North Korea’s Ballistic Missile Programme

Overview

North Korea’s interest in developing a ballistic missile capability appears to stem from its continuing efforts to establish and maintain robust military forces against South Korea, Japan and US forces in the region. As such, the reach of North Korea’s missile programme has expanded from, in the 1960s and 1970s developing and deploying tactical artillery rockets, to developing and deploying short-­range ballistic missiles in the 1980s and, in the 1990s, developing and deploying medium-range ballistic missiles. Systems capable of greater ranges are currently under research and development.

North Korea’s missile programme is based primarily on Soviet Scud missile technology, and Pyongyang is believed to have developed an infrastructure for missile research and development, testing, and production. This indigenous infrastructure has been supplemented by imports of specialised material and components. Although the primary motivation for North Korea’s missile programme appears to be related to security concerns, the export of missiles and missile­-related technology to numerous customers in the Middle East and South Asia for cash and barter has become an important secondary consideration.

Currently, North Korea has produced and deployed short-range Hwasong-5/-6 missiles (Scud-B/-C types), which can reach targets throughout South Korea, and medium-range No-dong missiles, which can reach targets across Japan. The exact size, disposition, and armament of these North Korean missile forces are unknown, although they are presumed to be modelled on Soviet doctrine, and organised into launch battalions with four to six mobile launchers per battalion plus support vehicles. (Normally, four missiles would be deployed with each launcher, one missile on the launcher and three on a reload missile carrier.) A number of different bunkers, shelters and tunnels associated with missile forces have been identified throughout North Korea. Conservatively, North Korea is estimated to deploy around five to seven Hwasong-5/-6 battalions and between two and three No-dong battalions – a capability, including reserves, equal to several hundred missiles. Armed with high-explosive warheads, North Korea’s missiles could serve as terror weapons against foreign cities or as long-range, though inaccurate, artillery against military targets. US and allied military commands assume that North Korea has armed some missiles with chemical and possibly biological warfare (CBW) payloads. Pyongyang certainly has this capability, but amounts and types of possible CBW agents cannot be determined. The No-dong missile is capable of accommodating a simple fission-type nuclear weapon, if Pyongyang has produced such a warhead.

In comparison to its deployed single-stage short- and medium-range missiles, even less is known about North Korea’s efforts to develop multiple-stage missiles. The August 1998 flight test of a Paektusan-1, or Taepo-dong-1 (TD-1) missile – configured as a three-stage space launch vehicle – was a mixed success. It achieved some technical milestones, but fell short on others. In any event, the TD-1 would not be capable of delivering a nuclear payload to intercontinental ranges. On paper, the larger Taepo-dong-2 (TD-2) is a much more capable missile, with a possible strategic capability against the continental US. However, it is not possible to determine how much progress Pyongyang has made on the TD-2 missile or even whether it is ready for testing. Since September 1999, North Korea has observed a moratorium on long-range missile testing, which imposes significant limits on Pyongyang’s ability to develop or deploy multiple-stage systems. Given the importance that North Korea has attached to missile development – whether as a strategic asset, bargaining chip, or both – it is plausible that such work continues. However, the true status of North Korea’s efforts to develop an intercontinental ballistic missile with a nuclear capability remains unknown.

From FROG to Scud-derivatives: a decade of short­range missile development

In 1968, the Soviet Union supplied North Korea with FROG rockets, Pyongyang was probably able to produce FROG copies soon after. The FROG system had strategic utility to Pyongyang because these solid­-propellant ballistic artillery missiles, with a range of 50–60km and with a 400kg high-explosive or chemical warhead, can hit Seoul from North Korean firing positions near the Demilitarised Zone (DMZ). This move presaged Pyongyang’s efforts, in the mid-1970s, to acquire or develop other short-range ballistic missiles. Meanwhile, education and training programmes designed to generate North Korean expertise in missile­related disciplines had begun before that time.

Pyongyang’s efforts to develop systems with a longer range than the FROG, with the intention of striking targets located deep in South Korean territory – presumably including cities as well as military-related-facilities – may have been partly spurred by a variety of South Korean efforts, beginning in the early 1970s, to develop surface-to-surface missiles derived from US-­supplied Nike-Hercules surface-to-air missiles. It is widely assumed that Pyongyang also sought the capability to strike targets in Japan, both to intimidate the Japanese government, so that it would not cooperate with the US in the event of a war on the Korean Peninsula, and to strike US military facilities in Japan. Presumably, North Korea’s air force capabilities – weak relative to the US and its allies – gave Pyongyang more incentive to develop a ballistic missile capability to strike distant targets, just as Arab states saw the acquisition of ballistic missiles as a strategic  counterweight to Israel’s air superiority after the 1967 Six-Day War.

By the early 1970s, the Soviet Union had begun to export short-range Scud ballistic missiles to its main Arab allies in the Middle East, and the missiles played a minor role in the 1973 Yom Kippur War. Strained political relations between Moscow and Pyongyang, however, apparently precluded similar exports to North Korea. Instead, Pyongyang turned to Beijing. During a visit by North Korean leader Kim Il Sung to Beijing in April 1975, North Korea reportedly proposed the development – with China – of a short-range ballistic missile, using the expertise China had gained in developing its medium- and long-range ballistic missiles. The resulting missile project, designated the Dong-feng–61 (East Wind-61) envisioned a single-stage mobile missile with a range of 600km and a one-tonne (1,000kg) payload. Probably based on Soviet Scud technology, which China had obtained from the Soviet Union, the planned missile was designed to use storable liquid fuels, turbo pumps, and inertial guidance. Work on the Dong-feng–61 may have started in 1977, but the project was suspended in 1978 when its main supporter General Chen Xilian was ousted from the Chinese government. Nonetheless, North Korean technical involvement in the programme may have assisted subsequent efforts by Pyongyang to build its own copy of the Scud missile.

About the same time that the Dong-feng-61 project collapsed, North Korea formed a strategic partnership with Egypt that allowed North Korea to reverse­-engineer Scud missiles. The relationship was borne out of mutual necessity: the collapse of Egypt’s military supply relationship with the Soviet Union following the 1978 Camp David accords had deprived it of the external assistance necessary to maintain and develop its ballistic missile force; Pyongyang had lost a potential ballistic missile source with the collapse of the Dong-feng-61 project. Between 1979–1981 – the exact date is unknown – Cairo provided Pyongyang with a small number of Soviet-supplied Scud-B missiles, along with mobile launchers, which North Korea could use as a basis for developing an indigenous capability to produce short-range ballistic missiles.

The Scud-B missile, a relatively simple and rugged system, was suitable because it both met Pyongyang’s military requirements and could be reproduced within North Korea’s technological capabilities. Based on German V-2 technology, the Scud-B missile is a single­stage liquid-fuelled ballistic missile with a range of 300km and a 1000kg payload – a significant improvement over the range and payload of the solid­propellant FROG artillery rockets in North Korea’s existing inventory. The Scud-B type provided to Egypt by the Soviet Union is designed to be transported and launched from a mobile launcher and carries a unitary high-explosive warhead of about 800kg, which does not detach from the missile body when the engines stop burning – thereby avoiding the technical complications of warhead separation that are typical of longer-range systems. The Scud steel airframe, a little more than 11m in length and 88cm in diameter, contains separate tanks of fuel (kerosene) and oxidizer (nitric acid), which are burned in the missile engine to generate thrust. These basic ingredients are easy to produce. The fact that the liquid fuels used can be stored also simplified design and operation over the use of cryogenic fuels, which had to be maintained at very low temperatures, employed in other types of liquid-fuelled missiles. The Scud-B is steered by four flaps, or jet vanes, that protrude into the exhaust plume and divert some of the thrust laterally – a method used by many early generation missiles. With a relatively unsophisticated on-board guidance system, the Scud-B has an accuracy of about 1km circular error probable (CEP).

Based on the missiles and mobile launchers provided by Egypt, Pyongyang embarked on a national programme in the early 1980s to build an indigenous version of the Scud-B missile, thought to be designated the Hwasong-5 (Mars-5). Logic dictates that Pyongyang must have built new facilities or converted existing facilities for the production of essential materials and components, including kerosene and nitric acid oxidizer, internal steel tanks and airframes, missile engines, guidance systems and warheads, although much of the publicly available information on production facilities claimed to be associated with ballistic missile production is unsubstantiated. Identifying facilities uniquely associated with the production of ballistic missile components is especially complicated because some components are probably manufactured at facilities primarily responsible for building tactical missiles or that are engaged in some other military or civilian function. Based on limited defector information and satellite surveillance, key facilities probably include the Sanum-dong research and development facility (also known as No.7 Factory), the Musudan-ri flight test facility, the Sungni Automobile Factory (for mobile launcher production), and the No.125 Factory near Pyongyang (for final missile assembly). North Korean production of the Hwasong-5 (and later missiles) was almost certainly supplemented by purchases of specialised materials, equipment, and components from foreign sources.

In 1984, North Korea conducted a series of tests of Hwasong-5 prototypes from the Musudan-ri flight-test facility, reportedly totaling three successful and three failed launches.5 Serial production of the Hwasong-5 probably began around 1985–86 and continued until around 1991–92, when serial production of the extended range Scud-C (Hwasong-6) missile began, most likely using the same facilities, materials and equipment previously used for the Hwasong-5. Early versions of the Hwasong-5 were delivered to Iran for use during the Iran–Iraq War, which provided the opportunity for North Korean engineers to collect valuable data on the operational use of this missile and to improve production. Based on essentially the same airframe as the Hwasong-5, the Hwasong-6 was designed to achieve a longer range (of 500km) by reducing the payload from 1,000kg to around 700–800kg, expanding the size of the internal fuel tanks, and slightly modifying the engine for longer burn time. The extended range was of particular interest to Pyongyang, since such missiles could reach all targets in South Korea and also meet the demands of Middle Eastern customers for a longer­range missile. The CEP of the Hwasong-6 is unknown, but the missile was probably less accurate than the shorter-range Hwasong-5 – perhaps in the order of 1–2km.14 Pyongyang reportedly conducted five tests of the Hwasong-6: in June 1990; July 1991; and May 1993 – the last a multiple test with three missiles. In addition to the Hwasong-6, Pyongyang may have worked on other Scud variants such as the Scud-D with even lighter payloads and longer ranges, but it is not known whether any of these were produced in significant numbers and deployed. Production of the Hwasong-6 (and other possible Scud variants) may have been phased out in the mid-1990s as North Korea began to build No-dong missiles.

There are no reliable figures for annual production capacity and overall production of the Hwasong-5 and Hwasong-6 missiles in the decade from the mid­1980s to the mid-1990s. No defectors claiming direct knowledge of production statistics have come forward, and intermittent satellite imagery of suspect production facilities cannot determine production rates. Estimates of annual production capacity range from about 50–100 missiles per year, and are based on estimates of North Korean deployments and exports – the presumed total is at least several hundred missiles. North Korea’s apparent ability to reverse-engineer sample Scud missiles and begin serial production of its own Scud-B copy within a few years, and then an extended range Scud-C – with a minimal test launch programme – is an impressive technical achievement. This success has led to speculation that the Hwasong-5/-6 missiles ‘produced’ in North Korea in this period were actually assembled from imported components.16 While it is clear that Hwasong-5/-6 production depended heavily on imported production equipment, raw materials, and electronic components, no evidence has come to light that major subsystems were manufactured abroad. One possible explanation for North Korea’s rapid progress is that China, which is believed to have obtained detailed blueprints and technical information on Scud-type missiles from the Soviet Union in the 1950s – before the schism between Moscow and Beijing – may have passed this technology to North Korea in the 1970s, thus accelerating North Korea’s ability to produce its own version of the Scud.

In any event, North Korea probably began deploying Hwasong-5 missile units in 1985–86, most likely beginning with small numbers and then establishing larger units as more missiles were produced. Presumably, as Hwasong-6 missiles came on-line in the early 1990s, they were deployed in the field in place of some Hwasong-5 missiles, but since the Hwasong-5 and Hwasong-6 missiles are virtually identical in external appearance and use very similar types of mobile launcher, the mix of missile types in North Korea’s missile forces cannot be exactly determined.

Since North Korean missile technology is based on Soviet technology, the organisation of North Korea’s missile forces probably replicates the structure of Soviet Scud forces. In this organisational structure, a North Korean missile regiment would be divided into several launch battalions, with a battalion divided into two or three firing batteries, each consisting of two launchers, reload missiles, and support vehicles. Thus, assuming that North Korea follows typical Soviet doctrine, each battalion would control four to six launchers, with each launcher accompanied by four missiles, one on the launcher itself and three in reserve carried by a supply vehicle. Conservatively, North Korea is estimated to maintain a single Hwasong-5/-6 missile regiment of about 30 mobile launchers making for a total of about 120 missiles deployed within the regiment, not counting missiles in reserve.

Although the total number of Hwasong-5/-6 missiles (including those stored in reserve) is unknown, they probably total several hundred. In 2000, the US Department of Defense estimated that North Korea’s ‘ballistic missile inventory now includes over 500 Scuds of various types’ – an assessment shared by the South Korean Ministry of National Defense. However, this number is based on rough assumptions, rather than direct information. For example, it is possible to count the number of shelters and bunkers assumed to be associated with Hwasong-5/-6 missile launchers, but it is impossible to determine how many of these shelters are in use at any one time or indeed how many may be decoys.

The Hwasong-5/-6 regimental headquarters is believed to be located in the Chiha-ri area (about 50km north of the DMZ), in a location where satellite imagery has detected a nest of hardened bunkers and tunnels. These are thought to store mobile launchers, missiles and associated equipment. The regiment has reportedly deployed launch battalions and firing batteries in various locations across the country, but, as with the location of missile production facilities, much of the publicly available information from defectors and press stories cannot be confirmed. Washington and Seoul have identified a number of locations, including command centres, bunkers, tunnels and maintenance facilities near the DMZ, that are assumed to be associated with Hwasong-5/-6 units. A number of pre-prepared launch sites have also been identified. In wartime, Hwasong-5/-6 mobile launchers would be scrambled from hide locations to these sites for firing and then moved as quickly as possible to another location for reloading.

Targeting doctrine and armament for Hwasong-5/-6 missiles is uncertain. Some Hwasong-5/-6 missiles are equipped with unitary high-explosive warheads and perhaps submunition bomblets, intended for delivery against cities, and military-related command locations, ports, and airports throughout South Korea. Inaccuracy and missile defences would limit the missile’s military effectiveness against any target. However, the political impact of, and the civilian terror generated by a number of missiles hitting cities on a daily basis, could create tremendous political pressure on leaders. Along with unitary warheads, North Korea may have been able to develop submunition bomblets to blanket a small target area. Military planners in the US and South Korea assume that chemical and possibly biological warheads are also available, although this cannot be confirmed. North Korea is almost certainly capable of building unitary CBW warheads with various types of agents and impact fusing; whether it has developed more sophisticated means of delivering CBW involving proximity fuses and bomblets is unknown. Hwasong-5/ -6 missiles are generally thought to have too small a diameter to deliver a first generation nuclear warhead.

Although South Korea is presumed to be the primary target area for North Korea’s Hwasong-5/-6 missile force, Seoul has typically viewed these missiles as a less serious threat than Pyongyang’s more extensive short-range artillery and rocket inventory, which is capable of hitting Seoul with large numbers of high-explosive or chemical rounds. Nonetheless, Hwasong-5/-6 hide locations and launchers are likely to be high priority targets for US and South Korean forces in a conflict. As a consequence, analysts argue that Pyongyang would likely scramble its missile forces as as soon as possible before fighting begins and use its missile forces early in a conflict to avoid pre-emption. It is impossible to accurately estimate how many North Korean missile launchers and crews would survive potential pre-emptive action, but at least some missiles would likely escape such strikes and be launched. Although US and South Korean forces currently deploy Patriot Advanced Capability-3 (PAC-3) missile-defence batteries around key military sites in South Korea, at least some missiles that are launched are likely to penetrate these defences. On balance, however, assuming Hwasong-5/-6 missiles are not armed with nuclear weapons, the damage they could inflict on major South Korean population centres is probably less than the damage Seoul could suffer from the short-­range battlefield and artillery-type systems deployed near the DMZ. The Hwasong-5/-6 threat against US military force locations in South Korea is also limited. As a result, Washington and Seoul have not placed the highest priority on North Korean short-range missiles in their negotiating strategies with Pyongyang. However, as US forces along the DMZ are gradually deployed further south, out of North Korean artillery range, Pyongyang may have an incentive to build up its short-range missile forces even further.

Decade of No-dong MRBM development

By the early 1990s, the deployment of Hwasong-6 missile units gave North Korea the ability to strike targets throughout South Korea with high-explosive or CBW warheads. A new missile design would be needed to reach targets in Japan and to deliver a nuclear warhead. To meet these requirements, North Korea embarked on a programme in the late 1980s to build a new missile, known as the No-dong, with a range of 1,000–1,300km and payload of 700–1,000kg.40 In addition to meeting North Korean security requirements, the missile proved a popular export item, both for existing customers, such as Iran, and for new customers, such as Pakistan.

The No-dong missile is essentially a scaled-up Scud, using similar or identical liquid propellants with a larger engine. Like the Scud, the missile is single stage, but it is roughly 50% larger in length and diameter – with a length of approximately 15–16m and diameter of about 1.2–1.3m. At 16 tonnes, it is also two and half times larger in terms of mass. To achieve its maximum range, the No-dong must reach speeds 50% greater than the Scud-C, and doing so with its greater mass requires a new engine that can produce roughly four times the thrust of a Scud engine. Interestingly, the design of the No-dong is very similar to that of the ‘fat al-Abbas’ or S13 project begun by Iraq in 1989 to develop a missile that could deliver a 1,000kg payload to 1,200km. As far as is known, there was no direct contact between Iraqi and North Korean designers, but both may have been working from old Soviet designs for a scaled-up Scud, sometimes referred to as the Scud-E project. The concept design was generated by the Makayev design bureau, but it was never developed in the Soviet Union. North Korea may also have had direct assistance from Russian designers in developing the No-dong missile, although this cannot be confirmed.

Little is known about the development, production, and deployment of the No-dong missile. Development work is assumed to have begun at some point in the mid- to late-1980s. According to press reports, US satellites detected a No-dong-type missile at the Musudan-ri test launch facility site in May 1990, and subsequent imagery showed scorch marks on the launch pad, suggesting that the missile had exploded on take off. In May 1993, however, a No-dong was successfully tested from the Musudan-ri launch site to a distance of 500km into the Sea of Japan (East Sea). No other No-dong tests have been conducted by Pyongyang, although North Korea’s missile export agreements with Pakistan and Iran have apparently included provisions for North Korea to participate in tests and to obtain test data from Pakistan and Iran – both of which have conducted a series of missile tests of No-dong copies since 1998.43 There have not been enough No-dong tests to provide an accurate assessment of the missile’s accuracy: estimates suggest a CEP of 3–5km, but this could be considerably larger if the warhead spirals or tumbles during reentry.

North Korea is believed to have produced No-dong units in the same facilities used for the production of Hwasong-5/-6 missiles. As with the Hwasong-5/6, the line between pilot and serial production of the No-dong and the line between partial and full deployment is hazy. Small numbers of No-dong missiles may have been produced and deployed in 1994–95, but full production may not have begun until 1996–97, and the system was not considered operationally deployed by South Korea’s defence ministry until 1997. Although most of the production capacity is indigenous, North Korea has reportedly obtained some imported materials and components for No-dong production such as special steel alloys from China and electronic components from Japan. The production rate for the No-dong missile is unknown, but probably never exceeded an average of about half a dozen missiles per month, with some missiles slated for deployment in North Korea and some for export. Like the Hwasong-5/-6, the No-dong is designed for delivery from mobile launchers, which North Korea is believed to produce at the Sungni Automobile Factory.

The size, deployment and armament of North Korea’s No-dong missile force are unknown. Conservatively, The Military Balance estimates around ten No-dong missile launchers, organised into two or three battalions, but the number could be greater. This would imply at least 40 deployed missiles (assuming that North Korea adopts the Soviet practice of deploying four missiles per launcher in the field), and perhaps a total of 100-200 missiles including reserves. As with estimates of Hwasong-5/-6 launchers and missiles, estimates of North Korea’s No-dong force are very approximate, based on various calculations and assumptions concerning production rates, numbers of shelters and firing points, and military organisation and practice, though the actual number of No-dong launchers and missiles cannot be precisely determined. Nevertheless, the size of North Korea’s No-dong missile force essentially represents Pyongyang’s calculation of what it requires in order to meet its military requirements and political needs. In principle, there is no reason why North Korea could not expand its No­dong force if it believed it was necessary.

Unlike Hwasong-5/-6 missiles, which needed to be deployed near the DMZ in order to reach their presumed targets, No-dong missiles can be deployed further north. According to various press reports, satellite intelligence has detected a series of underground launch sites near the Chinese border and the East Sea, which are apparently constructed to allow No-dong launchers to enter tunnels and launch through openings at the top of underground bunkers. Given its range, the primary targets for the No-dong missile are presumed to be in Japan. Armed with a high­explosive warhead, the No-dong is probably not accurate enough for effective use against military targets, such as US military bases in Japan, but it could serve as a terror weapon against Japanese cities, including Tokyo. North Korea is assumed to be capable of arming No-dong missiles with CBW warheads, but there is no reliable information on whether it has done so or on what type of warhead and delivery system it may have employed. The potential effectiveness of a chemical or biological warhead against civilian targets varies significantly depending on such factors as the type of agent, the type of delivery mechanism, and the effectiveness of civil defence efforts. As discussed in the chapter dealing with North Korea’s nuclear programme, it is possible that North Korea could arm the No-dong missile with a nuclear warhead. There is, however, no confirmation that it has done so.

The ability of No-dong missiles to successfully attack targets in Japan depends upon the survivability of North Korean missile forces against offensive action, as well as the ability of No-dong missiles to penetrate whatever missile defences they encounter. Given that No-dong missiles could be armed with nuclear weapons, US and allied forces are likely to make every effort to destroy bunkers and facilities suspected of hiding these missiles and their launch-infrastructure before they can be mobilised and used. On the assumption that suspected No-dong bunkers and launch sites would be a high priority target for US and allied forces in wartime North Korea is credited with taking a number of active measures – for example, the construction of decoy shelters and redundant bunkers – to conceal the size and disposition of its missile force.

There may also be undetected No-dong hiding locations. It is presumed that, should a conflict occur, North Korea would plan on scrambling its No-dong forces before pre-emptive action could be taken against them, and that Pyongyang might fire missiles early in a conflict if the survival of the force were jeopardised.

Nonetheless, given its low numbers and poor accuracy, the No-dong is probably more a political weapon than an effective military instrument. In one wartime scenario, for example, North Korea could fire a small number of conventionally armed missiles and threaten to escalate to unconventional warheads if the US and its allies did not accept a cease-fire. Presumably, Pyongyang recognises that the actual use of a nuclear warhead (if it has one) would be suicidal, but it calculates that the threat that it might take desperate measures in extremis serves an important deterrent function.

After the May 1993 No-dong test, the US proposed a joint Theater Missile Defense [TMD] development project with Japan, although Tokyo was reluctant to commit itself, in part because of concerns about antagonising China. Following the August 1998 Taepo-dong-1 test, however, which aroused a strong public reaction, the Japanese government reached agreement with the US in December 1998 on co-development of a sea-based TMD, called the Navy Theater Wide (NTW) defence system. The NTW system envisages a new Standard Missile-3 (SM-3) with a kinetic, hit-to-kill warhead designed to achieve midcourse intercepts outside the atmosphere during an intermediate-range ballistic missile’s flight. Given a number of technical issues, the US Navy does not plan initial operational testing of the NTW system until 2010. In December 2003, Tokyo announced that it would proceed with plans to develop an integrated missile defence system, including both sea-based SM-3 and land-based Patriot Advanced Capability-3 (PAC-3) missiles.

The Taepo-dong-1 shock

By the late 1990s, the No-dong missile gave North Korea a credible missile with which to threaten Japan. The No­dong aroused particular concern given the possibility that it could carry a nuclear warhead. However, if Pyongyang wanted to pose a threat to US cities with a missile launched from North Korea, or wanted to develop a space launch capability, it would need to develop a missile with considerably greater thrust and range than that of the No-dong. Presumably, from Pyongyang’s standpoint, a capability to attack US cities might undermine the credibility of US security guarantees to South Korea and Japan in the event of a North Korean attack. Privately, senior North Korean military officers have warned visiting American officials that North Korea is seeking to develop missiles that could attack the US in order to retaliate against any US attack on North Korea. From a technical point of view, the development of the No-dong’s engine – more powerful than those used in the Hwasong series – was an important step towards developing missiles with even longer range. The other key technology required to attain longer ranges is multiple staging.

The two Taepo-dong-family missiles that are assumed to be currently under development both utilise multiple staging. On 31 August 1998, Pyongyang attempted to place a small satellite in orbit by launching a three-stage Taepo-dong-1 (TD-1) missile. Although the third stage failed, the test demonstrated for the first time North Korea’s technical capability to launch missiles with multiple stages, since the second and third stages both successfully separated. After it seperated, however, the third stage failed to boost the satellite into orbit. By the time it failed, the third stage reached a high enough speed that debris from its breakup reportedly travelled some 4,000km. What caused this third stage failure is unknown. Some analysts have speculated that its engine exploded, while others have suggested that the stage broke apart after tumbling caused by guidance and control failure.

It is not known when North Korea began development work on the TD-1, but in February 1994, US intelligence satellites reportedly detected two new types of missiles or missile mock-ups at the Sanum­dong missile research and development facility. After imagery analysis, estimates were made of the likely staging combinations for the two missiles along with calculations of possible range and payload parameters, based on technical information on the performance characteristics of the No-dong and Hwasong missiles. US analysts said the press have dubbed these new missile types as the Taepo-dong-1 and the Taepo-dong-2, although the Taepo-dong-1 is apparently known as the Paektusan­1 in North Korea. It was originally believed that the TD-1 would be a two-stage ballistic missile, consisting of a modified No-dong missile as the first stage and a Scud-C as the second stage. Such a combination was estimated to have a range of about 2,000km with a 700–1,000kg payload, and with an overall mass of 20–25 tonnes. Over the next several years, US analysts sought to follow North Korean efforts to develop these new missile types by, for example, evaluating satellite evidence of static engine tests that appeared to be associated with engines larger than those for the Scud missile, but Pyongyang took a number of concealment and deception measures to hide its activities from US intelligence satellites, especially after press leaks of classified information.

Nonetheless, US analysts anticipated, on the basis of estimated development times, that North Korea would be able to test the TD-1 by the late-1990s. As a result, Washington was not surprised when satellites detected preparations for the launch of a TD-1 missile at the Musudan-ri launch site in the summer of 1998, well before the actual launch. Despite warnings from US diplomats not to test the missile, North Korea launched the TD-1 on 31 August 1998. To the surprise of US analysts, the TD-1 included a third stage designed to launch a satellite. The first stage appears to have been a modified No-dong missile, which performed in a manner characteristic of a ballistic missile launch. The second stage Scud engine, however, appears to have been adjusted for a long burn time at relatively low thrust. Such performance, unlike that of a ballistic missile, is designed to allow the third stage to reach a high altitude before it fires its engine and attempts to put the satellite in orbit.

The third stage consisted of a small solid-fuel rocket motor that carried a small satellite with a mass of probably a few tens of kilograms. The satellite, named Kwangmyongsong (Bright Lodestar), apparently carried a small transmitter that was intended to broadcast political songs to demonstrate its presence, in a similar manner to the first Chinese satellite in April 1970, which played ‘The East is Red’. However, exact information on the origin and technical details of the third stage are unknown. North Korea is not known to have previously developed solid-fuel engines for space launches, but North Korea is capable of building small, solid-fuel rockets for tactical missiles, and it might have obtained assistance from China or other countries in satellite and space launch technology.

The TD-1 launch was perceived as highly provocative, since the launch trajectory was directed to the east of the Korean Peninsula, passing over Japan. However, a launch in that direction is consistent with an attempted space launch, since it allows the missile to use the Earth’s rotation to increase its speed. The launch also occurred within days of the 50th anniversary of the founding of the North Korean state and a consitutional revision that formally transferred power to Kim Jong-Il, suggesting that it was intended to garner prestige and commemorate these events. Citing South Korean and Japanese space launch and satellite programmes, Pyongyang argues that it has the right to develop a space launch capability to orbit its own satellites for telecommunications, meteorological, and other peaceful purposes. In addition, Pyongyang’s proclaimed pursuit of a ‘peaceful’ space launch capability is intended to provide a less provocative political cover for the development of long-range ballistic missiles with a military mission, since the technology involved in space launchers is essentially the same as ballistic missile technology.

In any event, the TD-1 launch caused international uproar, with Japan suspending support for the Agreed Framework and the US threatening to cut off humanitarian food assistance if Pyongyang tested another long-range missile. It also boosted missile defence efforts in the US and Japan, in turn alarming China and Russia, both of whom reportedly warned North Korea against any additional firings. Facing concerted international pressure, Pyongyang embarked on negotiations with the United States, which resulted in agreement in September 1999 on a moratorium on long-range missile tests, in exchange for US agreement to lift a number of economic sanctions. The moratorium is understood to include missiles with, and above, the range of the No-dong. Pyongyang has unilaterally extended the moratorium on several occasions, most recently in September 2002 during the visit of Japanese Prime Minister Koizumi, although North Korea has also threatened to resume ‘satellite launches’ since it broke the nuclear freeze in late 2002.

The current status of the TD-1 is uncertain. Presumably, Pyongyang has built or could assemble additional TD-1 missiles for firing on fairly short notice, but it is not known whether North Korea has deployed the TD-1 (in a two- or three-stage configurations) to military units. Unlike the No-dong, the TD-1 is not designed for launch by mobile launcher. Some analysts have speculated that it could be deployed in ‘silo fashion’ in the underground firing bunkers that North Korea has built near the Chinese border. Arguably, however, there would be limited military value in deploying the TD-1 because it contributes little to the strategic role already played by the deployed No-dong force, which effectively covers all critical targets in Japan with a warhead capable of delivering a nuclear weapon. In two-stage configuration, the TD-1 can deliver a payload comparable to that carried by the No­dong to a greater range, but the extra distance does not encompass any key targets of significant value to North Korea. Moreover, since the second stage of the TD-1 is essentially a Scud missile, the diameter is probably not large enough to accommodate a simple fission-type nuclear warhead, unlike the No-dong.

Using three-stages, the TD-1 is theoretically capable of delivering a small payload of 100–200kg to targets as far away as the continental US, but this payload is considered far less than the amount needed to deliver an early generation nuclear weapon. Even were the missile to carry a chemical or biological warhead, a large fraction of this payload, perhaps two-thirds, would be needed for the structure of the warhead and the reentry heat-shield. In other words, the amount of chemical or biological agent that it could potentially carry would be very small, and probably not enough to pose a significant mass-casualty threat. Moreover, at this maximum range, assessments are that the accuracy of the TD-1 would be very poor. Given these limitations, the TD-1 cannot be considered a serious military threat to the continental US, even if it is deployed. Indeed, it may be the case that the TD-1 was intended primarily as a space launch vehicle and a ‘technology demonstrator’ to develop North Korean capabilities for the development of longer-range systems capable of striking the US with a militarily significant payload.

Missile defence and the US debate on the Taepo-dong-2

Given the technical limitations of the Taepo-dong-1, the US has long believed that North Korea is much more likely to develop a larger missile, designated the Taepo-dong-2 (TD-2), as a potential intercontinental­range ballistic missile (ICBM) capable of effectively threatening the US with a nuclear-sized payload. While the TD-2 uses components from North Korea’s previous missiles, it is significantly different from any missile North Korea has built or tested before and poses a number of new technical challenges. As a result, a number of flight tests would probably be required to develop the TD-2. The TD-2 would use the same engines as the No-dong and the same (or a similar) propellant, but it would require a new, larger first stage, consisting of a cluster of four No-dong engines. This would substantially increase the project’s technical complexity and create more opportunities for failure if the four engines do not operate in concert. The TD-2’s second stage is thought to use a single No-dong engine, modified for use at high altitudes. Rather than using a No-dong missile body for the second stage, North Korea is more likely to build a stage that is shorter and greater in diameter than the No-dong, since that configuration can reduce the structural mass of the stage, which is crucial for attaining the high speeds needed to reach long ranges.

To achieve maximum range, the TD-2 would need to include a third stage similar to the solid rocket third stage that failed in the August 1998 TD-1 test. North Korea would also need to develop and flight test a re­entry heat shield for a long-range missile before it could use it to deliver a warhead. Overall, the TD-2 missile would be significantly larger than the TD-1, with a maximum diameter (2.4m) nearly twice that of the TD-1 (1.25m). At 75–80 tonnes, it would weigh four times as much as the TD-1 and would generate greater thrust, leading to greater mechanical stresses on the missile’s airframe than those seen on previous missiles. As noted above, keeping the structural mass low is crucial to achieving long ranges. However, making the structure sufficiently strong to deal with greater stresses, while minimising overall mass, is a complex engineering challenge.

In the US, assessments of North Korea’s TD-2 threat became intertwined with the debate over National Missile Defense (NMD). In general, opponents of missile defence tended to emphasise the technical limitations of the TD-2 and the hurdles that North Korea would face before it could develop and deploy an operational ICBM capable of delivering a nuclear warhead to the continental US. Proponents of missile defence, on the other hand, tended to emphasise the danger of North Korea surprising the US by demonstrating a capability much sooner than expected. In 1995, the public version of US National Intelligence Estimate (NIE) on missile threats concluded that no country, other than the declared nuclear powers, would develop or otherwise acquire a ballistic missile in the next 15 years that could threaten the continental US. Turning to North Korea, the NIE assessed that Pyongyang was unlikely to achieve the technological capability to develop an operational ICBM longer than the 4,000–6,000km-range Taepo-dong-2, which could theoretically reach Alaska (but not the rest of the continental US or Hawaii) with a payload large enough to accommodate a nuclear warhead. According to the report, ‘For such an ICBM, North Korea would have to develop a new propulsion and improved guidance and control systems, and conduct a flight test program.’ Congressional critics of this intelligence estimate believed that it underestimated the potential ballistic missile threat, and Congress passed legislation to form a  ‘Commission to Assess the Ballistic Missile Threat to the United States’. In July 1998, the Commission concluded that North Korea could develop an ICBM within five years of a decision to do so.

The August 1998 TD-1 space-launch attempt was seen as supporting the Commission’s argument that Pyongyang could make significant progress towards the development of a missile that could threaten the US with little or no warning, especially given the inherent difficulties of collecting information on North Korea. Even though the third stage of the TD-1 was unsuccessful, the failure to anticipate or detect North Korean work on third-stage technology led the US intelligence community (already under criticism for underestimating the missile threat) to incorporate ‘worst case’ scenarios in their assessments of North Korean missile developments. The intelligence community adopted a looser definition of operational deployment, recognising that Pyongyang might be willing to ‘deploy’ a new missile system even before a full series of flight tests. In addition, US assessments of the possible range and payload performance of the presumed TD-2 have postulated greater capabilities over time taking into consideration possible techniques for enhancing performance. The original 1995 US government assessments of the TD-2 estimated that it had a range of 4,000–6,000km, sufficient to reach Alaska, but not the main Hawaiian Islands, nor the rest of continental United States. The 1999 NIE on Foreign Missile Developments and the Ballistic Missile Threat Through 2015 – the first official assessment after the August 1998 launch – concluded that a two-stage TD-2 could carry a ‘nuclear weapon sized’ payload, defined as ‘several-hundred kilograms’ to Alaska and Hawaii (7,500 km). In 2001, the official US estimate concluded that ‘The Taepo Dong-2 in a two-stage ballistic missile configuration could deliver a several-hundred-kg warhead up to 10,000km’ – sufficient to strike some coastal areas in the western continental US.56 Assuming that the TD-2 is designed to carry a third stage similar to that used in the 1998 TD-1 launch, the 1999 NIE estimated that a three-stage TD-2 ‘could deliver a several-hundred kilogram payload anywhere in the United States’. In 2001, the NIE specified the range at 15,000km, with a payload of several hundred kilograms, which would be sufficient to allow it to reach all of North America.

The greater capabilities attributed to the TD-2 by more recent US estimates reflect a combination of several possible measures that North Korea could take to enhance TD-2 performance in comparison to that observed during 1998’s TD-1 launch. Manufacturing the missile body from significantly lighter materials, such as aluminum magnesium rather than steel – or improving engine performance – are two possible measures. It is not known whether North Korea is pursuing these routes to enhance TD-2 range, but recent US government estimates are intended to cover such potential ‘worst case’ contingencies. If it used known North Korean capabilities, the TD-2 could have a CEP of tens of kilometers at long range, which could limit its effectiveness against anything except large targets such as cities.

Though it is presumed that North Korean scientists and technicians are working on improved guidance and control systems, there is little information on such efforts. Improving the accuracy of a ballistic missile is not just a matter of improving its guidance system, which is used to aim the missile during the first few minutes of flight, when it is under power. Once the engines burn out, the warhead falls through space, and a significant part of the missile’s CEP results from intense atmospheric buffeting as the warhead reenters the atmosphere in the last few minutes before impact. For example, if ablation does not take place symmetrically over the heat shield that is used to protect the warhead from re-entry heating, uneven wear may result in lateral forces pulling the warhead off-course. Moreover, the difficulties in achieving a low CEP increase as the range of the missile increases. For instance, errors in aiming the missile through its initial burn phase leads to greater errors in hitting the target as target range increases. And, since atmospheric forces increase rapidly with speed, the higher speed of longer-range missiles leads to considerably greater re­entry errors. From Pyongyang’s standpoint, however, high accuracy would not be required for the TD-2 to perform its presumed strategic function – that of threatening US cities with nuclear attack.

While range, payload, and warhead type are typically discussed as the primary characteristics of a ballistic missile, another key consideration is whether or not the missile includes countermeasures designed to defeat missile defences. If Pyongyang is indeed determined to pursue long-range missiles, it has both aerospace skills as well as a motivation for developing mechanisms intended to defeat missile defences, such as releasing balloon decoys while disguising the warhead by enclosing it in a similar balloon. It is not known whether North Korea is seeking to develop such mechanisms.

The TD-2 has never been flight-tested and its development status is unknown. Since 1998, the US government has estimated that the TD-2 ‘may’ be ready for flight testing and could be deployed in a ‘few years’, but this assessment is based on a combination of assumptions about North Korea’s missile development capabilities and fragmentary evidence. For example, in 1999, North Korea reportedly modified the Musudan-ri site (from where the TD-1 was launched) by extending the launch gantry so that it could accommodate the taller TD-2 missile, suggesting that the TD-2 was ready for flight testing, or at least that Pyongyang wanted Washington to believe this to be the case. According to press reports, intelligence satellites have also detected several static engine tests since 1999 that might have been associated with development of a TD-2 engine, although it is difficult to determine the type of engine being tested or to evaluate the results of the tests from the information available. In contrast to US government assessments, the Russian government has tended to regard the TD-2 as a ‘paper missile’, emphasising that North Korea faces significant technical hurdles in developing a missile that could threaten the US. However, these Russian assessments may be influenced by a desire to undercut the justification for missile defence.

In reality, it is impossible to make a confident judgement about the status of the TD-2 because too little information is available. Even North Korea cannot be certain about the capabilities of the TD-2 until it is flight tested, and since September 1999, Pyongyang has continued to observe a moratorium on long-range missile flight tests, though it is unknown whether this is for technical or political reasons. Since the collapse of the Agreed Framework in late 2002, Pyongyang has periodically issued threats to resume satellite launches – for example after Japan launched a reconnaissance satellite in March 2003 – but no preparations for another missile test have been reported at the Musundan-ri launch site. Even if the TD-2 is successfully tested, some experts argue that a series of flight tests would be required to provide a meaningful estimate of reliability and therefore to consider the missile operational in the normal sense. Other experts cite the No-dong programme and argue that North Korea might ‘deploy’ the TD-2 based on one or two successful tests, calculating that the mere possibility of the missile operating would serve a deterrent role.

New developments?

In September 2003, according to press reports, intelligence satellites observed several copies of a new type of intermediate-range missile (or missile mock-up) and mobile launcher near Pyongyang. Presumably based on its observed dimensions and external appearance, the new missile was said to resemble the SS-N-6, an old Soviet-era single-stage, liquid fuelled submarine-launched ballistic missile developed in the 1960s and designed to deliver a payload of about 700kg to a range of 2,400–3,000 km. The SS-N-6 is shorter and fatter than the No-dong, with a length slightly under 10m and diameter of 1.65m, compared to the 15–16m length and 1.2–1.3m diameter of the No-dong. Of course, it cannot be determined from satellite images whether the new North Korean missile is actually derived from the SS-N-6 or even whether it is real. Some analysts speculated that the missiles were mock-ups intended to be displayed in a military parade marking the 55th anniversary of the foundation of North Korea, but the missiles did not make a public appearance. If it is a new missile project, the estimated range would not be great enough to strike the US. However, the larger diameter would make it easier to deliver an early generation nuclear device, when compared to the No-dong.

In conclusion, since the early 1990s, North Korea has actively pursued an interest in developing space­launch vehicles and long-range ballistic missiles, which requires the development of technology beyond that used in the No-dong. Most notably, the August 1998 launch of a three-stage TD-1 demonstrated North Korean achievement of some important technical benchmarks necessary for development of intercontinental missiles, such as stage separation, but failure to achieve others, such as a successful third stage powered by a solid-fuel motor. Deployment of a TD-1 would serve little greater military utility than that served by North Korea’s existing No-dong missile force. However, Pyongyang could presumably resume TD-1 testing on fairly short notice, whether to renew efforts to orbit a satellite, or develop ICBM technology, or make a political statement – or indeed all three.

In contrast to the TD-1, the Taepo-dong-2 is a more credible candidate for an intercontinental missile, although it poses a number of additional technical hurdles. In either a two-stage or three-stage configuration, the TD-2 could theoretically deliver a nuclear weapon-sized payload to cities in the United States. There is some evidence that North Korea has continued to work on the TD-2 or other types of long-­range missiles since the August 1998 TD-1 test, but information on the level of effort and the level of success is very fragmentary and elusive. As a result, a firm judgement on the status of the TD-2 cannot be rendered because too little information is available.

North Korean missile exports

North Korea has become the world’s most prolific exporter of ballistic missiles and related equipment, materials and technology – especially as other potential suppliers, such as China, have gradually withdrawn from the market. Over the past two decades, North Korea has sold at least several hundred Hwasong-5/-6 or No-dong missiles, as well as materials, equipment, components and production technology, mainly to countries in the Middle East, such as Egypt, Iran, Libya, Pakistan, Syria, the United Arab Emirates (UAE) and Yemen. During that time, North Korea’s missile export business has probably earned several hundred million dollars – a significant portion of North Korea’s hard currency earnings. Missile deals have also probably included barter arrangements for oil (with Iran) and nuclear technology (with Pakistan), and have provided North Korea with opportunities to test missiles off­shore. In recent years, however, revenues from missile sales may have fallen off. Some of North Korea’s longstanding customers, such as Iran, have come close to achieving an independent production capability, reducing their need for North Korean imports and even presenting competition to North Korean sales. Other customers, such as Pakistan, Yemen, Egypt and the UAE and most recently Libya, have come under political pressure from Washington to sever their missile relationship with Pyongyang. Nonetheless, North Korea remains the world’s leading exporter of ballistic missiles and related technology.

As part of the original missile partnership established between Pyongyang and Cairo more than 20 years ago, North Korea probably sold spare parts to help maintain Egypt’s Scud-B force throughout the 1980s. Cairo, however, focused its missile development efforts on a joint project with Argentina and Iraq to develop the solid propellant Condor missile, a project which started in 1985. US pressure on Argentina and European prosecutions of companies involved in the project brought an end to the Condor project in 1988–89. After this, Egypt apparently turned back to North Korea, which sold production technology, raw materials and key components for extended range Scud-C missiles throughout the 1990s. Under pressure from Washington, following a 1996 incident in which Swiss authorities intercepted a large shipment of Egypt-­bound North Korean Scud-C missile components at Zurich airport, Cairo made commitments to limit missile cooperation with Pyongyang. In 2000, Egypt reportedly approached North Korea in a bid to acquire engines and other components for No-dong missiles, though US diplomatic intervention with Cairo apparently blocked the complete transfer.

North Korea’s missile exports to Iran began during the Iran–Iraq War when North Korea shipped Scud-B missiles (designated the Shahab-1 by Iran) and mobile launchers to Iran in 1987 for use against Iraq in the ‘War of the Cities’. In the early 1990s, North Korea provided Iran Scud-C missiles (designated the Shahab­2) and helped Iran to establish an indigenous missile production infrastructure, in exchange for money and oil. In 1993, North Korea negotiated with Iran for the sale of No-dong missiles, but the exports were delayed, perhaps because of warnings from Washington that No-dong transfers to Iran could derail negotiations for the Agreed Framework, which were taking place at the same time. In 1995 however, after the conclusion of the Agreed Framework, North Korea began exporting No­dong missiles to Iran and helped Iran develop its own version of the No-dong, which Tehran designated the Shahab-3. The North Korean sale of No-dong missiles to Iran prompted Washington to begin a series of negotiations with Pyongyang, seeking an agreement to end Nort, h Korean missile exports, but the two sides were never able to reach agreement on the amount and type of ‘compensation’ that North Korea would receive for ending missile-related exports. North Korea may have provided Taepo-dong technology to Iran, but this cannot be confirmed.

Reports of Syrian negotiations with North Korea for missile sales surfaced in 1989, apparently after China, under pressure from the United States, withdrew from arrangements to help Damascus upgrade its ageing force of Soviet-supplied Scud-B missiles. In the early 1990s, North Korea shipped Scud-C missiles and production equipment to Syria, but there have been few reports of substantial North Korean sales to Syria since then, perhaps in part because Syria and Iran developed a close missile cooperation relationship during the mid­1990s. North Korean missile sales to Libya came to light in June 1999, when Indian customs officials seized the North Korean ship Ku Wol San after they suspected the ship was delivering missiles to Pakistan. In fact, the Ku Wol San was destined for Libya, carrying raw materials, test equipment, machine tools, missile components, and blueprints for Scud missiles. Since 2000, there have been several unconfirmed reports that North Korea has sold No-dong missiles or components to Libya. Tripoli’s December 2003 announcement that it will renounce its weapons of mass destruction programmes and ballistic missiles of over 300km range presumably limits future North Korean sales to Libya.

During the 1990s, North Korea also made minor sales of Scud missiles to the United Arab Emirates (UAE) and Yemen. These exports attracted US attention, and Washington negotiated agreements with both the UAE and Yemen to end further missile purchases from North Korea. In December 2002, however, Spanish and US naval vessels intercepted the North Korean ship Sosan off the east coast of Africa – the ship was found to be carrying 15 Scud missiles to Yemen. The ship’s cargo was also carrying conventional warheads and 85 drums of ‘inhibited red fuming nitric acid’, an oxidizer for Scud missile fuel. The ship was stopped ostensibly for being ‘unflagged’, but was later released when Yemen protested to Washington. However, Yemeni President Ali Abdullah Saleh said that the missiles would be used only for national defence, and that it would be the last shipment. Based on evidence uncovered since the US invasion of Iraq, it also appears that Iraq began negotiations with North Korea in 1999 to purchase No­dong missiles and production technology, although the deal fell through in 2002 when North Korea decided it was too risky to begin shipments to Iraq via Syria. Baghdad’s downpayment was, though, retained by Pyongyang.

Strategically, North Korea’s most significant missile customer has been Pakistan, which reportedly agreed around 1997 to trade uranium enrichment technology to North Korea in exchange for No-dong missiles and production technology. According to various reports, the missile components and technicians were transported by aircraft from Pyongyang to Islamabad in late-1997, and in April 1998, Pakistan flight tested its version of the No-dong, which it designated the Ghauri.64 North Korean technicians apparently continued to work with Pakistan on development of the Ghauri, which was further tested in April 1999 and May 2002. Following the US accusation in October 2002 that North Korea was pursuing a clandestine enrichment programme, Washington has increased pressure on Pakistan to end its missile cooperation with North Korea.