IFR FaD Nuclear

IFR FaD 7 – Q&A on Integral Fast Reactors – safe, abundant, non-polluting power

Back in 2001, Dr George Stanford conducted an interview with the National Policy Analysis Center on the IFR. Nearly ten years later, in September 2010, George has updated this material, which is pitched squarely a layman audience. I post the updated version below (7-page printable PDF version here), since it fits like a glove with the IFR Facts & Discussion series that I am gradually developing. It’s probably the single best introduction to the IFR for the totally uninitiated — but there’s also real value here for the nuclear veteran, so everyone, read on!

Q&A on Integral Fast Reactors – safe, abundant, non-polluting power

By Dr George S. Stanford. George is is a nuclear reactor physicist, part of the team that developed the Integral Fast Reactor. He is now retired from Argonne National

Laboratory after a career of experimental work pertaining to power-reactor safety. He is the co-author of Nuclear Shadowboxing: Contemporary Threats from Cold War Weaponry.

What is the IFR?

IFR stands for Integral Fast Reactor. It was a power-reactor development program, built around a revolutionary concept for generating nuclear power—not only a new type of reactor, but an entire new nuclear fuel cycle. The reactor part of that fuel cycle was called the ALMR—Advanced Liquid Metal Reactor. In what many see as an ill-conceived move, proof-of-concept research on the IFR/ALMR was discontinued by the U.S. government in 1994, only three years before completion.

You might also see references to the AFR, which stands for “Advanced Fast Reactor.” It’s a concept very similar to the IFR, with some improvements thrown in. GE-Hitachi has the plans for a commercial version they call PRISM.

How was the IFR idea different from the concepts underlying traditional nuclear-power fuel cycles?

All of those fuel cycles were derived from technologies developed to meet special military needs: naval propulsion, uranium enrichment, weapons-plutonium production, and plutonium separation. Waste disposal has been approached as “someone else’s problem.” The IFR concept is directed strictly to meeting the needs of civilian power generation. It is an integrated, weapons-incompatible, proliferation-resistant cycle that is “closed”—it encompasses the entire fuel cycle, including fuel production and fabrication, power generation, reprocessing and waste management.

Do we need a new kind of reactor? What’s wrong with what we have now?

IFRs could reduce or eliminate significant difficulties that beset thermal-reactor fuel cycles—problems or concerns with:

* production and build-up of plutonium

* short-term management of plutonium

* disposition and long-term management of plutonium

* plutonium in national and international commerce

* other proliferation concerns

* long-term waste management

* environmental effects

* resource conservation

* long-term energy supply

* safety

What is a fuel cycle? What fuel cycles are there?

“Fuel cycle” refers to all the steps involving nuclear fuel that are needed to generate electricity: mining, milling, enrichment, fuel fabrication, reactor operation, reprocessing and waste management. Depending on the fuel cycle, some of those steps might not be needed. The three major fuel cycles of current interest are: thermal without reprocessing (“once-through,” or “throw-away”), thermal with reprocessing and IFR. The IFR will eliminate the need for mining (for centuries), and milling and enrichment (forever).

Who was working on the IFR? How far along was it?

The idea of the IFR originated at Argonne National Laboratory, which was about three years from finishing a study that was expected to establish firmly the technical and economic practicality of the concept. Progress had been spectacular. Design of the ALMR was being done at General Electric in San Jose, California. However, construction of a full-size prototype was not to be approved until Argonne’s research study had been completed and a current need was demonstrated.

What sort of reactor was the ALMR?

The ALMR was to be a “fast” reactor (one in which the chain reaction is maintained by high-energy neutrons)—so called because the energy spectrum of the neutrons is said to be fast.

Is there a slow reactor?

Yes, in concept, but it’s not called “slow,” it’s called thermal. Almost exclusively, current reactors are of the thermal variety: their chain reaction relies on thermal (slow) neutrons. In most of the thermal-spectrum reactors, the neutrons are moderated (slowed) by light water. Such reactors are called LWRs.

What is the most important difference in capabilities?

Probably this one: Inherently, thermal reactors are copious producers of plutonium, while IFRs can consume plutonium. In fact, two IFRs could consume the plutonium output of five LWRs of the same size, while generating electricity and bringing in revenue.

What’s so important about plutonium?

High-quality plutonium is the preferred bomb material for a sophisticated nuclear weapons program. It is even possible to make a nuclear explosive with low-quality plutonium, such as is found in power reactors.

What else can IFRs use for fuel, besides plutonium?

Their fast spectrum permits IFRs to burn any and all actinides from thorium on up. This is because in a fast neutron spectrum there are enough extra neutrons to convert the actinide isotopes that don’t fission easily into ones that do. The most important actinide elements are uranium (atomic number 92), plutonium (94) and, to a lesser extent, thorium (90). Since currently there is a growing glut of plutonium continuing to pile up from nuclear weapons and from thermal-reactor operations worldwide, the first IFRs will undoubtedly be fueled primarily with some of that plutonium.

How is that different from thermal reactors?

In a thermal neutron spectrum, many of the fission products and actinide isotopes absorb neutrons readily without undergoing fission (they have a high “capture cross section”), and the chain reaction is “poisoned” if too much of such material is present. Thus a thermal reactor cannot be a net burner of transuranic actinides. The main starting fuel for thermal reactors is a mixture of the fissile isotope U-235 (Pu- 239 can also be used), along with the fertile isotope U-238.

What in the world are “transuranic actinides?”

They are the elements beyond uranium—that is, their atomic number is 93 or greater: neptunium, plutonium, americium, curium and more. All of them are man-made elements, since they are so radioactive that the naturally created ones have long since decayed away in our little bit of the universe. They are also called higher actinides.

And what do you mean by “fissile” and “fertile?”

An isotope is called “fertile” when the addition of a neutron changes it into a fissile isotope—one that, like U-235, has a very high probability of undergoing fission when exposed to thermal neutrons. Both fissile and fertile isotopes are fissionable—it’s just that fertile ones require a high-energy neutron to make them split.

Burning and Breeding

What is a “breeder?”

A breeder is a reactor that is configured so as to produce more fissile material than it consumes. A fast reactor can be designed and operated to be either a net breeder or a net burner. A thermal reactor is a net burner of nuclear fuel, but—and this is very important—all thermal reactors are prolific breeders of plutonium.

What do you mean?

A thermal reactor starts out with no plutonium at all, and soon has a lot of it. In the process, though, it burns more fuel (mainly uranium) than it gives back as plutonium, and therefore is not called a breeder.

If IFRs can be either breeders or burners, why do some people insist on calling them breeders?

Partly for historical reasons (originally, fast reactors were investigated because of their potential to breed), partly because of genuine confusion, and partly for the emotional impact, since “breeder” carries the subliminal connotation of runaway plutonium production. The central fact that those people are missing is that with IFRs you can choose not to breed plutonium, whereas with thermal reactors you make plutonium whether you want it or not.

Then it is today’s reactors that are runaway producers of plutonium, and IFRs could put a stop to it.


What about the high-grade plutonium from dismantled nuclear weapons? Can we get rid of it?

Depends on what you mean by “get rid of it.” I suppose you could say we would have gotten rid of it if it were degraded to the point where it is as hard to deal with as the poor-quality plutonium in the used fuel from thermal reactors. That’s called the “spent fuel standard” for disposition of weapons plutonium.

How do you do that?

One straightforward way is to incorporate it in fuel for today’s thermal reactors. The fuel would consist of a mixture of the oxides of uranium and plutonium, called MOX. That process is now being started in the U.S. and Russia.

How long will it take?

There’s maybe 200 metric tonnes (1 metric ton equalling 1.1 standard tons) of weapons-grade plutonium in the world, most of which, we hope, will gradually become available for disposal over the next two or three decades. To process that much plutonium in 30 MOX-burning thermal-reactor plants (1000 MWe) would take approximately 20 years. Thus the MOX approach should be able to deal satisfactorily with the weapons plutonium.

Can IFRs help with this?

Eventually, when we get IFRs. When they do start up, they will of course begin consuming the accumulated plutonium— weapons-grade first, probably, if there’s any left, and then reactor-grade.

Commerce in Plutonium

You explained why the IFR is “fast.” Now, why is it “integral?”

“Integral” refers to the fact that the fuel processing facility can be an integral part of the IFR plant.

Is that important?

Very, if you are concerned about shipments of plutonium and spent fuel, or if you want to minimize national and international commerce in plutonium.

I think it is U.S. policy to discourage commerce in plutonium.

Yes, it is. And these days there certainly is commerce in plutonium—witness the controversy over shipments of plutonium from France to Japan a few years ago, and the recent controversy in England over a reprocessing plant at Sellafield. For the foreseeable future, and beyond, there will be no plutonium shipped out of IFR plants. The only shipments will be into them, from dismantled weapons and thermal reactors. Those are not extra shipments, but ones that otherwise would be to repositories. Thus the IFR all but eliminates commerce in plutonium.

How can that be?

An IFR plant will be a “sink” for plutonium: plutonium to be disposed of is shipped in, and there it is consumed, with onsite recycling as needed. Only trace amounts ever come out.


How safe are IFRs?

While the safety record of commercial reactors of Western design is superb, Three Mile Island notwithstanding, it would be desirable to have reactors that rely more on inherent safety features and less on engineered ones. ALMRs do that.

What is an “inherent safety feature?”

A safety mechanism that does not depend on human or mechanical intervention. For instance, ALMRs use metallic fuel rods, whereas LWRs use oxide fuel (as the Clinch River Breeder Reactor [CRBR] would have done).

Why are metallic fuel rods an inherent safety feature?

Metal is a good heat conductor, while oxide is a poor one. That means the interiors of the metal rods stay much cooler, which means that there is far less heat stored in an operating ALMR, which means that if there were a loss of coolant flow there would be much less heat present to raise the temperature of the fuel, which means that the consequences of a hypothetical accident would be much less severe.

Why is that?

Briefly, there’s a phenomenon called the “resonance Doppler effect,” which causes the reactivity to change somewhat with temperature. Because in an ALMR the temperature does not change much in a hypothetical accident, the reactor is much more stable.

O.K. What else?

ALMRs use liquid sodium for cooling and heat transfer, which makes the system intrinsically safer than one that uses water. That is because the molten sodium runs at atmospheric pressure, which means that there is no internal pressure to cause the type of accident that has to be carefully designed against in an LWR: a massive pipe rupture followed by “blowdown” of the coolant.

Also, sodium is not corrosive to steel like water is.

But doesn’t sodium burn in air and react violently with water?

Yes it does, and this of course requires prudent design, involving inert atmospheres and multiple barriers.

Not so fast! Seems to me there was a serious sodium leak and fire at a Japanese fast reactor.

You’re right. In December, 1995, at the Monju reactor, a temperature sensor broke and sodium leaked from a secondary sodium loop and caught fire. The plant was shut down, and has not yet been restarted.

How many people were hurt?


Was radioactivity released?


Was the reactor damaged?


Was there any damage at all?

Yes. Some minor damage was caused by the burning sodium, and combustion products were spread through a portion of the building; cleaning them up took almost a year. The accident was classified as Category 1 on the international scale of 0 to 7 (with 0 being the least serious) by a committee of independent specialists.

So the sodium isn’t so safe after all.

When you think about it, it is pretty safe. There have been sodium fires, and undoubtedly there will be more. The Monju fire was a public-relations disaster, but did not even come close to being a public health threat. There is a great deal of industrial experience with liquid sodium, and there have been very few problems.

Well, I suppose that’s a risk we can tolerate, since we need electricity.

I think so.

What’s become of the Monju reactor?

Last I heard, it had commenced initial operations in May 2010, after a string of administrative and other non-technical snafus. But it’s not scheduled to go on the grid until 2013.

We were talking about inherent safety features. Are there any others?

The ALMR core sits in a pool of liquid sodium. In combination with the low heat content of the metal fuel rods, this means that, if there were to be loss of control power, the core would be cooled passively by convection.

Is this different than for other liquid-metal-cooled reactors?

Almost all the earlier fast reactors were of the “loop type”—relying more heavily on forced coolant flow—and also had oxide fuel, making passive cooling more problematic.

Wasn’t passive cooling tested in a prototype ALMR?

Yes, it was. All control power for the operating reactor was cut off. Coolant pumps stopped, control rods did not move, and the operators did nothing. The core temperature rose slightly, causing the reactor to go subcritical and shut itself down without incident. Unassisted convective cooling then prevented overheating.

Conservation and the Environment

What are the environmental considerations?

We already mentioned waste management. In addition, it can be argued that the major environmental problems with nuclear power are the consequences of the mining and milling operations. Because IFRs can use not only the surplus plutonium, but also the uranium (including U-238) that has already been mined and milled, they can eliminate for centuries any further need for mining or milling.

And of course, in common with all nuclear reactors, operating IFRs emit no carbon dioxide.

Do they put out any atmospheric pollutants?

None worth mentioning.

Then there some that aren’t worth mentioning?

Extremely small amounts of radioactive gas.

How small?

So small that there’s a lot more radioactivity from coal-burning plants.

You’re pulling my leg.

No I’m not. In coal there are trace amounts of radium and uranium, for instance, that come out of the smokestacks.

Then there’s dangerous radiation from coal plants?

No, there isn’t. It’s far below natural background levels. But nuclear plants put out even less.

Then I won’t worry. How do IFRs help conserve natural resources?

Thermal reactors are incredibly profligate with the earth’s endowment of potential nuclear fuel. The once-through, “throw-away” cycle in favor in the U.S. uses less than a hundredth of the energy potential of the mined uranium. Even with recycle, probably less than 1% can be extracted. IFRs can use almost all of it.

Wait a minute—less than 2% with recycle? I thought you could get nearly all of the energy that way.

Sorry, but you can’t. After two or three passes through a reactor, the fuel has gotten so contaminated with isotopes heavier than Pu-239 that various technical and operational problems arise. The only way to consume all of it is in a flux of fast neutrons.

I’ll be darned! Well anyway, with uranium so cheap, why do we care about conservation?

For the same reason we care (or should) about conserving petroleum, even though oil is now cheap. The current worldwide glut of reactor fuel is strictly temporary. Particularly with the U.S. throw-away cycle, the economically available U-235 is not predicted to last much longer than the petroleum reserves—a few decades.

Reprocessing & Proliferation

Thermal reactors with reprocessing would do at least a little better.

Recycling (it would be with the PUREX process, or an equivalent) could stretch the U-235 supply another few decades—but remember the consequences: growing stockpiles of plutonium, pure plutonium streams in the PUREX plants, and the creation of 100,000-year plutonium mines.

If you’re going to talk about “PUREX” and “plutonium mines” you should say what they are. First, what’s PUREX?

It’s a chemical process developed for the nuclear weapons program, to separate plutonium from everything else that comes out of a reactor. Weapons require very pure plutonium, and that’s what PUREX delivers. The pyroprocess used in the IFR is very different. It not only does not, it cannot, produce plutonium with the chemical purity needed for weapons.

Why do you keep referring to “chemical” purity?

Because chemical and isotopic quality are two different things. Plutonium for a weapon has to be pure chemically. Weapons designers also want good isotopic quality—that is, they want at least 93% of their plutonium to consist of the isotope Pu- 239. A chemical process does not separate isotopes.

I see. Now, what about the “plutonium mines?”

When spent fuel or vitrified reprocessing waste from thermal reactors is buried, the result is a concentrated geological deposit of plutonium. As its radioactivity decays, those deposits are sources of raw material for weapons, becoming increasingly attractive over the next 100,000 years and more (the half-life of Pu-239 being 24,000 years).

You listed, back at the beginning, some problems that the IFR would ameliorate. A lot of those problems are obviously related to proliferation of nuclear weapons.

Definitely. For instance, although thermal reactors consume more fuel than they produce, and thus are not called “breeders,” they inescapably are prolific breeders of plutonium, as I said. And that poses serious concerns about nuclear proliferation. And proliferation concerns are even greater when fuel from thermal reactors is recycled, since the PUREX method is used. IFRs have neither of those drawbacks.

Why does it seem that there is more proliferation-related concern about plutonium than about uranium? Can’t you make bombs from either?

Yes. The best isotopes for nuclear explosives are U-235, Pu- 239, and U-233. Only the first two of those, however, have been widely used. All the other actinide isotopes, if present in appreciable quantity, in one way or another complicate the design and construction of bombs and degrade their performance. Adequate isotopic purity is therefore important, and isotopic separation is much more difficult than chemical separation. Even so, with plutonium of almost any isotopic composition it is technically possible to make an explosive (although designers of military weapons demand plutonium that is at least 93% Pu-239), whereas if U-235 is sufficiently diluted with U-238 (which is easy to do and hard to undo), the mixture cannot be used for a bomb.

High-quality plutonium is the material of choice for a large and sophisticated nuclear arsenal, while highly enriched uranium would be one of the easier routes to a few crude nuclear explosives.

So why the emphasis on plutonium?

You’re asking me to read people’s minds, and I’m not good at that. Both uranium and plutonium are of proliferation concern.

Where is the best place for plutonium?

Where better than in a reactor plant—particularly an IFR facility, where there is never pure plutonium (except some, briefly, when it comes in from dismantled weapons), where the radioactivity levels are lethal, and where the operations are done remotely under an inert, smothering atmosphere? Once enough IFRs are deployed, there never will need to be plutonium outside a reactor plant—except for the then diminishing supply of plutonium left over from decades of thermal-reactor operation.

How does the IFR square with U.S. policy of discouraging plutonium production, reprocessing and use?

It is entirely consistent with the intent of that policy—to render plutonium as inaccessible for weapons use as possible. The wording of the policy, however, is now obsolete.

How so?

It was formulated before the IFR’s pyroprocessing and electrorefining technology was known—when “reprocessing” was synonymous with PUREX, which creates plutonium of the chemical purity needed for weapons. Since now there is a fuel cycle that promises to provide far-superior management of plutonium, the policy has been overtaken by events.

Why is the IFR better than PUREX? Doesn’t “recycling” mean separation of plutonium, regardless of the method?

No, not in the IFR—and that misunderstanding accounts for some of the opposition. The IFR’s pyroprocessing and electrorefining method is not capable of making plutonium that is pure enough for weapons. If a proliferator were to start with IFR material, he or she would have to employ an extra chemical separation step.

But there is plutonium in IFRs, along with other fissionable isotopes. Seems to me that a proliferator could take some of that and make a bomb.

Some people do say that, but they’re wrong, according to expert bomb designers at Livermore National Laboratory. They looked at the problem in detail, and concluded that plutonium-bearing material taken from anywhere in the IFR cycle was so ornery, because of inherent heat, radioactivity and spontaneous neutrons, that making a bomb with it without chemical separation of the plutonium would be essentially impossible—far, far harder than using today’s reactor-grade plutonium.

So? Why wouldn’t they use chemical separation?

First of all, they would need a PUREX-type plant—something that does not exist in the IFR cycle.

Second, the input material is so fiendishly radioactive that the processing facility would have to be more elaborate than any PUREX plant now in existence. The operations would have to be done entirely by remote control, behind heavy shielding, or the operators would die before getting the job done. The installation would cost millions, and would be very hard to conceal.

Third, a routine safeguards regime would readily spot any such modification to an IFR plant, or diversion of highly radioactive material beyond the plant.

Fourth, of all the ways there are to get plutonium—of any isotopic quality—this is probably the all-time, hands-down hardest.

The Long Term

Does the plutonium now existing and being produced by thermal reactors raise any proliferation concerns for the long term?

It certainly does. As I said earlier, burying the spent fuel from today’s thermal reactors creates geological deposits of plutonium whose desirability for weapons use is continually improving. Some 30 countries now have thermal-reactor programs, and the number will grow. To conceive of that many custodial programs being maintained effectively for that long is a challenge to the imagination. Since the IFR can consume plutonium, it can completely eliminate this long-term concern.

Are there other waste-disposal problems that could be lessened?

Yes. Some constituents of the waste from thermal reactors remain appreciably radioactive for thousands of years, leading to 10,000-year stability criteria for disposal sites. Waste disposal would be simpler if that time frame could be shortened. With IFR waste, the time of concern is less than 500 years.

What about a 1994 report by the National Academy of Sciences? The Washington Post said that the NAS report “denounces the idea of building new reactors to consume plutonium.”

That characterization of the report is a little strong, but it is true that the members of the NAS committee seem not to have been familiar with the plutonium-management potential of the IFR. They did, however, recognize the “plutonium mine” problem. They say (Executive Summary, p.3):

Because plutonium in spent fuel or glass logs incorporating high-level wastes still entails a risk of weapons use, and because the barrier to such use diminishes with time as the radioactivity decays, consideration of further steps to reduce the long-term proliferation risks of such materials is required, regardless of what option is chosen for [near-term] disposition of weapons plutonium. This global effort should include continued consideration of more proliferation-resistant nuclear fuel cycles, including concepts that might offer a long-term option for nearly complete elimination of the world’s plutonium stocks. [Emphasis added.]

The IFR, obviously, is just such a fuel cycle—a prime candidate for “continued consideration.”


You mentioned safeguards a while ago. Are you saying that IFRs need to be safeguarded?

Of course. Any kind of nuclear fuel cycle needs safeguards procedures, the most important job being to make sure that a power reactor is not operated so as to produce high-quality plutonium. The IFR is no exception, although it might be more easily safeguarded than other cycles.

Are there now any reactors that are not safeguarded?

Unfortunately, yes. A number of countries, such as India, Pakistan, and Israel, have not yet signed the Nuclear Non- Proliferation Treaty and do not permit all of their reactors to be inspected by the IAEA (International Atomic Energy Agency).

Why should there be IFRs if a country could expel the inspectors and make bombs?

Because a country could do that with any kind of reactor, and overall the IFR is by far the most proliferation-resistant nuclear fuel cycle.

Better than the thermal-reactor throw-away cycle?

Near-term, it’s comparable—quite possibly better. But when you factor in the long term (no plutonium mines), it’s the clear winner. And the IFR is far better than anything involving PUREX—which will inevitably be used increasingly in at least some countries, unless they go to IFR-type reactors.

Has a bomb ever been made that used reactor plutonium?

That is unclear. A U.S. weapons lab conducted an explosion in 1962 that made use of what was stated to be “reactor-grade plutonium,” the definition of which was different in 1962. While the details of that test and its results are still classified, some of the difficulties that would be encountered in making a weapon composed of reactor-grade plutonium are formidable and predictable. Dr. Peter Jones, who was Director of the Aldermaston weapons research establishment in England when they too tested some low-grade material, says the job is extremely difficult. We do know that, even with sophisticated weapons design, the explosive yield would be seriously degraded.

But suppose a country wanted to make some plutonium bombs, and had nothing but an IFR?

If their IFR plants were safeguarded, the material in the processing stream would be highly undesirable, as I explained earlier, and their chances of diverting it undetected would be slim indeed. If not safeguarded, they could do what they could do with any other reactor—operate it on a special cycle to produce good quality weapons material. But in either case, most likely they would do what everyone else has done: construct a special production facility, including a PUREX plant. Detecting such a clandestine facility is probably the main, immediate challenge facing international safeguards, and has nothing to do with whether a country has IFRs or LWRs.

But the uranium route to a simple bomb is so much easier that that’s probably the way a wannabe nuclear power would start out.

The Downside

There must be a downside. What is the single best technical argument against the IFR?

Isotopic composition of the plutonium. This is not at all a forceful argument, but it is probably the only one with any technical validity. (There are arguments, largely nontechnical, about whether the IFR is needed or whether it would be economical.)

How does that isotopic argument go?

In outline, like this: Designers of military weapons demand plutonium that is at least 93% Pu-239, although it is technically possible, with difficulty, to make an explosive with plutonium of almost any isotopic composition. Plutonium from LWR spent fuel runs around 60% or less Pu-239, while that from IFRs tends to be in the 70-80% range, and thus is somewhat closer to what weapons designers want.

Why isn’t that a forceful argument?

First, isotopic contamination is only one of many obstacles between a proliferator and a weapon from IFR fuel. I mentioned some of them a while back.

Second, having material that is 80% Pu-239 instead of 60% does not greatly lessen the difficulty of designing and building a bomb.

Third, and most important, remember that there are far easier ways to get fissile material for weapons—high quality material, at that—than from spent reactor fuel. Iraq, for instance, chose uranium enrichment. No country has ever used reactor-grade plutonium to make weapons.

Strongest Point

You mentioned the best argument against the IFR. What is the best argument for it?

Proliferation prevention. Near-term, the IFR makes PUREX illegitimate and plutonium inaccessible. Long term, it relieves future generations of the responsibility to guard the plutonium mines, and of the risks of not guarding them adequately.

There’s another huge benefit, of course. If nothing better comes along, the IFR can supply the world with pollution-free energy for as long as civilization lasts. Uranium becomes just as inexhaustible as wind or solar energy.

Since the IFR has so much going for it, research should be steaming full speed ahead, right?

Wouldn’t you think so? Nevertheless, at the Clinton administration’s urging, Congress terminated the research on October 1, 1994. The Senate voted to continue it, but the House prevailed in conference.

Well, I suppose at least we saved some of the taxpayers’ money.

Wrong again. Termination cost as much over the ensuing four years as finishing the research would have done, especially since the Japanese were all set to chip in $60 million.

You’re kidding. Why would our government do what it did?

Combination of factors, but the main one is plain misunderstanding of the facts I have just explained to you. Well-meaning but ill-informed people claiming to be experts confused pyroprocessing with PUREX, and convinced so many administrators and legislators that the IFR was a proliferation threat that the project was killed.

Isn’t it time to revive it?

I think it is. Other countries are working on their own fast rector designs—inferior, in my opinion, but this country abdicated its technological lead in 1994.

What should we do to get that lead back?

Needed first is one or two commercial-scale demonstration plants to finalize the details of the pyroprocessing techniques and to permit a better estimate of the probable cost of production-model IFR plants.

Add to FacebookAdd to NewsvineAdd to DiggAdd to Del.icio.usAdd to StumbleuponAdd to RedditAdd to BlinklistAdd to TwitterAdd to TechnoratiAdd to Furl


By Barry Brook

Barry Brook is an ARC Laureate Fellow and Chair of Environmental Sustainability at the University of Tasmania. He researches global change, ecology and energy.

152 replies on “IFR FaD 7 – Q&A on Integral Fast Reactors – safe, abundant, non-polluting power”

Dougla Wise, you said:

“Similarly, Rubbia’s spallation approach to U233 production is one that was not considered in the papers you cited, but clearly seems to have the potential to allow fast thermal breeder startups if economic considerations were set aside.”

Rubbia’s idea is criticized because it creates fissile material which currently is so abundant that we have concerns about weapons of mass destruction. However, we could run out of fissiles if a rapid expansion of NPP capacity was attempted.

Has anyone calculated what it would cost to produce fissile Uranium using ADRs compared to more conventional means such as gas centrifuges?


Douglas Wise,
As an after thought to my previous quote, the cost of neutrons is coming down. The SNS at ORNL produces a 1 MW proton beam and L.N. Chang (Virginia Tech) says he can get more than 30 neutrons for every proton.

Charlie Bowman (ADNA) built a prototype ADR back in 2002.


Tom Keene wrote:

I’d be interested to know what Dr Stanford thinks of [the MIT] assessment.

. Tom, I haven’t had time yet to read and digest it. But I’m struck by this excerpt:

For the next several decades, light water reactors using the once-through fuel cycle are the preferred option for the U.S.

. The meaning of “several” is not clear, but it’s all-important. If it means two, then that’s just a statement of reality — we’re won’t be ready to build much of anything other than LWRs for the next 20 years (although there will be some fast reactors going by then in forward-looking nations like China, Russia, and India).
. But if “several” means five, that’s a different kettle of fish, particularly from a global perspective. For a simplistic example, suppose every reactor that comes on line before 2060 is an LWR, and every subsequent reactor to come on line is an IFR. And suppose also that the global population ,stabilizes at perhaps 11 billion by 2110, and that our global goal is to have enough clean, carbon-free energy to provide everyone with 50-60% of the per-capita energy now enjoyed by Americans.
. Data: To fuel a 1-GWe LWR for one year, ore containing about 140 tonnes of natural uranium has to be mined, and each LWR can be expected to last at least 60 years.
. To meet our global goal, after allowing considerable growth in wind and solar power, nuclear capacity would have to grow at a compounded rate of about 5.4%, starting with the current capacity of about 300 GWe. This puts the nuclear capacity in 2060 (5 decades hence) at ~4000 GWe. (Most of the numbers here are very approximate, but in the ballpark.)
. That means that between now and 2060 there will be ~94,000 GWe-years of operations requiring 94,000 x 140 = 13.6 million tonnes of uranium to have been mined and fed to enrichment plants.
. And say that the entire 4000 GWe is provided by LWRs, which in 2060 will have an average life expectancy of something like 40 more years. Then the used-plus-committed natural uranium will amount to 13.6 million + (4000 x 60 x 140) = 47 million tonnes of uranium.
. Undoubtedly that much uranium exists, but not at today’s prices (~$140/kg). The European Nuclear Society says, “The global uranium resources with mining costs up to US $ 130 per kilogram amount to about 3.3 million tonnes.”

. But that’s not all. In 2060 the annual enrichment capacity needed would be 4000/300 = 13 times today’s enrichment capacity — not a situation to please those who worry about proliferation of uranium weapons. And the amount of spent fuel would be almost 20 times what we have today and are wondering what to do with.
. In this simplistic scenario, we are assuming that, starting in 2061, all new reactors coming on line are breeders that get their initial fissile charge existing supplies of plutonium and require no further increase in the rate of uranium mining or enrichment. Whether there’s enough plutonium to accomplish this is a topic for another discussion.
. For a really unrealistic exercise, consider the case where the breeders (or reactors such as LFTRs with unity conversion ratio) don’t come on line until nuclear expansion ceases in 2110. By then, there will have been about 1.4 million LWR-years accumulated and about 2.8 million more LWR years committed, for a net requirement of 4.2 million x 140 = 588 million tonnes of uranium. The enrichment capacity would be about 250 times what we have today.
. But if, on the other hand, we commit ourselves now to deploy breeders as soon as possible, the global future uranium requirements will total something like 16 million tonnes, an amount that is probably available at prices not many times today’s. Also, the need for any mining or enrichment capacity will have disappeared by 2050 or so. And the entire global inventory of plutonium could be safely sequestered in operating nuclear plants by the end of the century.
. Caveat: Some of the above numbers come from an algorithm I have been working on, and it has not been thoroughly tested. If anybody can independently confirm or refute any or all of the above, please do so.

— George


Charles Barton,

You suggest that ships should be powered by small MSRs. That sounds like a great idea to me but is anyone offering such power plants. What about using power plants derived from military designs?


George Stanford:

Your response to Tom Keene was very informative. It is good to hear from authorities on the subject. I would really appreciate answers to a few follow up questions.

1) You suggest that 4000GW nuclear capacity will be needed by 2060 after allowing for considerable growth of wind and solar power. How much power have you allowed from the renewables? I would suspect that a more economic route might be to increase the proportion of nuclear should such be possible within the time frame. Is it?

2) You suggest that we are stuck with LWRs for 2 to 3 decades. Does this have to be the case? Could we move quicker by using transitional technology?

3) Is it correct to state that the ARC-100 is to the IFR what the uranium converting DMSR is to the LFTR and that both could be deployed more quickly than full breeders? If so, how much more quickly?

4) Is it correct to think that the above will defer the requirement to demonstrate and hence licence the appropriate reprocessing technologies for two to three decades and that both are designed to be constructed as small, factory built units such that one might hope that they would be cheaper to build and capable of more rapid rollout? If so, are they more likely to facilitate or delay the deployment of IFRs and LFTR iso breeders?

5) What, in essence differentiates the IFR from the ARC-100 anyway other than onsite reprocessing in the case of the latter? If small factory built modular units are going to be cheaper and quicker to build, wouldn’t it be more pragmatic to have regional reprocessing centres rather than one with each plant?

I am acutely aware that a little learning is a dangerous thing and I therefore hope for some expert advice.


Douglas, below are my responses to your questions. Be aware, however, that I don’t claim to be the horse’s mouth, and that this a fanciful scenario with no predictive value whatever.

1) You suggest that 4000GW nuclear capacity will be needed by 2060 after allowing for considerable growth of wind and solar power. How much power have you allowed from the renewables? I would suspect that a more economic route might be to increase the proportion of nuclear should such be possible within the time frame. Is it?

. My scenario is only partly baked at present. I’m assuming that the non-nuclear portion of the energy economy remains at its current value (in quads per year), leaving unspecified the split between fossil fuel and non-nuclear renewables. If the latter were to come anywhere near the current fossil-fuel contribution, that would be a considerable expansion.

2) You suggest that we are stuck with LWRs for 2 to 3 decades. Does this have to be the case? Could we move quicker by using transitional technology?

. I said *two* decades — we can hope that the pace of change will be picking up during the third. And I do expect that “transitional technology” will start to be deployed within the next two decades — but it will mainly be non-breeding fuel cycles for niche applications (e.g., non-breeding SMRs), and it will be make only a minor contribution to the overall energy supply by 2030. But yes, their deployment will provide valuable base-line info. Things will move faster in countries other than the U.S.

. Incidentally, I’m using “breed” in its intuitive sense, namely, implying that the conversion ratio CR — the number of fissile atoms created per heavy-metal atom fissioned — is significantly greater than unity. The customary term for a reactor with CR of unity or less is “converter.”

3) Is it correct to state that the ARC-100 is to the IFR what the uranium-converting DMSR is to the LFTR, and that both could be deployed more quickly than full breeders? If so, how much more quickly?

. The ARC-100 is a SMR based on IFR technology, and its deployment should advance the state-of-the-art. Whether it is deployed before the PRISM is hard to predict. Regarding the DMSR & LFTR, that technology seems to be a decade or more behind the PRISM, in terms of readiness for commercialization.

4) Is it correct to think that the above will defer the requirement to demonstrate and hence licence the appropriate reprocessing technologies for two to three decades and that both are designed to be constructed as small, factory built units such that one might hope that they would be cheaper to build and capable of more rapid rollout? If so, are they more likely to facilitate or delay the deployment of IFRs and LFTR iso breeders?

. No. The urgent need now is to show how to bring pyroprocessing up to commercial scale, so that breeders can be coming on line by 2030 or so. It would be short-sighted to delay finalizing development of the IFR technology — which, after all, can do just about everything that needs to be done — in the hope that something marginally better will come along if only we do more R&D.

5) What, in essence differentiates the IFR from the ARC-100 anyway other than onsite reprocessing in the case of the latter? If small factory built modular units are going to be cheaper and quicker to build, wouldn’t it be more pragmatic to have regional reprocessing centres rather than one with each plant?

. The ARC is a small (100-MWe) modular reactor with a CR of unity. It is intended for long-life “nuclear battery” applications, for which breeding would not be appropriate and centralized reprocessing of the returned cores would be.

. The PRISM version of the IFR is somewhat larger (~300-GWe modules), for use in multi-unit power plants. Keep in mind that pyroprocessing is done in batches, which means that economy of scale is less important than you might think. One advantage of in-plant processing is that it can be integrated with the reactors, and commerce in plutonium is all but eliminated. The annual input to a 1-GWe IFR plant operating in breeder mode consists of ~1.5 tonnes of uranium. The output is heat and electric power, one tonne of fission products (with trace amounts of actinides), and some very radioactive fuel containing 0.5 tonne of Pu, to be delivered to a new IFR that is starting up somewhere else.

— George



Very helpful. Many thanks. The comment that mademost impact on me was your opinion that the DMSR would be a decade behind the IFR.



George, thanks for all that information in your response to Douglas, particularly on the ARC-100 reactor.

From my reading of the ARC white paper and the EBR-II report, it looks like the ARC-100 is almost exactly the EBR-II. In other words, ARC is basically commercialising EBR-II, while GE is taking the scaled up version (PRISM) to market. Its the same reactor concept targeted at different markets, differentiated by power outputs.

Is that a fair reading of the situation?

If so, what differentiates the isobreeder (ARC) from the breeder reactor? Could the ARC reactor be run as a breeder by using a different fuel configuration? Or does breeding require changes to the reactor design?



. The ARC-100, Toshiba 4S, PRISM, and TWR are all variants on the IFR theme, which Argonne developed using EBR-II as a test bed. IFRs can be designed to have a conversion ratio anywhere between approximately 0.5 and 1.5, depending on the loading configuration. I suspect, however, that the reactor vessels for the 4S and ARC-100 would too small to accommodate the somewhat larger core that would be needed for appreciable net breeding (although maybe it could be done by reducing the interval between core changes). And the TWR has other, unique constraints.

. On the Web there’s considerable info on the ARC-100 — for example, see
, and

. — George


What I’ve learned is. don’t use angle brackets — they don’t come through, and neither, apparently, does anything between them.
— George


While the full LFTR iso-breeder definitely needs some more research, I don’t think a 10-year gap between IFR and DMSR is fair. There is a detailed design, based on extrapolating results of smaller scale experiments (MSRE). A prototype at 100 MW(e) scale, preferably bigger, is needed. That’s not very different from the IFR case. My main reservation about DMSR is that it is too good as a stopgap. By reducing uranium needs by a factor of five vs PWRs, it allows us to get well into the 2nd half of the century (on George’s scenario above) without running into severe uranium constraints. This ought to be a good option to have, but if it is simply seen as an excuse to do nothing about developing real solutions that need no fissile input and leave no long term (actinide) waste output, then it might not be. DMSR waste also makes terrible start-up fuel – lots of non-fissile Pu isotopes and heavy actinides. The DMSR design priority was proliferation resistance, not waste minimisation.


Douglas Wise:

Can I therefore assume that it should be technically possible to generate, say, 3 times the power that the paper contemplated by mid century without hitting constraints relating to nuclear fuel (a level necessary to put a serious dent in AGW)?

Yes, at least. My next post on BNC, which I’ll put up in a few days, outlines some scenarios for this. It’s a work in progress, and I’d hoped to do more on it before posting, but I think it would be better to just put it up and invite comments/improvements.

Short answer is I can see a pathway to 12 TWe of nuclear (a thermal/IFR/LFTR combo) by 2050 without any extreme assumptions regarding fuel supply or breeding rates.



Luke, I was interested in your DMSR comments. Could you elaborate?

1) You thought George was being unfair to suggest that the DMSR would be a decade behind the IFR (while agreeing that this would be the case for the full LFTR iso-breeder). I don’t have the wherewithal with which to judge whose view is correct. However, you were not clear as to whether you thought the DMSR would be ready, say, a decade before the IFR or at about the same time. One of its potential attractions would disappear unless it could be brought on stream sooner than the IFR.

2) It would appear that the term, DMSR, is generic. When you say that its waste makes terrible start up fuel, does this apply to all designs? I was thinking specifically about those that relied on uranium only and didn’t blend it with thorium. While this approach might be less efficient than a version using blended fuel, it would nevertheless seem to be considerably more efficient than LWRs in terms of fuel economy. If one then considers the potential relative economic advantages, it would seem to have virtue unless those same advantages were true for the ARC-100 or unless the DMSR would post date the IFR in its rollout , as George suggests.



I shall look forward to your next post. I hope that it may touch on economic as well as technical determinants of rate of rollout, but that might be asking for too much in one post.


Sorry to have been unclear. I should also point out that my opinion is in no way expert. Look up David LeBlanc’s articles on DMSR, and judge from them. To clarify my own opinion, it is that the gap in technical readiness between the IFR and DMSR is less than 10 years, but whether it’s 2 or 5 I’m not qualified to determine. However, the major problem for IFR deployment is not technology but politics. IFR development was shut down for political reasons due to misplaced fears about proliferation. That political opposition still exists. The DMSR was designed specifically to allay these concerns, at the expense of other features. These design decisions – like removing most of the reprocessing system – also make it simpler. Which is the better course, to try to defeat political opposition, so we can build the technically best reactor , or to accept that we aren’t going to be allowed to do that and build what we can? That might be DMSRs, or ARC100s.

There are lots of possible variations on the DMSR theme, but the only one with real design work done is the graphite moderated, thorium + uranium mixed fuel version. No version produces much Pu-239 for starting up IFRs or LFTRs. Any that is formed is left to stew until it fissions or transmutes in to something else.



You make an important distinction between whether it is better to plot the way forward with the technically best or politically most acceptable nuclear designs. This, I think, is a decision that is only of real concern in liberal democracies, but is nonetheless of supreme importance to their denizens. I think that, maybe, one should add in economic concerns which are not necessarily related to political ones in the short term.

Having tried to digest and integrate all the answers to the questions I have been asking on this and other recent threads, I am coming round to the view that the ARC-100 approach might be the best way forward in the short term for the following reasons:

1) It should be capable of deployment sooner than the IFR.
2) Its development is more likely to expedite than delay deployment of the IFR.
3) It is, to all intents and purposes, an iso-breeder.
4) It can be deemed to be more proliferation resistant than the IFR.
5) It has the capability of being factory built and rapidly rolled out.

The potential disadvantage might relate to economy of scale. Will the advantages of modular construction outweigh each module’s relatively small electrical output? Would 10 modules/GW be competitive with PRISM’s 3 modules/GW?

I can fully appreciate why it is somewhat pointless for an ignoramus such as myself to worry about such matters when there are others with a great deal more knowledge. However, if it is one’s intention to try and exert influence to bring about nuclear solutions to the peak oil and AGW problems we face, there would seem to be advantages in being well briefed. It seems to me that many of the experts working in the nuclear industry are not driven by a sense of urgency. For example, is GE maximising its propaganda effort towards rapid deployment of the PRISM or is it happy to go along the ALWR route or any other route that will generate the best returns in the short term?

While we are stuck with ALWRs to produce all our nuclear power, nuclear critics will probably continue to flourish until such time as our economic state is so parlous that we can’t move fast enough to prevent catastrophe. The sooner we can demonstrate the means to make nuclear power sustainable and more proliferation resistant and show that the waste problem is of no practical consequence, the more quickly opposition will wilt.

I think it possible/probable that molten salt technolgy will eventually provide a more economic way forward, but I think the priority now is to go with the fastest demonstration we can get of new generation nuclear in the liberal democracies.


George Stanford and I have sorted out our disagreement on LFTR breeding, but other issues remain to be settled. For example the notion that the IFR is ahead of MSRs in the race to deploy. However, a review of IFR development related documents on the information bridge suggests that all IFR development to date was directed to low IFRs that offered only a low breeding ratio. I believe that 1.07 was the highest I saw. The ARC-100, based closely on the EBR-II IFR prototype, is a converter, with something like am 0.83 or 0.87 conversion ration. Given the much larger fissionable inventory which fast reactors require, the number of low ratio IFR breeders that can be deployed will be quite limited, and that number will grow relatively slowly.

High breeding ration IFRs can increase in deployment numbers fairly rapidly. But it will require quite a lot of R&D, and perhaps more than 1 prototype, to bring the high breeding ration IFR on line. I have pointed to the Indian metal fuel fast reactors as an example. The indians already have a commercial breeder program, but expect to take at least 15 years to develop their metal fueled FBR. It seems unlikely that a high breeding ratio IFR can be developed to the point of commercial production within a shorter period of time. Given equivalent R&D resources, it is more probable that the relatively simple DMSR will be developed before the high breeding ratio IFR.


Barry, point me to the report on the successful operation of a high breeding IFR prototype, or at least an advanced design scheme for a high breeding IFR prototype. My understanding is that he only IFR prototype was the EBR-II that was a converter not a breeder. The only prototype IFR designs I have been able to find on the Information Bridge have a maximum breeding ratio of 1.07. My interpretation of this is that the high breeding ratio IFR will need considerable R&D. The results of the EBR-II IFR experiments are documented, and the Indians have read all of the reports. If the Indians need an extensive development program it was because they want to know a lot of things that were not included in IFR R&D reports. Chances are American reactor developers also lack that information.


Charles, you are of course free to make whatever interpretations you will — but it doesn’t make it right, or correctly in context. As such, I respectfully suggest that you have relatively limited knowledge on this matter and are drawing erroneous inferences, whereas the members of the IFRG, with hundreds of cumulative man years worth of experience in operating and researching fast reactors, know what they’re talking about when they speak of high potential BRs and commercial-demo-ready designs. Indeed, the principles of breeding rates in a hard spectrum, and how this is influenced by core/blanket design, smear density, etc. are well established, and as such, were not a major research question that the EBR-II team needed to pursue. That team instead focused on matters like SFR operational/safety performance and metal fuels testing.


Douglas Wise, on 27 September 2010 at 17.00 — Since thermal efficiency isn’t much of an issue, the lower manufacturing and site construction costs+delays for smaller units probably makes the 100 MWe units preferable to larger ones.

Also siting issues are surely lessened.


Barry, I think that in general, I play a useful role by playing the Devil’s Advocate here. You do not point to the sort of evidence that a rational person would find convincing, in fact you do not point to any evidence at all. You sat “that members of the IFRG, with hundreds of cumulative man years worth of experience in operating and researching fast reactors, know what they’re talking about when they speak of high potential BRs and commercial-demo-ready designs.” But I don’t know what they say about R&D programs. ORNL went through the exercise of documenting in detail the R&D requirements for the creation of a MSBR prototype (ORNL-4812 This document shows that there were no huge technological gaps standing in the way of MSBR prototype development in 1972. it would be rational to look for similar assurance that the high breeding ration IFR could be accessible with a similarly modest effort. Since I lack documented estimates from the IFR community as to the high breeding ration R&D project size, I have to go with the nearest analogy, which is the very large Indian R&D project with similar goals.

Note that I am not questioning the potential for high breeding ratio. Rather my issue is the size of the research project that would be required to produce a commercial prototype.

You claim, “the principles of breeding rates in a hard spectrum, and how this is influenced by core/blanket design, smear density, etc. are well established, and as such, were not a major research question that the EBR-II team needed to pursue. That team instead focused on matters like SFR operational/safety performance and metal fuels testing.” Is it your position then that the IFR breeding ratio can be increased from a maximim of 1.07 to 1.65 with no technological issues to be resolved? I would like to have confirmation from a published source if you maintain that.


Charles, I agree that you play a useful devil’s advocate role.

Is it your position then that the IFR breeding ratio can be increased from a maximim of 1.07 to 1.65 with no technological issues to be resolved? I would like to have confirmation from a published source if you maintain that.

Yes, it is — I will work through this material on the blog in the coming weeks, with reference to published sources where possible. But most of this is in ANL reports, such as C.E. Till, Y.I. Chang, et al, “Fast Breeder Studies,” ANL-80-40 (1980). This report documented the Argonne contributions to the International Nuclear Fuel Cycle Studies (INFCE).



You refer above to “commercial-demo-ready designs”. Other than money, what is needed before such demonstration units can actually be built? Presumably, regulators get involved, but are they likely to slow things down as much as in the case of fully commercial units? Is there, for example, an accelerated path to take for a small one off designs? Alternatively, one might even envisage even longer delays for new technologies.


Douglas, in the US it (S-PRISM) would need to be built by government as a demonstration project. This would fast track later NRC approval for commercial rollout. The DOE need to make the decision to do it. Or, internationally, someone (probably Russia), needs to decide to build the first one, and the US needs to agree to a tech sharing arrangements — perhaps with a consortium of interested countries (e.g. Russia, Japan, South Korea, China, India etc.). Also, a facility for converting LWR spent fuel into PRISM feedstock is an essential first step, and should be done in the US. I think the international PRISM and US LWR spent fuel facility are the most likely pathways.



Thanks for the link. It seems that the manager of the DoE’s Savannah River Site has an excellent plan to allow construction of several different demonstrator reactors. However, it seems that NRC will make things as difficult as possible and none of the power could be released to the grid without involving them. They also seem to be intimating that they won’t be considering licensing any commercial designs except those already on its books for many years. Unless something changes dramatically, it seems that the NRC is committing the citizens of the States to energy/economic poverty.

Any chance that the UK could steal a march and host a range of demonstrator units? We have potential sites and plenty of fissile material. Do we have enlightened leadership to force through the necessary decisions?


I have a question for George about the metal fuel in the IFR.

The safety advantage of the metal fuel is that due to properties of thermal expansion, in a loss of coolant situation, the fuel initially heats up and then the reactor goes subcritical. so the metal fuel prevents a criticality accident.

How does the passive safety of the metal fuel compare to the negative temperature coefficient of a water moderated reactor? The latter also prevents a criticality accident, thus TMI only had a partial meltdown.

It’s my impression though that the metal fuel negative feedback mechanism is safer than the water moderated one–that in the EBR 2 demos, there was no partial meltdown as in TMI.

Is this impression correct? and just out of curiosity, is the negative temp coefficent of water moderated reactors technically considered a passive safety feature?


. Let’s not overemphasize the importance of the IFR’s safety advantages. While they are indeed real, the LWRs (and HWRs) already built have proven to be so safe (TMI hurt nobody), that further improvements are really just gilding the lily.
. In comparing oxide (or carbide) fuel with metallic, the most important consideration is not safety-related at all — it’s the much superior breeding potential of metallic fuel (the neutrons are less moderated).
. However, here’s a couple of the technical safety considerations.
– During an unprotected loss-of-flow, the coolant temperature rises rapidly, which introduces negative reactivity feedbacks due to radial expansion of core, control rod drive-line expansion, etc. This brings the power down.
– The high thermal conductivity of metal means that the fuel has far less stored heat to be dissipated in the event of loss of coolant flow, greatly reducing the temperature swings, making passive convective cooling much more feasible. In the safety demos at EBR-II in 1986, completely passive effects brought the coolant (and core) temperature down to ~200 degrees F below the boiling point of the sodium coolant.
. You ask,

How does the passive safety of the metal fuel compare to the negative temperature coefficient of a water moderated reactor?

What counts is whether the negative thermal-expansion compensates for the positive Doppler effect, and with metal fuel the thermal expansion comfortably overrides the Doppler-induced reactivity increase — not necessarily the case with a water-moderated reactor. (If you don’t happen to know what “Doppler” means here, don’t worry about it. It’s an effect that tends to increase the fission rate as the temperature of the fuel rises.)

Is the negative temp coefficient of water moderated reactors technically considered a passive safety feature?

Since this causes something to happen without human or mechanical intervention, it’s a passive feature.


george: thank you very much.

when I first read about the IFR, the passive safety features of the metal fuel really stood out (they stood out due to my ignorance of the safety features of current reactors).

so when I read about what seemed (to my untrained eye) a similar feature in water reactors, I was just curious.

at any rate, the negative feedback feature of the metal fuel does indeed seem an improvement over the water moderated reactor, though as you note, and I agree, we should not make so much of the gen 4 safety features as that might imply that most reactors currently in use are not safe.

can you explain radial expansion and control rod drive line expansion? are these negative reactivity feedbacks properties of IFRs specifically?

I (just) found a 2009 INL article by R Wigeland that discuss it but have not read it yet.


Reactor dynamics is not my specialty, but here’s how I see it. When the whole core expands, everything gets farther apart, which means that “neutron leakage” increases — i.e., more neutrons can escape from the core without causing fissions — decreasing the reactivity.
That would apply to any reactor. The IFR’s advantage lies in the fact that the (positive) Doppler effect is considerably smaller than the (negative) expansion effect..
The control-rod assemblies in an IFR (such as GE’s PRISM) would be anchored at the top of the reactor vessel, quite far above the core, while the core itself rests on the bottom of the vessel. When the coolant temperature increases, the heated drive trains expand more than the less-heated vessel does, resulting in a net insertion of neutron absorber into the core.
Perhaps the Wigeland paper you refer to is the one here:
which has a considerably more detailed discussion.


Geroge S Stanford,

Thank you for your clear statement:

Let’s not overemphasize the importance of the IFR’s safety advantages. While they are indeed real, the LWRs (and HWRs) already built have proven to be so safe (TMI hurt nobody), that further improvements are really just gilding the lily.

That seems reasonable and realistic to me.

Taksing thsi into account, I believe the reasons for developing Gen IV are:

1. To reduce the cost of electricity (this is by far the most important reason from my perspective, and should be the number one stated objective)

2. to improve the rate at which nuclear power can be implemented all over the world, so we can avoid building new, and replace existing, fossil fuel plants

3. to improve reliability of supply over the longer term (a long term, not a short term, requirement).


More on the ‘balloon storage’.

Some across in another science forum had this reply. This is all a bit confusing for me, because being non-technical I hate it when ‘experts’ I respect disagree with each other. I don’t have the time to dig deep enough into the physics to decide one way or another. All I can do is play ‘Devil’s Advocate’ and ultimately admit I don’t know if reasonable enough objections are mounted either way.

“When I read the title, I immediately gave a bit of thought to the thermodynamic issues but I don’t quite agree with those objections. If the deep water is no warmer than the environment at the surface, good design could make it efficient. Further, if the temperature is significantly lower down below, the difference could even be exploited to draw a bit of thermal energy from the environment, while you’re at it. It all depends on how feasible a good design would be. ”

Then this…

“according to this (i can’t vouch for its accuracy though, it just seemed reasonable to me), the temperature at depth is between 3 and 6 degrees, and on the surface between 4 and 28 degrees. so if done on the equator, there could be some potential to that idea.

personally i’d be more worried about friction and viscosity losses over such a large length of piped air, i think.”


EN, without seeing the context of the commenter you quote, I can only guess, but it looks like they are just thinking in terms of the temperatures of the two different heat sinks (surface and deep water) and ignoring the large pressure difference.

I put some numbers on this in this comment. Here’s another way to think about it.

Imagine pushing a large beachball under water. It takes a lot of effort, working against the bouyancy force. You do more work as you push it deeper. Part of the work winds up as thermal energy in the air in the ball. Quite a lot of it in fact. The work you do to push it down to 750 m would heat the air to ~730 C, if the ball were insulated and the heat couldn’t escape. But its not insulated, and the ball doesn’t get hot, because the heat just gets transferred to the ocean, and lost.

Think about this at the scale of a grid connected storage system. Unless the whole submerged system is insulated, you’re throwing away a lot of the energy you used to push the air underwater. This energy is “recovered” by using gas to reheat the air.

The energy associated with water temperature differences of 10 or 20 degrees which your correspondent is thinking of are irrelevant in the context of energy losses associated with many hundreds of degrees. Without calculating it out, I would also expect it would dwarf the pumping losses.

I hope that helps conceptualize whats going on here.


[…] Finally, I was a bit disappointed by how at the end of the film, the issue of waste transmutation was dismissed almost out of hand as being a pipe dream. It was also claimed that uranium would run out ‘just like oil’ in around 100 years. It just so happens that transmutation of nuclear waste and extension of nuclear fuel supplies are both accomplished in the same way – by using an advanced type of nuclear power station called a fast reactor.This type of reactor uses the actinides discussed earlier as fuel, meaning that only the fission products remain (see first graph) and the supply of fuel can be extended by up to 100 times. I probably will not be discussing this type of reactor on this website, as it already has been done excellently elsewhere. […]


Leave a Reply (Markdown is enabled)

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s