Integral Fast Reactors

Notes by Jim Green

See also:

Why would anyone want to know about IFRs?

Because well-known climate scientist James Hansen is promoting them (and an Australian scientist, Barry Brook). Australian nuclear lobbyist Ben Heard (whose lobby group ‘Bright New World’ accepts secret corporate donations) led a united push to develop IFRs during the 2015-16 SA Nuclear Fuel Cycle Royal Commission. To its credit, the Royal Commission flatly rejected their arguments, stating that “fast reactors or reactors with other innovative designs are unlikely to be feasible or viable in South Australia in the foreseeable future. No licensed and commercially proven design is currently operating. Development to that point would require substantial capital investment. Moreover, the electricity generated has not been demonstrated to be cost-competitive with current light water reactor designs.”

Barry Brook, Tom Blees et al.

Some of the comments below (re Barry Brook, Tom Blees and George Stanford) refer to comments posted at the Brave New Climate BNC blog/website:



The second of those webpages is a critique of an earlier version of this FoE webpage. Sadly, there’s nothing in the critique which allays concerns about IFR and WMD proliferation and nothing on proliferation risks that hasn’t already been addressed in this FoE webpage.

Here’s a letter which sums up some concerns:

Old-style spin
Letter published in The Advertiser, 18 Nov 2009
BARRY Brook promotes what he optimistically labels “next generation” reactors with old-style spin (“Follow Britain’s lead on nuclear power”, The Advertiser, 10/11/09).
For example, he repeatedly has claimed the non-existent “integral fast reactors” he champions “cannot be used to generate weapons-grade material”. Unfortunately, that simply is not true. Worse still, Brook persists with that claim although he knows it has been contradicted by, among others, a scientist with hands-on experience working on a prototype integral fast reactor in the US.
Brook and other promoters of “next generation” reactors have another credibility problem. They acknowledge the need for a rigorous safeguards system to prevent the use of peaceful nuclear facilities to produce weapons of mass destruction, and they acknowledge the existing safeguards fall well short of being rigorous.
None of them, however, is willing to get off his backside to support important, ongoing efforts to strengthen safeguards. This simply is irresponsible. Moreover, it is hypocritical for Brook to criticise Friends of the Earth and other groups which have worked long and hard to strengthen safeguards – with absolutely no help from such people as him.
Brook also berates Friends of the Earth for failing to acknowledge “technological developments that solve the long-lived nuclear waste problem”. Those developments, however, involve another non-existent technology, called pyroprocessing.
South Korea recently announced its intention to embark on a research and development program which aims to provide a “demonstration” of the viability of operating reactors in conjunction with pyroprocessing by the year 2028. That is almost 20 years – just to demonstrate the concept.
Brook offers nothing but false and extravagant claims based on non-existent technology. We deserve better.
Jim Green, Friends of the Earth, Melbourne, VIC.

What are IFRs?

* reactors proposed to be fuelled by a metallic alloy of uranium and plutonium. ‘Fast’ because they use ‘fast’ unmoderated neutrons.

* coolant: liquid sodium

* electrolytic ‘pyroprocessing’ to separate actinides/transuranics (inc plutonium) from spent fuel and to re-irradiate (both as an additional energy source and to convert long-lived waste products into shorter-lived, less problematic radioactive wastes).

Pyroprocessing is troubled / failed technology – see this 2017 article by physicist Dr Ed Lyman.

Here is one description of pyroprocessing:

“Pyroprocessing differs completely from conventional spent fuel reprocessing (and its associated proliferation dangers) because it doesn’t produce a pure stream of separated plutonium. In pyroprocessing, spent fuel is cut into pieces, heated, and turned into a powder. This process burns off volatile fission products such as krypton and xenon as well as some semi-volatile fission products such as iodine and cesium. (The hotter the process, the more is burned off.) The powder is then transformed into a metal and placed in a molten bath of lithium and potassium chloride salts. An electric current is run through the bath to dissolve the radioactive metal and to separate its elements in several stages, beginning with the recovery of uranium. This operation continues until the concentration of transuranics (plutonium, neptunium, americium, and curium) in the molten salt reaches a level where they also can be separated from the bath, along with a significant amount of fission products (cerium, neodymium, and lanthanum). It then can be directly refabricated into metallic fuel for use in fast reactors without any further processing or purification.”

Here is another description of pyroprocessing:

The pyrometallurgical process (“pyro” for short) extracts from used fuel a mix of transuranic elements instead of pure plutonium, as in the PUREX route. It is based on electroplating— using electricity to collect, on a conducting metal electrode, metal extracted as ions from a chemical bath. Its name derives from the high temperatures to which the metals must be subjected during the procedure. Two similar approaches have been developed, one in the U.S., the other in Russia. The major difference is that the Russians process ceramic (oxide) fuel, whereas the fuel in an ALMR is metallic.

In the American pyroprocess, technicians dissolve spent metallic fuel in a chemical bath. Then a strong electric current selectively collects the plutonium and the other transuranic elements on an electrode, along with some of the fission products and much of the uranium. Most of the fission products and some of the uranium remain in the bath. When a full batch is amassed, operators remove the electrode. Next they scrape the accumulated materials off the electrode, melt them down, cast them into an ingot and pass the ingot to a refabrication line for conversion into fast-reactor fuel. When the bath becomes saturated with fission products, technicians clean the solvent and process the extracted fission products for permanent disposal.

Thus, unlike the current PUREX method, the pyroprocess collects virtually all the transuranic elements (including the plutonium), with considerable carryover of uranium and fission products. Only a very small portion of the transuranic component ends up in the final waste stream, which reduces the needed isolation time drastically. The combination of fission products and transuranics is unsuited for weapons or even for thermal-reactor fuel. This mixture is, however, not only tolerable but advantageous for fueling fast reactors.

Although pyrometallurgical recycling technology is not quite ready for immediate commercial use, researchers have demonstrated its basic principles. It has been successfully demonstrated on a pilot level in operating power plants, both in the U.S. and in Russia. It has not yet functioned, however, on a full production scale.


Complete IFR systems don’t exist.

Blees cites five reactors with some IFR characteristics.

Brook gives this summary of the state of development of IFR components: “IFRs are sodium-cooled fast spectrum nuclear power stations with on-site pyroprocessing to recycle spent fuel. Fast spectrum power reactors exist … Indeed, even sodium-cooled fast reactors (a type of Advanced Liquid Metal Reactor, ALMR), the type an IFR facility would likely use, already exist (others include lead- or gas-cooled). Metallic alloy fuels (uranium-plutonium-zirconium), operating within a reactor, existed, in the Experimental Breeder Reactor II at the Argonne National Laboratory. Just because they are not currently used in any operating nuclear power plant doesn’t mean they don’t (haven’t) existed). The only thing that doesn’t currently exist is the full systems design of the integrated plant.”

In short:

* Fast neutron reactors (breeders) exist but experience is limited and they have had a troubled history (accidents, and their WMD proliferation potential).

* The pyroprocessing and transmutation technologies intended to operate as part of IFR systems are some considerable distance from being mature. See the references below for further discussion.

* South Korea is investigating IFRs but plans to spend the next 18-19 YEARS just to ASSESS their viability.

For a properly functioning IFR system, the individual components would need to work and the components would need to be integrated, with potential technical and social obstacles. For example, there’s no point having the capacity to irradiate significant quantities of fissile material from outside sources if states and/or nuclear utilities won’t surrender fissile material or if IFR operators don’t want to irradiate outside sources of fissile material. And its no good overcoming those potential social obstacles if the technology doesn’t meet its proponents’ expectations.

The possibilities are endless, e.g.:

* Pyroprocessing is scrapped in favour of conventional reprocessing.

* IFRs are rolled out in the absence of rigorous international safeguards.

* The potential non-proliferation benefits of IFR are not realised because they are not used to irradiate outside sources of fissile material to any degree.

* IFR proponents envisage each IFR reactor having on-site pyroprocessing (thus minimising transportation of nuclear materials and the attendant risks of accidents, terrorism etc) but one can readily imagine centralised processing facilities being preferred on economic grounds.

The MOX plant and the THORP reprocessing plant at Sellafield (UK) provide two recent examples of nuclear plants which have been conspicuous failures despite considerable historical experience with the basic technology, despite the UK’s lengthy and extensive experience with many facets of nuclear technology, and despite the UK’s relative economic strength and relative technological/industrial strength.

Potential advantages of IFRs

IFRs would breed their own fuel (plutonium) and would therefore not be dependent on outside fuel sources (e.g. uranium) except for the initial fuel load. Hence less demand for uranium with its attendant problems (finite resource, social and environmental impacts of uranium mining) and less demand for enrichment and thus enrichment plants.

Recycling of plutonium extracts more energy, and gets rid of the plutonium with its attendant proliferation risks.All the above could potentially be achieved with conventional reprocessing and plutonium use in MOX (uranium/plutonium) reactors or fast neutron reactors. IFR offers one further potential advantage: irradiating long-lived waste radionuclides, which wouldn’t produce any extra energy but it would convert (some/most) long-lived radionuclides into shorter-lived radionuclides.



IFRs can be used to produce plutonium for weapons in the same ways that conventional reactors can:

1. Production of weapon grade plutonium in the fuel, using a shorter-than-usual irradiation time. As George Stanford notes, proliferators “could do [with IFRs] what they could do with any other reactor – operate it on a special cycle to produce good quality weapons material.”

Conventional PUREX reprocessing can be used to separate the plutonium from irradiated fuel/targets/blanket. Blees notes that: “IFRs are certainly not the panacea that removes all threat of proliferation, and extracting plutonium from it would require the same sort of techniques as extracting it from spent fuel from light water reactors. The bottom line is that fissile material has to be subject to oversight …”

Another option is to separate reactor grade plutonium from IFR fuel and to use that in weapons instead of weapon grade plutonium.

2. Production of weapon grade plutonium by irradiating a uranium or depleted uranium targets/blanket, and separation using PUREX reprocessing. Unlike research reactors, power reactors aren’t generally designed to facilitate the insertion and removal of targets/blankets, but where there’s a will there’s a way.

As with conventional reactors, IFRs can be used to produce large quantities of fissile material for nuclear weapons, which must weigh very heavily against them in any rational comparative assessment of energy options. Whether IFRs are somewhat more or less proliferation resistant than conventional reactors is a marginal debate.

IFR advocates propose using IFRs to draw down global stockpiles of fissile material, whether derived from nuclear research, power or WMD programs. Well and good, but reprocessing/MOX/breeders promised the same thing but have demonstrably increased rather than decreased proliferation risks (discussed later). Some specific problems:

* WMD proliferators won’t use IFR to draw down stockpiles of their own fissile material let alone anyone else’s. They will use them to produce plutonium for nuclear weapons.

* The proposal confronts the familiar problem that the countries with the greatest interest in WMD production will be the least likely to forfeit fissile material stockpiles and vice versa.

* The proposal may (or may not) also face practical limitations. Numerous states/utilities etc would gladly get rid of their stockpiles of spent fuel (and perhaps other nuclear materials), but what is the incentive for the operators of IFR plants to irradiate/transmute/destroy nuclear materials produced elsewhere and what are the costs/risks of so doing? Presumably the incentive is financial, in which case what’s the cost and who’s paying?

Brook says “The net effect of the IFR will be reduced availability of bomb material worldwide. What is your solution to eliminating the existing stockpiles if it is not via fission transmutation?” In response:

* IFRs could be used to get rid of fissile material from outside sources but that doesn’t mean they necessarily will (except for their initial fuel load).

* Whatever benefits arise from the consumption of outside sources of fissile material must be weighed against the problem that IFRs could themselves be used to produce fissile material for weapons.

* If it seems unduly pessimistic to be suggesting a neutral or negative effect on non-proliferation grounds, witness the increased proliferation risks from reprocessing/MOX/breeders – systems which were also meant to reduce proliferation risks by consuming fissile material.

* Stopping / minimising the production of fissile material is obviously the single most important step forward. As for stockpiles, all the options are problematic.

* Plutonium should be left in spent fuel because spent fuel provides the best protection against diversion (radioactivity and heat).

The intention is to avoid separating plutonium from irradiated IFR fuel except in a stream that incorporates the plutonium with a waste stream (which is preferable to conventional PUREX reprocessing). This would be unsuitable for nuclear weapons. However:

* the plutonium could be separated from the waste radionuclides with further reprocessing (using conventional PUREX reprocessing).

* the plutonium/waste stream would be suitable for use in ‘dirty bombs’.

Blees says: “Spent LWR [light water reactor] fuel can be put through a PUREX process to extract virtually pure plutonium, though its isotopic composition will be far less than ideal for weapons.” Reactor grade plutonium, whatever its source (LWR, IFR etc), can be used for weapons even though it is less than ideal.

This paper:

Proliferation Resistance Assessment Of The Integral Fast Reactor

Harold F. McFarlane, Argonne National Laboratory

includes the acknowledgment that

“The reactor … could be used for excess plutonium consumption or as a breeder if needed …”

and acknowledges uncertainties and proliferation risks:

“The key to objectively assessing the proliferation resistance of the IFR concept is to recall that much of what Bengelsdorf and Wymer said years ago still pertains in large measure today, i.e., that some elements of the technology still remain to be developed and demonstrated. The reactor aside, neither the recovery of transuranics from the molten salt system nor the remote fabrication of fuel has been demonstrated. Even the concept for transuranic recovery has evolved through two generations since those early assessments were done. For every chemist worried about degradation of proliferation-resistant characteristics, there is another worried about obtaining a product sufficiently decontaminated to be useful in fuel fabrication. The assessment of this fuel cycle should be an ongoing analysis that keeps up with the research rather than one based on the presumptions of either the advocates or the critics.”

Proponents of IFR paper over the cracks in their arguments by imagining, in Blees’ words, “rigorous international oversight” to prevent misuse of IFR for WMD production. But there is no rigorous international oversight. The Director General of the International Atomic Energy Agency, Dr. Mohamed El Baradei, has noted that the IAEA’s basic rights of inspection are “fairly limited”, that the safeguards system suffers from “vulnerabilities” and it “clearly needs reinforcement”, that efforts to improve the system have been “half-hearted”, and that the safeguards system operates on a “shoestring budget … comparable to that of a local police department “.

IFR advocates acknowledge the need for a rigorous safeguards system (and implicitly or explicitly acknowledge the WMD potential of IFR), but there’s no evidence of them getting off their backsides to engage in the laborious task of trying to bring about improvements in safeguards.

Do IFR advocates accept the need for a rigorous safeguards system to be in place before a large-scale IFR roll-out? What is their timeframe for the establishment of a rigorous safeguards system? How do they propose to hasten progress, which has to date been painfully slow?

Another argument from IFR advocates is to explicitly or implicitly acknowledge the WMD potential of IFR but to argue that proliferators would most likely find a simpler method to produce fissile material for bombs. Blees says: “In point of fact, anyone hoping to make a bomb from plutonium will likely try to obtain an isotopically more pure plutonium by creating it from U-238 at a small research reactor.”

But IFR can be used to produce isotopically pure (weapon grade) plutonium, either in the fuel or targets/blanket. As for using a research reactor instead of IFR, proliferators might do just that, depending on the options available to them. Historically, would-be weapons states have simultaneously pursued multiple different methods/technologies e.g. R&D into both plutonium and highly-enriched uranium production, and diversifying fissile material production sources. For example, India has for decades operated research reactors to produce a reliable supply of fissile material for weapons, but India nevertheless uses its uranium/plutonium/thorium power program to further its WMD program (as evidenced by its refusal to allow safeguards to be applied to numerous reactors).

Another variable is how much fissile material is wanted – if a large amount, then power reactors (inc IFR) will be favoured over research reactors (see the debate between Fainberg and Holdren in the Bulletin of the Atomic Scientists, 1983, January and May editions).

As an aside, a large number of research reactors exist which are too small (<1MWt) to produce Significant Quantities of plutonium – though perhaps over a long period of time they could produce enough plutonium for one weapon.

Blees says: “As for breeding high-quality … plutonium, virtually any reactor (including research reactors) can do that by wrapping a U-238 blanket around the core and letting it get bombarded with neutrons for a while, then removing it and extracting the Pu with the PUREX method. It requires relatively brief exposure, which is NOT what one would have in a reactor core operated for power purposes.”

Thus we have an (implicit) acknowledgement from Blees (if any was needed) that IFR can produce high purity, weapon grade plutonium. And we have the problem that weapons proliferators simply won’t use IFR as Blees would want them to, i.e. they will use IFR to produce plutonium for weapons regardless of the implications for power generation. The same argument applies to irradiated fuel – IFR proponents want plutonium to be separated from irradiated fuel in a mixed plutonium/waste stream but proliferators will use standard PUREX reprocessing to separate the plutonium for use in weapons. The same argument applies to breeders vs burners, as mentioned above – IFR proponents would use IFR to draw down plutonium stockpiles, proliferators will use them to increase plutonium stockpiles. So on at least three counts, IFR proponents are making implausible claims about the likely use of IFR by WMD proliferators, and papering over the remaining cracks in their arguments by imagining “rigorous international oversight”.

Blees offers this: “Almost 80% of greenhouse gas emissions come from nuclear-capable countries anyway …” Brook says: “If you deploy IFRs … first in nuclear club countries – those that already possess, or are capable of making, nuclear weapons, then there is no additional proliferation risk.” However:

* in weapons states or weapons-capable states, IFR could facilitate vertical proliferation, which in turn motivates horizontal proliferation.

* why would Brook’s proposal not be rejected or seriously curtailed as with every other proposal for selective deployment of (‘sensitive’) ‘civil’ nuclear technologies?

* are we to assume that the current weapons capable/incapable status of all countries is locked in forever (which is both unlikely and problematic)? Will countries that obtain weapons-capable status then be given the option of IFR technology? Couldn’t that (if only marginally) encourage states to develop a nuclear weapons capability? Will countries that move in the opposite direction be asked/forced to abandon their IFR programs and couldn’t that (if only marginally) discourage disarmament?

So we begin with IFR rhetoric including the claim that they are proliferation resistant, but we’re left with the argument that, to paraphrase, IFR can in fact be used to produce fissile material for nuclear weapons – but so can other reactor types, and in any case some countries already have nuclear weapons.

Blees trots out the tired old lie that: “Every country that’s developed nuclear weapons has done so separately from, and usually prior to, nuclear power program development.” In a few countries, power reactors have been directly involved in WMD programs. In numerous countries, power programs have indirectly facilitated WMD programs – these links are no less important for being indirect. The links between nuclear power and weapons are detailed at To give some sense of the scale of the problem, of the 60+ countries to have developed a nuclear industry of some significance (including power and/or research reactors), over 20 have used their ‘peaceful’ nuclear facilities for some level of WMD research and/or production. Of the 10 states to have built nuclear weapons, five did so on the back of their ‘peaceful’ nuclear programs:

* Pakistan and South Africa by misusing enrichment expertise/technology acquired ostensibly for their power programs.

* India using research reactors to produce plutonium and, later, using its uranium/plutonium/thorium power program in support of its WMD program.

* Israel using a research reactor.

* North Korea using a so-called ‘Experimental Power Reactor’ to produce plutonium for bombs.

Conventional reprocessing with plutonium and uranium use/reuse in MOX and/or breeders held the same promise as IFR – reducing proliferation risks by getting rid of plutonium once and for all (as well as other potential advantages, namely reducing demand for uranium and partially addressing waste management problems). In practice, the result has been very little use of plutonium or recycled uranium, but very extensive stockpiling of plutonium separated from spent fuel. Proliferation risks have been increased not reduced because the separated plutonium can be used directly in weapons. Global stockpiles of separated civil plutonium amount to 270+ tonnes and counting – enough for 27,000 nuclear weapons. That problem is alarming because of its scale, because it is so completely unnecessary and indefensible, and because its resolution could hardly be simpler – suspending or reducing the rate of reprocessing such that plutonium stockpiles are drawn down rather than continually increasing.

IFR advocates demonstrate little or no understanding of the realpolitik imposed by the strife of commercial, political and military interests responsible for, amongst other things, unnecessarily creating the problem of 270+ tonnes of separated civil plutonium and failing to take the simplest steps to address the problem.

Blees says “Nuclear power isn’t going to go away, like it or not. Wouldn’t it be far more effective to work for an international regime and the safest, most proliferation-resistant design of nuclear power plant to be standardized and deployed under comprehensive and dependable supervision.” Anti-nuclear people/NGOs have a long history of working to improve safeguards even though that work implicitly acknowledges ongoing operation of nuclear facilities. IFR advocates have, by and large, abstained from that laborious work and are therefore, by and large:

* irresponsible in their actions (promoting dual use technology but failing to take responsibility for the WMD proliferation risks)

* hypocritical in their criticisms of opponents of nuclear power / IFR.

As for Blees’ comment that nuclear power isn’t going to go away, everyone would agree that nuclear power is likely to be around for some decades at least, like it or not. Beyond Blees’ banal observation, the future of nuclear power is of course contested and uncertain.


IFR still produces radioactive waste – albeit (in theory) a more manageable waste stream than conventional reactors. But to lessen the long-term hazards, the short-term public health, environmental and proliferation risks are increased through reprocessing, plutonium recycling etc. The Massachusetts Institute of Technology Interdisciplinary Study states that:

“Decisions about partitioning and transmutation must … consider the incremental economic costs and safety, environmental, and proliferation risks of introducing the additional fuel cycle stages and facilities necessary for the task. These activities will be a source of additional risk to those working in the plants, as well as the general public, and will also generate considerable volumes of non-high-level waste contaminated with significant quantities of transuranics. Much of this waste, because of its long toxic lifetime, will ultimately need to be disposed of in high-level waste repositories. Moreover, even the most economical partitioning and transmutation schemes are likely to add significantly to the cost of the once-through fuel cycle.”

Ansolabehere, Stephen, et al., 2003, “The Future of Nuclear Power: An Interdisciplinary MIT Study”,


Brook says IFR reactors would be “safe from melt down”. But technologies fail, and well-intentioned humans err. And even if we generously assume that safety mechanisms will certainly prevent a serious accident, there’s still the problem of sabotage or outside attack resulting in a large release of radioactivity.

Easy to make wild claims about non-existent reactors since such claims cannot be tested or disproved. As a nuclear industry representative has noted about non-existent reactor types: “We know that the paper-moderated, ink-cooled reactor is the safest of all. All kinds of unexpected problems may occur after a project has been launched.”

Australian nuclear engineer Tony Wood notes that probabilistic risk assessment failed to anticipate the world’s worst reactor accident (Chernobyl) and the worst reactor accidents in the UK (Windscale) and the USA (Three Mile Island).

In response, Blees says: “And this one’s a doozy: “…probabilistic risk assessment failed to anticipate the world’s worst reactor accident (Chernobyl)…”” But as mentioned it is simply a statement from a nuclear engineer – a statement Blees does not dispute.


* briefing papers on GNEP and new reactor types at

* Hisham Zerriffi and Annie Makhijani, August 2000, The Nuclear Alchemy Gamble: An Assessment of Transmutation as a Nuclear Waste Management Strategy,

* Arjun Makhijani, Hisham Zerriffi and Annie Makhijani, “Magical Thinking: another go at transmutation”, Bulletin of the Atomic Scientists, March/April 2001, Vol. 57, No. 2, pp. 34-41.

* The more ambitious aspects of GNEP were deprioritised under the Bush presidency and that will continue under Obama: Past, present and future: Who’s voting for GNEP?, August 01, 2008,;storyCode=2050691

* The future of GNEP