Nuclear and Missile Update

As the Iranian negotiations move into their next phase, it is useful to review some recent estimates of the status of the North Korean nuclear and missile programs. In particular, we highlight reports from SAIS’s North Korea Nuclear Futures series (SAIS’s US-Korea Institute). The core of this debate is not only about reprocessing, enrichment and the generation of fissile material and weapons but also about the linked question of the missile program and miniaturization: whether the North has the capability to mount a device on a delivery vehicle and of what range.

Anyone interested in this issue needs to start with the study by David Albright at the Institute for Science and International Security called “Future Directions in the DPRK’s Nuclear Weapons Program: Three Scenarios for 2020.” The paper generated a lot of traffic, pro and con, which Albright and Joel Wit review in an open-minded rejoinder at 38North.

It is puzzling why the report generated so much grumbling, but some of it appears to come from a misunderstanding of scenario analysis. These exercises start with a particular baseline of capabilities (such as the amount of fissile material; their guess is about 32 kilos), which itself is subject to some uncertainty. The exercise then outlines possible ranges for future developments based on more or less optimistic (or pessimistic, as the case may be) assessments of the ability of the North Koreans to overcome a myriad of technical constraints: in the production of fissile material, in weaponization, in miniaturization. The bottom line is offered up in the diagram reproduced below; each of these paths is also graphed out more precisely over the time period. But it is important to understand that the probabilities of these three scenarios sum to 100%; no given scenario is certain by any means, and analysts will differ on the probabilities attached to each.

So what are the points of controversy? A few bear mentioning as they relate to the Iran debate:

• The HEU program is subject to particular uncertainties “including whether there is a
secret centrifuge plant in addition to the Yongbyon plant, how many P2-type centrifuges have been deployed successfully, and how well have these centrifuges operated.” (More detail on these estimates is contained in an earlier report by Albright and Christina Walrond here in .pdf; Albright and Wit also believe North Korean capacity to be higher because of direct contact with the Pakistani program). Albright’s modeling in this part of the paper is particularly impressive in my view because it outlines two HEU scenarios very clearly—with one facility at Yonbyon and with a second at a covert site—and offers up a distribution of possible estimates, with medians at 100kg and 240kg of weapons-grade uranium over the time period respectively.

• The Iran debate in particular has focused largely on fissile material as the determinant of break-out time. But fissile material is not the same as devices or deliverable weapons. Even if the North Korean program keeps churning out weapons grade uranium, the North Koreans still have to engineer bombs. There are more replete technical analyses, but the New York Times has a useful interactive graphic on what might be called the warhead supply chain, and the list is long: detonators, lenses, neutron initiators, fusing, arming and firing systems and so on. Clearly, North Korea has demonstrated a capacity to solve these problems given its three tests. But sanctions on dual-use technologies have almost certainly tightened and a few tests do not necessarily demonstrate a reliable capacity to repeat with ease, particularly where there is uncertainty about the capacity to undertake computer modeling of these processes.

• Then there is the debate about the ability to miniaturize a device adequately to mount it on a delivery vehicle (a question separate, obviously, from the range, accuracy and reliability of those systems, which I take up below). Albright assumes that the North Koreans are currently capable of mounting a warhead on a Nodong and that if they can there is no reason that one could not be mounted on an intermediate-range or intercontinental missile. Jeff Lewis at 38North reaches a similar conclusion but provides more detail on the reasoning. The answer to the miniaturization question hinges on whether the North Koreans can build a device that is small enough; whether it can weather ballistic missile flight; and particularly whether it can survive re-entry. His answer to each of these three questions is “yeah, probably,” but with one important caveat: the challenges rise with distance.

• The scenarios actually lump together developments that may not covary. Each scenario assumes a similar, roughly linear development (with different slopes) in each of the main components of the nuclear program, namely the generation of fissile material (from two sources) and effective weaponization. But the world does not work this way. Each of these programs is subject to different constraints and therefore uneven growth. Lagging in one weak link—including the missile program--means effectively lagging in final capability.

• The problem of intent. We need to take the byungjin line—the commitment to both economic development and the pursuit of nuclear weapons—seriously. But future developments depend heavily on the capacity to test, which may be constrained by China, as well as the resources the regime chooses to pour into the program and whether sanctions matter for any portion of it.

Which brings us to the second big debate; the question of the missile program. In a post at Arms Control Wonk in mid-February, Jeffrey Lewis offers up his count of tests since January 2014: no fewer than 21 tests of rocket artillery and missiles, with each test consisting of from one up to 30 discrete launches depending on the system. But if we set aside the artillery tests, the missile tests are entirely Scuds, Nodongs and perhaps a new version of the short-range, solid-fueled road-mobile KN-02 “Toksa.” The longer-range Musudan (intermediate range or IRBM) and KN-08 (solid-fueled intercontinental) missiles that were on parade in 2012 and 2013 have not been seen since, and neither has been flight-tested. According to John Schilling there does not appear to be much visible work on the Unha-3, which put a satellite into orbit in December 2012 and is considered by some—but not all—as a stalking horse for the country’s Taepodong-2 ICBM program.

Thus the considerable to and fro when James Clapper testified in February that North Korea had in fact taken steps to deploy the KN-08 and that it will be capable of targeting the U.S. (In 2013, the DNI was quick to cast doubt on North Korean capabilities in this regard; see the earlier statement here. Views have clearly changed). The controversy was heightened when David Stilwell, DoD’s Deputy Director for Politico-Military Affairs for Asia, linked the KN-08 threat with the deployment of the THAAD missile defense system in South Korea. A week later, Admiral William Gortney, commander of the U.S. Northern Command, was even more unambiguous stating “our assessment is that they have the ability to put a nuclear weapon on a KN-08 and shoot it at the homeland.” Again, this is a shift and at odds with a February assessment by the ROK Ministry of Defense. A report on North Korea's nuclear roadmap by Peter Hayes and Roger Cavazos for Nautilus brought the following assessment to our attention from last October which probably gets it about right. General Curtis Scaparotti, head of US Forces Korea, stated that North Korea did probably have the technology to make a small nuclear warhead and put it on a missile. But he admitted he did not know if they had done so and that “something that’s that complex, without it being tested, the probability of it being effective is pretty darn low.”

The military obviously has intelligence that the open-source community does not. But John Schilling and Henry Kan offer up a complementary report to Albright’s on delivery systems and are more cautious; the missile estimates in the diagram are from their report, which highlights a range of likely constraints. At the core of their argument is the necessity of testing. Neither the existing IRBM’s or ICBM’s have been tested at all (although first-stage engines probably have) and the Unha “space launch vehicle” has been launched four times, but successfully only once. Without testing, it is hard to improve range, accuracy and reliability. While the country’s Nodong’s and KN-08’s enjoy the benefits of mobility—and thus survivability—the Unha-Taepodong-2 effort has relied on the Sohae test site--where there was tremendous building over the last year according to another report at 38North by Jack Liu and Nick Hansen--but which remains unhardened. And all of this is in addition to the uncertainty surrounding miniaturization noted above, and particularly with respect to ICBMs.

The main points I take away from these recent reports are the following:

• Although fears of an imminent test have abated, the North Koreans appear serious about the byungjin line of simultaneously pursuing both nuclear weapons (and missile) capabilities and economic growth.

• From all we can gather about activity at the Yongbyon site, both the plutonium and HEU paths to fissile material are actively in play and will continue to expand North Korea’s stockpile of fissile material.

• We know very little about weapons production, but the North Koreans do not even bother to tout civilian uses in the same way the Iranians do (#byungjin). 20 additional weapons to the few they probably now have seems a plausible lower bound over the next five years.

• Whether they have mastered the ability to mount a nuclear warhead on a missile—and of what range—remains uncertain but it is safe to assume they probably can, at least at shorter ranges.

• The current missile inventory gives North Korea pretty good coverage of the Northeast Asian theatre, including Japan.

• The longer-run capability, extending to US forces elsewhere in the Pacific such as Guam or to the homeland, is much less certain and subject to constraints. Moreover, the uncertainties with respect to their ability to mount a warhead on a missile at that range are higher. The bad news is that KN-08 has the advantage over the Unha/Taepodong of higher survivability, making them more difficult to take out during a crisis than anything sitting on a launchpad.

What is surprising given these capabilities—however uncertain—is the lack of urgency when compared to Iran. One reason is the lingering doubts about the utility of reconvening the negotiations. Views among the Six Parties on the preconditions for restarting the Six Party Talks have narrowed in recent months, but North Korea remains defiantly uninterested. The most obvious reason for the priority given to Iran is that the Korean peninsula and Northeast Asia look a lot more stable than the Middle East--despite the developments just outlined--with fewer risks of major surprises.

Figure 1. Delivery system projections: three nuclear forces for 2020

Source: Directly reproduced from Schilling and Kan (2015), Figure 15.