Estimating North Korea’s Nuclear Stockpiles: An Interview With Siegfried Hecker
38 North recently interviewed Dr. Siegfried S. Hecker, renowned expert on North Korea’s nuclear program and a senior fellow emeritus at the Freeman Spogli Institute for International Studies at Stanford University, about the status of North Korea’s plutonium and uranium enrichment production and implications for the North’s nuclear arsenal. Below are excerpts from that interview.
38 North: Recent assessments of North Korea’s nuclear program estimate that they may have up to 90 nuclear weapons. What do you make of this estimate?
Siegfried Hecker: That’s much too high. I think 20 to 60 is possible, with the most likely number being 45. These numbers are based on estimates of how much fissile material—that is, plutonium and highly enriched uranium (HEU) bomb fuel—North Korea has produced. In other words, it may have enough fissile material for 45 nuclear weapons, but that does not necessarily mean it has produced that many at this time.
Plutonium production can be estimated quite accurately. It is produced in the North’s 5 Megawatt-electric (5 MWe) nuclear reactor, and satellite imagery allows us to monitor when the reactor is operating. Prior to 2008, the best signature was a plume emanating from the reactor’s cooling tower. The North blew up the cooling tower as a political goodwill gesture in June 2008. When the reactor was restarted in August 2013, they decided not to rebuild the cooling tower but rather go to the river for cooling—essentially creating a heat exchange mechanism using river water. Now, reactor operation is more difficult, but not impossible, to monitor.
My current estimate is that North Korea has a plutonium inventory in the range of 25 to 48 kilograms. Based on what we have learned about reactor characteristics, including from my visits to the Yongbyon nuclear complex, North Korea can produce at most six kilograms per year at full operation. My inventory estimate is based on production estimates, production losses and estimates of amounts expended in nuclear tests.
What do you think about recent reports that North Korea is reprocessing plutonium now, perhaps as a strategic political move to ratchet up pressure on the Biden administration?
SH: Once plutonium is produced during reactor operations, it must be chemically separated (that is, reprocessed) from the rest of the reactor fuel products—what is called spent fuel. Satellite imagery has shown signs of operations in the Yongbyon reprocessing facility during the past couple of months. That means plutonium is either being separated from the spent fuel generated during the last reactor run, or the North is treating the nuclear waste from the last reprocessing campaign. At this point, we don’t know which it is. The more important point, however, is that we believe the reactor has not operated since at least December 2018. Therefore, if plutonium is being separated, it is from the previous reactor run—that is, it’s old, not newly generated plutonium.
Whichever the case may be, none of this is done for political reasons. These are strictly technical decisions. The reactor has not operated for over two years because we believe they are having problems with the cooling system. If they have waited to reprocess the plutonium produced before 2018, it is because they have additional technical problems. If they can overcome the technical issues, they will surely produce more plutonium, separate more into weapons-grade bomb fuel and also make more tritium for hydrogen bombs unless Washington reaches some diplomatic agreement to prevent that.
How about monitoring uranium enrichment activity? How accurately can that be assessed?
SH: Uranium enrichment operations are very difficult to estimate because the centrifuge facility signatures are so small. The North Koreans showed our Stanford delegation—John Lewis, Robert Carlin and me—the centrifuge facility at Yongbyon during our trip in 2010. It’s the only enrichment facility they have declared. That building housed 2,000 centrifuges at the time. The size of the building was doubled by 2013, so we assume they have had 4,000 centrifuges spinning since. And, by the way, centrifuges pretty much operate 24/7.
Based on what we saw at Yongbyon and on our previous visits there, we were convinced that North Korea had at least one other centrifuge facility outside of Yongbyon. We don’t know how large it is or where it is located. I am not convinced that open-source reports suggesting such a facility exists at Kangson, just outside Pyongyang, are correct. However, the other facility (or facilities) must have been large enough to provide sufficient operational experience for the North Koreans to construct the centrifuge hall at Yongbyon by 2010.
It is difficult to estimate the total enrichment capacity because we don’t know how large the other one or possibly two facilities are. That said, our team at Stanford developed a probabilistic analysis based on the likelihood of Pyongyang’s ability to procure key centrifuge materials from outside sources or produce them domestically. We estimated that North Korea’s capacity to produce HEU for bomb fuel is on the order of 175 kg per year. Based on these assumptions, our estimate is that North Korea likely has around 600 to 950 kg of HEU as of the end of 2020.
Let me stress, however, that all estimates of uranium enrichment capacity in North Korea are highly uncertain. The best way to reduce the uncertainty is for US or International Atomic Energy Agency (IAEA) inspectors to get back into Yongbyon to see the operations. No other outsiders have seen the centrifuge facility, and no one has been there since our team’s visit in 2010. Unfortunately, the two previous US administrations have missed opportunities to get back in.
How many bombs can North Korea make with those inventories of plutonium and highly enriched uranium, and can they make hydrogen bombs?
SH: The plutonium bomb that destroyed Nagasaki in August 1945 used around six kilograms. The Hiroshima bomb used HEU, but it was of a primitive design. How much plutonium or HEU the North Koreans need for a bomb depends on how good their scientists are and what kind of bomb they want to build. A reasonable estimate is five kilograms for plutonium bombs and 25 kilograms for HEU bombs. Using the plutonium and HEU inventories I mentioned leads me to believe the most likely number of bombs is 45. The recent estimates in a RAND/Asan Institute report of 67 to 116 today and 151 to 242 by 2027 are much too high. They estimate that North Korea has the capacity to add 12 to 18 bombs per year; ours is closer to six.
As for hydrogen bombs, these need fusion fuels, namely the heavy hydrogen isotopes deuterium and tritium. Deuterium is easy to produce. Tritium has to be produced in reactors. Looking at the North’s reactor operations over the years, I believe they have produced small amounts of tritium, perhaps enough for a few hydrogen bombs. The real question, of course, is, do they know how to design and build a hydrogen bomb? We are not certain, but the sixth nuclear test was large enough to have been a hydrogen bomb. It likely used a plutonium fission device to drive the fusion reaction. Since the production of plutonium and tritium requires reactors, it is very important to stop reactor operations in Yongbyon permanently.
You mentioned that North Korea likely has at least one enrichment facility outside of Yongbyon. Does your estimate include that potential production capability?
SH: Yes, our probabilistic model estimates North Korea’s total enrichment capacity regardless of where it’s located. The facilities outside of Yongbyon could possibly produce HEU by themselves or be operated in tandem with those in Yongbyon. I believe it is quite likely that the Yongbyon centrifuge facility produces low-enriched uranium (LEU) at 3.5 percent uranium-235, as the North Koreans told me—and they can make a lot of that. The LEU would then be sent to another facility to step up the uranium-235 content in stages: first to 20 percent, then 60 percent, and finally to the 90 percent HEU level.
It would make sense for the North Koreans to structure their enrichment program like that—use the plant at Yongbyon to make LEU and then send it off someplace else to make HEU. It seems likely then that they have at least one, maybe two other enrichment facilities, but they aren’t all doing the same thing. This makes it difficult to estimate how much of their enrichment capacity is outside Yongbyon. My guess is that it’s roughly half.
Let me stress again, however, that all of the plutonium and tritium production capacity is at Yongbyon.
Does the country’s enriched uranium all go towards its weapons program? Would there be other potential uses of the LEU produced at Yongbyon, such as for fuel fabrication for the nuclear reactors?
SH: Yes, indeed. The Experimental Light Water Reactor (ELWR) under construction requires LEU fuel in contrast to the natural uranium used in the 5 MWe Reactor. In fact, Pyongyang’s first admission of having a centrifuge program (which was not until 2009) was that they needed it to produce fuel for their new ELWR. The initial 2,000-centrifuge plant they showed us was capable of producing enough LEU to fuel the ELWR for continuous operation. I was told it would require four tons of LEU uranium oxide fuel. The initial centrifuge plant could have produced that much LEU in two years. The expanded plant can do it in one year. They have had plenty of time to produce sufficient LEU fuel for the reactor since it is taking much longer than anticipated to start up.
Speaking of the ELWR, do you think the North Koreans will ever get that running? Assuming they do, wouldn’t that mean less LEU for the weapons program? And, is it possible to make plutonium in the ELWR?
SH: I believe the North Koreans are determined to get the ELWR running. We have seen continued activity at the site, but they have yet to start operations. We have to keep in mind that light water reactors (LWRs) are a new technology for them. Almost every aspect of LWR operations and the safety requirements are more demanding than those for the gas-graphite reactor with which they have experience. They continue to struggle with the cooling systems for both reactors, especially given the unreliable water supply from the Kuryong River, which runs alongside the Yongbyon complex. In addition, both the fuel and the cladding for an LWR are different from that used in their gas-graphite reactor. It could be a whole host of issues that are delaying start-up, but they are a very determined people and will eventually get there.
As to how reactor requirements for LEU will affect the amount of enriched uranium available for the weapons program, I think they’ll be able to manage both. One of the unfortunate aspects of the nuclear business is that it takes much less uranium to build bombs than it does to fuel a reactor to make electricity. Until they build much bigger LWRs, they have most of their centrifuge capacity available to produce HEU bomb fuel.
I believe North Korea was serious about nuclear electricity. The ELWR was going to be their pilot project for larger LWRs for power generation. Light water reactors do produce plutonium as well, but it is less desirable for bomb production when operated in an electricity-production mode. However, the ELWR could be operated in a way that produces bomb-grade plutonium. It gives the North Koreans a good backup to the 5 MWe Reactor for plutonium production. If operated in such a mode, it could roughly double the plutonium production of the 5 MWe Reactor.
There has been a lot of talk in recent years that downplays the value of Yongbyon—saying it’s old and the facilities are obsolete. But from what you’re saying, it seems that it still plays a significant and critical role in North Korea’s ability to produce fissile material, even if it’s not the entirety of the fissile material program. Is that correct?
SH: Yes, this is what people need to understand. You shouldn’t just write off Yongbyon as old and obsolete. It isn’t. That’s quite apparent if you look at the amount of construction that’s occurred at Yongbyon since 2009 when the IAEA inspectors were expelled. They built a new centrifuge facility (the one they showed us), then doubled its size. They are building a new reactor. They built a new facility to make uranium hexafluoride needed for enrichment operations. They have built new fuel fabrication facilities and what appear to be new tritium separation facilities and more.
The 5 MWe Reactor has been operational since 1986. When I asked the director of Yongbyon how long he thought they could operate it, he replied for decades longer. They have recently experienced cooling problems with the reactor, and that is why it has likely not been operational since late 2018. The plutonium reprocessing facility has been operational since around 1990. It appears to be operational right now.
I can’t help but chuckle when people say Yongbyon is old. The TA-55 Plutonium Facility at Los Alamos National Laboratory, which has produced the plutonium cores for the American stockpile the past couple of decades, was built in 1978. So, most of Yongbyon is newer than much of the Los Alamos lab.
The nuclear weapons and the missiles are not in Yongbyon. And there are nuclear weapon facilities outside the complex as well. The bombs are produced elsewhere and stored elsewhere. But Yongbyon is the heart of North Korea’s fissile materials production complex. All of its plutonium and its tritium have been and will continue to be produced there. It houses most of the chemical facilities, such as those that convert yellowcake from the mining complex to uranium hexafluoride, for uranium enrichment and around half of its centrifuge capacity. Whereas HEU could still be produced if Yongbyon is shut down, its production would be greatly curtailed.
In January, North Korea listed the development of tactical nuclear weapons as one of its WMD objectives for the coming years. Given your estimates of the country’s fissile material stockpile, how realistic is the goal of building tactical nuclear weapons?
SH: One problem in assessing this is that there is no single definition of what constitutes a tactical nuclear weapon. If one defines “tactical” as shorter-range and “strategic” as longer-range, then North Korea already has short-range tactical nuclear weapons. I believe that with their extensive record of short-range missile testing and their nuclear test history, they are already capable of putting nuclear warheads on short-range SCUD and medium-range Nodong ballistic missiles. In other words, they can already reach all of South Korea and most of Japan with nuclear-tipped missiles. These warheads are likely fueled with HEU. And, the North already has plenty of HEU for several dozen short-range tactical nuclear weapons.
However, some people use different measures to define tactical nuclear weapons—such as their purpose or yield. For example, low-yield nuclear weapons are often considered tactical. My greatest concern is that the North will develop tactical nuclear weapons for the purpose of battlefield use. That’s a direction Pakistan has taken to combat India’s great conventional military superiority. These could be artillery-fired nuclear projectiles and nuclear landmines. Do they have enough fissile material for battlefield nuclear weapons? Of course. Could they make a battlefield nuclear weapon? Yes, I believe they could.
However, these types of weapons raise an additional set of concerns. The first is safety. It is questionable that the North Koreans are able to engineer nuclear battlefield weapons that are “one-point safe”—that is, they detonate only by design rather than be subject to accidental detonation. Early US bombs were not one-point safe. It required sophisticated science and engineering to get there.
The second concern is security. Nuclear battlefield weapons require a pre-delegation of launch authority to commanders in the field. For example, Pakistan may pre-delegate such authority to the field in case the Indian military had crossed its border, and the only way to stop the invasion would be to blow up a battlefield nuclear weapon. It’s not hard to imagine the North Koreans drawing up similar plans. At the same time, tactical nuclear weapons are mostly dangerous to the North Koreans themselves, because they can accidentally blow up or be diverted in some fashion, which can lead to domestic disasters.