Mark HibbsThe IR-40 and Diplomacy

Because neither side has disclosed details, we don’t know for sure whether the roadmap Iran proposed to the P5+1 powers in Geneva in mid-October includes the future of the IR-40 reactor project. Iran’s counterparts have good reason to make sure that the unfinished heavy water reactor is on a short list of must-resolve issues. Once it is completed, Iran will have an installation that can generate annually in its irradiated fuel between 5 and 10 kilograms of weapons-grade plutonium. That’s enough for one nuclear weapon a year.

The powers don’t want the IR-40 to make plutonium that Iran then separates. Iran wants a versatile and safe research reactor. These issues could be reconciled by diplomacy. The P5+1 could help Iran finish the reactor if Iran provides credible assurances it won’t access the plutonium.

Since 2006, the powers have served notice to Iran in United Nations Security Council resolutions that both the IR-40 project and Iran’s uranium enrichment program are proliferation threats.  The resolutions ordered Iran to suspend both endeavors. Iran has ignored the resolutions on both counts. It continues to enrich uranium and to build the IR-40.

Unlike Iran’s enrichment plants, which are operating routinely, once the IR-40 is completed, operation of the reactor will pose safety concerns. Some of these are related to a raft of technical challenges that Iran must overcome in manufacturing the nuclear fuel for the reactor.

In part because of these issues, the IR-40 will not be finished and operated for awhile, most likely not before 2015, and possibly not for as long as Iran remains under international sanctions.

This project has been delayed at least four times since it got underway during  the 1990s. The most recent delay was revealed to the IAEA in August during a physical inventory verification. A number of items were still not in place, including cooling pumps, instrumentation and control systems, and sufficient uranium fuel.

Iran is behind on this project because it is under sanctions enforced by global export controls on know-how, materials, and equipment. Without these measures, the reactor would probably be operating today. Nearly everywhere else worldwide, research reactor projects like the IR-40 have access to off-the-shelf nuclear-grade items conforming to international safety and quality specifications. Iran is cut off from these sources.

Fuel Fabrication and  Testing 

As of September, Iran had made about 1/5 of the amount of fuel it had predicted in June it would make by then.

Iran faces two main challenges here: fabrication of fuel to quality standards, and testing of the fabricated fuel to make sure it will perform safely and as expected.

Fuel fabrication is difficult for several reasons. Iran chose zirconium for cladding material instead of much softer and more malleable aluminum. This will permit Iran to operate the reactor at higher temperatures and with higher fuel burnup–but only if Iran masters the difficulties of extruding less-ductile zirconium tubes. Add to this the difficulties in making the fuel itself. The natural uranium fuel pellets are produced in Iran in batch mode using mixers, which must be precisely controlled to ensure homogeneity, as well as sintering ovens, which must be carefully operated to guarantee a constant temperature profile and other technical parameters. Precise control is essential to make sure that the thousands of fuel pellets loaded into the  IR-40’s core are identical. If they are not, or if tubing is not precisely machined and welded, slight irregularities in geometry can result in deformation and loss of fuel integrity, possibly leading to a serious core accident.

In virtually all nuclear programs today, national regulators would not permit a reactor to operate using specially-designed fuel, produced by people with relatively little experience, without first carrying out a comprehensive fuel testing program. Tests would measure the behavior of fuel under irradiation, focusing on such factors as high temperature, chemical interactions, and both thermal and mechanical stresses. The cladding would be tested for irradiation growth, bowing, swelling, and resistance to fission gas pressure.

The fuel assemblies for IR-40 are uniquely designed long-tube bundles that do not fit into any other reactors in Iran and represent a modification of a Russian fuel design. Iran currently has no access to Russian experts to assist it in making and testing the fuel. So far, Iran has conducted limited irradiation tests of fuel assembly components only, not complete assemblies. Because of sanctions, Iran can’t test the fuel abroad.

What to do?

In theory, Iran could rely on testing individual segments of the fuel in Iran. Given Iran’s current lack of access to foreign expertise, this would not be ideal, but Iran could obtain considerable data that would be useful in determining whether the fuel made in Iran would be safe and perform as calculations would expect.

In May, the IAEA reported to its Board of Governors that Iran disclosed its intention to begin hot fuel testing in Q1 of 2014, but provided no details.

Iran currently has no management system to carry out such tests, little competency and experience, and insufficient infrastructure. The IAEA is aware of these deficits because its personnel have been to the IR-40 project on numerous occasions, including in 2010 when the IAEA carried out a formal review of Iran’s resources for performing experiments and tests which would be relevant for this project. Iran has not disclosed to the IAEA or other outsiders where it plans to carry out testing of the IR-40 fuel.

A decision by the Iran Atomic Energy Organization (AEOI), which is in charge of the IR-40 project, to operate the reactor without prior comprehensive fuel testing would challenge the authority of the Iran Nuclear Regulatory Authority (INRA). Its mission includes “regulatory review and assessment… prior to authorization… to determine whether the operating organization’s submissions demonstrate that the facility complies with the safety objectives, principles, criteria, and regulations approved by INRA.” Presumably, that would include preparations for commissioning a reactor of unique design loaded with fuel that was produced by a fabricator without a demonstrated performance track record.

Were Iran to systematically carry out comprehensive quality-assurance tests for all the hand-made fuel it is producing for IR-40, detection of any irregularities should delay commissioning of this project further since Iran must first identify and rectify any quality control problems. Should Iran instead assume the risk of commissioning and operating the IR-40 without taking such care, that course of action should prompt objections–both in Iran and elsewhere–that the AEOI is proceeding without heeding advice from the IAEA, which is based on international good practices.

What should Iran and the P5+1 do? The best course of action would be to include IR-40 commissioning in the basket of issues subject to a negotiated settlement. The powers don’t want Iran operating a reactor to accumulate weapons-grade plutonium in spent fuel, which it could reprocess. Iran needs a research reactor that it can use for a variety of civilian applications and that Iran can be confident will be safe to operate.

As part of a negotiated comprehensive settlement with the P5+1, Iran could get access to foreign expertise to realize its ambitions to have a versatile research reactor. The P5+1 might agree to help Iran replace the IR-40 with a light-water reactor better suited to produce the medical isotopes Iran says it wants to produce. Should Iran–like a number of other countries–want instead to have a heavy-water reactor to maximize production of neutrons for research (and given Iran’s investment in heavy-water production infrastructure, Iran may well insist on this), the P5+1 might consent to help Iran finish the IR-40 project–provided that Iran agree that IAEA safeguards would apply in perpetuity on both the IR-40 and its inventory of nuclear material, and that the irradiated fuel will not be reprocessed.


  1. Rob Goldston (History)

    There is a fundamental problem with any deal that permits Iran to accumulate many years of spent fuel in country. The issue is that if Iran decides to break out from safeguards, there is no way, diplomatically or militarily, that it can be prevented from doing so. North Korea demonstrated this.

    It seems to me that even a muscular Additional Protocol would not resolve this problem, since Iran could construct a small underground reprocessing plant only after it had spirited away the spent fuel. Thus any deal that permits IR-40 to operate needs to include removal of its spent fuel from Iran. It would be ironic, but wise, for the U.S. to take on the burden of Iran’s spent fuel from IR-40, as Russia should do for Bushehr. Even for the case of fuel that is manufactured in Iran.

    • mark (History)


      By all means: If the P5+1 cuts a deal with iran to remove the irradiated IR-40 fuel from Iran as a condition for finishing and operating the reactor, that would without a doubt be the most robust of what I called in the post “credible assurances” that the plutonium won’t be separated.

    • FlamesInTheDesert (History)

      A better deal would be for the west to supply the fuel as russia does for bushehr,of course the sticking point here would be the surety of supply,but an agreement to swap spent fuel for new fuel plus a suitable stockpile on iranian soil might be a possible solution

    • kme (History)

      Doesn’t this position amount to an admission that safeguards are almost completely pointless? Why go to all the effort of implementing safeguards if it’s considered that they don’t present a significant barrier to any moderately advanced industrial NNWS weaponising?

    • mark (History)

      No. Far from pointless. The safeguards that would be in place would assure that the fuel is not being diverted for non-peaceful uses. Of course, if Iran were to throw out the inspectors, then that’s that. North Korea did that. Having Iran commit to perpetual safeguards on the reactor and the fuel inventory would not give Iran that legal option under NPT Article X. If there is such an agreement with Iran, quitting the NPT would not give Iran legal access to the fuel to use in an explosive device.

      Consider the situation in Bushehr. The IAEA is safeguarding all the Bushehr fuel inventory. Confidence that the Bushehr fuel is all accounted for is extremely high. The same would be the case for fuel which Iran would make for IR-40. If Iran decided to snatch the fuel and send the IAEA packing, they can do this and take whatever risks and consequences which follow. But it doesn’t mean that the safeguards are meaningless. Safeguards’ effectivness depend on states fulfilling their legal obligations under their safeguards agreements. If they don’t do that, then there are procedures which if followed to the letter would result in sanctions or other actions in the case of serious violations.

      The IAEA would follow the uranium trail from start to finish. To be extra sure, there would be an Additional Protocol in force for Iran, plus whatever else the P5+1, Iran, and the IAEA may agree to as part of a comprehensive solution to the crisis. The weaponization issue is a different track. There, Iran and the powers may agree on a list of specific activities related to weaponization that Iran would agree not to do. One last point: In addition to IAEA safeguards on the IR-40 and its fuel inventory, I would want to make sure that the facility attachment/subsidiary arrangements for this project would permit the IAEA access into the plant deep enough to make sure that there is no clandestine plutonium production possible via irradiation of targets or undeclared material inside the reactor or associated with irradition ports that the IAEA doesn’t know about. This is very important because Iran is not implementing its modified Code 3.1, meaning that the IAEA is not able right now to have enough design inforomation about the IR-40 to come up with a safeguards plan. Not being able to do that is putting the IAEA at a disadvantage and in theory could make it easier for Iran to hide diversion paths associated with the reactor.

  2. nukeman (History)

    It should also be noted that the irradiated heavy water moderator will be a significant source of tritium. Iran has the knowhow and the ability to reprocess spent reactor fuel. Two of their leading nuclear scientists studied in Germany and actually have patents on spent fuel reprocessing to produce plutonium.

    • Rob Goldston (History)

      In terms of neutron economy, a gram of tritium produced from Li-6 costs you about 80 grams of Pu produced from U-238. So if you want 20g of T, give up 1.6 kg of Pu. This is probably easier than extracting grams of T from tons of heavy water. But you are right, in principle you could get something like 10g/a from IR-40.

    • JO (History)

      Cryogenic tritium removal from heavy water is an economic way of producing tritium – India does it for example. The relative cost is orders of magnitude less for separation versus lithium targets (assuming heavy water reactors are in use). Tritiated heavy water is a nasty contaminant: two birds with one stone from a separation process.

    • Rob Goldston (History)


      Interesting comment about India’s tritium production. Do you have a reference I could chase down to understand this better? There is of course lots of interesting discussion about India’s nuclear weapons program, and their tritium production rate might be shed some light on what they are doing.

      – Rob

  3. nicoullaud (History)

    thanks a lot, quite informative…

    Is it true that the type of fuel envisaged by the Iranians for this reactor (uranium oxide) is less proliferation prone than fuel using uranium in metallic form? is the difference significant?

    What about producing plutonium of military quality by introducing natural uranium elements in the core or disposing them around the core?

    • mark (History)


      Uranium metal fuel would be somewhat easier to reprocess than UO2, but the difference is not a showstopper by any means. India’s weapons reactor uses UO2 fuel which India reprocesses without great difficulty (there were some problems in the early days in India related to reprocessing, having to do with fines from chopped up fuel rods, but that’s unrelated to separation of plutonium from UO2 fuel).

      All the fuel for this reactor will be natural uranium. See above my comments concerning what you call “disposing them around the core.” This is an issue. It would appear that one or more nuclear-armed states with an Infcirc-66 safeguards agreement for a reactor probably did clandestine irradition as I describe in my remarks to kme above.

  4. kme (History)

    Planning to operate the reactor at high burn-up appears to be a good sign from a proliferation perspective. What about high-temperature operation – is that proliferation-relevant?

    • mark (History)


      No. If the fuel is made without flaws, then the zirconium will permit higher burnups (which is good for nonproliferation) and higher core temperatures, which is proliferation-neutral.

    • Cthippo (History)

      On this thread, isn’t fuel lifetime also a non-proliferation tool? I know that the longer you leave the fuel in there the more PU-240 it ends up contaminated with, but is that enough to make it unusable in weapons?

  5. Tobias Piechowiak (History)


    nice article. I also enjoyed your essay about the potential risk for the URENCO sale on carnegieendowment.

    It probably depends on how much fissile material the Iranians already amassed, may it HEU or Pu.
    Then they could refrain from completing IR-40 as a sign of goodwill. Another things that strikes me is their new IR-3 centrifuge. That smell almost like russian technology transfer since they now seem to focus on subcritical centrifuges. Anybody know if there could be something to it?

  6. Mansoor (History)

    Excellent analysis of the “other side” of Iran’s nuclear effort. In some ways, Iran is following Pakistan’s pathway to the bomb in terms of public posturing of pursuing parallel routes to producing fissile material–with the world’s attention focused on the enrichment route while building plutonium capabilities quietly.

    Similarly, Pakistan had in the 1970s publicly raised the stakes by seeking to acquire a commercial reprocessing plant from France while it was surreptitiously but steadily developing the complete nuclear fuel cycle (including uranium processing, conversion and enrichment facilities along with a fuel fabrication and pilot reprocessing plant).

    Once these plants were completed by 1980-81, it subsequently shifted focus on completing work on the remaining elements for the plutonium route (production reactor, heavy water and tritium plant) while the world’s attention was diverted to Pakistan’s enrichment efforts.

    • Bahram Khodadoost (History)

      The analogy with Pakistan is perfect, with the minor difference that Iran does not seek to make nuclear weapons.

    • mark (History)


      Iran systematically hid activities from the IAEA for 18 years. Why should I have confidence that without more intensive verification Iran wouldn’t irradiate undeclared targets at IR-40 if they felt they could get away with it–leaving aside whether they want weapons or not?

    • pak (History)


      Iran systematically hid some activities in way to gain technical development, unfortunately with negligence and failures, but without diversion intention. There is no indication of diversion, except some allegations which have been continuously raised since more than 20 years ago.

      I don’t think that your lack of “confidence” is the result of Iran’s hidden activities for 18 years, because, your colleagues in the US government forced international companies, about 30 years ago, to not help Iran in completing Bushehr reactor, do not supply fuel for the Tehran research reactor, and so on.

      And now, we have to be accounted for some failures that have nothing to do with non-proliferation concern. Meanwhile, your colleagues in the US government lead 5+1, with persistent policy, to compel Iran forego its rights and accept new limits!

    • Bahram Khodadoost (History)

      Mark, A request for verification would be reasonable. Ultimately, verification could be part of a final deal that does not limit civilian applications and recognizes a civilian program that does not depend on outsiders for fuel, under a well-defined regime of inspection that doesn’t turn Iraqi-style. (As Scott Ritter charged and Hans Blix acknowledged, Iraqi inspections came to encompass matters related to state security that went well beyond the UN team’s WMD-related mandate.) The problems Iran has faced and continues to face are of a completely different order. For the most part, the problem has not been about verification; it’s been about the very existence of an autonomous civilian program.

      Iran faced active opposition and sabotage for activities that should otherwise have been considered within its rights under the NPT. In previous decades, if Iran tried to openly purchase something for its civilian program, the sale would be quashed. Consequently, in making innocuous purchases Iran conceals the destination and purpose of the goods.

      It was par for the course when the US prevented Iran from purchasing fuel plates for the Tehran Research reactor, effectively holding cancer patients as hostages, even though the fuel plates were not for military purposes. Previously, the US used to oppose the Bushehr power plant, too. Moreover, previously Iran would not have been allowed to import 5%-enriched uranium from Western European countries, even though it owned part of an European enrichment plant.

      Besides, after Iran began negotiating with Europeans in 2003, suspended enrichment, voluntarily implemented the AP, and raised the prospect of ratifying the AP as part of an agreement, it came to light that the intended endgame for Europeans did not include Iran’s right to a civilian enrichment. The Europeans thought of “suspension” as a face-saving euphemism for termination, not as a step towards verification.

      Iran derived the lesson that it needs a relatively autonomous program, one that outsiders cannot switch off at will. As long as Iran is granted this, it will allow verification.


  7. shaheen (History)

    Mark, useful as usual.

    But I wonder how much any arrangement that would leave – after IR-40 has operated for several years – Iran a few weeks away from producing one or several SQs would be acceptable to all parties. As you know it would be very difficult to ensure that there is not a small reprocessing plant somewhere in the region.

    An arrangement that ensures that the irradiated fuel is taken out of the country would be more reassuring.

    Alternatively, why should we expect that an operational IR-40 should be part or a deal? After all, we have never heard from Tehran any claim about a so-called “inalienable right to operate a HW-moderated NU-fulled reactor”? So face-saving would not be an issue. Suspending work on Arak (implicitly for eternity) could be part of a deal, provided that the international community guarantees Iran access to any civilian/scientific benefit that Iran could reasonably expect from the IR-40.

    As you know, all countries which operate such large, HW/NU reactors did it or planned to do it for more than mere civilian/scientific reasons.

    • mark (History)


      I proposed that the reactor be permitted under a deal to be finished and operated on the basis that there are effective assurances that the plutonium won’t be separated. I agree with you that if the reactor is completed and operated, removing the irradiated fuel would be the solution which would represent the least proliferation risk. See my above comment to Rob Goldston on this.

      Tehran has not claimed “inalienable rights” in this case, correct. I would make a difference in this regard between rights claimed for reprocessing (which is clearly a dual-use nuclear activity–mlitary and civil–in the same sense as uranium enrichment) and rights which might obtain to operating of the research reactor itself. You could argue that the reactor produces plutonium in its spent fuel in the process of conducting nuclear research or producing isotopes, and that that would qualify operation of the reactor also as dual-use. But that is a fundamental aspect of nearly all nuclear activity. So where to draw the line about where “rights” apply under Article IV and where they don’t? After all, the Bushehr PWR is also making plutonium in its spent fuel which is going back to Russia after it is cooled off.

      The fact that Iran has not claimed “inalienable rights” in this case makes it more likely that the fate of the IR-40 can be negotiated by all six parties involved.

      The heavy water in the design of course raises the proliferation profile of the reactor by facilitating neutron capture, but there are quite a number of HW reactors which are used purely for research, and I would note that the same properties of the moderator in these reactors are seen as an asset by nuclear researchers likewise because the HW serves to increase the neutron population of the system.

    • Cyrus (History)

      Iran has in fact explicity stated, repeatedly, that it has no interest in reprocessing. Without a reprocessing facility, whatever plutonium generated by this reactor, simply cannot be used.

    • mark (History)


      See my reply to your comment concerning Iran’s statement to me that Iran had “no interest” in uranium enrichment. So maybe they also have “no interest” in reprocessing. Or maybe they do.

      The point being that Iran hasn’t generated a lot of confidence over the last 20 years about the extent and scope of its nuclear problem. All the whining about the details doesn’t overcome this one single fact. That is the context in which the PMD allegations must be seen.

  8. Rene (History)

    Thank you Mark for this illuminating post.

    I have a question that’s not directly relevant to the IR-40 but to Dr. Kemp’s post, but it seems the comments thread for that post is closed, so I’ll ask my question here:

    When we speak of asking Iran to limit its enrichment capacity, what kind of concession are we talking about? Is it something like “we commit to not increasing our enrichment capacity until all outstanding issues and PMD questions are resolved” or “we commit to not increasing our enrichment capacity for 20 years and to renegotiate this agreement in 2033 based on mutual interests” or “we commit to never ever increase our enrichment capacity, even if the sky falls down”?

    • mark (History)


      In your question you covered the gamut of basic options. An acceptable compromise would most likely be something in the middle. Beyond the resolution of the important resolution of PMD issues, it is likely that simply because of the tinder-box nature of proliferation threats in the region, the powers will want Iran to volunteer to limit both the capacity and enrichment level, as you suggest, but also the number of locations where enrichment in Iran is carried out. The latter is important because there is a history of Iran preparing to do enrichment for a long time and then it becoming revealed without advance notice–both in Natanz and in Fordo. Anything Iran can do to assure people that there aren’t any enrichment plants secretly in the making that suddenly appear with tunnels, bunkers, and centrifuges being installed in them would help generate transparency and confidence.

    • Cyrus (History)

      Mark you mischaracterize Iran’s facilities at Natanz etc. They were not “secret” — undeclared is not the same thing as secret. Iran has a safeguards agreement in force, and it specifies when a facility has to be disclosed. Just because a site is not yet disclosed doesn’t make it a “secret” site or cause for suspicion. Whatever undeclared activities Iran had in the past were of LEGAL activities, and according to the IAEA involved no diversion of nuclear material and have since been resolved to Iran’s favor pursuant to the Iran-IAEA modalities agreement. The only thing that’s left is the “Alleged Studies” which the US has refused to provide the necessary documentation, and furthermore claims about Iran’s alleged “intentions to acquire capabilities” to make nukes in the indefinite future that do not amount to a legal challenge in any way to Iran.

    • mark (History)


      Your narrative is off the mark. This whole enrichment program was secret for years, in fact from its very outset until 2002-2003 when it was exposed.

      I asked Iran’s representative to the IAEA back in the early 1990s if Iran had any interest in fuel cycle research and development including uranium enrichment and his answer was: “No–we have no interest in this.” I printed that in an article in Nuclear Fuel. It elicited no comment–ever. Not by western officials because they knew it and they kept what they knew secret (well, almost secret). None by Iran.

      This statement by Iran however was, to put it frankly, a lie. It was not a misunderstanding. This guy had a PhD from University of California at Berkeley. He knew what he was doing and he knew what Iran was doing and he know what he was saying to me. It was a point blank answer. He knew me personally, we had had many discussions for several years prior to that Q&A. So he knew who he was talking to and in what capacity he answered my question.

      At about the same time, I was printing in other articles off and on various details about Iran’s centrifuge enrichment procurement activities based on allegations I had learned of from Western government documents. This was in the early 1990s. By telling me Iran had no interest in uranium enrichment and in fuel cycle research and development Iran was in effect denying my conclusions following my research. I look back at the record of my research from 20 years ago. I have nothing to retract.

  9. Gero von Randow (History)

    The rephrase what Cthippo wrote: What about guaranteed high burnup as nonproliferation measure?

    • mark (History)


      Increasing the burnup would reduce the quota of Pu-239 in the irradiated fuel. But it would not make it prohibitively difficult to reprocess the irradiated fuel. The reprocessing would require more time and effort, but Iran could still recover the Pu-239. That said, on the margin it would be easier for Iran to recover the Pu-239 if burnup was low. In this case I wouldn’t want to rely on this as a proliferation barrier. More important would be agreement by Iran not to reprocess the irradiated fuel and–as others have also said in comments to this post–a takeback of the spent fuel by the US or another state with skin in the game.

    • kme (History)

      Isn’t separating the Pu-240 contaminant from the Pu-239 essentially the same problem as separating U-235 from U-238?

    • Rob Goldston (History)

      Some technical info that most readers know, but maybe not all: You get a lot of Pu per MW-yr from a natural-uranium-fueled reactor. This is because natural uranium has a high ratio of U-238 to U-235. So the U-238 gobbles up a lot of neutrons and makes Pu, compared to the U-235 that gobbles neutrons and makes fission. (The gobbling tendency of all that U-238 is why you need non-gobbling moderators such as heavy-water or graphite.) Now the Pu does get converted to Pu-240 over time, like in any other fission reactor, and it is very unlikely anyone is going to separate Pu-240 from Pu-239. For one thing, the mass ratio is too small. For another Pu is nasty stuff. You prefer people you don’t trust to go to higher burn-up, since Pu-240 generates heat and neutrons that make it harder to turn it into bomb. But the folks who know say that this does not make it qualitatively harder. Hope this helps.

  10. Gero von Randow (History)

    It was my understanding that high burnup would make the isotope mixture “too hot to handle” (at least to a certain degree). But probably I got it wrong and I will reopen my textbooks 🙂