Errors in Analysis of the Hwasong-18

Errors in Postol’s Analysis of the Hwasong-18

Daniel Allen, Madeline Berzak, Michael Duitsman, Decker Eveleth, John Ford, Sam Lair, Jeffrey Lewis, Tricia White

Key points:

• Ted Postol’s The Transfer of a Russian ICBM to North Korea? is marred by many errors. Most of these errors are careless.  Correcting them, in some cases, actually may make his argument stronger. Other errors, when corrected, demonstrate that Postol’s conclusions are likely wrong.

• Postol places the stage separation points in the wrong places. The first stage and third stages are, in fact, longer than those of the Topol-M.
  • Smoke obscured the bottom of the missile in the single image Postol examined, leading him to incorrectly measure only the portion of the missile he could see. 
  • Postol fails to notice that the length of the full third stage–with its distinct checkerboard pattern–can be estimated from images after separation.

• Postol also misidentifies the spent second stage as a decoy canister, despite the fact that it is labeled as “second stage separation” in Korean.

• Postol does not consider the evidence that the third stage may use liquid propellant.

• Postol does not consider the evidence that the Hwasong-18 guidance system uses a different process to align the missile than the Topol-M.

🚀🚀🚀

Introduction

Ted Postol’s analysis of the Hwasong-18 is marred by significant factual inaccuracies. Many of these inaccuracies would have been caught by an analyst carefully looking at the totality of the evidence, rather than simply assuming the missiles are the same. Some of the inaccuracies, if corrected, would strengthen his argument. Other inaccuracies, when corrected, clearly show the missiles are different and, therefore, exclude Postol’s preferred explanation.

In general, Postol works backwards to create a model that fits his assumptions. All analysts modeling missiles must start with some assumptions to explain the performance shown by the missile in flight. But these assumptions must be examined in light of all the evidence as the analyst considers subsequent iterations of the model.  Many different models of the Hwasong-18 are possible that replicate the missile’s two test flights in April and June.  Some of these models of the Hwasong-18 are decidedly unlike the Topol-M, fit the available visual evidence better, and replicate the observed performance in past flight tests. This should be a serious caution about drawing extreme policy conclusions from any one model.

Postol claims that Russia made a complete “transfer of the Topol-M missile or its technology” to North Korea – not merely that some Russian entities may have provided assistance to North Korea. Postol repeatedly emphasizes what an extreme claim this is – calling such a transfer “far beyond previously forbidden political boundaries” and “a radical departure from past political practice by Russia,” “singular” and “a matter of greatest concern.” Extreme claims come with a heavy evidentiary burden.

North Korea’s Hwasong-18 ICBM does, at first glance, resemble Russia’s solid-propellant ICBMs like the Topol-M and the Yars. Those Russian ICBMs are a logical point of comparison for any analysis of the Hwasong-18. The Hwasong-18 clearly takes some design inspiration from Russian missiles, in this case Topol-M and Yars, just as many other North Korean missiles take inspiration from Russian designs. We do not rule out the possibility that Russian entities may have assisted North Korea in its development of this system. (There is considerable evidence that Russia and the former Soviet space in general, is a significant source of design information for North Korea, as a recent Reuters story about North Korean efforts to hack a Russian missile producer shows.)

A close examination of images, videos and the performance of the missile, however, shows clear differences that exclude the possibility that Russia transferred a complete ICBM system. The biggest differences relate to the size of the missile stages, as well as the missile’s third stage design and guidance system. 

🚀🚀🚀

Many of Postol’s errors are careless.  Correcting them, in some cases, actually may make his argument stronger. Other errors, when corrected, demonstrate that Postol’s conclusions are wrong.

Postol is wrong to refer to the “sudden appearance” of a solid-propellant ICBM in the DPRK. The DPRK has been publicly signaling its development of a solid-propellant ICBM since 2017. Based on the early US and Soviet experiences, the time between testing a large (more than 800 mm) diameter solid rocket motor and a solid-propellant ICBM is about three years.  North Korea had tested such a motor by early 2017 – six years ago. In 2020, several outside experts, including Jeffrey Lewis, provided an American Physical Society committee with evidence that “North Korea could test a solid-propellant ICBM relatively soon.” There is nothing sudden or surprising about North Korea’s continued development of large solid propellant rocket motors.

Postol makes many factual errors about Russian missiles, some of which if corrected might strengthen his conclusion.  For example, Postol appears not to know the difference between the Topol-M and the Yars. The image on page 2 is labeled “Topol M,” but the caption says “SS-27 Mod 2” which is the US designation for a similar but different missile, the RS-24 Yars. Postol misstates the warhead loading of the Topol-M. The RT-2PM2 Topol-M (SS-27 Mod 1) carries a single warhead. The RS-24 Yars (SS-27 Mod 2), however, is armed with multiple warheads. The fact that Postol confuses the two systems makes it difficult to understand which system he thinks Russia transferred to North Korea.  The Topol-M and Yars are difficult, if not impossible, to distinguish externally. We refer only to the Topol-M when discussing dimensions or other details for the sake of simplicity.

Postol also misstates the diameter of the Topol-M.  The diameter of the first stage of the Topol-M, which Russia declared under the START Treaty, is around 1.9 m – not 2.2 m.  If the Hwasong-18 is 2.2 m as Postol asserts, it cannot be the same rocket motor. The Hwasong-18 is probably not 2.2 m in diameter, however, and had Postol measured it, the result may have actually strengthened his argument. Measuring objects in photographs and videos is extremely difficult and there is always room for error. In this case, however, Postol seems not to have made any measurements at all.

Instead, he relies on another analyst’s measurement of a solid rocket motor tested on the ground in December 2022 to infer the diameter of the Hwasong-18 ICBM. (Postol refers to “films” of the ground test, but North Korea released only still images and not video suggesting that Postol did not do this analysis himself.) Postol then asserts that this motor, which North Korea claimed produces “140 tons” of thrust, is the same as the first stage of the Hwasong-18 ICBM.

The images from the December 2022 ground test are extremely difficult to measure. Our estimate of the motor tested in December 2022 and the Hwasong-18 first stage is that both are about 1.9 m in diameter – similar to the Topol-M. Had Postol known the correct diameter of the Topol-M and had he measured the Hwasong-18 and rocket motor tested in December, it may have strengthened his argument.  

However, Postol might also have noticed clear differences among the three relating to their length. The rocket motor tested in December 2022 appears to have a different length-to-width ratio than the Hwasong-18 – the December 2022 motor is much shorter than the Hwasong-18 first stage. (Again, the images of the December test do not produce high-confidence measurements so caution is in order here.) The two motors may be related — North Korea may have tested a partial or segmented Hwasong-18 motor – but is unlikely to have been a full Hwasong-18 first stage.

Postol’s estimate of the acceleration of the Hwasong-18 in the video is substantially different from ours. He claims the acceleration of the missile is consistent with the first stage motor producing 140 tons-force of thrust. We measured the acceleration of the missile at ~9.3 m/s², which implies a thrust of about 100 tf (1000 kilonewtons) for a 50-ton missile. (There is a much more complicated question of how the initial thrust during the first few seconds of burning relates to both the average and maximum thrust of the motor.) 

As in the case of the diameter, our estimate of thrust is more consistent with the performance of the Topol-M, which also has a reported thrust of around 100 tf. Again, a more careful measurement may have strengthened Postol’s argument. 

We do agree the first stage boosters are similar in performance, although this is unremarkable. Many missiles have a thrust to weight ratio at takeoff of about 2:1.  The missiles are similar in overall size; it is no surprise their thrust is similar as well. 

🚀🚀🚀

Postol places the stage separation points in the wrong places. The first stage and third stages are, in fact, longer than those of the Topol-M.

We used still and video images to assess where to place the separation planes between each stage of the missile.  Correctly placing the stage separation planes helps show that each of the three stages of the Hwasong-18 is probably different in length that their counterparts on the Topol-M. 

In general, we find that Postol misidentifies most of the stage separation planes to make the Hwasong-18 seem much more like a Topol-M than it really is.  The two most glaring mistakes are failing to measure the full length of the first stage and not noticing that the length of the third stage could be estimated from images released by North Korea.

Smoke obscured the bottom of the missile in the single image Postol examined, leading him to incorrectly measure only the portion of the missile he could see. 

The easiest error to explain is that the first stage of the Hwasong-18 appears to be longer than the Topol-M. Postol attempted to estimate the length of the first stage with a single image, but did not realize the bottom of the motor was obscured by smoke. (He stopped measuring at the line where the skirt begins, even though it is clear the rocket exhaust does not start until further down.) Once the additional length is added to the Hwasong-18, the similarity disappears between the Hwasong-18 and Topol-M first stages. The motors cannot be the same if they do not have the same dimensions.

The length of the full third stage–with its distinct checkerboard pattern–can be estimated from images after separation.

Postol assumes the third stage of the Hwasong-18 is identical to the Topol-M, but an examination of the photographs and videos suggests this is probably not the case. By counting the squares of the checkerboard paint scheme, it is possible to determine the length of the third stage. It is longer than the third stage of the Topol-M.

🚀🚀🚀

Postol also misidentifies the spent second stage as a decoy canister, despite the fact that it is labeled as “second stage separation” in Korean.

Postol also misidentifies a spent rocket stage as a “decoy canister.” The object is not a decoy canister, but rather the second stage falling away. The Korean text on the picture translates as “2nd Stage Separation.” In the video itself, one can see the stage detach and separate from the rocket. If North Korea chooses to display missile defense countermeasures, our judgement is that they will likely do so directly.

🚀🚀🚀

The third stage may use liquid propellant.

As noted above, photographs and videos suggests the third stage is probably longer than that of the Topol-M. But there is another difference from the Topol-M visible on the airframe.

The Topol-M third stage is a solid-propellant rocket motor with a carbon fiber airframe. Visual examination of the Hwasong-18, however, suggests the third stage is a liquid propellant rocket stage with a metal airframe. The surface of the Hwasong-18 third stage appears to show weld lines that correspond to internal tanks holding liquid propellants, although the images are poor.  (The third stage of the Hwasong-18 displayed at the defense exhibition for the Russian Defense Minister does not show such weld lines, possibly indicating a desire to develop a solid-propellant third stage.)

If the third stage is liquid-fueled, rather than solid like Topol-M or Yars, this would be a serious objection to Postol’s belief that Russia transferred a full ICBM. 

🚀🚀🚀

The Hwasong-18 guidance system may use a different process to align the missile than the Topol-M.

Postol hints the Hwasong-18 and Topol-M guidance systems are similar, but images and videos show important differences. An essential element in missile guidance is alignment – the missile must not only know where it is, but in which direction it points. The initial alignment of the missile can be accomplished in different ways. Russia uses an extremely accurate gyrocompass attached to the transporter-erector-launcher to determine the heading of the missile prior to launch.  The Hwasong-18 launch vehicle does not have such a feature, which should be prominently visible.

An alternative method, used by China for its modern mobile ICBMs like the DF-31, is to place an gyrocompass inside a vehicle which then communicates directional information to the missile’s onboard guidance system by shining a laser at a mirror mounted to the side of the missile while the missile is in an erected position. To speed this process, China constructs distinctive pre-surveyed launch sites that are identifiable by a 25 m radius circular ring road for the alignment vehicles.  North Korea used what appears to be a lightly camouflaged, semi-circular launch site for the two Hwasong-18 tests that is a close match to the pre-surveyed sites China uses for its mobile ICBMs. The lack of a gyrocompass attached to the launcher suggests North Korea does not have access to the Topol-M guidance system, while the launch site implies North Korea may be moving toward the using the Chinese method of alignment in operational contexts.

🚀🚀🚀

Modeling the Hwasong-18

Any missile model must be constrained by the performance of the missile in actual tests. We used the trajectories from Hwasong-18 tests in April and July 2023, including the stage drop points announced by North Korea in April, to constrain our model of the missile. Our model was able to match the observed performance in both tests, even though it differs significantly from the Topol-M and Yars ICBMs. This should caution analysts from drawing extreme policy conclusions from any single model, whether it is ours or Postol’s.

The fundamental problem with Postol’s analysis of the Hwasong-18 is that it appears to have less to do with North Korea’s missile program than with his long-standing opposition to existing US missile defense programs. Since 1999, Postol has modeled North Korean missiles almost exclusively in this context, always to make the same point over and over again – that a regional boost-phase system would provide a better defense against North Korean missiles than existing US programs of record. Over time, Postol has moved the interceptors from ships to UAVs, but his argument has otherwise remained remarkably consistent – that North Korea has no indigenous missile industry to speak of, but through substantial foreign assistance can acquire missiles and countermeasures that will defeat the existing missile defense system in Alaska, which should be replaced by his proposal for a regional boost-phase defense. He repeats this argument in this analysis as well. 

We take no position on the relative merits of a UAV-based boost phased defenses compared to the midcourse defense in Alaska. We raise this point to explain why Postol starts from his conclusions and works backwards. That might be a successful strategy to advance one particular missile defense system over another, but it’s no way to understand the evolution of North Korea’s missile capabilities