Jeffrey LewisSafeguarding Breeder Reactors

A pair of interesting stories appeared over the past week relating to negotiations about which facilities India will declare “military” (and hence, exempt from safeguards) under the US-India Nuclear Cooperation Agreement.

Siddharth Varadarajan reports that the Bush Administration has dropped its insistance that the CIRUS reactor be listed as a civilian facility. (For some background, see: CIRUS, December 20, 2005).

Varadarajan claims the new deal breaker is that India will not place its Fast Breeder Reactor (FBR) program, including the FBR at Kalpakkam (at right), under safeguards. The Indian Express carried a scathing op-ed (and, according to one reader, water for the PM’s office) lambasting India’s Department of Atomic Energy (DAE) for intransigence on the FBR:

The failure of the latest round of consultations between Foreign Secretary Shyam Saran and US Under Secretary of State Nicholas Burns last week has been blamed on DAE’s reluctance to put its fast breeder programme on the civilian list.

Forget the Americans for a moment. Indian public has a right to know the nature of the breeder programme — is it civilian or military? The DAE apparently wants it both ways: a peaceful facility with future military options. It is this twisted logic, backed by decades of political self-deception, that has landed India in a nuclear mess. It neither has a successful civilian nuclear power programme nor a purposeful weapons programme. Sanctimonious rhetoric over the decades from the Indian political leadership that the nation’s nuclear programme was entirely for peaceful purposes resulted in a mixed mandate for the DAE and the loss of operational clarity. Separating civilian and military programmes and making them both efficient has been a long-neglected national need. After claiming the lion’s share of the nation’s R&D money for nearly six decades, the DAE today produces barely 3000 MW of power. On the strategic front, instead of building the necessary plutonium production reactors, the DAE has got into the bad habit of using its civilian programme for military needs.


Safeguarding A Breeder Reactor

DAE, according to Varadarajan, argues that safeguards are infeasible for an FBR program:

Allowing IAEA inspections will seriously compromise the quality and scope of ongoing research, nuclear scientists who have worked closely on and led the breeder programme told The Hindu.

“Moving fuel from one section to another would then require informing the IAEA in advance, waiting for their inspector to arrive and approve, and then executing the task concerned,” said one former DAE scientist. Asked at what stage he would be willing to offer the breeder technology for inspections, another senior retired nuclear official said there was no reason to ever subject breeder reactors to safeguards.

I have my doubts about this technical description of safeguards—because the Japanese FBRs do have safeguards (and I don’t think they work remotely like that). As I understand the problem—and this goes way beyond my current knowledge of safeguards systems—FBRs have some technical quirks that require innovation but … Oh hell, why don’t you just read the same thing I did from the Japanese themselves:

In fast breeder reactors and advanced thermal reactors, there exist difficult-to-access areas where direct verification of the inventory of fuel in the reactor core is impossible. In such cases, it was agreed that dual containment and surveillance (c/s) measures to monitor the movement of fuel to and from the reactor core were to be applied in order to meet inspection goals.

The surveillance systems for the experimental fast breeder reactor JOYO and the prototype fast breeder reactor MONJU of PNC consist of Modular Integrated Video Surveillance (MIVS) systems and the radiation monitoring systems.

At JOYO, 4 radiation monitors have been installed in the fresh fuel storage, on the cask car, on the ex-vessel transfer machine and in the spent fuel storage pool as shown Figure 1 (Y. Hashimoto, 1994).

The entrance gate monitor (ENGM), for example, is a passive neutron coincidence collar, a type of neutron coincidence counter which can distinguish the spontaneous fission neutrons of Pu isotopes from the background. A total of 24 helium-3 detectors are housed in the detector area, and the detector can identify the direction of the fresh fuel passing through the entrance gate.

MONJU is the prototype fast breeder reactor in Japan designed to have the electricity output of 280 MWe. PNC started its construction in the autumn of 1985 in Tsuruga site and the MONJU reached initial criticality in April 1994 (S. Usami, 1995).

The verification of the flows is designed to be made with fuel flow monitors measuring radiations, which can abridge the inspector attendance during the fuel handling. The monitors consist of the ex-vessel radiation monitors (EVRM) and the exit gate monitor (EXGM). EVRMs are equipped to the ex-vessel transfer machine which charges and discharges fuel assemblies into and out of the core and the external vessel storage tank, EVST (a sodium tank located in an adjacent area to the reactor containment building and stores temporarily new and irradiated fuel assemblies). Flows of fuel assemblies into and out of the core and EVST can be monitored by EVRMs, and the flows of irradiated fuel assemblies into the spent fuel pond (water pool) can be monitored by EXGM, including the flow direction. These monitors have been developed through JASPAS and the joint R&D by PNC and US-DOE involving Los Alamos National Laboratory, Sandia National Laboratory and Oak Ridge National Laboratory.

Beyond the above monitors, IAEA installed the entrance gate monitor (ENGM) using neutron coincidence counter and several optical surveillance cameras. By using this equipment the transfer of fuel assemblies inside the facility can be monitored in an un-attended mode and routine inspections are made monthly.

For the advanced thermal reactor Fugen, the development of a fuel gate monitoring system is now under way.

In principle, I’d be willing to accept the same concession to India—assuming this is really about hunting bears, that is.

In the meantime, let’s actually find out how difficult it is to safeguard a breeder:

I’ve already ordered from ILL:

  • Hashimoto et al, “Development of Plutonium Fuel Monitors for the Experimental Fast Reactor JOYO,” Proceeding of a Symposium on International Safeguards, IAEA-SM-333/51, P427-438, Vienna, 14-18 March 1994 and
  • Usami et al, “Safeguards in Prototype Fast Breeder Reactor Monju,” 5th International conference on Facility Operation-Safeguards Interface, Jackson Hole, 24-29 September 1995.

Additional technical commentary about the challenges of safeguarding a breeder reactor is requested, particularly the PNC/DOE joint venture involving Los Alamos, Sandia, and Oak Ridge.


  1. Papa Ray (History)

    Question: Why should we even care about this?

    Explain please.

    Papa Ray
    West Texas

  2. Jeffrey Lewis (History)

    By “this” do you mean (1) the India deal in general? (2) The desire to put the FBR progam on the civilian list? (3) The technical challenges associated with safeguarding a breeder reactor?
    My answers:

    (1) The India deal is important because of its implications to the the nonproliferation regime.

    (2) The US wants to safeguard the FBR program because:

    At the heart of the U.S. insistence on safeguarding the fast breeders is its reluctance to accept India as a nuclear weapons state, scientists familiar with the programme’s potential weapons application say. Though India wants breeders for civilian purposes, a breeder reactor can also be used as a “laundry” to breed weapon-grade Pu-239 from reactor grade plutonium (Pu-240) generated by pressurised heavy water reactors (PHWRs). Placing the breeder programme under safeguards, then, ensures that the reactors are never used as a “laundry”, effectively limiting India’s ability to produce fissile material through this route.

    (3) The technical challenges involved in safeguarding a Breeder Reactor are inherently interesting from my point of view.

    I’ve been easy on the moderating of comments lately, but this is really intended to be a blog for wonks rather than laypersons.

  3. Arrigo (History)

    A convoluted question: what safeguards, if any, are in place at the SuperPhenix reactor near Lyon?

    The reason I ask is that the obvious answer “none, it is in France” does not completely cover the issue since it is owned in the measure of 1/3 by Italy via Enel and was due (is due?) to receive its proportion of waste and, I assume, the “bred” fissionable material.

    Of course since SuperPhenix never really produced anything having been plagued by reliability problems and accidents this is a relatively moot question but surely the protocols must have been in place before the control rods were raised.

  4. Anonymous

    I think this discussion is missing a few key point.

    Although breeder reactors can be used to produce weapons-grade plutonium, as the article Jeffrey cites above notes, doing so would be significantly more complicated than just lowering the irradiation time of some of the natural uranium UO2 fuel pellets used in India’s commercial CANDU power reactors.

    According to the below article (among others), India reportedly left the MAPS reactors off its proposed civilian list in addition to the FBR cycle:

    This seems to me to be a much bigger story than the FBR cycle issue. Safeguarding the MAPS reactors would’ve been simple. If the MAPS reactors remain unsafeguarded, why would India use its FBR reactors for weapon-grade production? It seems more likely that India is telling the truth regarding the difficulties it would have with safeguards on the FBTR and PFBR.

  5. Yale Simkin (History)

    “If the MAPS reactors remain unsafeguarded, why would India use its FBR reactors for weapon-grade production?”

    Each MAPS reactor would produce only(!) ~160 kilograms of WgPu per year (~40-50 bombs), and in doing that require the reactors to use almost 7 times as much uranium as a purely electricity production burn-up uses. (Which India has in short supply)

    The FBR, on the other hand, involves plutonium in multiple TONS (and is inherently “super-grade”)

  6. Anonymous

    Actually, the FBR would require equal or nearly equal amounts of uranium to produce the same amount of WGPu. Plutonium is produced in FBRs by irradiating natural or depleted uranium. “Breeding” WGPu requires again using about 7 times as much uranium.

    Plutonium production in any reactor is never inherently “supergrade.” The grade of the plutonium is solely a function of how long the uranium remains in the reactor. Reactor design (PHWR, FBR, etc.) doesn’t change the fact that a growing amount of Pu-239 will accept a neutron to produce Pu-240 instead of fissioning the logner the fuel is irradiated in the reactor.

  7. Jeffrey Lewis (History)

    I think some of you are overlooking the fact that the Indian breeders are designed to utilize thorium in the place of uranium.

    Pallava Bagla in Science wrote a nice summary of the Indian effort:

    India’s Homegrown Thorium Reactor
    Pallava Bagla

    KALPAKKAM, INDIA–For more than 5 decades, India has followed its own path on nuclear power. After refusing to join the Nuclear Nonproliferation Treaty and detonating a nuclear device in 1974, it was excluded from the international group that shares fission technology. In isolation, it launched an ambitious nuclear electric program that relies heavily on homegrown technology..

    What makes India’s strategy unique is its plan to build commercial reactors that run not on uranium but on a lighter element, thorium-232. India has one of the world’s largest reserves of thorium — about 225,000 metric tons — but little uranium ore. Thorium does not fission; when irradiated with neutrons from a source material such as uranium-235, however, some of the thorium becomes uranium-233 (U-233), which does fission and can sustain a nuclear reaction.

    In 1958, India announced that it was embarking on an ambitious, three-stage plan to exploit its thorium reserves. The first stage required building pressurized heavy-water reactors powered by natural uranium; they yield plutonium as a byproduct. Twelve are now operational. The plan called for stage two to kick in after sufficient plutonium had been extracted from spent cores; it would be used as a fuel in future fast-neutron reactors, which can irradiate thorium and produce U-233 as a byproduct. In the third stage, Advanced Heavy Water Reactors will burn a mixture of U-233 and thorium, generating about two-thirds of their power from thorium. Other nations–including the United States, Russia, Germany, and Israel–have studied the route but have not attempted to use it to generate electricity.

    Stage two of this grand strategy began officially last October. In the sleepy southern township of Kalpakkam, a government-owned company began building a 500-megawatts-of-electricity (MWe) fast-breeder reactor that will use fast neutrons to produce U-233. In its core, the reactor will use a “seed” fuel containing uranium and plutonium oxide; this source will send neutrons into a surrounding thorium blanket.

    Indian atomic energy officials are confident that this exotic fuel system can be scaled up from a smaller, 40-megawatt Fast Breeder Test Reactor (FBTR) that has been running in Kalpakkam without major problems since 1985. This reactor and other research projects at the Indira Gandhi Center for Atomic Research in Kalpakkam have demonstrated, IGCAR officials say, that India has mastered the new technology. In a “bold step forward,” says Anil Kakodkar, chair of the Atomic Energy Commission (AEC) in Mumbai, researchers at IGCAR in May of this year successfully extracted plutonium in high purity from the unique plutonium-rich mixed carbide fuel discharged from FBTR.

    AEC anticipates that the fast breeder at Kalpakkam will cost about $700 million and produce 500 MWe. The long-term goal, according to Kakodkar, is to increase nuclear electric output from 3360 MW today to “around 275 gigawatts” by the middle of this century.

    Construction at Kalpakkam ran into trouble early this year: The 26 December 2004 tsunami flooded the foundations of the reactor building and set the schedule back by 4 months, says Baldev Raj, IGCAR’s director. But he says that the work is now on track and predicts that the reactor will go critical as planned in September 2010.

    Mujid Kazimi, a nuclear engineer who studies thorium fuels at the Massachusetts Institute of Technology in Cambridge, says India’s approach to breeding nuclear fuel from thorium is “slightly more complicated” than fuel breeding planned elsewhere in the world. But he adds, “everything they have reported to date indicates they are on track.”

    India cannot go it entirely alone, however. It still requires uranium, including for two boiling water reactors it bought from General Electric in the 1960s, and that may be one reason it is interested in opening nuclear trade with other countries. At a meeting last month with Prime Minister Manmohan Singh, President George W. Bush called India “a responsible state” with “advanced nuclear technology.” The opening could lead to future exchanges of personnel and technology–and possibly fuel. Singh reassured Parliament, however, that the deal would not undermine India’s nuclear self-sufficiency.

  8. Yale Simkin (History)

    Jeffrey Lewis wrote:
    “I think some of you are overlooking the fact that the Indian breeders are designed to utilize thorium in the place of uranium.”

    Its somewhat more complex than that.

    The 500 MWe Indian PFBR will have a core loaded with TWO TONS of plutonium. The core will contain an average of ~25% plutonium and 75% uranium (both as oxides)

    This reactor will have two separate breeding blankets. One will be rods of thorium to breed U233 in (the radial blanket), and the other will be composed of depleted uranium (in the outer 40% of each core fuel pin) to breed Pu239 (the axial blanket).

    The reactor will thus “launder” the MOX plutonium into supergrade Pu AND produce vast amounts of U233 bomb material. The blanket supergrade Pu may also be blended with large amounts of the core’s reactor-grade Pu, producing enormous amounts of ordinary weapons-grade Pu – essentially a superpower-sized atomic arsenal from a single core load.

    The U233 produced from the thorium blanket is a superb bomb material. A weapon can be created just by DROPPING a subcritical mass from a few feet onto another subcritical mass.

    It is important to keep the U232 (a hard gamma emitter) minimized for health reasons, but in the FBR it is easy to keep it under 5 ppm.


  9. Yale Simkin (History)

    Anonymous wrote:
    “Actually, the FBR would require equal or nearly equal amounts of uranium to produce the same amount of WGPu.
    Plutonium is produced in FBRs by irradiating natural or depleted uranium.
    “Breeding” WGPu requires again using about 7 times as much uranium.
    Plutonium production in any reactor is never inherently “supergrade.”
    The grade of the plutonium is solely a function of how long the uranium remains in the reactor.
    Reactor design (PHWR, FBR, etc.) doesn’t change the fact that a growing amount of Pu-239 will accept a neutron to produce Pu-240 instead of fissioning the logner the fuel is irradiated in the reactor.”

    I think you are comparing apples to oranges. The events occuring in the core of a slow (or “thermal”) neutron reactor is completely different than the events in a breeding blanket of a fast neutron reactor.

    Key point one (supergrade Pu production):

    Capture versus fission ratios. A nucleus of Pu239 will absorb 2 out of 3 impacting thermal neutrons, producing Pu240 and other higher actinides. Only 1 in 3 neutrons produce fission. Therefore as burn-up continues the percentage of Pu239 decreases as the pu240+ increases.

    The opposite occurs in a fast neutron reactor. A nucleus of Pu239 will fission 2 out of 3 times from impacting fast neutrons.

    This dramatically lowers the production of higher actinides.

    Also these fissions eject 3 neutrons which may be captured by 238U producing more 239Pu or fission a higher Pu.

    At very high burnup a thermal reactor may have percentages of Pu239 as low as ~60%, while a fast reactor will have Pu239 at +95% (or supergrade)

    Key point two (needed uranium):

    In a thermal reactor the “breeding” of Pu239 normally occurs in the fuel itself. If the fuel is allowed to burn-up as much of the rare (and expensive) U235 as feasable, the Pu isotope blend will be “reactor-grade”, perfectly usable, but definately not optimal. To avoid buildup of higher Pu isotopes, the fuel must have low burnup. A very great amount of fuel must reprocessed.The uranium that must be remade into fuel is very poor quality, under-enriched, and contaminated with unwanted U isotopes and residual fission products, dramatically increasing expense and health/safety effects.

    As the UIC points out:

    Processing of Used Nuclear Fuel for Recycle
    Uranium Information Centre Ltd
    Nuclear Issues Briefing Paper # 72
    December 2005

    “In France… EdF has demonstrated the use of RepU (reprocessed uranium) in its 900 MWe power plants, but it is currently uneconomic due to conversion costing three times as much as that for fresh uranium, and enrichment needing to be separate because of U-232 and U-236 impurities (the former gives rise to gamma radiation, the latter means higher enrichment is required).”

    In the fast reactor, the fuel goes to (very) complete burnup (see point one above) without needing extraction or reprocessing. Secondly, the Pu239 is bred in a fertile blanket of depleted uranium, which converts an inexpensive waste material into high quality bomb material without compromising the co-generation of power.

    Here are some additional comments from:

    Plutonium Isotopics – Non-Proliferation And Safeguards Issues
    Australian Safeguards Office, Canberra ACT, Australia

    “In the future, another major source of low burn-up plutonium will be the blanket material from fast breeder reactors (FBRs). FBR blankets will contain plutonium well within the weapons-grade range, even of “super-grade” (around 3% Pu-240)…

    It is reported that France has obtained WGPu from reprocessing blankets from the Rapsodie and Phénix prototype FBRs at Marcoule ..

    In addition to current operating situations where production of low burn-up plutonium cannot be avoided, potentially there will be large-scale arisings of low burn-up plutonium in the blanket material from fast breeder reactors. Since in the future production of blanket material will be the major reason for operating FBRs (ie to obtain plutonium for recycle), obviously it is not practicable to proscribe the production of such plutonium in irradiated blanket material.”


  10. Peter Crail (History)

    Yale Simkin says:
    “Its somewhat more complex than that.

    The 500 MWe Indian PFBR will have a core loaded with TWO TONS of plutonium. The core will contain an average of ~25% plutonium and 75% uranium (both as oxides)

    This reactor will have two separate breeding blankets. One will be rods of thorium to breed U233 in (the radial blanket), and the other will be composed of depleted uranium (in the outer 40% of each core fuel pin) to breed Pu239 (the axial blanket).”

    Do you know of any sources that say that the PFBR will use two blankets, both thorium and DU? Everything I’ve seen says that the DU will only be used as part of the MOX fuel used in the core.

    I’ve also heard that the US and India have agreed to hold off on safeguarding the PFBR until 2010. Is there any way to ensure that only thorium would be used as a blanket once safeguards are in place (if they are in fact put in place then)?

  11. Satish (History)

    Guys what you are discussing here is correct. Our (INDIAN) breeder reactors (and its virtually infinate loop for us using thorium) can produce huge amounts of weapons grade material in comparetively very short amount of time.

    Now let me tell you this, the breeder reactors (FTBR), PFBR and Kamini (only reactor in the world that attained criticallity on U233 and operational with out any problems), ATBR (Advance thorium breeder reactor), All future test and prototype reactors, Advanced naval reactor, all future Naval reactors, Laser uranium enrichment center, gas centrifuge enrichment center, two of three existing plutonium reprocessing centers will NOT be under safeguards under any circumstances for ever. We are world leaders in breeder and very very highly advanced in reprocessing technology. Also we will want all the burned imported fuel from the imported reactors also reprocssed in INDIA itself under safe guards though, this way it will be very econoical for us as our advanced reprocessing technology gives lot of valuable by products useful further in second and third stage of our nuclear program. [YOU SHOULD REMEMBER THAT IN LAST 30 years INDIA HAS DEVELOPED SOME VERY UNIQUE TECHNOLOGIES]

    Our Indeginious Pressurized Heavy water reactors are very advanced and have very unique features so we should have intellictual property rights for them also along with the indeginious breeder reactors. And all indeginious reactors that will get under safeguards will NOT be uner safeguards in-definitely, if the national security demands INDIA will reserve the right to move them into military sphere.

    If you wont accept it we will simply walk away from the deal and nobody can do any thing about it (mind you we are no push over bananna republic). If you try to put restrictions on our research and deterrance capabilities this deal will fail and US will be dislodged from ASIA and East Europe for very very long time to come.

    You should decide if you want a alliance with rising super power (INDIA) againest another rising super power (CHINA) or you want to get isolated.

    If you won’t supply fuel for our power reactors no problem we can cope on our own for atleast another 25 years with the uranium/alternative oxide fuel avaiable in the country. by that time we will be having 4 operational 500MWe breeder reactors operational and we would have stockpiled enough plutonium from second stage (Plutonium breeder) to kick start our third stage of our nuclear program (thorium cycle).

    If we have a deal we will start our thorium cycle 10 years from now or else our third stage will kick in any way in another 20-25 years by which time we will be third largest economy in the world till then we can also tap alternative energy sources like bio-gas, wind, solar and any way we have pleanty of low grade coal (we will pollute the world like crazy till our third stage kicks in).

    Now choice is yours. Decide if you want the deal or not. Do you think US is a fool to give us an exception, their scientists know very well that we are having a dimond mine of future technologies.

    My assement is US will accept all major conditins set by INDIA and with few remaining they will accept with minor changes.

    And regarding CIRUS reactor (supplied by canada under atoms for peace) we cannot put it under safegurads simple because its in BARC complex (it was never used for weapons production as suspected), at the most we can shut down the reactor and scrape it.

    In short BARC complex in Trombay (Weapons Research + Reactor Research) and MAPS (Reactor Research + Weapons Research) Complex in kalapaakam will never be under safe guards. These are the two places where cutting edge research and development takes place.

  12. ad (History)

    india has 35% of the worlds thorium is better to import uranium than using our reserves.