North Korea has reportedly tried and failed to launch a Musudan missile for a third time in two months. It is not surprising that a new missile would fail on its first test, but previous North Korean practice has been to stand down for several months to a year before another attempt. Repeating a failed test again and again with no more than a month for analysis and troubleshooting will almost guarantee repeated failure. One of the tests apparently involved two simultaneous launches, and launching two copies of an unproven design just meant a double failure while learning nothing new. Whether this unrealistic tempo is driven by impatience or desperation, it may mark the end of the Musudan program—whose military utility is in any case increasingly questionable as North Korea’s other programs advance.
The South Korean Joint Chiefs of Staff (JCS) reported on May 30 that North Korea had attempted and failed to launch a Musudan intermediate-range ballistic missile (IRBM) from a mobile launcher. This follows similar reported failures on April 15 and April 28, with the latter possibly involving two missiles. We have no confirmation that these were Musudan missiles. Still, that is the only possibility that really makes sense here—the only other missiles the North Koreans could launch from a mobile platform are the proven Scud, Nodong and Toksa designs. Those usually work, and there would be little value in a test or demonstration. The Musudan desperately needs a successful test and such a success would have made for good propaganda.
A little background: The Musudan ballistic missile was seen in North Korea in 2003, but until this year had probably never been flown. The missile appears to be based on an early Soviet submarine-launched ballistic missile (SLBM) called the R-27 “Zyb” or the SS-N-6 “Serb” in the West that was first produced in the mid-1960s and retired by the late 1980s. Some of the missile’s designers are known to have made their way to North Korea in the immediate post-Cold War era and incomplete records suggest that some of the surplus hardware may have wound up there as well. But the subsequent development of the Musudan has been extremely slow, to the point that some people had suggested it might be little more than a hoax or a bluff.
The last thing a poker player will do in a bluff is reveal his cards when he doesn’t have to. The North Koreans didn’t have to do any of these tests so they’re clearly not bluffing. At least up until yesterday, they thought they had a working missile. Maybe they still do.
We know the North has a working engine. North Korean television released images of an engine test in early April that we assessed as a pair of Isayev 4D10 engines from the R-27 missile coupled to form an ICBM-class power plant. As these were only still images, we couldn’t be sure the test had been fully successful. But satellite images show no indication that the test stand had been damaged by an explosion, and if this ground test had failed in some other respect then presumably the North Koreans would not have been so eager to launch a missile using the same engine.
It would seem that a missile with a proven design and a proven engine simply ought to work. And maybe the reason the North Koreans hadn’t tested the system before now is that they were so confident it would work that they didn’t need to test it. So much for that theory. Engines that work on the ground don’t always work in flight, as the missile’s acceleration can produce instabilities in the propellant feed. The structure could collapse under flight loads. There are always surprises the first time a new system is flown no matter how much ground testing is done.
And working from old Soviet designs, even using old Soviet hardware, is no guarantee of success. The design has been visibly modified, at a minimum stretching the propellant tanks by 2-3 meters for additional range. That might not have been done right. Even the parts of the design that haven’t changed will have been based on the implicit assumption of Russian parts, Russian raw materials and Russian technicians reading blueprints to Russian conventions. This author has seen rocket engines fail simply because the manufacturing site was moved from England to Ireland—how many more opportunities for failure must there be between Russia and North Korea? As for any tested, proven Russian hardware the North Koreans might have obtained, such hardware will by now be at least 40 years old and perhaps not properly cared for given all that time. In short, there is no substitute for flight-testing to see if an engine still works.
Or more accurately, to find out why it doesn’t work and fix it. North Korea’s rockets almost never work on the first try. The first Nodong apparently blew up on the pad just like the most recent Musudan. The Taepodong 1 and 2 rockets, the Unha, and now the KN-11 SLBM, all failed on their first flight. To be fair, this is literally rocket science. It isn’t supposed to be easy, and even NASA failed on its first satellite launch attempt. Like NASA, North Korea has shown the ability to persevere, fix the problems, and eventually succeed.
The North Koreans have also, in the past, shown patience. This is critical. When a rocket fails, it takes at a minimum several months to figure out why it failed and fix it. Trying to repeat a failed test without taking that time almost guarantees additional failures. NASA understands this, but when the first Vanguard satellite launch attempt failed at a time of intense political pressure, it attempted a second launch a little more than a month later, which also failed. Ultimately, six launches were conducted at roughly one-month intervals, with only a single success. Modern space programs, facing less pressure, will usually stand down for a year or more after a failed test. Just a few weeks before the North Korean failures, an American Atlas rocket suffered a minor anomaly that forced an early engine shutdown. The mission was successfully completed but further Atlas launches have been postponed to give the engineers at least three months for troubleshooting.
The North Koreans have traditionally followed this approach. Yet after the April 15 failure they conducted a second test less than two weeks later, possibly involving two simultaneous launches. With embarrassing failure the almost certain outcome, why was the attempt made? At the time, we had speculated that the upcoming Party Congress was putting extreme pressure on the schedule, but the Congress is over and done with, and now we have yet another test that was probably doomed from the start. There has been speculation that the test was intended to send a message to China, as a senior North Korean official visited Beijing and met with the Chinese leader. But what kind of message was intended?
Another possibility is that this is driven by pressure from the top. While the test failures were not highly publicized, they were probably an embarrassment to a Kim Jong Un who was expecting success. He may still be impatiently demanding that success. If true, this does not bode well for the engineers on the project. If they can’t convince the Party leadership, even now that the Party Congress is done, to back off and give them the time they need to find and solve the Musudan’s problems, they will keep failing. How many failures will Kim tolerate, and what happens when he runs out of patience?
The other possibility is that the Musudan, and the people behind it, may be struggling for relevance. The Musudan is a missile with a single mission—to deliver nuclear or perhaps chemical weapons to the island of Guam. It could also reach targets in Japan or South Korea, but the DPRK’s more reliable Nodong and Scud missiles can do that as well. Guam is certainly an important target for North Korea—the only sovereign US territory it can reach without developing a true ICBM, and a critical logistical base for any US operations against North Korea. But with limited economic resources and with several ambitious new rocket and missile programs, is it worth the cost to develop a troublesome system devoted to a single target? Maybe the senior leadership has decided to leave Guam to the more capable and versatile KN-08 and KN-14 missiles now under development, or the GORAE-class ballistic missile submarine, and the Musudan engineers are desperately trying to prove their own worth.
In any case, these failures may mark the end of the Musudan program. If there are continued test attempts at the current rate, whether driven by impatience from above or desperation at the bottom, there will likely be continued failures. Eventually, patience and resources will run out, and a team of North Korean rocket engineers will find themselves unemployed. If, instead, there is a real commitment to make the Musudan work, we would expect to see anywhere from three months to a year or more of ground testing before any further launch attempts. Only time will tell.
 To be fair, the Irish engineers got the system working properly – largely because they took the time to do it right.