In December, the U.S. Department of Energy issued a loan guarantee solicitation for $12.5 billion for “advanced” reactors. While the solicitation is not specific about what kinds of advanced reactors, many of the designs expected to compete will be fast neutron based reactors, or “fast reactors,” for short.

But the federal government would be well-advised to proceed with caution. For decades, efforts around the world to commercialize fast reactors have ended in failure.  The expenditure of billions of dollars to date and the serious underlying technical problems with such reactors suggest that further pursuit of research and development of these so-called “advanced” reactors amounts to throwing good money after bad.


The push for fast reactors comes in the context of the current slowdown of nuclear power, both globally and within the United States.  Nuclear energy contributes barely 10 percent of global electricity generated, down from a historical maximum of 17 percent in 1993. In the United States, the number of operating nuclear power plants has slipped below 100 for the first time since 1986 with the shutdown of the Vermont Yankee plant.

The shaky premise underlying much of the advocacy of advanced reactors is that the slowdown of nuclear power is due to the kind of reactors that are operated or constructed today. If new kinds of reactors are constructed, it is argued, then nuclear power will be revived. But this argument ignores the fact that the primary challenge that nuclear power faces in the United States is economics: Nuclear reactors are simply not cost competitive with alternate sources of electricity generation, including wind and natural gas.

In theory, the big “plus” associated with fast breeder reactors is reducing the requirement for uranium used as fuel. When nuclear power plants were first built, uranium was thought to be scarce and there was widespread concern that global resources would be insufficient to support the anticipated large expansion of nuclear power.

We know better now. There hasn’t been an expansion of nuclear power. But, more importantly, geologists had concluded by the late 1970s that uranium ore is plentiful, especially if one were to mine poorer grades of ore. The other indicator that availability of uranium will not be a problem is that the price of uranium, when corrected for inflation, has hardly risen. If less reliance on uranium is the major selling point for fast breeder reactors, it is not a persuasive one.

Historical experience has also taught us that breeder reactors have persistent reliability problems—the French Superphenix reactor had a lifetime load factor, a measure of operational efficiency, of 7.9 percent.  Because breeder reactors generate a large amount of heat in a very small volume, they are forced to use molten metals, usually molten sodium, to remove this heat from the reactor core. Since sodium is opaque and burns on contact with air or water, breeder reactors are notoriously difficult to maintain and susceptible to serious fires and long shutdowns. The chemical interactions between sodium and the stainless steel used in various components of the reactor are a potential cause of persistent leaks. A sodium leak caused the shutdown of the Japanese Monju breeder reactor on 8 December 1995.

Breeder reactors are also susceptible to catastrophic accidents because of technical properties of sodium and the nature of the nuclear core. Having to deal with these properties naturally makes breeder reactors more expensive than water-cooled reactors. In other words, fast breeder reactors are more expensive to build and will operate with less reliability.

Yet another problem with breeder reactors is that the fuel required contains plutonium. Producing plutonium is an expensive process. Further, plutonium is tens of thousands of times more radioactive than the uranium used to fuel water-cooled reactors.  The expensive safety precautions required during fabrication of plutonium-bearing fuel increases the cost of fueling breeder reactors to much beyond the fuel-fabrication cost for standard reactors today. Given that many standard nuclear reactors facing shutdown in the United States simply because their operational and fueling costs are too high, fast breeder reactors would be a non-starter in competitive electricity markets.

And, finally, because of the use of plutonium to fuel breeder reactors, there is an obvious connection with nuclear weapons programs. An illustration of the connection is provided by India’s former Atomic Energy Commission Chairman Anil Kakodkar, who said in February 2006, that the Indian breeder reactor could not be go under international safeguards “because it hurts our strategic interest”.

The U.S. Congress affirmed the uneconomic nature of breeders, not to mention its implications for non-proliferation, it voted to terminate taxpayer support for the prototype Clinch River Breeder Reactor in Tennessee in 1983 and the entire DOE breeder program in 1994. No development in the past two decades offers any reason to revisit that decision. While it is understandable that the foundering U.S. nuclear power industry wants to maintain the appearance that it is just a breakthrough or two away from a new generation of “advanced” nuclear reactors, fast breeder reactors are likely to be the same dead end they already have been for decades. 

Ramana is with the  Nuclear Futures Laboratory and Program on Science and Global Security at the Woodrow Wilson School of Public and International Affairs at Princeton University.