Estimating How Many Pits DPRK Has, Part 1: Relevant Materials Production

Nearly a year ago, a report leaked from the DIA said that the DPRK had enough fissile material for 60 pits and was making enough fissile material for 12 more each year. A few days ago, the OSINT Team at MIIS had a slam dunk analysis of where DPRK has a second HEU enrichment facility and reporting from Ankit Panda confirmed the analysis. This site has been covertly been operating from maybe as far back as 2003. If that's the case, this place has been operating for 15 years producing HEU. This got me thinking, what kind of manufacturing capability does DPRK have when it comes to making the materials required for these pits?

The DIA report "assumes the use of composite pit core designs" so let's examine what composite can mean. It's an important distinction as it is an indicator that Pu-239 is scarce and as I'll lay out later, it is in very short supply for North Korea. Composite cores are made of a combination of HEU and Pu but I think with DPRK it is worth examining some other options given how low some estimates of North Korea's stockpile of Pu-239 is. I believe the DIA maybe using composite to mean a combination of fissile material with a second gas or solid used to "boost" the bomb. Below is a list of potential candidates available to North Korea for their pits;

  1. HEU/Pu composite
  2. HEU and/or Pu/3H composite
  3. HEU and/or Pu/DT(2H3H) composite
  4. HEU and/or Pu/6Li2H composite


With a list of potential candidates for their pits, let's have a look at the production capabilities for each of the listed materials.


To make HEU, you need to enrich U-235. The three main techniques are Gas or Thermal Diffusion, via Centrifuge or via Laser Separation. Diffusion is highly unlikely as both methods generate a lot of heat and would be readily visible in thermal imagery from satellites. North Korea does have a Laser Research facility that is suspected as a laser enrichment facility but little is known about it in the open source as to whether it has been successful or if it was, it's scale. This leaves North Korea with just Centrifuges which they use at both their known facilities in Yongbyon and Kangsong.


Sig Hecker is one of the few people to have ever been inside Yongbyon and he reported that they are P-2 centrifuge's and that there are 2000 of them at the complex, capable of producing 40kg of 90%.


Assuming that Kangson is using the same setup as Yongbyon, it's safe to say that the centrifuges are P-2 as well. P-2 centrifuges are similar to P-1 as seen at Natanz but the internal rotors are made of Maraging Steel instead of Aluminium. Since they are similar to the ones at Natanz we can assume that they took up 1sqm of floor space. And knowing that Kangson facility is 110m wide and 50m long, giving a total area of 5500 sqm. In that space, the maximum number of centrifuges, if the building is absolutely packed is roughly 5500. Iran wasn't using the total floor space of their facilities, only 83% of floor space, so I think we can assume that the DPRK is using a similar number so there is space for staff, control equipment, power etc, lets say 85% giving them roughly 4675sqm of centrifuges or 4675 centrifuges.


Estimating the capacity of 6 675 centrifuges at both facilities, given Sig Hecker's numbers would be 534Kg of 90% U-235 per year, at 100% capacity or 374Kg 90% U-235 at just Kangsong as all of the capacity from Yongbyon may be required to make LEU for the ELWR reactor at Yongbyon. These numbers are likely not correct as they are estimates if the facilities are running at 100% capacity and that is currently an unknown.


Pu-239 is made by breeding U-238 in nuclear reactors and then processing the fuel after unloading to separate the Pu-239 from the U-238. North Korea has access to two ways of breeding Pu-239. The first and what I believe is the less likely option is in their IRT-2000 Reactor and the second is the 5MWe Reactor, both are at Yongbyon. David Albright has a paper that estimates as of 2007, North Korea had 46 Kg to 64 Kg with 28 Kg to 50 Kg for use in weapons.

5MWe Reactor

Using the same paper from Albright, he estimates that North Korea could produce 10Kg to 13Kg every 20 months and in the 11 years since then, North Korea had the capacity to produce an astonishing 4092Kg IF there were core changes, but I haven't been able to verify core changes outside of 1994, 2005 and 2007, as listed in the paper. There MAY have been a fuel loading in 2016 or 2017 which may bring them up to as much as 56Kg to 77Kg though it's very hard to verify unless they are covertly loading and unloading fuel though there is no evidence of that.


It isn't really feasible to use the IRT-2000 to produce Pu-239 as it's "has operated only intermittently due to North Korea’s inability to obtain new fuel" since the collapse of the Soviet Union. Though it appears that North Korea has been able to manufacture it's own fuel to keep this reactor running. According to Albright's paper, it is capable of producing " at most a few hundred grams", though the U.S. Department of Energy estimates that it could have been used to produce up to 1 Kg to 2 kg of Pu-239. I find it is unlikely that the IRT-2000 is used for this purpose since it's not able to produce much, and it has two other purposes. One is in another paper by Albright and it is the production of iodine 131 for treating thyroid cancers while the second is something I will cover later, the production of 3H.


Assuming that the 5MWe reactor hasn't been loaded it is safe to assume that they have approximately 56Kg to 77Kg total Pu-239, not accounting for Pu-240 contamination. If as claimed in 2013, North Korea restarted operation of the 5MWe reactor, they would have 66Kg to 90Kg and if it was loaded again in 2016 or 2017 may have happened, they would have a total of 76Kg to 103Kg of Pu-239

A Note on the ELWR

North Korea has been building a 25 MWe to 30 MWe ELWR at the facility at Yongbyon. "While a light-water power reactor can be used to produce plutonium, it is not optimal for this purpose, so it must be assumed that the Experimental Light Water Reactor (ELWR) is intended to develop techniques and train personnel to lay the foundation for much larger reactor program ..." It is plausible that this is a peaceful use of nuclear power to generate electrical power, though it may also be used for the production of Pu-239. This is easier to monitor as LWR's have batch reloading which is harder to hide and though not optimal, "While the reactor seems designed to produce electricity for the civilian economy, it will have a residual capability to produce plutonium that can be used for nuclear weapons."


2H or Deuterium, sometimes abbreviated to D, is a stable isotope of Hydrogen used as heavy water in some nuclear reactors, like North Korea's 5MWe reactor at Yongbyon and also can be used to boost and store weapons as it is a stable gas. It is produced via distillation or electrolysis of seawater to extract naturally occurring 2H. This is also relatively easy to hide as it would just be a plant next to the sea, though lacking output. It can also be produced by with Lasers which North Korea could be looking into at their Laser Research Institute but there is no evidence of this.


Currently there is no known estimate of North Korea's production capability for 2H. Given their access to the sea however and their potential future need for 2H in a Pu production HWR reactor, it's safe to say that they have a reasonable capability that most likely is not a limiting factor for their pit designs.


6Li is a stable isotope of Lithium that naturally occurs and is around 7.5% of Lithium deposits. When a neutron splits it, it creates 4He but more importantly, 3H which can boost weapons. Information about the production of 6Li is extremely controlled information as it has very little use out of boosting fission weapons or as a neutron absorber in fusion weapons. ISIS has an article stating that mercury-based column exchange process (COLEX) and procured this around 2012. ISIS notes; "... the purchase of mercury in combination with lithium hydroxide is a strong indicator that North Korea is using the chemicals in a mercury-dependent lithium 6 production process."


Currently there is no known estimate of North Korea's production capability for 6Li. North Korea attempted to export 10Kg of 99.99% 6Li per month in 2016 and given those figures, it's safe to assume that they are processing at least 120Kg per year of 6Li and this is just for the export market and does not account for how much 6Li they may need internally for its own uses. This is a substantial amount of 6Li production and I think it puts to bed the rumour of the Lithium shortage in North Korea, especially when coupled with news reports that North Korea might have trillions of USD in untapped mineral wealth


3H is a radioactive isotope of hydrogen and crucially, it has a half-life of 12 years. There are two main ways for North Korea to access 3H, importing and breeding it in a rector. Importing is unlikely given that no producer of 3H would export it to them and China, who does produce 3H, has been cutting down on sanctioned exports. They can also 6Li2H as listed above but this is a less efficient method of for the production of 3H as it needs to be incorporated into the pit and during fission, the 6Li2H steals a neutron from the chain reaction slowing it down.


6Li is easily breed in the control channels of the IRT-2000, that run though the core and are exposed to the maximum neutron flux. To quote extensively from the above link;

Over an eight-month operating cycle, this cubic centimetre target of lithium-6 could generate about 80 milligrams of tritium. To generate three grams of tritium (an approximate amount used in modern boosted weapons), North Korea would therefore have to irradiate approximately nineteen grams of lithium-6 over an eight-month operational cycle.

A single target containing nineteen grams of lithium-6 (equivalent to a slug roughly 3cm in diameter and 5cm in length) could potentially generate enough tritium for a single nuclear weapon in one yearly operational cycle.

Assuming that only four of the IRT’s experimental channels are located within the reactors core, North Korea would be able to generate enough tritium for twenty ‘DT’ boosted nuclear weapons per year by irradiating five such slugs in each channel.


Given the data above, one 19g slug can generate 3g of 3H. Exposing 20 slugs as described above would allow for the production of 60g of 3H over the yearly operating cycle, giving North Korea a significant amount of 3H for use in pits.

Other boosting compounds


Is a solid, low density compound that is made by treating 6Li with 2H gas. The reaction can occur at as low as 29 Celsius, though it only has a yield of 60%, which may be ideal for North Korea as it hides the activity. Though more temperature and pressure do speed up the process. I can't find solid numbers for what mass of material can be processed per hour or per day, given that you can treat 6Li with 2H in 2 hours, at 600 Celsius and get a yield of 98%, it seems to be an issue of how much 6Li and 2H North Korea has and not how long it takes. All temperatures and yield are in this paper.


Being a gaseous compound, it's not difficult to manufacture when you have the raw materials and it just requires a form of sealed storage with a top up every 12 years to maintain the supply of 3H. It has been estimated that "... North Korea would be able to generate enough tritium for twenty ‘DT’ boosted nuclear weapons per year ..."

Lets just recap our total yearly production

Below are yearly production figures with the exception of Pu, that is their total stockpile of Pu.

Material Minimum Quantity Maximum Quantity
HEU 374Kg 534Kg
Pu 56Kg - 77Kg 76Kg - 103Kg
2H ??? ???
6Li 120Kg ???
3H 60g (IRT-2000 only) ???

In Part 2, I'm going to examine what designs North Korea is using for it's pits so I can better project how many pit's of what type they have in service and how many they are capable of making per year.

This post wouldn't have been possible without a great conversation that Peter started in the ACWP Community Slack as well as help from Peter, Andrew, Rethin, Chris, and Nathan. Massive thanks to you guys for the entertaining conversation, facts, fact checking and finding awesome sources!

Paddy Kerley

Paddy Kerley