Geoff FordenTRR Refueling and Proliferation Barriers

Much has been said about the decision to refuel Iran’s Tehran Research Reactor (TRR) and many have worried about possible consequences for proliferation that supplying Iran with uranium enriched to 19.75% U235 might imply. As I hope I made clear, I am a big supporter of this deal. However, we definitely want to make sure the details of the deal reassure the West, not increase their suspicions; something that has already occurred with John Bolton. (Ok, perhaps it is impossible to increase Bolton’s suspicions of either Iran or the Obama administration, but you get my point.) Unfortunately, the form of uranium that Iran is currently using, powered U3O8 pressed and perhaps sintered into a box, just about begs to be further enriched: it’s the starting point for the production of UF6. Perhaps it needs to be broken up a bit but, as I showed in my recent post on molybdenum contamination, if the West supplies Iran with fuel in that form, it could be enriched to weapons grade in one step through the centrifuges without worrying about damaging the apparatus.

Three Components to Effective Non-Pro Barriers

While considering steps to increase the proliferation resistance of the TRR fuel, it is important to remember that there are three components to any proliferation barrier and that all three must work together to produce an effective barrier. These components involve technical, legal, and organizational aspects. Without all three components, no barrier is effective. The example I like to give is the humble inspection seal: a piece of tape used to ensure, for instance, that a piece of equipment is not moved or used for illicit purposes between inspections. The seal itself is the technical component and quite a bit of work has gone into its development to ensure it is not surreptitiously replaced or tapered with. However, a seal is meaningless without the legal component that gives some agency (like the IAEA) the right to periodically inspect the seal. And, of course, you need the organizational component so that you have trained inspectors who actually go and inspect the seal. When we think about creating barriers around the TRR refuel load, we do not need to think of a technical barrier that works independently of other barrier components.

Let us start off with the technical component. In designing this portion of the barrier, we cannot impose onerous conditions on the use or form of the fuel. For instance, it must meet all relevant ASTM specifications. Anything else could, and should, be immediately rejected by the Iranians. This severely limits some of the ideas people have posted here. (It rules out “salting” the fuel with a radiation barrier such as U232 since ASTM standard C996 limits the uranium isotopics.) At the same time, if the West imposes conditions like requiring all the LEU feed stock for the TRR fuel to be shipped out at one time, it seems only equitable that the enriched fuel should be shipped back as expeditiously as possible. It is conceivable that the first reactor load fabricated might arrive first to allow the Iranians to install it and convince themselves that what ever form it is in satisfies their needs without redesigning their reactor. But after the Iranians give their go ahead, the future loads should be made and shipped back as fast as possible.

The first technical component should be a new form of the uranium. There are, however, a number of requirements it should meet. First, it should, as much as possible, have the same neutronics (i.e. the same density of uranium in the fuel rods) as the old fuel. It turns out that quite a bit of thought went into designing new types reactor fuel rods during the 1990s when there was a push on to switch from research reactors using essentially weapons grade uranium to “low enriched uranium.” Low enrichment is rather arbitrarily defined as less than 20 percent, presumably because the resulting critical mass is over 400 kg; an amount that just doesn’t seem weaponizable to most people.

Beryllium to the Rescue

Based solely on reactor uranium loading, the best match for Iran’s TRR fuel rod density (of 2.96 grams of uranium per cubic centimeter, gU/cm^3, according to the Iranian publications) is alloying uranium and beryllium in the form of UBe13 (i.e. thirteen atoms of beryllium for each atom of uranium). Perhaps not surprisingly, the resulting density of the alloy is very similar to the uranium oxide powder Iran current uses (4.37 g/cm^3 as opposed to the current 4.76 g/cm^3: not even the mechanical support structure of the rods will need to be changed. Replacing the oxide with a beryllium compound immediately increases the time it would take Iran, if it chose, to convert it to UF6 for further enrichment.

It is very doubtful that Iran already has much experience with industrial scale beryllium processing. And while it is conceivable that Iran could acquire lab-scale expertise at separating uranium from beryllium using natural uranium in its tests, it would be a major technological challenge to take experience and turn it into a process for efficiently separating 150 or so kilograms of uranium. After all, the Russian nuclear industry doubted its ability to reprocess uranium-beryllium alloyed spent fuel rods as late as 2006.

If UBe13 alloy fuel rods represent a very significant time and technology barrier to further enrichment, they do not represent an absolute barrier. That is where the other components to proliferation resistance come in. Clearly, an increased time barrier is most effective when it is accompanied by either periodic or continuous monitoring of the fuel supply; something that was already in place for existing fresh fuel stocks. For most of the time Iran had fresh, 19.75% enriched U3O8 fuel rods sitting around waiting to be used (since 1992) it did not have a significant enrichment capacity. Now that it has, it is only natural that the periodicity of the monitoring be increased.


The best situation would be if the fuel supply was remotely monitored by video cameras. However, the procedures for such continuous monitoring have not been worked out with Iran, even for Natanz, so we should at least insist on video monitoring of the fuel with periodic film pickup—typically once every two weeks. This together with the safeguards already in place at Iran’s other nuclear facilities—at the enrichment halls at Natanz and the conversion facility at Esfahan—should provide the organizational components.

Finally, Iran needs to agree to the monitoring of this fuel regardless of its future NPT status, i.e. placing the fuel under safeguards in perpetuity. This legal component is obviously necessary and fair. The P5+1 is agreeing to provide a service, the further enrichment of uranium, that will in all practical circumstances be irreversible with the understanding that it will never be used for weapons purposes; something Iran says it agrees with.


  1. Yale Simkin (History)

    I think assisting Iran with beryllium technology would be a bad idea.

    Also is the extra processing step required to convert UBe13 into a form suitable for fluorinating very burdensome? Particularly with the small quantity of feed material? The total volume would fit in four 2-liter Pepsi bottles – with space left over.

  2. Hairs (History)

    As I’m sure you all know, beryllium produces neutrons quite nicely when under alpha bombardment. So, given all the alphas that are emitted by the various fission products, the mixture described is going to change the core’s neutronics. Does anybody have any idea whether such a modified core would still be within its normal range of control without some change of enrichment, absorbers / poisons, or dimensions?

  3. Geoff Forden (History)


    No one is talking about assisting Iran with beryllium technology; quite the opposite. After all, the fuel rods would be manufactured in the West.


    UBe13 has been proposed for just such reactors.

  4. Yale Simkin (History)

    Geoff, let me expand and clarify my point.
    Right now, if Iran is found messing around with beryllium processing and machining in any way, but particularly if it involves neutrons, it would look very suspicious.

    If, on the other hand, their nuclear infrastructure becomes dependent on sophisticated application of beryllium nucleonics, chemistry, and materials processing, they may claim a right to have trained technologists, facilities, and most importantly, justifiable non-suspicious traces of beryllium contamination.

    Iran is not likely to simply accept “blackbox” technology with a “No Iranians Allowed” sign at the door.

  5. rs (History)

    Concur with Hairs –
    The question is whether the UBe13 fuel would work in the TRR. The neutronics are probably very different (sorry, my references are at work), and the TRR may not be able to accommodate the difference without significant redesign.
    This might be a good reason to propose the change, particularly if the redesign gives more international access to Iran’s program.

  6. Hairs (History)


    My apologies if I wasn’t very clear with my first, quickfire comment: I don’t doubt that a reactor can be built with UBe13, nor that using UBe13 has been proposed as a replacement fuel. What I have doubts about is whether UBe13 fuel could simply be loaded into the existing TRR and everything would work just as before.

    If such a radical change of fuel were suggested for a functioning reactor in the USA or UK, the NRC or NII respectively would have an awful lot of questions that they’d want answers to before it was ever licensed to commence operation. There would also be the obligatory re-training of the operators, new safety cases, new procedures, etc etc.

    The Argonne document you link to (and the references and appendices therein) is very heavy on fuel fabrication and metallurgical / irradiation testing of mini-plates, but it’s extremely light on the required changes in core design and, particularly, any associated changes in controls. In fact, as far as I can Google, no-one has yet converted an existing HEU reactor to UBe13.

    All in all I suspect there is every chance that you cannot simply load the new fuel into the TRR and expect it to work as before. And even if you could, the responsible thing to do (in fact, the legal obligation in most countries) would be to update the entire safety cases, procedures, training, etc; all of which takes time. Given the current shortage of Mo-99 production and the Iranians’ economic case for producing it at home (both very well documented here on ACW), Iran might not be inclined to accept the necessary delays to isotope production that a change of fuel would require. If we then add the paucity of worldwide experience in using, or converting to, this type of fuel, then the Iranians would have plenty of reasons to resist any proposal to change to UBe13.

    I’m not claiming that a change of fuel can’t be technically done (I’m sure that it can), but I don’t think it would be as easy as just supplying rods with a different composition. More importantly, there seem to be perfectly reasonable grounds for Iran to reject such a proposal, if they so wish.

  7. Yale Simkin (History)

    I re-read my original post and noted an error – it is the product that would fit in the 4 2-liter bottles. The feed would instead fill a kitchen wastebasket, larger, but not with a vastly greater materials requirement.

  8. Tom Bielefeld

    Very interesting post, Geoff!
    Any idea who can produce the proposed UBe13 alloy fuel rods (also, how fast) and how much (more) money they would cost?

  9. Nick (History)

    TRR is an IAEA safeguarded site, any modifications, tampering, or refinement will be noticed. For 30 years Iran has been sitting on TRR fuel rods without any attempt to take it to the next level, it is hard to imagine that they would do it now, given that they have a robust fuel cycle dealing with hex.

  10. Geoff Forden (History)


    Iran is not going to be allowed to reprocess that fuel. I’m sorry that I didn’t explicitly state that but I assumed it was understood.

    As to “industrial scale,” perhaps I should have stated “pilot plant scale.” It is still definitely bigger than laboratory scale.


    Russia has produced UBe13 fuel rods. They were used in some of their naval reactors. It seems, based on some of the papers I read, that it is easier to make the fuel rods than reprocess them. As to costs, well, who knows and who cares. If Iran is vastly undercharged for more proliferation-resistant fuel rods, it will be a well-spent opportunity cost for the West provided that UBe13 is really as resistant to proliferation as my initial estimates seem to imply.

  11. Yale Simkin (History)

    I was not referring to reprocessing.

    The beryllium issue is that Iran would likely demand to have it have expertise and materials capability to manage the fuel being placed in it’s reactor. On the fresh fuel side, not the reproc. end of the cycle.

    I suspect they would demand (and have every right) to become competent in both the knowledge of beryllium nuclear characteristics and quite possibly, capability to fabricate fuel plates. They certainly weren’t satisfied buying finished fuel assemblies from Russia nor staying ignorant of Zirc/UOxide nuclear properties.

    The completely separate second issue – the diversion of the non-irradiated fuel in a breakout I do see as less important. It is however a window where in extremis Iran could divert and extract enough uranium to build a bomb without further enrichment.

    If the Russians hand carry the fresh fuel, monitor its immediate loading into the core, and wait for it to be irradiated, while simultaneously exporting the spent fuel, then I don’t see a problem.

    There is of course the long term issue of them extracting irradiated fuel, reprocessing, and building a bomb, but if they are that gung-ho in breaking out, they are only weeks to HEU by directly enriching the ten tons of permanently un-irradiated reserve fresh LEU that our dear friends the Russians have supplied the Iranians.

    That is primo LEUO2 ready to be hammered to powder, have any rare earth poisons extracted, and converted to hex on the existing production line.

  12. Geoff Forden (History)


    I’s sorry but you are missing the whole point of this exercise. We want to prevent Iran from ever enriching to 19.75%. There is no way that we would give them any technology associated with the fuel. When you buy a PC (or Mac, I suppose) do you have a “right” to the technology for the CPU? Of course not. Technology transfer is not part of the deal. And it certainly is not part of the deal with Iran. If you need further evidence of this, consider Iran’s purchase of the current fuel in 1992. Argentina certainly did not give Iran any of the technology for producing anything about the 19.75% LEU fuel they sold Iran. So, lets not invent new “rights” for Iran.

  13. Bob Reed (History)

    I tend to agree with Yale on the provision he mentioned of having the Russians oversee this entire process, as opposed to the IAEA, who have already demonstrated their willingness to cover for Iranian violations of agreements…

    Although the Russians may be ale to be swayed by badly needed cash and military hardware sales to the Iranians, if we compensated them nicely for fulfilling this oversight role, the combination of cash and heightened international status might make for an effective program…

    I am inclined to agree with your assertion of the need for technological, legal, and organizational components. But at this point, in the wake of the IAEA/El Baradei scandal, along with other failed UN administered programs, I would prefer a third party, like the Russians, to middleman the deal…

    All the best

  14. Yale Simkin (History)

    Please be assured that I am aware of the reasoning behind the proposal. I am not trying to derail it.

    I am just being devil’s advocate, because its a truism that over extended time, unintended side effects become the dominant effect.

    I don’t want to stretch out the discussion, so I will leave it with the observation that Iran has, from 1967, been heavily involved in uranium oxide-based reactors. Except for the actual enrichment technology, there was nothing special that Argentina brought to the table when updating the TRR to use 20% fuel. The Iranians certainly knew everything there was to know.
    They could certainly create new fuel elements if supplied with u-oxide.

    Beryllium is a different bird, altogether. The reactor engineers must become intimate with the new fuel and the nuclear and physical properties of a weapon-enabling technology.

    Maybe, I’m over-heated in my concerns, but my trust of the Iranians would have to increase astronomically to reach up to zero.