Visiting Professor of Nagasaki University, Japan.

【Mar. 8, 2019】Special Essay on US-DPRK Summit in Hanoi (PSNA-WP-5)

Category : PSNA Activities

Applying diplomacy for the common good

Shen Dingli, Professor and Director, Program on Arms Control and Regional Security, Institute of International Studies, Fudan University

PSNA Working Paper Series (PSNA-WP-5)1

March 6, 2019

[Japanese version: PDF]

  The much expected second summit between President Donald Trump of the US and Chairman Kim Jung Un of the DPRK has just ended with no deal. Despite the fact that both advance teams had reached Hanoi for days-long pre-summit negotiation without reaching an agreement, Washington and Pyongyang still expected to deliver a deal at the summit level.

  However, on the surface it seems that both parties have had significant misunderstandings before and during their talks, which made it impossible to hammer out a deal with proper mutual compromise.

  Accord to President Trump in his solo post-summit press conference, Chairman Kim demanded that the US lift “all” economic sanctions in exchange for the DPRK’s verified closure of its Yongbyon nuclear complex, without addressing any further denuclearizing details including declaration, inspection and dismantling all DPRK’s nuclear weapons and missiles program, hopefully completely and irreversibly.

  Nevertheless, the DRPK Foreign Minister Ri Yong Ho said at a late press conference of his own that Pyongyang only asked for “partial” lifting of sanctions, contradicting what Trump said. DPRK has indicated that in principle it agrees with denuclearization for complete lifting of sanctions. But at present the DPRK only asked for partial lifting of sanctions, and offered to close its Yongbyon nuclear site.

  If both sides had expressed their positions unambiguously, they should not have such vastly different understandings, and their differences could be narrowed. The US side was opposed to entire lifting of sanctions for non-complete implementation of DPRK’s denuclearization. This is quite understandable. Actually the US current stance of possible partial lifting of sanctions for denuclearizing a significant part of the DPRK nuclear program has already been much more flexible than its previous position at the earlier summit in Singapore.

Though it is still hard to decipher why the two teams could not communicate and understand each other properly, one should keep in mind that they are willing to use diplomacy to settle their difference. In response to a journalist’s question, Kim Jung Un said he would not have come all the way to Hanoi without a genuine intention to denuclearize.

  It is possible that both sides have attempted high-pressure diplomacy. In this initial tangible high-stake give-and-take game, Pyongyang wanted to retain as much as nuclear flexibility by restricting the US freedom in verification scope, and meantime to maximize its potential gain through attempting a total lifting of sanctions, though signaling also a possible partial lifting as well. The US side, after offering a liaison office as its carrot, resolutely denied a “non-equal” deal. Given their similar character and temperament, their best intent of opting for diplomacy could end with applying pressure. As both feel “no rush” at this time, their talk quickly collapsed.

  If the above conceived scenario makes any sense, the lesson is clear. Each side has not applied sustained diplomacy. Instead, they could well end up with using diplomacy as a one-shot pressure. In light of this, their failure to strike out a deal this time is not surprising.

  Given the vast distrust between the two countries, any pursuant of a quick solution through denuclearization is unrealistic. The reason is pretty simple: the DPRK doesn’t feel secure and is keen to preserve certain nuclear capability in its early stage of “denuclearization”, even if Pyongyang would do it eventually. The US, for its part, has to address the DPRK nuclear and missile arsenal in its entirety despite its own softening of position from a “time-bound total denuclearization” to a “no rush” incremental process.

  Understanding the DPRK’s feel of geostrategic insecurity would help make explicable the collapse of this talk. Given the demise of the Soviet Union, the DPRK has lost one key ally. The other, China, still has a legally-binding alliance agreement with the DPRK on paper. However, Pyongyang has doubts that China would ever have honor its commitment. At a time China is branding itself as a “responsible power”, Beijing might accord its regional and global responsibility to stem nuclear proliferation a higher priority over its bilateral commitment to come to the DPRK’s defense, if the two objectives conflict with each other.

  In addition, Pyongyang would not fail to see that the US still has a hegemonic foreign policy. Without any hard evidence or UN Security Council endorsement, the Bush administration launched an illegal preemptive attack on Iraq, and described Pyongyang along with Teheran and Baghdad as a member of “an axis of evil”. President Obama, while receiving his Nobel Peace award, partly for his leadership of “global zero”, exercised his “strategic patience” to be distant with the DPRK. In addition, President Trump has ordered the building of miniaturized nuclear warheads, so as to deter those leaders of “rogue states” more effectively.

  Remembering President Trump’s “fury and fire” in 2017, his sudden shift of stance toward the DPRK could well be his display of “art of deal”. Once the DPRK would imprudently follow the US demand of a CVID approach to comprehensively, verifiably and irreversibly denuking itself while the US would do nothing during this process, Pyongyang could well commit political and strategic suicide without any parallel economic benefits. No sensible DPRK leader would ever embrace such a one-sided highly risky formula.

  Seven months after the first summit in Singapore in June last year, the White House has become more pragmatic, as it now seems to understand that virtually it is impossible to achieve a “US-wins-all, the DPRK-loses-all” denuclearization game. Lately, President Trump has sent a number of messages to show that based on his wonderful personal relations with Chairman Kim, he would not expect the second round of summit to conclude a total and immediate denuclearization. Instead, he would enlighten his DPRK counterpart with the success of Vietnam, the host country, that Pyongyang holds a tremendous opportunity of economic prosperity, if it is prepared to surrender the core of its nuclear and missile arsenals, while the US would offer some enticing olives including setting up an American liaison office in Pyongyang.

  While the US has softened its position to some extent, the rest of the world would expect the DPRK to reciprocate in return. Pyongyang may not only agree to extend its existing virtual nuclear and missile testing moratorium, but also offer a certain version of roadmap toward eliminating its fissile material and missile arsenal, though in a limited and incremental manner.

  A reasonable conclusion from Trump’s walkout from this summit is that high-pressure diplomacy has led to an impasse on relative cost-benefits on the table. This may indicate either their lack of skill in using diplomacy, or lack of intent to make deal so soon. From the DPRK side, indeed it is in no rush to yield.

  Militarily, after six rounds of nuclear weapons tests and tens of missile flight tests, Pyongyang’s nuclear deterrent should not be underestimated. With nuclear armaments, Pyongyang trusts that it is more secure rather than without. And, with its acquired data from previous tests, the DPRK may be confident in building up its nuclear tipped missiles without immediate further tests, and hence without annoying its neighbors. Economically, it may be hungry but not dying. With its testing moratorium, Pyongyang has already much improved its relationship with Beijing, Seoul and Moscow. It is likely to believe that the beleagued Trump may be thirstier to get a deal, given his presidential reelection need, at a time of America’s challenging relationship with China, Russia, and allies, as well as his many legal battles in the Congress.

  The world should not allow the US and the DPRK to squander time this way. Without making even an imperfect deal of denuclearization, Pyongyang’s nuclear and missile arsenal could keep expanding, adding its weight in any future talks with the US. If the two sides intend to meet again, they should be better prepared than this time, communicating well and accommodating each other’s legitimate concerns. They should trust and apply diplomacy for common goods, rather than use it merely as a platform. Based on improved diplomacy, the US and the DPRK shall present to the world that they are able to achieve both denuclearization and the opening of a new peace era with official ending of the Korean War.


1 This paper was commissioned by RECNA on behalf of Co-chairs of Panel on Peace and Security of Northeast Asia. The views and opinions expressed here do not necessarily reflect those of PSNA.

【Mar. 8, 2019】Special Essay on US-DPRK Summit in Hanoi (PSNA-WP-4)

Category : PSNA Activities

After the 2nd US-DPRK Summit in Hanoi, Vietnam: Sustain the Inter-Korean Momentum

Mark Byung-Moon Suh, Council Member, Pugwash Conferences on Science and World Affairs, Member of the German Council on Foreign Relations and
Elisabeth Imi Suh, Research Assistant, German Institute for International and Security Affairs

PSNA Working Paper Series (PSNA-WP-4)1

March 6, 2019

[Japanese version: PDF]

  The much expected and awaited second meeting between Donald Trump and Kim Jong Un abruptly ended without any agreement and will go down in history as the 2019 no deal-summit. Surely, the lack of an agreement is disappointing; it does not equate, however, to the end of diplomacy with North Korea. What went wrong and what needs to be done to keep the momentum of peace process in Korea?

Two Track Diplomacy vis-à-vis North Korea

  After the tension-loaded year of 2017, diplomacy on the Korean peninsula quick-started after Kim Jong Un’s New Year’s speech at the beginning of 2018. The progressive administration in Seoul had iterated offers for inter-Korean dialogue since the inauguration of Moon Jae-in in May 2017. Moon was very keen on bringing North Korea and the US to negotiating tables, and careful to set the right tone as he did in his Berlin speech July 2017. His genuine interest in dialogue with the North – important not for the sake of accelerated reunification, but in the name of joint Korean ownership of peace and stability on the peninsula – has translated into the re-establishment of crisis communication channels, the institutionalization of working-level dialogue through e.g. the liaison office in Kaesong as well as leader-to-leader meetings that resulted in the Panmunjom Declaration of April 2018, the Pyongyang Declaration and Military Agreement of September 2018. Seoul’s continuous endeavor to constructively engage Pyongyang and to find mutually acceptable steps towards political reconciliation, military confidence-building and economic cooperation has spilled into a relatively stable track of inter-Korean dialogue(s).

  US-North Korean diplomacy, however, stands in stark contrast to this: Donald Trump’s initial willingness to meet directly with Kim quickly fell into oblivion with his 2017 maximum pressure campaign and “fire and fury” rhetoric. While first inter-Korean talks and Pyongyang’s high-level attendance of the Pyongchang Winter Olympics in February 2018 were met with US discontent, nevertheless, Trump gratefully accepted Kim Jong Un’s invitation that was extended by Seoul’s special envoys in March 2018. Diplomacy between Washington and Pyongyang since then, however, appears rather as “off-again on-again” momentum of dialogue and summitry: First high-level talks between the Trump administration and Kim regime were followed by the cancellation and then resumption of preparations for their first summit meeting, finally resulting in the Singapore Summit and its consequent declaration of intentions in June 2018. Besides frictions in July, August and November, high-level meetings between Washington and Pyongyang took place in July and October. High-level meetings in the run-up to the Hanoi Summit, and especially working-level talks in January and February, were cause of high hopes for the second encounter of Donald Trump and Kim Jong Un. The ups and downs in US-North Korean interactions as well as the no deal-summit reveal the shortcomings of the US administration’s approach to Pyongyang.

The Singapore Summit in June 2018: Start of a new relationship?

  The first-ever leadership interaction between the US and North Korea supposedly marked the beginning of a fundamentally new bilateral relationship and the start of an innovative top-down approach to diplomacy. The Trump administration explicitly distances itself from previous administrations’ North Korea policies – which is correct in terms of its willingness to directly and personally engage Pyongyang’s leader(ship) without explicit preconditions. In terms of content, however, US positions remain the same and rather increasingly emphasize the threats posed to its national security by North Korea’s ICBM and uranium enrichment capabilities.

  On a rhetorical level, Donald Trump has played with ideas of withdrawing all US troops from South Korea, of opening liaison offices and declaring the end of the Korean War. The no deal-summit of Hanoi, however, bluntly revealed the actual stakes being discussed and the narrow room for maneuver: While Pyongyang insists on the partial lifting of sanctions and suggests the focus on confidence- and relationship-building measures, Washington shows no flexibility nor creativity in allowing for compromise in terms of North Korea’s nuclear weapons and ballistic missile programs. Although details of the failure to reach agreement in Hanoi remain to be seen, the Trump administration has passed up the opportunity to put Pyongyang’s nuclear testing and missile moratorium into writing, to sign a narrow deal consisting primarily of allowing inspectors into Yongbyon and to declare the end of the Korean War. These three milestones would have put negotiations on track towards building actual confidence, and towards achieving peace and disarmament realistically step-by-step.

The right lessons to be learnt from interactions with Pyongyang

  Moon’s administration is not only building on the fundaments of the Sunshine Policy (1998-2008), but rather developing it. Moon has apparently drawn the right lessons to prioritize political and military confidence-building measures in dialogue with Pyongyang. The prospect of economic cooperation remains an important element in inter-Korean relations, but is understood to not automatically spill into the desired change of North Korea’s foreign policy behavior.
The current administration in Washington, however, has not drawn any lessons from past successful and failed negotiations with Pyongyang. International and unilateral sanctions remain the preferred (and only) tool; a tool that is merely being utilized as ‘sticks’ instead of ‘carrots and sticks’. Besides insisting on the implementation of the complete sanctions regime against North Korea, the Trump administration overestimates the effectiveness of its maximum pressure campaign. Sanctions do have an effect on North Korea’s economy and population; the abrupt changes in its foreign policy behavior and willingness to engage in talks, however, derive from Pyongyang’s own strategic developments and Moon Jae-in’s genuine interest in resuming inter-Korean dialogue as well as facilitating US-North Korean diplomatic interactions.

  As a general lesson to be drawn from past negotiations, willingness to compromise and creative persuasion are more likely to result in constructive dialogue and agreements, than mere coercion. Moreover, a declared end of the Korean War would not only represent the ultimate security guarantee of respecting mutual state sovereignty and the beginning of actual non-adversarial relations, but also alleviate the entire peninsula’s population from the seven decades-long state (and threat) of war, and put a dent in the military’s power within North Korea.

Outlook and Recommendations

  North Korea and the US remain (at least rhetorically) interested in dialogue; both of their leaders insist on having and wanting to extend their positive personal relationship. It is imperative for Seoul to double its efforts and resume its role of facilitator and mediator, continuing the positive momentum of inter-Korean relations and helping to bridge the gaps between Washington and Pyongyang. After having practically agreed to terminate the Korean War in September 2018, the two Koreas must commerce economic cooperation and expand exchanges in all areas to grow together. Extensive inter-Korean relations can encourage US-North Korean relations and help to reduce political and military tensions on the Korean peninsula.

  Independent from the no-deal of Hanoi, there is an urgent need to follow-through on the de facto inter-Korean end of war and declare the end of the Korean War multilaterally. Instead of coercing North Korea into unilateral disarmament, a more realistic step-by-step approach needs to be adopted; a comprehensive approach that conceptualizes steps of a freeze, capacity reductions and then dismantlement, steps reciprocated through selective sanctions lifting. Additionally, military confidence-building measures and institutionalizing bilateral interactions serve to mitigate present and future tensions.

  Most importantly, it is imperative to understand the roots of North Korea’s conviction regarding its possession of an indigenous nuclear deterrent. While the US consideration of nuclear weapons during the Korean War sparked this conviction, Washington’s handling of Iraq and Libya as well as its nuclear-capable strategic assets stationed in Guam, bomber overflights as demonstrations of force and decapitation plans have intensified it. Without substantial changes in the security environment on the Korean Peninsula therefore, there will be little substantial changes in Pyongyang’s nuclear weapons or ballistic missile programs. It is therefore laudable to continue the indefinite suspension of US-South Korean joint military exercises, which was mostly expected after the Singapore Summit in June 2018. The absence of large-scale military drills serves to build the fundament of normalizing state interactions and constructive dialogue towards a negotiated solution to the deep-rooted security dilemma on the Korean Peninsula.


1 This paper was commissioned by RECNA on behalf of Co-chairs of Panel on Peace and Security of Northeast Asia. The views and opinions expressed here do not necessarily reflect those of PSNA.

【Mar. 8, 2019】Special Essay on US-DPRK Summit in Hanoi (PSNA-WP-3)

Category : PSNA Activities

Kim–Trump summitry: Neither breakthrough nor breakdown

Ramesh Thakur, Emeritus Professor, Crawford School of Public Policy, Australian National University and Co-Convenor, Asia-Pacific Leadership Network for Nuclear Non-Proliferation and Disarmament

PSNA Working Paper Series (PSNA-WP-3)1

March 6, 2019

[Japanese version: PDF]

  CANBERRA— The first summit in Singapore last June gave Kim Jong-un legitimacy as the head of a de facto nuclear-armed state engaging with U.S. President Donald Trump as an equal. The second summit in Hanoi on February 27–28 has normalized that status but accomplished little else.

  Trump had successfully managed expectations downwards since last year. Gone were the boasts about the nuclear threat from Pyongyang having ended. Instead Trump shifted the “transformational goal” of total denuclearization to the “transactional goal” of limiting Kim’s nuclear capability. The U.S. walked back from the insistence on total, verified denuclearization as a precondition for improved ties and normalization. Instead it has embraced the principle of simultaneous and parallel steps toward denuclearization and peaceful relations.

  The Hanoi summit offered neither a breakthrough nor a breakdown. White House press secretary Sarah Sanders said that although “No agreement was reached at this time,” the two leaders had “very good and constructive meetings” and “discussed various ways to advance denuclearization.” Trump said the impasse arose over Kim’s demand for a lifting of sanctions in their entirety in return for a promise to dismantle the Yongbyon nuclear complex. North Korea’s Foreign Minister Ri Yong Ho disputes this account, saying it only asked for limited sanctions relief.

  The Kashmir crisis that flared up dangerously with exquisite timing during the Hanoi summit underlines the logic of Trump’s moves on Korea. Abandoning a demonstrably failed policy over a quarter century, of insistence on a total and irreversible denuclearization of North Korea, is no big sacrifice. Engaging with Kim personally to establish a working relationship that can dispel misperceptions, build confidence and trust, deepen inter-Korean relations and in other ways greatly reduce the risks of a war with catastrophic consequences: now that is a big deal.

  For all his strategic illiteracy, Trump may have a surer intuitive grasp of this underlying big-picture reality than most of the devotees of the Washington playbook of increasingly militarized responses to foreign crises. Communications channels are now active between North and South Korea, and between North Korea and the U.S., at summit, high and working levels. This is no bad thing.

  That said, Trump was right to walk away from the demand to lift all sanctions now in return for dismantling just one key nuclear facility. Perhaps Kim misjudged Trump’s eagerness to make a deal, any deal, in order to claim a victory to offset the worsening domestic situation for the president, particularly with his former lawyer’s testimony to Congress. Trump has not reached that point of desperation yet.


1 This paper was commissioned by RECNA on behalf of Co-chairs of Panel on Peace and Security of Northeast Asia. The views and opinions expressed here do not necessarily reflect those of PSNA.

【Feb. 26, 2019】Special Essay on upcoming US-DPRK Summit in Hanoi (PSNA-WP-2)

Category : PSNA Activities


Leon V. Sigal, Director of Northeast Asia Cooperative Security Project, Social Science Research Council (USA)

PSNA Working Paper Series (PSNA-WP-2)1

February 26, 2019

[Japanese version: PDF]

      Concrete commitments, not the hyperbole of President Trump’s defenders or detractors, will determine how successful the Trump-Kim summit in Hanoi should be judged by objective observers.

      From the U.S. vantage point, four commitments matter. First, will North Korea completely halt the production of plutonium and enriched uranium and commit to the dismantlement of all its fissile-material production sites, not just those at Yongbyon? Second, will it stop making intermediate- and intercontinental-range ballistic missiles? Third, will it permit inspections at its reactor and reprocessing plant at Yongbyon and all its enrichment sites, as well as its nuclear test sites, uranium mines and sites where the uranium ore is refined and turned into a gas for enriching in order to bound uncertainty over how much fissile material it may have produced and used in the past? And fourth, has it committed in writing to its moratorium on nuclear and missile testing?

      From the DPRK vantage point, will the United States move away from enmity by declaring an end to the Korean War, opening the way to a peace process that can culminate in a peace treaty? Second, will the United States relax some sanctions by exempting the delivery of humanitarian aid, use of the Mount Kumgang resort and reopening of the Kaesong Industrial Complex in North Korea, as well as lifting U.S. Trading with the Enemy Act sanctions? Third, will the two sides be willing to open liaison offices in each other’s capitals?

      Critics will claim that the Yongbyon facilities are old, as if shutting them down is not worth much. That is nonsense. Those “old” facilities, some of which have been operating for less than a decade, can produce three or four bombs’ worth of plutonium and highly enriched uranium a year, as well as the tritium without which the North’s thermonuclear weapons will no longer function after some dozen years or so.

      Critics will also object that the North still retains an unknown quantity of fissile material and nuclear weapons and the summit did not yield a complete declaration of the North’s nuclear inventory including how much it has made. But the Trump administration is right to phase in that inventory declaration, starting with the location of its plutonium reactors, reprocessing and enrichment sites. Before seeking an accounting of fissile material and number of weapons, it is prudent to seek access to these locations as well as the North’s nuclear-weapons test sites, its uranium mines, its ore refining plants, and its uranium hexafluoride plant to take various measurements. This nuclear archeology will reduce uncertainty and better enable it to assess how much fissile material the North could have produced. U.S. intelligence estimates vary widely so any number the North would turn over is certain to be controversial, as it was in the initial declaration to the IAEA in 1992, which is now nearly forgotten but for years complicated efforts to contain the growing security threat posed by North Korea’s continued fissile material and missile production.

      It is essential to understand that verification is a political judgment in technical guise. Verification is sometimes confused with playing “gotcha,” seizing on a suspected breach— however minor—as evidence of cheating and using it to discredit a deal. While no agreement can be absolutely verifiable and any breach takes on political significance because of what it implies about a violator’s intention to some, to say that an agreement is adequately verifiable is to assert that residual uncertainties are less consequential than the benefits of keeping the agreement. Absolutism in verification may pose as great a risk to U.S. and allied security as some North Korean violations.

      Getting most, if not all, of the above commitments would be a remarkable achievement. Implementing them will take the two sides further down the road to denuclearization than they have ever gone before. Critics will no doubt carp that such an outcome stops short of complete denuclearization and question whether Kim Jong Il will ever give up his nuclear weapons, but the only way to find out is to continue the negotiations while keeping U.S. commitments, and see how far they can get.


1 This paper was commissioned by RECNA on behalf of Co-chairs of Panel on Peace and Security of Northeast Asia. The views and opinions expressed here do not necessarily reflect those of PSNA.

【Feb. 26, 2019】Special Essay on upcoming US-DPRK Summit in Hanoi (PSNA-WP-1)

Category : PSNA Activities

The North Korea–United States summit and possibilities for new security-oriented thinking

Kyoji Yanagisawa, Former Assistant Chief Cabinet Secretary, Japan

PSNA Working Paper Series (PSNA-WP-1)1

February 26, 2019

[Japanese version: PDF]


      Since the North Korea–United States summit took place in June 2018, the United States and China have been in discussions regarding North Korean denuclearization. To break through the current stalemate in these discussions, a second summit meeting is anticipated in late February. Whatever results these talks may produce, the fact that nuclear disarmament is a topic signifies a new age in the history of nuclear weapons.

      We have been accepting the existence of nuclear arms as “a necessary reality” to prevent wars, while maintaining policies on how to avoid using them. Despite taking a long time to achieve the denuclearization of North Korea, if we find that it is possible to eliminate a nuclear program through negotiation, it would demonstrate that an approach other than deterrence could be successful. As a real-life case study, the North Korea–United States talks offer an opportunity to consider new possibilities and create experimental models for these.

      Speaking as a former defense establishment insider during the age of mutual nuclear deterrence, I find this degree of change rather shocking, and this has prompted me to try to contribute to a better understanding of this new era by identifying some of the tools to do so. No one can be certain where the North Korea–United States negotiations are heading, but what is certain is that change is coming. We are at the end of an era and the beginning of a new one.

Why is the Japanese Prime Minister unable to respond to the current situation of the atomic bomb victims?

      On August 9, 2018, 73 years after an atomic bomb was dropped on Nagasaki, Prime Minister Shinzo Abe, visited Nagasaki City and met with the victims. A representative of the victims asked why the Prime Minister did not mention the Treaty on the Prohibition of Nuclear Weapons in his speech. In response, the Prime Minister didn’t answer the question but reiterated the Government’s position that countries have diverging views on nuclear weapons, and it is important [for Japan] to act as a bridge between these.

      The Japanese government’s position is that as long as nuclear weapons exist and pose a threat to national security, Japan needs the nuclear umbrella of the United States. If Japan supports a treaty that considers nuclear weapons unlawful, Japan’s security policy could no longer be based on the nuclear deterrence capacity of the United States.

      The government believes that the U.S. nuclear weapons that once nearly destroyed our nation are “a necessary evil.” Indeed, there exists no weapon greater than a nuclear bomb. Therefore, there is a good reason to believe that the existence of a nuclear program is necessary to prevent the threat posed to us all. However, this line of thinking does not stop at a passive acceptance of nuclear programs as an unavoidable reality, but actively justifies their existence as part of our national security policy.

      During the Obama presidency, the Japanese government opposed the No First Use declaration by his administration. They believe that the ability to retaliate against nuclear weapons is not sufficient in and of itself; rather, there must be room to use nuclear weapons when attacked otherwise war cannot be prevented. However, if the use of nuclear weapons is not limited to retaliation against a nuclear attack, no threshold restricting the use of nuclear weapons remains.

      That is, indeed, the logic of nuclear deterrence theory. In response to an attack, one retaliates with the ultimate weapon, one that has the potential to cause annihilation. If there is always a risk that attacking Japan could provoke such a response, who would attack Japan? This is a convenient logic if we are only concerned about our own well-being. However, the problem with this logic is when it is applied to other countries as well. For the United States, using nuclear weapons to prevent an attack on its ally, Japan, is a difficult decision because the US itself could in turn suffer major damage. This is because the only countries with the ability to attack Japan on a scale that would require a nuclear response would be major military powers such as China and Russia, not North Korea.

      If Japan is under the U.S. nuclear umbrella, it follows that when Japan is attacked, the United States will use nuclear weapons in response. There is great irony in the only nation that suffered atomic bombings, Japan, maintaining national security assumptions of the use of nuclear weapons by its major ally. The oath “this mistake shall not be repeated” on the Memorial Monument in Hiroshima, the City of Peace, forces us to wonder if “mistake” does not include the use of nuclear weapons against enemies of Japan.

      The North Korea–United States summit held last June had no impact on the attitude of the Japanese government. The “Outline of the National Defense Program” that was adopted by the government last December states, “the extended deterrence of the United States, including nuclear deterrent force, is essential against potential nuclear threats.” Regarding North Korea, “Although it declared its intention to denuclearize, its nuclear and missile programs have not changed,” and as such, North Korea “remains a serious and imminent threat to Japan.”

      In any case, a path toward the total abolition of nuclear weapons will never be found under assumptions that nuclear weapons are necessary. It is easy to criticize the government’s attitude, but what the government really needs is clear logic that will prevent war, and if no other assurances than nuclear weapons are found, the government can hardly be expected to change its thinking.

      If we are to “act as a bridge between countries with diverging views” as the Prime Minister suggested, we must repeatedly examine the meaning of changes in North Korea, the extent to which the idea that nuclear weapons will prevent war holds water, and the potential of any other prospective approaches. Here if anywhere is a rift in viewpoints that requires bridging.

The historical significance of the North Korea–United States summit
From pressure to conversation

      In 2017, North Korea pursued a policy announced by Kim Jong-un at the beginning of that year that allowed for repeated nuclear tests and firing of missiles. In response, the United States stepped up economic sanctions and applied military pressure on the country by deploying multiple aircraft carriers and bombers in South Korea and around Japan. However, North Korea did not implement any changes in its nuclear program, leading to a stalemate. As more pressure was exerted, the risk of accidental war increased, and the concerns of nearby countries, including Japan, were heightened.

      In 2018, North Korea declared that its nuclear arms development was complete and changed the focus of its policies to economic development. North Korea explored the idea of discussions with South Korea in conjunction with the 2018 Winter Olympic Games in Pyeongchang. Moreover, during the April 2018 Inter-Korean summit, the two countries jointly declared denuclearization of the Korean peninsula and an end to the state of war. Kim Jong-un requested a meeting with the United States through South Korea, and President Trump met with him at the North Korea–United States summit on June 12 in Singapore. They agreed on security guarantees by the United States in exchange for the denuclearization of North Korea.

      Subsequently, North Korea and the United States have been bargaining intermittently in an effort to implement the agreement reached in principle at the summit, but it remains unclear how the USA, which wants to verify the status of North Korea’s nuclear weapons and facilities, and North Korea, which seeks to alleviate economic sanctions, can find a middle ground. Clear steps toward officially ending the Korean War are yet to be made, which would represent an important form of progress toward security guarantees.

      The USA and Japan suspect that North Korea has no intention to abolish its nuclear program but is simply attempting to get sanctions lifted by compromising with the Trump administration. However, the logic behind international negotiations is not that simple. The parties came to an agreement because benefits were to be gained from doing so, and to protect such benefits, they are obliged to comply with the agreement.

      If getting the sanctions lifted is North Korea’s primary aim, and their driving concern is therefore economic in nature, this implies that more diverse and flexible negotiations could be more successful than relying strictly on military means.

      The outlook for negotiations is not necessarily grim as long as there is a will to compromise.

Leading with force and benefits to achieve goals

      War is the use of violence by a nation to exert its will. The root of the problem with North Korea is the conflict between North Korea’s desire to retain its nuclear program and the USA’s unwillingness to accept it. To change North Korea’s position, the USA has made threats in the form of sanctions and military power.

      The intent behind sanctions was to cause enough economic loss that the North Korean economy would collapse and the regime would lose its power. These changes would be brought to North Korea by inciting fear that the regime itself would be overthrown through war. However, North Korea never changed its stance.

      The benefit for North Korea of retaining its nuclear program is enhanced security against a U.S. attack. Therefore, as the conflict intensified, the benefits related to having nuclear weapons also increased. On the other hand, the main benefits for the USA are to remove the threat to its own safety posed by North Korea’s offensive capability and to increase its legitimacy as the nation that maintains order by achieving the abolition of other countries’ nuclear weapon programs and preventing further complication of power relationships in the region. For these reasons, increased North Korean nuclear capacity is disadvantageous to the USA.

      The USA defined Iran, Iraq, and North Korea in 2002 as “the axis of evil” and declared a policy aimed at altering the situation, potentially through a preemptive attack. The axis of evil referred to “countries with weapons of mass destruction, or that wish to acquire weapons of mass destruction in defiance of the USA.” The USA accomplished regime change in Iraq but failed to build a stable postwar order there. In North Korea, the USA has attempted denuclearization by force and made threats that hinted at direct regime change, such as through the murder of Kim Jong-un. However, when ultimately the DPRK refused to change its position, the USA did not feel that openly declaring war was viable.

      The U.S. military probably has enough power to destroy the Kim Jong-un regime. However, a war to overthrow a regime because it is developing a nuclear program, when that country has not declared war against another nation, cannot be justified under international law. Should a war ensue, South Korea (which is within range of North Korean long-range artillery) and Japan (within range of intermediate-range missiles from North Korea) would almost certainly sustain damage. Furthermore, North Korea would fall into a state of anarchy if its current regime were overthrown. To establish a new order out of such confusion, North Korea would have to be occupied for a long period by about a million soldiers. In view of the high costs and limited benefits, war was simply not an option for the USA. In the event that the North Korea–United States negotiations collapsed while these conditions remained unchanged, war would still not be a viable option.

      If war is not an option, the threat of military force is ineffective as a means of forcing North Korea to bend. If anything, it unites North Koreans and strengthens their will to keep their nuclear program. This situation led to the 2017 stalemate.

      The ultimate goal for the USA is to change North Korea’s desire to keep its nuclear program. There are two ways to change someone’s will: forcibly or by offering incentives. If the threat of war is ineffective in changing North Korea’s position, the only option left is to entice the regime with benefits. What North Korea wants more than anything is a guarantee that the USA will permit its regime to exist, in exchange for which it would not attempt to destroy the USA. Similarly, what the USA wants is for North Korea to abandon its nuclear program, in exchange for which the USA would not seek to overthrow the North Korean regime.

      The agreement reached at the North Korea–United States summit constitutes a promise to offer these concessions to each other. Having both countries grant what the other side desires without the use of force is the surest and most effective method of avoiding a war. If countries can obtain what they want without a war, they would never choose war, which has very high costs and no assurance of a favorable result.

Both cannot return to the past

      Of course, the best scenario would be for North Korea and the USA to achieve their mutual goals through negotiation. As long as they perceive an opportunity to negotiate, there is no reason for either country to take a course of action that would jeopardize such an opportunity. With the June 12 North Korea–United States agreement as a good point of departure, it is imperative that these nations move toward resolution of the issue.

      For North Korea, nuclear abandonment equals military disarmament. This is because North Korea does not have the ability to threaten USA’s safety with conventional military power alone. This situation has made North Korea cautious. North Korea cannot ignore the risks related to the potential identification by the USA of the locations of its nuclear weapons; that is, the U.S. military would immediately intervene to destroy them. Thus, the next challenge for the USA is how it can gain its counterpart’s trust in its assurances that it would not seek to overthrow the regime.

      The origin of the North Korea–United States conflict dates back to the 1950 Korean War. Despite a cease-fire being declared under the 1953 Korean Armistice Agreement, the countries are still technically still at war. It is a natural course of action for countries at war to pursue military superiority. To build trust between North Korea and the USA, the Korean War must be officially ended. Once the conflict that persists due to the ongoing state of war is eliminated, the fear of an attack by the USA should also be eliminated, and there would no longer be any reason to maintain nuclear weapons.

      The achievement of these ends, however, requires the involvement of the major parties to the Korean War and the establishment of the postwar order on the Korean peninsula. This includes not only the USA but also China. Such a multilateral agreement is vital for ending the war, and it would mark the start of a postwar framework for new talks in northeast Asia including South Korea, North Korea, the USA, China, Japan, and Russia. Here again, formally ending the Korean War is a prerequisite for achieving denuclearization. Only after a multinational framework is established to handle regional security issues in northeast Asia can the denuclearization of North Korea be finalized.

      Meanwhile the North Korean regime is negotiating for guarantees. It is deeply committed to improving its economy through increased access to investment opportunities brought by the lifting of sanctions. Since North Korea has no bargaining chips other than its nuclear weapons in its negotiations with the USA, the road to the country abandoning its nuclear weapons is a long one. Its two motivations, achieving a guarantee of stability of its regime and economic improvement, are closely interrelated.

      This does not mean that the situation will necessarily progress at the pace set by North Korea. The more pressing its economic needs become, the more flexible with its incentives the USA can be. For this reason, North Korea is attempting to push the USA toward compromise by involving China.

      By the time this paper is published, the second North Korea–United States summit, scheduled for late February, will already have taken place. Though steps for North Korea to take toward denuclearization and for the USA to lift its nuclear weapons-related sanctions are to be discussed, things might not turn out as expected. At this point, even forecasting the results of the negotiations is not feasible; the most we can do is discuss the logical path toward resolution of the issues, nothing more, and nothing less.

      In any case, what is important to recognize is that it is possible to solve the North Korean nuclear weapons issue through incentives rather than force. There is a path to solving problems without war. The question is how we can generalize, customize, and systematize this potential.

From deterrence to incentives: the possibility of a new mind-set

      The core incentive for Kim Jong-un is to ensure the stability of his regime. He has no prospect of winning a military resistance against the USA. Being sanctioned for opposing the USA will weaken his regime, and if the conflict leads to war, his regime will collapse. Beyond the military situation, the second key to a stable North Korea is a sounder economy.

      To rebuild the North Korean economy, economic sanctions must be lifted. North Korea appears to be attempting to outsmart the sanctions, but in a nation of 20 million people, smuggling and backroom deals cannot provide economic growth. To grow the economy, economic participation in the international market is essential; this means that North Korea must abandon its nuclear program, which is the cause of the sanctions. Kim Jong-un must be aware of this. At the same time, for it to start to feel confident about the discontinuance of its nuclear program, North Korea must first remove the risk of military conflict with the USA.

      If Kim Jong-un axes his strategy based on incentives such as these, the increased availability of incentives presents an opportunity. Of course, this is only a hypothesis that will be put to the test in the actual negotiation process.

      What makes this process important is that both North Korea and the USA have transitioned from conflict and deterrence, backed by their ability to attack each other, to negotiations based on offering each other incentives. In other words, instead of seeking to force each other to bend through threats or actual war, they are opting for persuading each other through incentives. Herein lies potential to drastically change both parties’ perceptions and approaches to attaining assurances of security.

      The main framework for security assurance has changed over time. In the 19th century, the primary approach was to solve international conflicts through war. In the 20th century, the mainstream approach was to use force as a deterrent to prevent war. What we are experiencing today could be the first step toward a new approach of solving problems through mutual compromise based on incentives.

      To effectively transition from deterrence to incentives, I would like to examine the future of the North Korea–United States agreement, which I believe contains the greatest potential significance for North Korea–United States reconciliation efforts.

Can we prevent missiles with deterrence?

      In his address to the National Diet on February 14, 2018, Prime Minister Abe stated, “When North Korea launches a missile, the only nation protecting Japan is the United States. If they miss intercepting the missiles, the only country that can retaliate is the United States. Unless [North Korea] takes this to heart, it may take reckless action going forward.” Abe’s statement is a summary of the concept of deterrence, in which the intention to attack is countered by threat of even more powerful retaliation.

      For this logic to be borne out in facts, three prerequisites must be met. The first is that the United States would indeed retaliate, ignoring considerations of its own benefits and risks; the second is that North Korea would hold back from launching missiles due to fear of retaliation by the United States. Since the “if they miss” scenario implies missiles potentially landing in Japan, a third unstated possibility is that Japan could withstand such a missile attack and request retaliation by the United States, including the use of nuclear weapons.

      None of these assumptions is certain. If North Korea were to begin a war that involved the use of missiles, it would be under a definite state of confrontation with the USA and would be willing to target the USA (or perhaps its ally Japan) with nuclear weapons and ICBMs. The following questions arise: Would the USA be prepared to sacrifice itself to retaliate against North Korea to avenge Japan? War sometimes begins out of fear, and North Korea might consider that preemption is the only way for them to survive, just as Japan Imperialists did in 1941.

      Furthermore, would Japan be able to withstand missile attacks on Tokyo or its nuclear plants? The logic of deterrence with force may appear solid, but it presumes the power to prevail in an armed conflict and, in war, one must be ready to withstand damage.

      In the end, PM Abe’s “deterrence through the USA’s retaliation” on Japan’s behalf is a weak proposition. Personally, I would give each of the three assumptions a 50 percent rating for reliability. When the three probabilities are multiplied, we get 0.5 × 0.5 × 0.5 = 12.5%. Can a security assurance policy with a 12.5 percent likelihood of success survive? This is why an alternative to the idea of deterrence is necessary.

      The most certain way to stay safe from missiles is not to “strike down a missile and retaliate,” but to eliminate the willingness to use missiles. Making military threats ends up increasing the ability and willingness of the counterpart to strike. If a nation cannot strike back in response to a missile attack, the most realistic defense is to eliminate the willingness to use missiles.

      The North Korea–United States negotiations seek to ensure North Korea would not launch its missiles under any circumstances before the actual abolishment of its nuclear weapons program takes place. What is necessary going forward are continued efforts in this direction and monitoring to ensure it is final and verifiably irreversible.

Is deterrence with an unusable weapon logical?
Why are nuclear weapons unusable?

      War is an act of confronting another country with violence to achieve one’s political goals. Deterrence involves discouraging violence by demonstrating the ability and resolve to use even more violence to suppress other actors’ willingness to go to war. To achieve realistic deterrence, one party requires a means of violence and must convince the other party that they are determined to actually use such violence.

      In other words, nuclear weapons’ function as a deterrent is based on the thought process by which the other nation recognizes the possibility of actual use of nuclear weapons, which prevents it from attacking. However, a nation that fears nuclear weapons must plan for the possibility of an attack. Nuclear weapons have the power to destroy not only opposing military forces but also whole cities. If troops who are seeking to force another country to bend to their will are destroyed, the enemy might retaliate by destroying cities, thereby seeking to destroy the core of its opponent’s resolve, the nation itself.

      If a war goes beyond its initial purpose of forcing another country to bend to its will in this way and ends up seeking the annihilation of the nation itself, the war becomes suicidal and loses its significance as a means to achieve a political goal. During the Cold War, the common understanding of “mutually assured destruction,” which signified recognition that the USA and USSR could both be destroyed by nuclear weapons if they engaged in a blind pursuit of victory, kept them from actual war.

      The reason the Japanese government believed that extended deterrence under the US nuclear umbrella, considered the foundation of its security assurance policy, was effective is because if one of the two nuclear superpowers during the cold war—the USA or the Soviet Union—attacked the other, allies would retaliate, creating a potential risk of nuclear retaliation. This assumption resulted in the deterrence of military aggression even against allies.

      In other words, nations were prevented from going to war because their prospective combatants owned nuclear weapons. Since they could not go to war, the USA and USSR had to coexist, leading to a strategic stability that was called détente. Nuclear weapons were acknowledged to have an ability to deter war simply by their existence, and their existence was justified because they had not yet been used.

      However, this relationship is based on the assumption that both countries are still willing to use nuclear weapons if deemed necessary.

      The East-West conflict during the Cold War did not involve mutual economic interdependence, and there was no room to budge due to the opposing ideologies characterizing the two sides. In this conflict structure, in which each side firmly rejected the other’s ideology, the perceived superiority of one side would cause fear in the other. The three causes of war as defined by Thucydides—wealth, honor, and fear—coexisted in this conflict. War could happen at any time, and once a war was launched, there would be nothing to stop it from spreading. If it did not stop spreading, it could escalate to the use of nuclear weapons. A country under attack may have retaliated with nuclear weapons, resulting in the risk of its own annihilation. In this manner, nuclear weapons became unusable because both countries’ willingness to use them was recognized. This is the essential concept of nuclear deterrence through mutually assured destruction, as well as of extended deterrence under the nuclear umbrella.

Differences in motivations for nuclear weapon use between the Cold War and the present

      The conflict between liberalism and communism or totalitarianism during the Cold War was a conflict between modes of existence. Both sides had strong beliefs—“I would rather die than lose my freedom under totalitarianism” and “we cannot allow ourselves to be exploited by capitalism”—that justified each side’s regime. This was a conflict with no compromise, and it was considered worth pursuing even at the risk of being destroyed by nuclear weapons. In other words, political motivations that created the willingness to use nuclear weapons existed on both sides.

      Today, no such absolute ideological conflict exists between any two nations. The world has become one market controlled by economic competition. Enormous amounts of capital move across national borders, and each country’s economy is dependent on that of other countries. Destroying another nation through war is now the equivalent of destroying part of one’s own economic foundation.

      Dictatorial political systems and their domestic restrictions on human rights may sometimes become a political issue, but this is a problem within a single nation, not a challenge to global governance as the communist world revolution was. Presently, no threat exists that must be prevented to avert a nation’s downfall. There is no “life-or-death” conflict that could lead to irrational decisions.

      Today, there is concern over a possible war between the USA and China over supremacy according to the “Thucydides trap.” Competition over trade and advanced technology continues to intensify. However, the true nature of this conflict is over economic and technological supremacy, and the fear that the USA experiences as it contemplates the possible loss of its economic supremacy is the exacerbating factor. This problem cannot be solved by either forcing the other side to surrender by war or by destroying the other in war.

      The current concern, therefore, is not the physical destruction of the world due to nuclear retaliation by the USA or China, but rather “mutually assured economic destruction” in which the global economy becomes dysfunctional.

      The logic of mutually assured destruction appears to persist between major nuclear powers. However, the big differences between the Cold War and today’s situation is the lack of motive to resolve conflicts through war and the unwillingness to destroy one’s opponent with nuclear weapons at the risk of one’s own destruction, even if one has to put up with some antagonism or harassment as a result.

      Without a motive, there is no will to use nuclear weapons. If this is the case, does the argument that nuclear weapons without the willingness to use them effectively represent a situation of deterrence still hold water? Even if there is no longer any motive for war between major states with nuclear weapons, some wars between a major nuclear state on one side and a medium-size or small state without nuclear weapons on the other—e.g., the annexation of Crimea by Russia, the capturing of islands in the South China Sea by China, and the Iraq war waged by the USA—were not prevented.

      Could these wars have been prevented if these small nations had nuclear weapons? Perhaps. This is exactly the motivation underlying North Korea’s drive to keep its nuclear weapons program. However, even if these lesser nations had armed themselves with nuclear weapons, they would not have reached parity with major states that have larger nuclear arsenals, and they would have had no hope of winning a war. Acquiring nuclear weapons may even increase the risk of being attacked by major states with nuclear weapons. The actual problem was that these nations were not under the nuclear umbrella of major states with nuclear weapons, and thus the attackers had no concern about provoking a war with another major state that possessed nuclear weapons. In other words, there was no motive for major states to use their nuclear weapons, and thus the existence of nuclear weapons did not prevent war in these cases.

      What about the allies? If medium-sized and small nations are under the nuclear umbrella of major states with nuclear weapons, could they avoid being attacked by other major states that have nuclear weapons? This logic seems to be moderately persuasive. However, today, as the logic of mutually assured destruction between major states with nuclear weapons is wavering—in other words, now that it is unclear whether major nuclear states are willing to go to war with another state that possesses nuclear weapons even at the risk of their own demise—the root question of whether the nuclear umbrella is secure cannot be answered.

      Crimea, the South China Sea, and the Iraq war were all major incidents that affected the international order, yet major states with nuclear weapons showed no will to go to war in these cases. If a nation is closely allied with a major power, an attack on that nation might have a greater chance of leading to war, but whether it should be protected with nuclear weapons is a different question.

      At present, the main factor that deters war between major nuclear states is not the existence of nuclear weapons, but the lack of perception that going to war would resolve any problem, compounded by the realization that the use of nuclear weapons would lead to the loss of an opportunity to become involved in negotiating incentives between major states. In our contemporary situation, no clear strategies exist for offering or negotiating incentives as power continually shifts between major states. Therefore, problems cannot be solved through the short-term destruction of the balance of power that we call war; that is the nature of the times we are living in.

      In any case, the logic that the existence of nuclear weapons prevents a war between states is useless in solving the problems of this day and age.

      On the other hand, armed groups called terrorists are supported by ideologies that deny other people’s very right to existence, and they are willing to pursue their ideologies even if it means their own deaths. Their objective is murder, not rule. That is why they are called terrorists. Since they do not have a state, land, or people to rule, they have no state or culture to worry about losing through nuclear attacks. Therefore, it is difficult to stop terrorism with military force.

      Today, the great global risk is the possibility of nuclear weapons, which are no longer used as a tool of war between states, falling into the hands of terrorists, who might then use them as a means of murder, not deterrence.

      This is why the management of nuclear weapons by states, as well as of nuclear technology and of the transport and disposal of related materials, has come under criticism. Terrorist groups, which are not states, would never ratify the Treaty on the Prohibition of Nuclear Weapons. Hypothetically, even if a world without nuclear weapons were achieved in the future, the threat of terrorists developing and using nuclear weapons would remain an issue for international society. Conflict among human beings, not just among states, must be resolved if we wish to eliminate all forms of violence and destruction. To that end, nuclear weapons and military power are not the solution.

      A new approach that moves away from deterrence by nuclear weapons is thus needed.

The U.S. nuclear strategy in confusion and the security of Japan
The U.S. objective of enhanced nuclear capacity

      The Trump administration did not rule out the preemptive use of nuclear weapons during its Nuclear Posture Review of February 2018, which stated that the United States aims to develop “usable nuclear weapons” with limited objectives, such as low-yield nuclear weapons and cruise missiles. In October, the USA declared its intention to withdraw from the Intermediate-range Nuclear Forces Treaty (INF treaty), and China, which has not signed the treaty, announced its intention to develop new intermediate-range nuclear weapons to counter intermediate-range missiles, as a means of countering U.S. military actions in Asia. In January 2019, in its Missile Defense Review, the USA announced plans to develop new technology that could shoot down all missiles aimed at it, fully utilizing sensors in space and all means of interception.

      This represents an attempt to devise nuclear options applicable to all war-related purposes, from the destruction of limited military targets to the annihilation of an enemy nation, and to gain absolute superiority to win all wars involving nuclear weapons by making all other countries powerless to mount effective attacks. Considering that the true nature of deterrence is the power to win a war, such steps would enhance the deterrence capacity of the USA.

      Of course, there remains the dilemma of security assurance. Actors who do not want to see the USA gain unquestioned superiority may act to further strengthen their own nuclear military power, resulting in the start of a new cycle of the nuclear arms race.

      During the Cold War, mutual deterrence based on nuclear weapons was grounded on the assumption of mutual vulnerability, that is, the belief that neither combatant could survive a nuclear attack. If either party is assured of its ability to avoid damage and achieve victory, mutual deterrence falls apart. This would no longer be deterrence but a unidirectional coercion based on the threat of violence, and if it were ever actually put into practice, it would constitute massacre without justification.

      USA’s opinion is that if the only nuclear weapon that the USA owned was large long-range missiles launched from its mainland, while the enemy possessed smaller nuclear heads along with intermediate-range missiles with the capability of forward deployment against the USA, the U.S. nuclear weapon would be too large to retaliate against limited nuclear attacks by their enemy, thereby making it unusable. In this case, the USA would lack any way of retaliating against limited nuclear attacks.

      From Japan’s standpoint, if the USA has diverse military options including intermediate-range nuclear weapons, its deterrence capacity would increase. However, there is a serious problem with this line of thinking.

Negative impact on Japan’s safety

      First, the USA’s aim is to have usable nuclear weapons. Low-yield nuclear weapons with limited power would be deployed against conventional military weapons of the enemy. If nuclear weapons are used against conventional weapons, the generally accepted threshold for the use of nuclear weapons would disappear.

      Second, low-yield nuclear weapons must be available at the front lines in a usable form. Equipping the front lines with powerful military force is effective as a deterrent in a crisis, but it could also strike fear in enemies’ hearts since it implies an ongoing state of crisis, possibly leading to a preemptive attack.

      Third, if low-yield nuclear weapons are used in the battlefield, retaliation with intermediate-range nuclear weapons from behind the front lines can be expected. Japan, where the U.S. military’s command and supply functions for northeast Asia are centered, would be an ideal target for intermediate-range nuclear weapons.

      Fourth, intermediate-range nuclear weapons with a range of 5,000 km could reach Beijing if launched by U.S. military forces stationed in the western Pacific Ocean. On the other hand, intermediate-range nuclear weapons owned by China could reach Tokyo but not the USA. What this means is that the USA could enter a nuclear war with intermediate-range missiles without fearing an attack on its own soil. Therefore, there is a major difference between Japan and the USA in the degree of threat from intermediate-range nuclear weapons, and this is leading to a new age of “decoupling.”

      The 1987 INF treaty was concluded because NATO nations were concerned about such decoupling. In that instance, Soviet intermediate-range nuclear missiles could not reach the USA but did threaten Europe, rendering the USA retaliation strategy dysfunctional. Accordingly, the USA purposely equipped Europe with intermediate-range nuclear weapons, which was followed by the USA and USSR abolishing these weapons simultaneously.

      Along the same line of thinking, Japan could actively house U.S. intermediate-range nuclear weapons on its own soil while simultaneously demanding that China eliminate its intermediate-range nuclear weapons. However, present-day China is not like USSR of the late 1980s, which was at risk of imminent collapse, and it could withstand an arms race. From the perspective of nuclear balance, China will demand a balance in ICBMs between Beijing and Washington, and it will want submarine-launched missiles in the western Pacific Ocean as a second force to counter the USA. The USA cannot do much about this complex equation, and Japan is not likely to gamble on its own security.

What Japan should consider

      What Japan needs to consider now is whether nuclear weapons or conventional weapons are to be used, as the battlefield for the USA will be on Japanese territory. Despite the tight military connections between Japan and the USA, the two countries would not suffer commensurate damage in such a war.

      After North Korea conducted a missile experiment in February 2017, in the joint US-Japanese press release, President Trump stated, “we stand behind Japan.” The response of Prime Minister Abe to the Diet also stated, “if they miss, the USA will retaliate.” He did not rule out that if the USA entered a war, Japan would suffer the damage.

      The foundations of security assurance policy are now being questioned. Should we accept this situation as necessary for deterrence? Should our priority be to prevent missiles from hitting Japan? Should Japan continue to depend on nuclear weapons as the ultimate deterrent? Alternatively, should we aim to create a world where nuclear weapons are not used and do not need to exist?


1 This paper was commissioned by RECNA on behalf of Co-chairs of Panel on Peace and Security of Northeast Asia. The views and opinions expressed here do not necessarily reflect those of PSNA.

【Jun. 1, 2018】Statement and Recommendations of the Co-Chairs of the 3rd Panel on Peace and Security of Northeast Asia (PSNA) Workshop

Category : PSNA Activities

Moscow, May 31 – June 1st, 2018
Sponsored by the Research Center for Nuclear Weapons Abolition,
Nagasaki University (RECNA)

[PDF version]

New Prospects for Dialogue and Denuclearization

We were heartened by a return to dialogue on Korean denuclearization following the 27th April Panmunjom meeting and Declaration between the South Korean President Moon Jae-in and the North Korean Chairman, Kim Jong-un. Following the alarming exchange of nuclear threats between US and North Korean leaders earlier this year, we welcome the Panmunjom Declaration’s commitment to reducing tensions, establishing permanent peace on the Peninsula, and achieving the “common goal of realising, through complete denuclearisation, a nuclear-free Korean Peninsula”. We also note that both Koreas have taken several unilateral confidence building steps towards these ends, including North Korean undertakings to halt nuclear and missile testing and dismantle their nuclear weapon test site, South Korean/US willingness to temporarily postpone joint military exercises till after the 2018 Pyeongchang Winter Olympics, and the steps to reduce confrontation across the demilitarized zone.

Equally, however, we would be dismayed if the Trump-Kim June summit meeting were not to go ahead as planned, and by any further exchange of threats and counter threats. A nuclear conflict in this densely populated region would be an unthinkable humanitarian catastrophe for the Koreas, Japan and across the whole Asia Pacific region. With the presence of neighbouring nuclear powers, China and Russia, such a conflict could escalate into a wider nuclear war engulfing the whole world. Dialogue and diplomacy, rather than resort to war, have never been more urgent if we are to achieve peace on the Peninsula.

We remain deeply concerned that the risks of devastating nuclear conflict in this region are so grave that any initial dialogue will need to be rapidly solidified into substantive internationally-recognized verified agreements on comprehensive measures to create regional Northeast Asian peace, security and denuclearization. The proposed Trump-Kim summit will be vital to achieving the necessary political will for a step-by-step peace settlement process with its own verification requirements.

Studies conducted by PSNA experts have identified a number of steps that would be necessary in this process, including, most recently a Roadmap for Nuclear Diplomacy in North Korea prepared by Morton Halperin, Peter Hayes, Thomas Pickering and Leon Segal. As the experts detail, key elements of this Roadmap (modelled on the previous 2005 Six-Party Talks Joint Statement) are to:

The first phase of this Roadmap would involve initial commitments by North Korea to suspend all nuclear and missile tests, and fissile material production (including enrichment) in return for the US and ROK scaling back joint exercises (especially those using nuclear capable systems); and to provide energy and humanitarian assistance to DPRK.

A second phase would involve a resumption of Six Party talks without preconditions, confidence building measures, verification of dismantlement of relevant test sites, and negotiations commencing on a new peace and regional security arrangement.

The final phase would include: the declaration and implementation of a legally binding and internationally verified nuclear-weapon-free zone treaty (as in the case of five other regions in the world together with the single state Mongolian NWFZ); a final peace treaty agreed for the ending of the Korean War; and negative security guarantees provided by the recognised nuclear-armed-states to all parties to a regional NWFZ.

Vital to successful implementation of such a Roadmap to Northeast Asian Peace and Security will not only be the position of the two Koreas and the United States, but also other regional states, particularly Japan and Mongolia, and neighbouring nuclear powers, China and Russia, as well as the UN and wider international community.

In its role as a key regional state, we urge that Japan offer strong and substantive support to the new peace diplomacy inaugurated in the recent Moon/Kim Panmunjom Declaration. In particular, we call on its leadership to pursue patient and considered diplomacy in joining six-party declarations on the principles and goals of a Northeast Asian peace process; support a comprehensive Korean peace settlement; resolve outstanding issues in normalizing relations with North Korea; and offer the same kind of practical diplomatic support for a Northeast Asian Nuclear Weapon Free Zone as it did for the successful establishment of the 2006 Central Asian Nuclear Weapon Free Zone.

Continued nuclear risks and dangers in Northeast Asia

We continue to hold grave fears about nuclear risks and threats in this region. These risks include: the arms race consequences of missile defence systems in the region, particularly the THAADS system and planned Aegis Ashore systems; increased deployment of potentially nuclear capable vessels in Northeast Asian waters; tensions in neighbouring regions, such as the South China Sea and the Taiwan Straits; and the long term threats posed by increasing regional stockpiles of fissile materials. Some of these risks are analysed, and solutions proposed, in specialist research presentations at this PSNA Moscow Workshop.

In the case of THAADS missile-defence (MD) interceptor-rocket launchers already positioned in South Korea and Aegis Ashore systems planned in Japan, we continue to be deeply concerned that, while at first sight these might seem purely defensive, such systems also have a dual role in accelerating regional arms races because of a perceived need by targeted adversaries to overwhelm any missile defence system by deploying increased number of missiles and adding multiple warheads to each missile. In addition, such systems may well prove destabilizing given their long-range radar surveillance capabilities extending into Chinese and Russian territories, and potentially posing a pre-emptive strike risk serving to undermine China’s and Russia’s second-strike nuclear capability.

We also continue to be highly concerned about the potential for miscalculated or accidental nuclear war as a consequence of previously expressed (excluding China which maintains a no-first-use policy) preparedness to engage in pre-emptive strikes by some nuclear-armed states. There is also the risk of nuclear war resulting from early warning system computer errors, and from cyber attacks on nuclear weapon systems. We are equally concerned about the development of new types of nuclear-armed intermediate range and cruise missiles, which, even if conventionally armed, might appear to be, and mistaken for, nuclear armed missiles.

Further, given the existing presence of nuclear weapons in North Korea, and potentially nuclear-weapon-related facilities in other parts of the region, an effective verification scheme and arrangements will need to be developed and implemented to ensure all sides have confidence of compliance with agreements reached. Such a verification scheme would not only draw upon appropriate safeguards agreements with the International Atomic Energy Agency (IAEA) but also warrant establishment of a regional verification agency with a more extended mandate to investigate compliance within the region.

The new UN Nuclear Weapon Prohibition Treaty (NWPT)

The recent July 2017 adoption of the UN Treaty on the Prohibition of Nuclear Weapons (NWPT), supported by 122 non-nuclear UN Member States, will serve to outlaw nuclear weapons in a similar way to how chemical and biological weapons have been stigmatized and prohibited under international law. The new treaty seeks to mobilize the world community in applying normative pressure on states still possessing or relying on nuclear purported deterrents. It appeals to these states to rethink the global humanitarian, economic and environmental consequences of even a limited nuclear war. Such globally catastrophic impacts would extend far beyond the borders of those states who justify their continued nuclear reliance on the basis of national rather than global security interests. The nine current nuclear-armed states have largely sought to ignore the wider humanitarian and global threats posed by nuclear weapons, whether launched deliberately, accidentally or by miscalculation. We call upon all states, including nuclear “umbrella” states, to move towards reducing reliance on nuclear weapons as part of their defence or military postures, and to sign the NWPT treaty at the earliest opportunity.

Nuclear Non Proliferation Treaty (NPT) Initiatives and NPT Action Plan

While the 1968 Nuclear Non Proliferation Treaty does not include all nuclear-armed states (Israel, India, Pakistan and North Korea stand outside it), the NPT does oblige the five NPT-recognised nuclear-weapon states to reduce reliance on nuclear weapons, and to move towards total nuclear disarmament, particularly under Article VI which requires states to pursue negotiations in good faith on “cessation of the nuclear arms race and nuclear disarmament”, and under the agreed Action 5 of the 2010 NPT Review Conference committing nuclear-weapon states to “promptly engage with a view to…diminishing the role and significance of nuclear weapons in all military and security concepts, doctrines and policies”. The five nuclear-weapon states, China, France, Russia, UK and US, are all embarked on programs to modernize their nuclear armaments and delivery systems, while at the same time arguing that the NPT is the proper forum for disarmament negotiations. As we approach the 2020 NPT Review Conference, we call upon the nuclear powers to take seriously their agreed obligations towards reducing the role of nuclear weapons and move more decisively towards nuclear elimination.



Co-Chairs of Panel on Peace and Security of Northeast Asia (PSNA)

Michael Hammel-Green
Emeritus Professor, College of Arts and Education
Vitoria University

Chung-in Moon
Asia Pacific Leadership Network for
Nuclear Non-proliferation and Disarmament (APLN)
Republic of Korea

Hiromichi Umebayashi
Visiting Professor, Former Director of RECNA