Jeffrey LewisJust How Big Was Al Kibar Again?

I just love my job. I am like a kid in a candy store with all these photos of Al Kibar. A couple of things don’t add up just yet. Not saying they won’t eventually, but I am trying to work through the evidence with some rigor.

Despite early press reports that the fuel channels atop the Al Kibar reactor core were identical to Yongbyon, I and others — including Geoff Forden, Cheryl Rofer and Richard Wendland — see some pretty significant differences that suggest Al Kibar might have been quite a bit smaller than its North Korean cousin.

To be clear, I don’t doubt that Al Kibar was a reactor and, although I think the evidence of North Korean involvement is less impressive than early press reports suggested, that’s my working hypothesis too.

But I don’t understand the claim that Al Kibar is a copy of Yongbyon in the strict sense — in particular, I don’t understand how the IC concluded that Al Kibar is the same size as Yongbyon.

Are the Fuel Channels Identical, Similar or Dissimilar?

Early reports suggested that the Al Kibar reactor had virtually identical configuration and number of channels to lower fuel rods into the core as at Yongbyon. Robin Wright wrote in the Washington Post:

Sources familiar with the video say it also shows that the Syrian reactor core’s design is the same as that of the North Korean reactor at Yongbyon, including a virtually identical configuration and number of holes for fuel rods. It shows “remarkable resemblances inside and out to Yongbyon,” a U.S. intelligence official said. A nuclear weapons specialist called the video “very, very damning.”

The actual claim during the video was much more careful — that there were similarities in the configuration and in the size and capacity of the reactor.

This photograph shows the top of the reactor vessel in the reactor hall before concrete was poured around the vertical control rod and refueling tubes. Note the similar arrangement of vertical tube openings in the top of the Syrian reactor on the left and North Korea’s Yongbyon plutonium production reactor on the right. We assess the Syrian reactor was similar in size and capacity to this North Korean reactor.

In fact, the number and configuration of fuel channels are different in some ways. Cheryl Rofer, over at Whirled View has been sifting through the images with care and meticulous attention and noticed that Al Kibar has only about 60 percent of the number of fuel channels as Yongbyon. Richard Wendland made a similar observation in a comment on the blog. Then Geoff Forden sent me a note stating:

I’ve been counting fuel tubes and it appears to me that the Syrian reactor is considerably smaller than the North Korean reactor.

The’ve got a point.

Take another look at the image of the fuel channels and count them.

One can see quite clearly that the reactor at Al Kibar had 52 56 fuel channels in a 4-6-8-8-8-8-6-4 configuration like this (my arithmetic is especially poor on Sundays):

(There are a few onsies here and there, maybe for control rods, but that doesn’t dramatically alter the picture.)

That’s the same arrangement in the computer model released by the IC, so it seems quite plausible that no additional channels were to be installed. Here is a screen capture:

Yongbyon has considerably more fuel channels — 97, configured in 5-7-9-11-11-11-11-11-9-7-5. Like this:

If you model the core of a reactor as a sphere, the volume (and hence capacity) of Al Kibar would be about one-fifth that of Yongbyon — sixty percent cubed.

The implication of a smaller reactor is smaller plutonium production — roughly, while Yongbyon could produce 5-7 kg of plutonium per year, Al Kibar could only produce about 1 kilogram of plutonium per year.

That’s still not good, but it also invites comparison’s to the fuss over Algeria’s reactor, which was resolved with safeguards not airstrikes.

On the other hand, maybe there are design differences in the size of the channels or the rods that we aren’t taking into account. But it doesn’t seem to be a copy of Yongbyon, in the strict sense.

Whiskey Tango Foxtrot?

The possibility that Al Kibar is smaller might explain why the “copy” of Yongbyon doesn’t have a secondary cooling tower as Yongbyon does (although, to be clear, Yale Simkin and others caclulate that Syria could river-cool even a Yongbyon-sized reactor without unreasonable pumping requirements or boiling fish).

A smaller reactor would also require less natural uranium fuel — something that might matter if one plans extract uranium from phosphates.

Finally, a smaller reactor would also, presumably, reduce the design throughput for whatever reprocessing facility the Syrians intended to build or have squirreled away. That question — what about the reprocessing facilities — is a big one that I suspect we will be talking about a lot in the coming weeks.

But for now I just want to know if the IC really judges that Al Kibar was going to be exactly the same size as Yongbyon and, if so, on what basis was that judgment made given the different number of fuel channels?

I mean that as an honest question.


  1. Jeff B. (History)

    From the DEBKAfile:

    “In an interview Sunday, April 27 with the Qatari daily al Watan , Syrian president Bashar Assad said: “We don’t want a nuclear bomb, even if Iran acquires one.” DEBKAfile’s military sources say that was only half true.

    What he omitted to mention was the division of labor agreed between Damascus and Tehran in a potential war against Israel: The Syrian reactor Israel destroyed last September would produce “dirty weapons,” while Iran would go for a nuclear bomb. Tehran therefore funded the North Korean reactor in Syria. The radiological weapons made there were to be distributed to the terrorist organizations fighting Israel and used as leverage to control them.” see

    Might that explain why a smaller reactor was acceptable?


  2. CKR (History)

    From the video (via the briefing transcript):

    “We assess the Syrian reactor was similar in size and capacity to this North Korean reactor.”

    Not from this analysis.

  3. Murray Anderson (History)

    4+6+8+8+8+8+6+4=52 ,not 56.
    The size goes as the 1.5 power of the number of control rods, if the rods are proportionately shorter in Al Kibar, so it would produce 2 to 2.7 kg of plutonium a year.

  4. Stuart (History)

    Jeffrey-I very much relate to your appreciation of precise veribage, however it appears your concern is with Robin Wright’s perspective. The bottom line: Assad was building a reactor with the help of the NK’s. The comparison with the Algerian’s facility is academic—different political forces at play. Keep up the good work.

  5. FSB

    The question is so much Whiskey Tango Foxtrot, but rather Bravo Foxtrot Delta.

    i.e. regarding the differences.

    The Intel. folks, as you say, never said it was identical, only similar.

    The obvious point they would make is that it is a reactor based on the Korean design.

    In the final measure, whatever it was, it was still illegal for Israel to bomb a sovereign nation. Can they now have the gall to complain when Syria facilitates attacks within Israel?

    The 60th anniversary of Israel (aka. the 60th anniversary of the ethnic cleansing of Palestine ) is coming up, and probably the Israelis ought to wary of some Syrian fireworks to commemorate the occasion.

  6. FSB

    I am wondering if you or your readers have more information on the Iranian Directorate in the Pentagon. It has been compared with the Office of Special Plans . As the real import for the Syrian reactor bombing may be related to the Iranian Directorate I think it will be useful if you or your readers could help us understand its activities and intentions.

  7. yale (History)

    I think that directly counting the fuel channels in the image is a bit misleading.

    The Syrian image of the top of the bioshield appears to be of a work in progress.

    The channel tubing appears to be in the process of being installed.

    Note the stacks of unused tubes around the concrete pour area. It appears that the Syrians are adding tubes in rings from center and the outer ring is not yet in place.

    Notice the way the pattern seems to truncate unnecessarily far from the boundary. Add one futher ring and everything looks better.

  8. Shual (History)

    “The same footprints” [ISIS] should solve the problem of the room-dimension. The RCMs show a Yonbyon-room, but the Israeli photo [vessel] show a smaller room. In my point of view the “fuel channel”-photo is a fake for the public to show what happened between spring 07 [Israelis showed the vessel-photo to the CIA] and September 07. It tries to deal with the real dimenisons of photo “vessel” and the 56 fuel channels are only corresponding with the dimensions of the faked fuel channels.

    I see a bigger question still unanswered. When did they know what. John E. Pike said that “American spy satellites and analysts had, in fact, watched the site for years” and know they say: “received indications in ’05 that the Syrians and North Koreans were involved in a project in the Dayr az Zawr region of eastern Syria, but again, no specific information on the nature or the exact location of the work.”

    The video-presentation raises more questions than it answers.

  9. Rwendland (History)

    Another misleading aspect of the presentation was the strong suggestion the BOE building was similar to the Yongbyon 5MWe reactor. Comparing the two from somewhat similar vantage points makes clear the Yongbyon reactor building is vastly taller:

    The explanation is that the Syrian pressure vessel is (almost?) entirely underground, with the reactor top roughly at ground level. The building height is mostly for the fuel loading/removal machines, which for Magnox style vertical fuel channels need quite a lot of height.

    But the presentation seemed to use some smoke and mirrors to suggest the building were of similar shape:

    But here the Syrian reactor building is shown with a temporary structure on top to lift the pressure vessel sections into the pit under the building. That temporary extra height is taken away to make the final BOE lower flattish roof.

    Putting the pressure vessel underground would lead to some plumbing and cooling fun, as the cool CO2 entry point would be well underground, and ideally you’d want the gas circulator motors down there as well, as well as the heat exchangers. This probably means physical evidence remains underground, if the IAEA can ever get to it.

    I find it hard to believe that the construction of the pressure vessel in the open under a minimal canvas roof, and its movement and lowering into a pit, would not have been seen by satellites of a fair few intelligence agencies. I guess the Syrians hoped it would be analysed as a chemical-weapons plant.

    I also am not convinced by the suggestion that the spent fuel pond would be within the building, as that puts it rather close to the reactor. I doubt having so much water so close to a reactor is good for safety reasons, and also the pond water should be as cool as possible to minimise Magnox fuel corrosion, and having it so close to a big heat source could make that tricky. But maybe needs must.

    NB I think the Yongbyon reactor building tall chimney is to filter and safely exhaust the high pressure CO2 reactor cooling gas – not sure what the Syrians would do for that, perhaps its absence is a sign the reactor was not finished.

  10. Russell (History)


    Sorry if this question seems dumb – but shouldn’t you be modelling the core as a cylinder rather than a sphere?

    You don’t appear to have any evidence of a change in core height, only of a change in surface area – the conservative calculation would be 36% (60% squared) not 20% (60% cubed).

    While there have been some oddities, every heterogeneous reactor core I have seen has been a cylinder.


  11. upyernoz (History)

    the question raised by this post is extra funny because “al-kibar” means “largeness” (i’m assuming “al-kibar” is a transliteration of الكبر)

  12. Rwendland (History)

    For those trying to compare sizes, the clearest Yongbyon size estimates I’ve come across are in Table 2 of the SAND 2005-1981P paper.

    That estimates the Yongbyon 5MWe pressure vessel at 16.8m height and 8.8m diameter. Amazingly containing 600t of graphite, which compares unfavourably with the UK 50MWe 1140t core mass. I guess the smaller the reactor is, proportionately more graphite reflector is needed.

    The “video” shows the Syrian pressure vessel, but with nothing good as a size comparison. I we assume the CO2 inlet is 0.5m, that makes diameter about 7.5m, end-section height about 2.5m and mid-section about 3.5m height. That makes a 3 section pressure vessel a bit small, so assuming 4 sections we have about 2.5 + 3.5 + 3.5 + 2.5 = 12m height, and 7.5m diameter.

    Assuming perfect cylinders, that makes Yongbyon pressure vessel volume 1022 m^3 (pi * 4.4^2 * 16.8), and Syrian vessel 530 m^3 (pi * 3.75^2 * 12.0). Assuming the smaller vessel volume can be used less efficiently for the core, about a third to a half the thermal power of the Yongbyon 5MWe does seem about right for a rough estimate.

  13. Jeffrey Lewis (History)


    Not, as you obviously know, a dumb question.

    I wondered about it myself, but Forden voted that the nucleonics would be different.

    This would be the appropriate place to have that discussion.


  14. Geoffrey Forden (History)

    Is either the Syrian or the North Korean cores spherical or are cylindrical? I don’t know but since I feel strongly that they would keep the same aspect ratio (length to radius ratio) for cylindrical cores, it doesn’t mater: they both volumes scale as R^3. First, let me explain why I think the aspect ratios would be the same if they are cylindrical then run through the volume calculation for both geometries. (The final result does assume that the plutonium production scales as the volume of fissionable uranium. I’m going to have to think some more about Murray Anderson’s post.)

    The physics of reactors is determined by their neutron density, which in turn is determined by their geometry. A given geometry will determine, for instance, how it responds to control rods or new fueling or the build up of fission products. Changing the geometry, by changing the aspect ratio of cylindrical cores, will result in a completely different neutron density distribution and, perhaps most importantly, how it responds to control rods. It appears that the radius of the Syrian reactor has decreased by a factor of 8/11 (~0.72). If, for some reason, they wanted to have the same load of uranium inside a cylindrical reactor, they would have to increase the length of the cylinder by the same factor. That means that the aspect ratio would change by about a factor of 2 greater; that would seem to completely change the running of the reactor. Keeping the length the same as the Yongbyong reactor results in a change in the aspect ratio of almost 1.4. Would that be sufficiently different that North Korea’s experience with its reactor would not be directly transferable? I would think so; it probably means that withdrawing control rods an equal amount would be 40% less effective etc. Certainly, some things will be different for a smaller reactor but they would have to change so much for such a large change in aspect ratio that I believe (but I certainly cannot prove it) that they would keep the same aspect ratio.

    Let me now run through the volume calculation for a spherical and a cylindrical core to show that they both scale as R^3 if the aspect ratio is the same. The total volume of a spherical fissionable core goes as R^3 (if it is spherical, than the cells on the outer edges only have uranium say in a small portion of its length, the center portion.) If its cylindrical instead, then if it has the same aspect ratio of Yongbyong reactor the length of the fuel rods is L=c*R where L= length of the rods, c is a constant determined empirically from the aspect ratio, and R is the radius of the core. The volume of the cylindrical core is then V=pi*R^2 * c*R which still goes as R^3 if you keep the same aspect ratio, c.

  15. Al (History)

    1. Satellite photos of the alleged reactor building show no air defenses around the building; no AA gun emplacements (of the type seen in Natanz in Iran), and no AA missile batteries/radar.

    2. The photos do not even show any military checkpoints on roads leading to the building.

    3. The photos show no power lines coming into the building and no electrical substations near the building.

    4. Here is a photo of the North Korean facility (not the circle, the whole area):

    Notice how much of the “infrastructure” is missing from the Syrian photos – basically everything is missing, except for one building. There are no facilities to
    store fuel, no place for plant workers to live in.

    5. Quite a few of the photos presented are computer enchanced to say it politely (or doctored). Example:

    Notice how the lower part of the building (the attachment) and the “windows” pointing south are so much sharper than the rest of the photo.

    Or in this photo, the sky is heavily overcast. Is this normal for Syria?

    Or compare these photos:

    Notice how a rectangular building has turned almost square in the second photo.

    6. If you recall at the time of the attack, there were reports of Israeli fuel tanks being dropped in Turkish territory and in Syria:,8599,1660477,00.html

    This may be evidence of the Israeli jets being faced with an unexpected challenge from Syria either on their way to the target, or on the way back. The aircraft
    jettisoned the tanks to gain altitude, become lighter, and get out of there FAST.

  16. James (History)

    Is it really impossible that the Syrians built this thing without North Korean help? We are dealing with 60-year-old technology, after all. The North Koreans certainly didn’t get British help in building their reactor and this Syrian project has been underway for quite some time. Long enough that maybe they really did dope it out all by themselves.

    Why no reprocessing plant? All comments on this site seem to agree that Magnox-style fuel rods need to be reprocessed fairly soon after being removed from the reactor. If the intent was to produce significant amounts of weapons plutonium, would they bother to fuel the reactor without having that plant ready?

    If they did receive assistance from North Korea, why a smaller plant? I had thought Yongbyon was pretty much the minimum useful size for a weapons program…surely the Syrians did not believe they could conceal the plant once it was active. They would not have had years to slowly build up plutonium.

    Biggest question of all: where is the supposed fuel?

    DEBKA’s deranged conspiracy theories are just that: deranged conspiracy theories. No one goes to the trouble of building a reactor just to make a “dirty bomb,” especially when Syria already has much more lethal chemical weapons in its stockpile. And if Iran is truly making real bombs, dirty bombs would be irrelevant. I don’t think the analysts who assert a weapons program should be allowed to skirt the the issue of absent weaponization facilities with that kind of handwaving.

  17. FSB

    Joe Cirincione is quoted extensively in this Counterpunch piece as saying that this may have just been a small research reactor.

    “In attacking Dair el Zor in Syria on Sept. 6, the Israeli air force wasn’t targeting a nuclear site but rather one of the main arms depots in the country. Dair el Zor houses a huge underground base where the Syrian army stores the long and medium-range missiles it mostly buys from Iran and North Korea. The attack by the Israeli air force coincided with the arrival of a stock of parts for Syria’s 200 Scud B and 60 Scud C weapons.”

    Cirincione says that there is a small Syrian nuclear research program, which has been around for 40 years and is going nowhere. “It is a basic research program built around a tiny 30 kilowatt reactor that produced a few isotopes and neutrons. It is nowhere near a program for nuclear weapons or nuclear fuel,” he said. Over a dozen countries have helped Syria develop its nuclear program, including Belgium, Germany, Russia, China and even the United States, by way of training of scientists, he said.

    So what is really going on here? Cirincione told the BBC that “This appears to be the work of a small group of officials leaking cherry-picked, unvetted ‘intelligence’ to key reporters in order to promote a preexisting political agenda.”

    This would jive with the analyses above saying that power of this small research reactor would have been substantially less than its Korean cousin’s.

  18. Rwendland (History)

    Geoffrey, both the Syrian or the North Korean pressure vessels are cylinders. There was an image of the Syrian vessel in sections before assembly in the “video”:

    (Note the minimal canvas satellite-view/sun shield.)

    Not all of the pressure vessel is used for the core, there are gas voids below and above to allow the CO2 to distribute from a few inlets across the core bottom, and similar on the way out. So you cannot work out core size precisely from pressure vessel (PV) dimensions.

    In the UK spherical PVs were introduced for the larger power stations, as CO2 pressure was increased (from 6.9 to 27 bar) and core size increased, for strength reasons. However the cores remained cylindrical in the PV space, for simple fuel/rod access paths from above only. (Though the first large concrete and steel cable PV went back to a cylinder.)

    There seems no overall consistency on Magnox core aspect ratios. SAND 2005-1981P says Yongbyon 5MWe core is wider (6.43m diameter) than high (5.9m). Wider than high for UK Magnox cores as well, but there is no standard exact ratio, some have nearly twice the diameter as height (see Table 3 of this ).

    Note that SAND 2005-1981P says half the graphite used (300t) is for “Graphite Reflector”. I presume such use could help in using non ideal core shapes.

    NB There are some diagrams of the last UK Magnox core at the end of this report (This Magnox uses the more modern idea of having the heat exchanger inside the PV.)

  19. CKR (History)

    RWendland: I have been concerned about the dimensions of the building as well. My first inclination is the same as yours, that the height of the alleged Syrian building is much less than that of the North Korean building.

    I have seen the suggestion (but lost where I saw it; reading too many posts on this subject) that the Syrian reactor was built further into the ground.

    We also do not know the scale of the Syrian building. If the door is of a normal size (say 7-8 feet), then the building is probably not as tall as the North Korean building.

    But the dimension we know is of the roof, and I believe someone here calculated the height from the shadows some long time ago. So those are the numbers we would have to scale from.

    I suspect that, as in any industrial building, the doors are larger than the doors we are accustomed to in our homes.

    Another concern I have about the building is that some of the overheads almost look like the ground level is different on different sides of the building. It may be that during the disassembly of the building, the Syrians built up berms to facilitate moving the debris. Again, I haven’t had time to check this out in detail; it’s just from rough observations.

    If you want to check this out, be sure to use the photos in their proper orientation rather than the orientations provided in the video. I’ve done that here:

    Yale – I noticed the additional tubes last night. Another look at the reactor photo shows additional smaller-diameter tubes installed in the right-hand portion of the reactor. These may be for control rods. The stock could be for installation in the left-hand part of the reactor.

    It may be useful to try to scale the tubes from the photo, although it will be more difficult than for the building because of the perspective involved.

    I probably won’t be able to get to any of these investigations today, so I invite any of you to try.

  20. Andy (History)

    Lots of good info here, hopefully I’ll have more time later to make some detailed comments.

    First though, I did some simple analysis of reactor size which hopefully may be of use to some of the experts here.

    The following image is rotated and cropped from one of the post-strike images in the presentation. Over this image I layered a 1meter grid which is based on the original ISIS measurements of the building with the short side being 47meters. If the ISIS dimensions are accurate, then the reactor bioshield is about 16 meters in diameter.

    In that same ISIS brief, the roof over the reactor hall is measured at 22 X 24 meters. With the roof “off” my measurements put the size of the hall at about 20 X 22 meters, which is fairly consistent when one considers the supporting columns and structure will make the roof dimensions a bit larger (similar to how a house’s roof covers a larger footprint than the actual interior wall-to-wall space).

    In the picture below, I took the computer model of the reactor building Jeffrey provides in the original post above and then scaled and rotated it to fit over the post-strike image. Because of some skewing issues (the computer model is slightly oblique) it’s not a perfect fit, but is pretty good. For example, the “heat exchanger” structures match exactly with the imagery.

    Now, if the relative size and spacing of the channels in the model is accurate, and provided the other measurements above are accurate, then the reactor vessel is 10-12 meters in diameter. Finally, because this last picture is kind of confusing, I made it into a little animated gif:

  21. Andy (History)

    One more thing. There are channels in the reactor room which are shown on the computer animation that you don’t show in your configuration graphics. There’s one in the very center of the reactor, and then 2 along each of the four outside edges. If these are actual channels, which may be disputable since some do not fit the neat geometric layout, that would add 9 to the total.

    Additionally, much is being made of the lack of reprocessing facilities and some have gone so far as to claim this “proves” the reactor isn’t meant for weapons. I would point out in response, however, that Dimona went critical in 1962 and its reprocessing facility wasn’t finished until 1964 or 1965. Likewise, Yongbyon went critical in 1985 and the reprocessing facility was begun sometime between 1985 and 1988. The IC says this Syrian reactor was probably “weeks to months” away from becoming operational, so it’s hardly surprising a reprocessing facility doesn’t seem to exist yet.

  22. J House (History)

    Assuming the images released are accurate, it does appear to be smaller scale.
    What is the diam of the inflow/outflow pipes? Size of the cooling storage tank?

  23. J House (History)

    Some of ‘Shuals’ comments do raise further questions-

    When was the WH notified that the Israelis would attack the facility? Surely they had drawn some red lines with the Bush administration long (months?years?) before the attack.
    If ‘good to go’ and ‘weeks or months’ is the red line, where does that leave the ‘Iranian problem’?
    The question of the Syrian source for enriched uranium goes unanswered.Was it part of the NK deal?
    Isn’t it true Saddam had 550 tons stored at Al Tuwaitha before the war? (under IAEA seal)
    Everyone should agree that whether the Pu yield was 1-3 Kg/yr,Syria making bomb fuel is a serious threat to Isreal’s security.
    ‘Illegal’ is a legal term they care a wit about when it comes to a nuclear-capable Syria.

  24. FSB

    Steven Aftergood sends this in his email list:

    An extensive, frequently updated collection of open source materials on
    the subject — including foreign and domestic news reports, satellite
    imagery and analysis — has been compiled by Allen Thomson in “A
    Sourcebook on the Israeli Strike in Syria, 6 September 2007” (currently
    812 pages in a 15 MB PDF file):

    An updated bibliography of Syrian nuclear science research, from
    reactor safety to laser isotope separation, was prepared by researcher
    Mark Gorwitz. See “Syrian Nuclear Science Bibliography: Open
    Literature Citations,” April 2008:

    A list of all cooperative agreements between the International Atomic
    Energy Agency and the Atomic Energy Commission of Syria, also compiled
    by Mr. Gorwitz, is here:

    The web site of the Atomic Energy Commission of Syria is here:

  25. Jeffrey Lewis (History)


    You are right — it is nine more pipes. I can only see that now that I have better images.

    I will post them when I get a chance.

    That significantly alters the configuration and changes the ratio to 97:62.

    But that still means it is only 25-40 percent as large, assuming spherical or cylindrical models.

    I was just wishing I had a damn copy of the three volume Reactor Handbook.

  26. FSB

    Thanks for your detailed work to figure out the dimensions of the Syrian reactor-to-perhaps-be.

    Quite separately, yes, Dimona (claimed to a desalination plant at the time!!! ) went critical between ’62 and ’64. It is too bad that it was not bombed then so that Israel would have been forced to keep its word about not introducing nuclear weapons in the middle east.

    Instead we have a lying belligerent regime that has proliferated WMD into the middle east, all with the aid of the US. (Although sometimes when aid was not forthcoming the Israelis simply stole nuclear material from the US.)

    Do people in the arms control community think for a moment that such a double standard is sustainable? If so, think again.

  27. abcd (History)

    “Additionally, much is being made of the lack of reprocessing facilities and some have gone so far as to claim this “proves” the reactor isn’t meant for weapons. I would point out in response, however, that Dimona went critical in 1962 and its reprocessing facility wasn’t finished until 1964 or 1965. Likewise, Yongbyon went critical in 1985 and the reprocessing facility was begun sometime between 1985 and 1988. The IC says this Syrian reactor was probably “weeks to months” away from becoming operational, so it’s hardly surprising a reprocessing facility doesn’t seem to exist yet.”

    Good points, Andy. But I wonder (admittedly out of ignorance, mostly): wouldn’t there also be indications that Syria was mining uranium, producing graphite, and readying a fuel fabrication facility?

  28. blowback (History)

    So where is the graphite? The reactor was ready to be fueled so it would have contained say 400t of graphite. Hit that with high explosives and most of it would be reduced to dust and blown on the wind. Therefore, one would expect to see a large area covered in black dust, but there is nothing.

    BTW, nobody has revealed where the photos in the presentation came from so how does anybody know that the incriminating photos (the top of the alleged reactor and the one unadorned external shot) are of Syrian origin. They could just as well be a misinformation exercise by the Israelis. All it would take are a few tons of concrete, some rebar, a few bits of tubing and a few workmen. If anybody claims that the US and Israeli government don’t lie then I have $100M deposited in a bank in Lagos you might like to make me an offer for.

    Finally, do the Syrians have the welding skills necessary to manufacture the pressure vessel. One reason the British resorted to concrete wound with steel cable for their pressure vessels was a shortage of welding skills – my father was an engineer and worked for the CEGB on nuclear reactor design and this was one of his pet peeves.

  29. Andy (History)

    To add a bit to Yale’s theory, if another “ring” is added, that would change the configuration to 4+6+8+10+10+10+10+8+6+4 for a total of 76 – 77 if one includes the single one in the center of the reactor.

  30. Rwendland (History)

    Andy, a difficulty in assessing reactor size/power from imagery is differentiating the thick biological shield from the pressure vessel. Yongbyon 5MWe pressure vessel is 8.8m diameter plus concrete radial biological shield 3m thick, making total diameter 14.8m, according to SAND 2005-1981P Table 2 So the shield is about 40% of the diameter.

    UK Magnox use biological shields about 3m thick as well. If the Syrian reactor is largely underground, as I suggest above, perhaps they could get away with a thinner biological shield , which could confuse this measurement a bit.

  31. Yossi, Jerusalem

    * The Norks are the first suspects for providing the fuel. Is it possible they sneaked it past the IAEA? According to Hecker NK fuel fabrication facility is in a bad shape but they have stored fresh fuel for about 2-3 BoE loadings. Were the Syrians waiting for this?

    * Looking at data on the big English Magnox reactors I see that the lattice pitch is almost constant at 20 cm. If this applies also to the NK designs it could be used in measuring the reactor.

    * It’s still not certain BoE was a nuclear reactor.

  32. Andrew Foland (History)

    J House—rwendland estimated above the inlet port (visible in the reactor pressure vessel image) is about 0.5 m, which seems reasonable from the image.

    I wonder if the gas ports might not be the reason the vessel may have been half-buried? I’ve seen it said (and it stands to reason but I can’t find an reliable original reference source) that significant radiation escapes through the imperfectly-shielded gas ports, and a few meters of earth might do a lot to attenuate that. (Presumably, in this scenario the concern would be more with external detectability than health effects on the local populace.)

  33. Jeffrey Lewis (History)

    Andy and Rwendland:

    I note the IC diagram shows a 4-6-8-8-8-8-6-4 with the additional nine channels.

    It seems unlikely to me that additional channels were to be added.

  34. Andrew Foland (History)

    Given this image of the west wall , and this image of the exterior (same as rwendland’s above), do we still believe the height of the BoE was 15 m, compared to the 47 m lateral dimension?

    Also, does anyone have any idea what construction of such a structure might cost? I think that might be an interesting number.

  35. Rwendland (History)

    Jeffrey, we have a terminology problem with “channel” we ought to nail down. I call the things we are counting (97 or 56ish) “access ports” (not an official name), they are not “fuel channels”.

    Yongbyon 5MWe has 812 fuel channels, and I assume each of the 97 access ports usually give access to 9 fuel channels (97 * 9 = 873). Some of the access ports will have fewer than 9 due to control rods, instrumentation, core edge, etc.

    If you want to take a gander at some UK Magnox pile caps for comparison, here is a selection of links to some images:

    Calder Hall pile cap, being inspected

    Wylfa pile cap with fuelling machine, showing its size

    Wyfla pile cap close-up

    Oldbury pile cap showing stand pipe assembly and control rod actuator

    NB there was never a UK Magnox standard design, so almost every power station was different, as you will see in those images.

  36. Carey Sublette (History)

    Regarding the reactor volume scaling, if the two reactors have spherical cores and the channel count is 52 and 97, then the volume ratio is (52/97)^1.5 = 0.39 (the tube ratio is a 2-D cross section).

    If the cores are cylindrical, then we have to guess the L/D ratio which might or might not be the same. If they are cylinders with the same L/D ratio then the computation is the same as the sphere. If the absolute core height is the same then the volume ratio is simply 52/97 = 0.54.

    So Syrian reactor/DPRK reactor size: 39% to 54% on the basis of visible installed tubes.

  37. Andy (History)


    You’re right and my calculated size figures are subject to substantial uncertainty because they’re based on the 47m measurement from ISIS. Still, if the the bioshield is 16m in diameter and those walls are 3m thick, then that gives an “inside” diameter of 10m (plus or minus, I’d estimate, 10%). I don’t know enough about this reactor design to know if there is anything between the interior of the bioshield and pressure vessel (more graphite maybe?), so the 10m figure would be a maximal diameter – again provided the 47m measurement is accurate.


    The graphite would be inside the bioshield and it appears from the imagery that this component was damaged, but not significantly enough to throw graphite about. Being underground seems to have left it relatively unscathed.

    Also, the photos I believe are completely genuine. One can do a bit of comparison with the terrain in each photo. Additionally, if you go to the excellent 15mb pdf file FSB linked to above, there’s a digital globe satellite image in there (page 731) that’s slightly oblique and the visible wall matches the ground photos.


    The questions on where uranium and graphite would come from are good ones, but it seems unlikely that Syria would go to the expense of building this reactor if it did not have a reliable source (either foreign or domestic) for both. Maybe this is the bone the US will throw the IAEA or maybe, if the source was/is/will be the DPRK, it will be used as leverage in negotiation. Just a theory.

  38. Rwendland (History)

    Andrew Foland, you are right about the small radiation hazard from the heat exchangers being outside the pressure vessel. But it is a small problem as the CO2 gas flow is not very radioactive; on-site people don’t stay in the worst areas, in the UK even unfortunately situated nearby houses were (just) within permitted UK general public “gamma shine” levels. Certainly no problem for military staff, but maybe a worry about passing spies with Geiger counters, so you might want to shield for that reason!

    I’ve often wondered if anyone worked on satellite detection of gamma from nuclear reactors? Be a neat tool if it could be made to work, by integrating across time.

  39. FSB

    Blowback raises an important point: besides the (ahem) word of Mossad, what real evidence is there that the alleged interior photograph is actually of the interior of the BoE?

    Secondly, where in the overhead satellite photos are the tarp-covered regions?

    Curious US taxpayers want to know.

  40. Jeffrey Lewis (History)

    Wow. Not exactly the coverage I was looking for.

    Anyway, in terms of the provenance of the photographs, Senior Official 2 (from now on, “Deuce”) said the inside images were mapped to the overhead:

    And just to hit a point I said earlier, you see the kind of crawlspace back there? If you have access to the wealth of photographs that we had, you can work from the crawlspace to the wall to the windows to the ventilation duct to the duct coming out the window, and now you’re looking at the overhead photography of that window in the right place with the duct coming out of it. See what I’m trying to describe for you that we are very confident that that on the left is inside that building that we showed you in the overhead imagery.

    That’s the method.

  41. CKR (History)

    Just out:

    “CIA Director Michael Hayden said Monday that the alleged Syrian nuclear reactor destroyed by an Israeli airstrike in September would have produced enough plutonium for one or two bombs within a year of becoming operational.”

    Guess he’s not reading ACW.

  42. FSB


    with due respect, I do not doubt the work of “Duece” or his co-workers. What I — and I think “Blowback” — is saying is that it is only the word (and/or photos) from the Mossad that connect the interior to the exterior.

    Though some interesting duct tracings may pan out, all these interior photos have been (presumably) provided to the US by the Mossad. They know who they are dealing with and can easily concoct such a trail of interior photos.

    This may appear paranoid, but remember these are the same guys who managed to build Dimona by telling people it was a desalination plant.

    In any case, are we to believe that a Mossad mole managed to climb inside the crawlspace with a camera?

    I tend to go with Joe Cirincione’s quotes above.

    It may well be an incipient low power research reactor, but I think the evidence from DNI (or rather Mossad) should be more carefully vetted.

  43. Yossi, Jerusalem

    Yongbyon data from Sandia and Lewis:

    812 fuel channels
    6.43m core diameter
    97 access ports

    A simple calculation:

    * Exactly 20cm channel lattice pitch (just like in the UK!)

    * 9 channels per port (actually 8.37 but there are also control rods etc). That means a 3×3 array at each port.

    * 60cm inter port distance

    If you assume similar design in CIA photo you can use the inter port distance for scaling.

  44. FSB

    Separately, does anyone know where the tarp-covered regions (shown by Rwendland · Apr 28, 06:07 AM above) may be in any epoch of overhead photos?

  45. blowback (History)

    Andy – when the Syrians “blew up” the remains of the “reactor” before burying them, any graphite would have been released then or are you going to suggest that they vacuumed it all up.

    Also, the photos I believe are completely genuine.

    Without knowing the source and seeing the originals, no-one can know how genuine the photos are. Since there are reports of Dick Cheney being involved in the preparation of the presentation, I think it only safe to remain highly skeptical of the information and pictures supplied.

    Why are there no satellite photos of the hole in the ground or of the reactor building before the curtain walls were added. Digging a hole and constructing a building on that scale would take several weeks and if the National Reconnaissance Office does not have the imagery, then some one should be sacked for incompetence.

    When are they going to publish all the other photos? Any source is now totally compromised so there is no further cost to publishing them. As to the claim that the internal layout matches the external layouts, well they would if the fakers were any good – they would take the real external layout and derive the fake internal layout from that so they would match.

    Why would the Syrians put so many windows and doors (fifteen that can be seen in the before photo) in place only to fill them all in to camouflage the building – that takes a very special kind of stupid.

    If it really was a reactor, the Syrians would know that the US and Israel, who both have reconnaissance satellites, would be interested in it so it would also require a special kind of stupid to build the original structure looking like the Yongbyon site only to draw attention to it by camouflaging it. After all they put in screens to shield the pressure vessel components from reconnaissance satellites.

    The photos look as though they are from two sources – one showing the “reactor internals”, the other showing the outside of the BoE – why weren’t there any photos included in the presentation taken through the substantial entrance in the east wall of the BoE towards where it is claimed the reactor is located.

    This could be a disinformation exercise similar to Operation Quicksilver

  46. abcd (History)

    Ah, more good points Andy. Here is what the 1996 Sandia-led Proliferation Vulnerability Red Team report to say on the timeline for reprocessing facility construction:

    “The team estimated the length of time that would be required to prepare and test a processing facility, and the time after acquiring the plutonium material to produce 1 SQ [Significant Quantity] of plutonium metal. The preparation lead-time varied from six months for a facility requiring shielding and remote operations to about three months for a facility requiring only containment. The time to produce 1 SQ varied from eight weeks for materials with a radiation barrier to four weeks for plutonium oxide that required purification, and less than one week for pure plutonium oxide that could be directly reduced to plutonium metal. The PVRT recognized that there are considerable uncertainties in attempting to estimate teams that would be required to produce 1 SQ of plutonium metal from various forms. The time would depend strongly on the skill and knowledge of the operators, the complexity of the process chosen, the reliability of the equipment, and the time spent in preparation and and testing.”

    Granted, that’s probably a technical baseline higher than what would be applicable in the case of Syria, but it’s a reliable comparison of the timeline for reactor construction versus that of a repro facility. I hadn’t thought of quite this way just yet. It’s certainly plausible, if Syria’s true destination was a limited nuclear arsenal. Plunging full-speed ahead with both simultaneously or concurrently runs plenty of risks in the immediate sense.

    Alternatively, I’d like to consider something else. Scattered reports, including the IC testimony, place the initiation of this project back nearly a decade ago, putting its launch under Assad sr. I wonder, then, how committed to this project his son was when he inherited it, if he was not given to relinquish it up for fear of appearing weak to the inner circle and state apparatus he also inherited, and if that all explains the lack of supporting infrastructure.

  47. SQ

    J House:

    Magnox reactor fuel is not enriched. Which is not to say that it is widely available, either. Westinghouse produces it for UK reactors, and the Yongbyon facility is shut down. So far as I know, that’s it.

  48. kme

    Andy: Adding another ring would actually make it 6+8+10+10+10+10+10+10+8+6 = 88. (It doesn’t really look like there’s space for that, though).

  49. kme

    Andy: Adding another ring would actually make it 6 + 8 + 10 + 10 + 10 + 10 + 10 + 10 + 8 + 6 = 88. (It doesn’t really look like there’s space for that, though).

  50. yale (History)

    I don’t think gross core volume is quite the correct measure for estimating Pu output.

    Remember that the bulk of the volume is simply cooling ducts and graphite moderator.

    If you add more fuel pins you have to separate then with an appropriate quantity of neutron moderator. This added graphite volume is way disproportionate to the added power output. Thus, adding a single ring of fuel multiplies the volume of the core – but not necessarily the power output.

    Way, way oversimplifying, the output is better related to the mass of fuel.

    So, crudely speaking, comparing sonotube counts of NK to Syria

    62/97 or 64%


    76/97 or 78%


    This is the top of the reactor where I skewed the image to see the tube layout better and to see if there is room for more tubes:


    Here is an image which I made B&W, pushed the contrast.

    I marked tube piles with blue arrows.

    I used red arrows to emphasize the width of the tubes:


    This is an excellent satellite image of the reactor outer shell, showing the side wall very well:


    Here is a google earth image accross the river down the wadi. Notice what an excellent concealment job the placement of the reactor plus the added berm created:

  51. Andrew Foland (History)

    By doing a perspective transform on one of the exterior images, I find that the height might be more like 25 m.

    rewendland—you answered one of my questions—whether the source of “shine” is scattered radiation from the core bouncing out along the cooling path, or radioactivity generated in the coolant itself. If the latter, then I’m not sure if burying it helps much, since the heat exchangers are largely above ground, at least according to the model.

  52. Hairs (History)

    Just got back online after being away for a few weeks, and boy is there a lot of interesting stuff to read!

    CKR: Based on the building’s shadows, my final “guesstimate” for the height of the original BOE was around 26m +/- 5m (posted on 26 Jan 2008). This still seems plausible in light of the photos we now have, though perhaps I’d err towards the 21m – 26m side of the uncertainty.

  53. Allen Thomson (History)

    > Why are there no satellite photos of the hole in the ground or of the reactor building before the curtain walls were added. Digging a hole and constructing a building on that scale would take several weeks and if the National Reconnaissance Office does not have the imagery, then some one should be sacked for incompetence.

    At least since the demise of the KH-9 program in the 1980s, the NRO has been out of the uncued area search business. Even cued search is probably undertaken with reluctance. The reasons for this are a) the satellites are designed for high resolution, with a consequent decrease in area coverage rate (“looking through a soda straw” is the way it’s usually described) and b) area search has to be performed by image analysts and mostly comes up empty — which can easily be regarded as a waste of valuable analyst-hours.

    I find it entirely believable that the Box went undiscovered by the US until the Israelis provided some reasonably specific location data in early 2007.

    And yes, I agree that NRO should have kept an area search capability and NPIC should have figured out a way to allocate analyst-hours to the generally unrewarding area search task. Fortunately, with Google Earth and the various commercial providers, matters are improving outside the bureaucratic sphere.

  54. Dave (History)

    Do y’all mind a question from a nuke layman (although I’m an engineer in a different discipline)?

    What do you think of this theory:

    I might conclude that this is a training reactor. That would be my reasonable assumption the size and lack of defensive measures. If The Box is part of a very long-term project started by Hafez al-Assad (where did I see that?), this would make a logical, staged approach for independent operation of a reactor.

    If they were to jump right to a strategic size, they would be beholden to DPRK staff, perhaps forever. This way, they learn and build a bigger, practical one without assistance.

    If there is sufficient background intel to draw that conclusion, it still makes this a valid target for interrupting Syria’s weapons program.

  55. Carey Sublette (History)

    * Exactly 20cm channel lattice pitch (just like in the UK!)

    I don’t think gross core volume is quite the correct measure for estimating Pu output.

    Remember that the bulk of the volume is simply cooling ducts and graphite moderator.

    If you add more fuel pins you have to separate then with an appropriate quantity of neutron moderator. This added graphite volume is way disproportionate to the added power output. Thus, adding a single ring of fuel multiplies the volume of the core – but not necessarily the power output.

    Way, way oversimplifying, the output is better related to the mass of fuel.

    The lattice spacing in a graphite moderated reactor is fixed within fairly narrow limits by the physics of moderation. In all the various reactors built at Hanford over the years the spacing varied from 19.05 to 21.59 cm. Likewise in a natural uranium fueled graphite reactor, the moderator/fuel ratio does not vary much. Again, at Hanford the atom ratio varied from 109 to 77.

    So – the core volume is a good proxy for the fuel load, and a good metric for the reactor size.

    Regarding the power output – this is dependent on the effectiveness of cooling and the fuel mass (for which we can substitute core volume for scaling purposes). The power output per ton of fuel can potentially vary over a huge range (roughly 20 to 1) depending on the sophistication of the fuel element fabrication but mass production of failure-free fuel elements is challenging. The DPRK and Syria they are probably stuck with low performance elements similar to U.S. WWII capabilities. So scaling from the Yongbyon to the BoE based on core volume is reasonable.

  56. abcd (History)

    “I find it entirely believable that the Box went undiscovered by the US until the Israelis provided some reasonably specific location data in early 2007.”

    But, it was known in 2003 where this building was and that undeclared construction was underway…

    @ Dave: the Syrians already have a basic, research-level infrastructure.

  57. yale (History)


    I think the difference between the numbers is based upon whether the length of the fuel rods is increased proportionately with the number of the elements (resulting in a 3D expansion of the core.)

    I make no assumption about that (I have no data).

    I am assuming that the lengths are the same and the only thing increasing is the number of elements (which then scales linearly with the fuel mass), and that power level per fuel element is constant. If so, cubic scaling does not apply.


  58. Allen Thomson (History)

    > But, it was known in 2003 where this building was and that undeclared construction was underway…

    The article says the site was imaged by Ikonos on 16 Sep 2003. GeoEye found the image subsequent to the 24 Oct 2007 publication of the August 2007 QuickBird image by ISIS. There is no indication that anybody noticed the site in the 2003 image before that.


    “A senior American intelligence official said yesterday that American analysts had looked carefully at the site from its early days, but were unsure then whether it posed a nuclear threat.”

    The recent background briefing by three officials (apparently McConnell, Hayden and Hadley) throws some light on that (excerpts):

    “Now, as early as 2003, we judged that the [Syria/NK] interactions probably were nuclear-related, again, because of who it was we were seeing in these interactions. But we had no details on the nature or location of the cooperative projects. We assessed the cooperation involved work at sites probably within Syria. But again, we didn’t know exactly where.”

    “We received indications in ’05 that the Syrians and North Koreans were involved in a project in the Dayr az Zawr region of eastern Syria, but again, no specific information on the nature or the exact location of the work.”

    “Imagery searches of the region revealed a large unidentified building under construction in a remote area near the Euphrates River near a point that we call al Kibar. And there you see the photo. The first time we saw it was after this evidence – look out there – remember ’05, ’06 timeframe – take a look there. We identified the facility. And once again, sometimes the present illuminates not just the future but can illuminate the past. We looked back on historical imagery that found that the only high-quality imagery we had was of a building that looked pretty much like this. It was externally complete.”

    “We acquired information, though, in the spring of ’07 that enabled us to conclude that this non-descript-looking building in al Wadi, near the Euphrates River in eastern Syria was indeed a covert nuclear reactor.”

    So I was wrong about the Box going unnoticed until early ’07: cued imagery search apparently did find it sometime in the “‘05, ’06 timeframe.” Whether the search involved taking new pictures or just examining the “historical imagery” isn’t clear to me. Nor is it clear whether the imagery came from NRO’s Finest or third party/commercial sources. The latter wouldn’t surprise me, but that might be discernable from the video. Unfortunately, I’m stuck on a dial-up line for the next few days, but maybe somebody else could take a look.

  59. blowback (History)

    I find it entirely believable that the Box went undiscovered by the US until the Israelis provided some reasonably specific location data in early 2007.

    — Allen Thomson · Apr 29, 09:56 AM ·

    Global Security has an article about the 16 Sept, 2003 photo which suggests the the US IC knew about the EoB from the beginning.

    The senior intelligence official said that American spy satellites and analysts had, in fact, watched the site for years.

    “It was noticed, without knowing what it was,” the official said. “You revisit every so often, but it was not a high priority. You see things that raise the flag and you know you have to keep looking. It was a case of watching it evolve.”

    BTW, the 16 Sept. 3003 image was taken while the large apertures at the corners of south wall of the building were open as the following images from Fox News and Moon of Alabama show.

  60. Allen Thomson (History)

    > BTW, the 16 Sept. [2]003 image was taken while the large apertures at the corners of south wall of the building were open as the following images from Fox News and Moon of Alabama show.

    Good catch. It would be nice to have a 1-2 day imagery analysis conference amongst the DC wonkery to sort out what is known, what is unknown, what is known to be available but unanalyzed, etc.

  61. Allen Thomson (History)

    More on the timing of things. Caution: I’m not entirely comfortable with the credibility of this report, but it might be significant if the information is confirmed.


    CIA’s Hayden: Syria was on verge of becoming nuclear power

    Wednesday, April 30, 2008

    WASHINGTON — The U.S. intelligence community, in an about-face from an assessment of less than a year ago, has concluded that Syria was close to becoming a nuclear power.


    Officials acknowledged that the U.S. assessment marked a near reversal of that in July 2007 when Israel provided aerial photographs of the plant and a video of the North Korean scientists inside. At the time, the officials said, the CIA and State Department said the North Korean facility — destroyed by the Israel Air Force in September 2007 — was years away from being completed and even tested.

    “Much of the revision of the CIA assessment came after the Israeli bombing when evidence of nuclear material was found,” an official said. “We also learned a lot from the Syrian refusal to the International Atomic Energy Agency to visit the site.”


    By June 2007, officials said, the CIA received information from Israel that North Korea was building a suspected nuclear reactor. Officials said the Israeli aerial and ground-based photographs overcame doubts that stemmed from satellite images of the facility.


  62. Bill Lenner (History)

    Foreign Policy reported that the particular design of the Syrian reactor is a poor man’s nuclear energy path and is in the public domain. That is why the size of the reactor is important. Not only that, but the Bush administration keeps repeating that the reactors are the same size. That shows much of their assertions re: the connection of Al Kibar and Yongbyon depend on similarity in size.
    So, yeah, size does matter. The Bush administration says so itself.