Joshua PollackWhy Does North Korea Have a Gas-Graphite Reactor?

A close-up view of the decommissioned Tokai Power Station — Japan’s first power reactor

The usual answer to that question is the one that the North Koreans have supplied: gas-cooled, graphite-moderated reactors can be operated without enriched uranium or heavy water, and are therefore ideal for a country with limited or no ability to get these things. They’re ideal “starter reactors.”

(The North Koreans originally said they were intent on producing electricity, of course, not plutonium.)

Thus, the DPRK set out to build a trio of gas-graphite reactors at Yongbyon and Taechon in the 1980s. Of these, only the 5 MWe reactor at Yongbyon — usually described as an indigenous copy of the UK’s Calder Hall design — was completed.

But maybe there’s another answer as well.

One of the more striking things about the Yongbyon reactor is how little it actually resembles a Calder Hall, at least on the outside. The basic technologies are the same, but the buildings themselves are not. Let’s start with the overhead views from Google Earth.

Here’s one of the surviving Calder Hall structures at the Sellafield site in the UK. It’s roughly cross-shaped, about 88 m at its longest. There are four scaffold-like structures at the corners. Update: These are the heat exchangers. The tall central structure flanked by two stacks holds the reactor vessel. Its footprint is about 20 × 40 m.

And here, of course, is the Yongbyon reactor. It’s a squarish, multi-tiered structure, about 53 × 60 m. It features a tall central structure measuring about 24 × 33 m and a single stack, offset to one side.

And now for something you probably haven’t seen before.

This is the Tokai Power Station in Ibaraki Prefecture, Japan. It was the country’s first power reactor, operational from 1965 to 1998. It’s a modified Calder Hall, designed and built by General Electric Corp. (UK) to suit Japan’s seismic conditions. Until Yongbyon was built, in fact, Tokai was the only gas-graphite power reactor outside of Europe.

It’s a multi-tiered structure, about 51 × 67 m. It features a tall central structure measuring about 26 × 47 m and a single stack, offset to one side.

Update: All of the above images are oriented with the north at the top. I don’t know what to read, if anything, into the apparently identical orientation of the Tokai and Yongbyon reactors.

Obviously, this reactor, which had 166 MW electrical output, is no dead ringer for Yongbyon. But it’s much closer than the original Calder Hall type. Have a look at this ground truth comparison, with views of Tokai on the top, Yongbyon on the bottom left, and Calder Hall on the bottom right:

Even the park-like settings are similar.

Construction at Tokai began in 1960. Japan’s acquisition of a reactor suitable for making both electricity and plutonium, just as North Korea’s own nuclear research program was getting underway, would not have gone unnoticed in Pyongyang. So it would not be surprising if Yongbyon was modeled on Tokai, even though it was built 20 years later.

For comparison, recall the striking resemblance between North Korea’s first artificial satellite, Kwangmyongsong-1, allegedly launched in 1998, and China’s Dongfanghong-1, actually launched in 1970.

Think of it as a statement of sorts, a declaration of national equality.

Update. I plan to add some further discussion of North Korea’s motives for building reactors in the 1980s. Watch this space.


  1. Josh (History)

    Here’s the .kmz file of the three reactors shown above.

  2. Yale Simkin (History)

    Josh wrote:
    “Think of it as a statement of sorts, a declaration of national equality.”

    I think you may be drawing to much from the resemblance.

    The reason for the similarity is that they made a copy of the type of reactor they had access to that was good for bombs and that they were technically capable of.

    DPRK certainly had their people in Japan, and it would not surprise me if they did not also have agents in with the construction and operation.

    The only other Magnox outside of UK is in Italy.

    It would be surprising if the reactors and buildings were not virtual clones (altho I do not know if Tokai 1 was a first generation Calder-hall-type Magnox, or a later version Magnox).

    In any event, if you box in the heat exchangers, (which later UK Magnoxs did)
    they would look almost the same.

  3. Josh (History)


    Yes, it should go without saying that the North Koreans built the reactor they were capable of building. The interesting question is whether they had outside help — most obviously from the Soviet Union — stole the technology from others (e.g., the UK or Japan), independently replicated it, or some combination of these.

    And yes, the North Koreans have long had their people in Japan, most notably in the form of the Chongryon.

    Just for completeness, we should also note that France built a fleet of gas-graphite reactors, based on a somewhat different technology than Magnox, and exported one to Spain. The Soviet RBMK reactors — including the notorious Chernobyl No. 4 — are also graphite-moderated (but not gas-cooled).

    I admit that I haven’t surveyed all the Magnox (or similar) reactors out there, and I’m not sure which ones are in the picture you’ve posted in the comment above. But it only tends to reinforce my view that Yongbyon looks more like Tokai than anything else I’ve seen.

    What I’m trying to get at most of all, I guess, is the question of inspiration.

  4. Yale Simkin (History)

    I have to check, but I think the picture is Bradwell.

    Dungeness A looks even more Tokai/Yongbyonesque.

    The image was to show that the exchangers can be enclosed, giving a triple box appearance.

    Josh, I don’t disagree in any way that Yongbyon resembles Tokai. The point was that it was a natural local design for DPRK to follow (copy, steal, whatever), but not necessarily, as you seem to be saying, a psychological/political expression, a “statement of sorts, a declaration of national equality”.

    It might be, but it just seems to me that Occam’s Razor slices it down to them copying something that was known and accessable, and in their neck of the woods.

    That being the case, the fact that the Syrian reactor appears to be a Mini-Me of Yongbyon, which appears to be a Tokai clone, points a rather incriminating arrow at North Korean involvement.

  5. Josh (History)


    I’ve now gone to the trouble of locating every Magnox reactor I could in Google Earth. (Chapelcross was in a bad patch and is not included.) Here’s the .kmz file.

    I think you’ll agree that each Magnox reactor (or pair of reactors, as is typically the case in the UK) has a unique look, although there are some common factors.

    I’m not sure that the identification of Tokai as playing a role in the story of Yongbyon (specifics unknown) tells us anything we didn’t already know about al-Kibar, whose (undisguised) proportions were closer to those of Yongbyon than any other facility, Tokai included. But all this peering at Magnoxes leads me to notice that all the others have something al-Kibar did not: a turbine hall. Evidently, it was a purpose-built plutonium factory, without so much as a turbine to spin.

  6. Yale Simkin (History)

    I’ll give it one more try…

    1) I do not disagree that Tokai is the baby-daddy for Yongbyon. The only thing I took (mild) issue with was your emphasis that DPRK cloned at least the external structure for anything beyond merely technical reasons. They may have done it (also) for psychological/political reasons, but I see that as an unnecessary stretch.

    2) My pointing out UK Magnoxs (particularly Dungeness A.)
    was to show how Magnoxs look with enclosed heat exchangers. Again I do think that Yongbyon is a Tokai clone, and argue why it should be.

    3) I think that Tokai does play a role in telling us about al-Kibar. If as you discovered, DPRK created an at least superficial clone of Tokai – which is unique in specific detail – and if al-Kibar appears to be an at least superficial clone of this unique Tokai/Yongbyon structure, then… it is logical (but not bulletproof) to assume that DPRK had a hand in the al-Kibar design (I rule out Japan).

  7. Eric Rauch (History)

    My master’s research was based on the fuel cycle in Yongbyon, and I think theres more here than just looks. Due to the nuclear properties of carbon, graphite moderated and gas cooled reactors will all share some common aspects such as fuel pin pitch, power density, and plutonium production rate per mass of fuel. These things do not change with respect to total reactor power (to achieve greater power the size of the reactor is increased).

    So, all of the work needed to design a graphite moderated and gas cooled reactor was done for Calder Hall and most of the relevant information was released for anyone to see. You could say that there was obviously outside help, because the problem had already been solved. But did the North Koreans have a design given to them from the USSR or steal a design from Japan? I would say no for 2 reasons. The first being that the Calder Hall design was already available and the second that their problems getting the reactor to operate reliable indicates they were learning the operating ranges the reactor was capable of.

    More on the second point, it is known that the 5 MWe reactor could not be run a full capacity for the first few years of its operating lifetime. Starting in 1991, the operators finally seemed to grasp what to do with the reactor, and it ran at a high capacity when not shut down during the AF and other negotiated pauses. That shakedown period should have been shorter if they had a design they were following from Japan, because they would have been able to get the correct operating ranges and should have been able to duplicate them on their reactor.

  8. Josh (History)


    I understood your various points. (I think.) I’m coming around to your view on point #1; the imitation of Tokai may not have had any political content — or at least, no political content specific to Japan that was intended by the senior leadership. (For the designers of the reactor, who knows?) More on this point shortly; I’ll update the main posting with some interesting materials.


    These are some fascinating details, but I have to ask, if the Calder Hall design was simply available to all, why do you suppose there such a close resemblance between these two Magnox reactors — and only these two?

  9. Yale Simkin (History)

    I agree with Eric that the reactor itself could likely be directly based upon the Calder Hall blueprints that are in the public domain (just as Saddam built his calutrons from applying public domain designs of the US WW2 machines).

    Tokai provides the specific architecture and layout (and possibly detailed training , operating experience, and detailed plans from moles) that converts engineering diagrams to a real breathing plant.

    As I pointed out in my first comment:

    “It would be surprising if the reactors and buildings were not virtual clones (altho I do not know if Tokai 1 was a first generation Calder-hall-type Magnox, or a later version Magnox).”

    The DPRK “cloning” may be just the “macro” structure, rather than a duplicate reactor. Tokai showed the DPRK how you put all this stuff together (with help possibly from our dear friends the Russians).

  10. Yale Simkin (History)

    I left this quote out:

    “Technical information about the Magnox reactors at Calder hall, Sellafield has been in the public domain for over 25 years. We continue to urge the Democratic People’s Republic of Korea to fulfil her obligations under the non-proliferation treaty and sign an agreement with the International Atomic Energy Agency which would place all of her nuclear facilities under safeguards”.
    UK Cabinet Minister Douglas Hogg 1992

  11. Eric Rauch (History)


    I would expect that Yale is most likely right, with the DPRK taking what it needed from the Calder Hall design and then engineering the rest with some knowledge either from the USSR or from contacts inside Japan.

    The similarities are striking, but if they had the full plan for Tokai why would they bother with the 50 MWe reactor. Perhaps the 200 MWe reactor, which was slightly different in its design (I believe the fuel loading was different, and some structural differences in the pile) was to be more like the Tokai plant, but if you have a detailed plan, why not go full scale once you iron out the operational parameters on a pilot plant (which the 5 MWe could be described as being).

    The imagery is certainly similar though, and its a great discussion to have.

  12. SMH (History)

    “For example, the [IAEA] diplomat points out, the North Koreans took the design plans for an early-generation British plutonium Magnox reactor, built a 5-megawatt reactor, and were in the process of building 50- and 200-megawatt reactors. The Magnox had design flaws that the North worked out on its own.”

    What was this design flaws? I think there are some implication.

Pin It on Pinterest