Wednesday, February 6, 2008

How the Fast Breeder Won the Great Breeder Sweepstake

From Chapter 6, "The First Nuclear Era: The Life and Times of A Technological Fixer," by Alvin Weinberg.

As must be clear to the reader by now, the earliest ideas about commercial nuclear power were dominated by the mistaken belief that uranium was very scarce. Moreover, in America at least, electricity generated in conventional coal-fired plants was very cheap. Thus we could hardly imagine that nuclear energy based on burning scarce 235U could compete with coal, or if it could, that the necessary uranium ore would last long. As Eugene Wigner put it, the ultimate purpose of nuclear energy was not to replace coal when fossil fuel was abundant; rather it was to substitute for fossil fuel when the latter became scarce. Energy would then be very expensive: nuclear energy based on breeders would be an inexhaustible energy source, and its cost would be perhaps two times the present cost of fossil energy, rather than ten or fifteen times that cost. (Were energy cost to rise so steeply, a large share of our GNP would be spent for energy, and this would reduce our standard of living.)

About this time I wrote two essays on the role of the breeder. The first, "Power Breeding as a National Objective," appeared in Nucleonics in 1958 (vol. 16, no. 8, pp. 75-6): in this piece I argued that current economics alone should not be the sole basis for choosing which reactor system to pursue. Efficient use of the raw materials of nuclear energy—uranium and thorium—was equally important. Indeed it would eventually be more important than estimates of the cost of power. Once inexpensive ores were used up, nuclear power from nonbreeders would be very expensive. This general argument remains valid today, except for one point: if the capital cost of the breeder is too high, the price of uranium ore at which the breeder, with its more efficient utilization of ore, becomes competitive may be extremely high—say, $300 per kilogram. At this price, uranium resources are vast: the breeder can price itself out of business if its capital is too high. Unfortunately, this seems to be the case at the moment, but with Japan and France actively pursuing "power breeding as a national objective," capital costs may come down within the next fifty years.

The other essay, "Energy as an Ultimate Raw Material, or Burning the Rocks and Burning the Sea," appeared in 1959 in Physics Today (vol. 12, no. 11, p. 18). In this essay I speculated on the very long-range future-hundreds, even thousands, of years in the future. Where will our energy come from at that distant time when coal, oil, and natural gas have been used up? Solar energy is one obvious inexhaustible source. Another, if it works, could be controlled thermonuclear energy based on deuterium from the sea (thus "Burning the Sea"). My main point, however, was to stress what Phil Morrison and then Harrison Brown had already noticed: that the residual and all but infinite uranium and thorium in granite rocks could be burned with an energy yield larger than the energy required to mine and refine the ore—but only if breeders, which could burn nearly all the fertile material, are used. I spoke of "Burning the Rocks": the breeder, no less than controlled fusion, is an inexhaustible energy system. Up till then we had thought that breeders, burning 50% instead of 2% of the uranium, extended the energy derivable from fission "only" 25-fold. But, because the breeder uses its raw material so efficiently, one can afford to utilize much more expensive—that is, dilute—ores, and these are practically inexhaustible. The breeder indeed will allow humankind to "Burn the Rocks" to achieve inexhaustible energy!

Until then I had never quite appreciated the full significance of the breeder. But now I became obsessed with the idea that humankind's whole future depended on the breeder. For society generally to achieve and maintain a living standard of today's developed countries depends on the availability of a relatively cheap, inexhaustible source of energy. (As I write these words, I realize that until recently I tended to dismiss solar energy as too expensive, and fusion as probably infeasible. I really don't know whether this will always be the case.)

The breeder became central in my thinking about nuclear-energy development. And, with Glenn Seaborg's becoming the chair of AEC in 1960, the breeder acquired ever-increasing status with AEC—especially recognition as an essentially inexhaustible source of energy.

In 1962, the AEC issued a report to the president on civilian nuclear power. Lee Haworth, a superbly responsible physicist-administrator, was in charge of drafting the report. He projected a nuclear deployment by 2000 of about 700 gigawatts (compared with the actual deployment in 1993 of 102 gigawatts), which seemed at the time quite reasonable. Both the fast breeder based on the 239Pu-238U cycle and the thermal breeder based on the 233U-232Th cycle figured prominently in the report. Indeed, the report implied that both systems should be pursued seriously, including large-scale reactor experimentation. It particularly favored molten uranium salts for the thermal breeder. But the molten-salt system was never given a real chance. Although the AEC established an office labeled "Fast Breeder," no corresponding office labeled "Thermal Breeder" was established. As a result, the center of gravity of breeder development moved strongly to the fast breeder; the thermal breeder, as represented by the molten-salt project, was left to dwindle and eventually to die.

The fast-breeder project in the United States centered around the Clinch River Breeder, a 250-megawatt sodium-cooled breeder to be built in Oak Ridge by Westinghouse. But, by this time, objections to the breeder were being voiced, ostensibly because the breeder, with its coupled chemical reprocessing system, lent itself to the clandestine diversion of plutonium for nuclear weapons. But in my view the real aim of some of the more dedicated opponents of Clinch River was the extirpation of nuclear energy. The Clinch River Breeder was a handy and vulnerable target, particularly since it could not produce power at a competitive cost. And the opponents eventually won—Clinch River was killed in 1975.

Although the molten-salt system was never allowed to show its full capability as a breeder, a 233U-232Th thermal breeder was demonstrated in Admiral Rickover's Shippingport reactor. Operating with 233U fuel and a thorium blanket, this reactor actually demonstrated a breeding ratio of 1.03—i.e., for every 233U burned, 1.03 new 233U was produced. This accomplishment has gone unnoticed since the cost of power from Shippingport is much higher than from other sources. Whether, as cheap uranium becomes scarce, other reactors will be fueled with 233U and thorium remains to be seen. Thus, as Wigner once said, breeders may emerge from incremental improvements of existing light-water or heavy-water reactors, or may spring from entirely new technologies specifically designed for the breeder. As for fast uranium breeders, the latter path is being followed in France, Japan, India, and Russia. (The French fast breeder PHENIX has demonstrated a breeding ratio of 1.13.) But as for thermal thorium breeders, it seems that these will emerge from the existing nonbreeder LWR or CANDU rather than from molten-salt technology.

Why didn't the molten-salt system, so elegant and so well thought-out, prevail? I've already given the political reason: that the fast breeder arrived first and was therefore able to consolidate its political position within the AEC. But there was another, more technical reason. The molten-salt technology is entirely different from the technology of any other reactor. To the inexperienced, molten-salt technology is daunting. This certainly seemed to be Milton Shaw's attitude toward molten salts—and he after all was director of reactor development at the AEC during the molten-salt development. Perhaps the moral to be drawn is that a technology that differs too much from an existing technology has not one hurdle to overcome—to demonstrate its feasibility—but another even greater one—to convince influential individuals and organizations who are intellectually and emotionally attached to a different technology that they should adopt the new path. This, the molten-salt system could not do. It was a successful technology that was dropped because it was too different from the main lines of reactor development. But if weaknesses in other systems are eventually revealed, I hope that in a second nuclear era, the molten-salt technology will be resurrected.

Afterword: For some reason Weinberg is off by 8 years on the shut down date for the Clinch River Reactor.  That project was shut down by congress in October 1983. 


Sovietologist said...

"It particularly favored molten uranium salts for the thermal breeder. But the molten-salt system was never given a real chance. Although the AEC established an office labeled "Fast Breeder," no corresponding office labeled "Thermal Breeder" was established. As a result, the center of gravity of breeder development moved strongly to the fast breeder; the thermal breeder, as represented by the molten-salt project, was left to dwindle and eventually to die."

I personally think this was probably the biggest mistake in the history of nuclear technology- in retrospect, the MSR seems a much more promising technology than the LMFBR. But France, Japan, and the USSR all made the same decision, and Russia is still pursuing the LMFBR in the form of the BN-800.

Charles Barton said...

The choice of the LMFB has always been based on breeding ratio. The LMFB delivers a superior breeding ratio. The MSR basically delivers only a little more than a self sustaining breeding ratio.

The problem is that the LMFB is a technological monster, that no one has been able to tame. No one has come up with a good way to manage molten sodium in a reactor system.

Think of the LMFB as a race horse that is incredibly fast, but is genetically predisposed to break its leg bone in the third quarter mile of the race. The MSR is a Morgan. It will get you where you want to go, but not so quickly.

Sovietologist said...

I recall seeing Kirk Sorensen mentioning the possibility of a fast MSR. If this could be made workable, it'd have huge advantages over the LMFBR, while potentially achieving a better breeding ratio that the thermalized MSR. Do you know how much this concept has been studied?

Charles Barton said...

As I understand it, there are significant and undesirable tradeoffs for upping the breeding ratio with a MSR.

Sovietologist said...

Drat! Well, could a fast MSR have other applications? Like, say, waste transmutation?

Charles Barton said...

Kirk would be a better person to ask these questions of.

Kirk Sorensen said...

Fast-spectrum MSRs are definitely possible using chlorides as the basic fuel form instead of fluorides. They can have extraordinary breeding ratios in uranium/plutonium, but I would favor using them to destroy transuranic waste and breed U233 from thorium to start fluoride reactors, which would thereafter have essentially break-even breeding ratios.

My main concern about widespread use of chloride reactors is proliferation.

You can read many papers on the subject of chloride reactors here:

Go near the bottom of the page under "chloride reactors".

Charles Barton said...

Thank you Kirk.

Fifi said...

Sorry to crash the party a bit but the MSBR is not all nice and roses.

There is the issue of the positive temperature coefficient on a wide range of the fuel/moderator ratio.

See for quick intro,
More detailed paper here,

Those are fairly recent results based on very fine grid simulation rather than the usual homogeneous assumption that indicates a negative coefficient.

It's either highly moderated or fast/hard-epithermal. Very strong moderation is detrimental to breeding and to moderator aging and the fast spectrum corner also has problems as noted by Kirk.

But in the middle, where it looks very promising, the MSBR is probably unsafe.

My personal preference goes to EBR-II, liquid sodium + sodium bonded metal fuel with a high fuel/coolant ratio for positive void coefficient + off course, the strongly positive temperature coefficient inherent to the EBR-II design. And nearly 30 years of proven operation. Granted though, the electro-refining part is still pretty much lab scale.

Charles Barton said...

Actually, this like much good science, this research closes some doors, but still leaves other options open. LPSC did not explore the use of DU as a moderator for example.

No one who grew up in Oak Ridge during my generation believed that Argonne had the slightest idea what they were doing. Argonne does not realize that it was the devil who tempted people to put sodium inside a reactor.

The EBR-II type reactor actually serves a different purpose than the MSR. The MSR was from the start envisioned as a self sustaining reactor with a 1 to 1 breeding ratio, while the liquid sodium reactor is expected to have a much higher breeding ratio.

The MSR is a candidate for replacing the LWR.

Charles Barton said...

Fifi, this post on Kirk's blog deals with the issue.

Kirk concludes that the LPSC study supports a two fluid appraoch to MSR breeding.

Fifi said...


Make that strongly negative temperature and void coefficient for EBR-II. Sorry about that.

My understanding is that the whole thing was a surprise for the LPSC team.

Beyond the specific result for MSBR designs, it seems to me that the really bad news in this result is that it exhibits a strong difference between the fairly simple calculations that can be done with an homogeneous model and what comes out of a truly detailed simulation.

It puts a big kibosh on the ability of independents to evaluate proposed advanced designs.

Charles Barton said...

I could have told you Fifi. If it came from Argonne, thee had to be something seriously wrong with it. Only those who listen to the voice of the Devil are going to build liquid sodium reactors.

Fifi said...

What's wrong with sodium? EBR-II never had problems with sodium during its nearly 30 years of operation. But the trick was that EBR-II was designed with safety in mind from the very beginning, not as an add-on tacked on the design at the last minute.

For design details and an insight into the engineers' mind, I strongly encourage you to read the reference book on EBR-II by L.J. Koch himself if you haven't done yet.

The real sin of most fast reactors was the choice of oxide fuel rather than metal, allegedly to work well with aqueous reprocessing. That was the BIG mistake. It induces all kind of design compromises that result in poor breeding, poor control and poor safety.

In other words, fast reactors shouldn't be designed by LWR engineers. They are a technology of their own. Funnily, that fundamental fact was recognized early with EBR-II then the lesson was forgotten for Clinch River and the rest.

Charles Barton said...

If the secondary coolant of a liquid sodium reactor is water, you potentially have a very very serious safety issue. Leakage is always a potential problem with liquid cooled reactors. When hot molten sodium metal comes in contact with water it can cause a violent explosion Thus accidents with liquid sodium reactors poses serious dangers.

Aside from a super breeding ratio, there appears little that would commend any LMFB reactor over a MSR. The Coolant/fuel of the MSR is far safer than Liquid Sodium. There is virtually no danger of radioisotopes escaping from the molten salt complex in the event of a leak. The chemical processing of the molten salt was an tremendous advantage over fuel reprocessing systems used wit LSRs. Xenon is easy to remove form a MSR. Fuel can be processed while the reactor is operating. New fuel can be added without shutting down the reactor.

Charles Barton said...

Fifi, Some more problems with sodium cooled reactors include the requirement for rapid reloading cycles for fuel. Some liquid sodium reactors have had fuel cycles are as short as 50 days.

A second problem is the rapid corrosion of pipes valves and other reactor parts under intensive fast neutron bombardment, and the corrosive effects of hot sodium.

In at least some LMFB's breeding has occurred much slower than anticipated. The Soviet BN-600 appears to have taken 20 years to double the amount of Plutonium, and It was claimed that the BN-600 had a better breeding record than British and French LMFB.,M1

Fast breeders also require very large and expensive plutonium reprocessing facilities. In contrast, MSR require relatively simple chemical processing techniques to extract protactinium from molten fuel or a molten salt blanket.


Blog Archive

Some neat videos

Nuclear Advocacy Webring
Ring Owner: Nuclear is Our Future Site: Nuclear is Our Future
Free Site Ring from Bravenet Free Site Ring from Bravenet Free Site Ring from Bravenet Free Site Ring from Bravenet Free Site Ring from Bravenet
Get Your Free Web Ring
Dr. Joe Bonometti speaking on thorium/LFTR technology at Georgia Tech David LeBlanc on LFTR/MSR technology Robert Hargraves on AIM High