An impressive team of scientists were assembled at Argonne under the leadership of Dr. Charles Till, who coordinated his exceptionally talented group in a multi-faceted project to solve all the issues of public concern over nuclear power generation simultaneously: safety, nuclear waste, proliferation, economics, fuel supply and fabrication, and construction. By the time a woefully shortsighted administration shut down the IFR project in 1994, all the problems had been successfully solved and all that was left was to demonstrate the commercial-scale fuel recycling system that would be an integral part of each power plant (hence the “I” in IFR).
The main difference between a fast reactor and a light-water reactor is the speed at which the neutrons move when liberated by the splitting of an atom. In LWRs, water acts as a moderator, slowing the neutrons and thus increasing the chance that they’ll encounter another fissile atom and cause it to split, perpetuating the chain reaction. In a fast reactor, the neutrons are allowed to move at a considerably higher speed, and for this reason the fissile content of the fuel must be higher (in an IFR it would be about 20% as opposed to the 3.5-5% in a LWR).
LWRs operate with water under pressure, hence the concern about pressure vessel leaks, coolant system leaks, etc, as well as the industrial bottleneck of only a single foundry in the world (though more are being built) capable of casting LWR pressure vessels. Fast reactors, on the other hand, usually use liquid sodium metal as the coolant, at or near atmospheric pressure, thereby obviating the need for pressure vessels. Because the boiling point of sodium is quite high, fast reactors can operate at a considerably higher temperature than LWRs, with outlet temperatures of about 550ºC as opposed to the 320ºC of Gen III reactors. Here is a simplified rendering [x] of a sodium-cooled fast reactor to convey the design features:
As can be seen from the picture, the heat exchanger loop, immersed in the reactor pool, contains non-radioactive sodium, which is piped to a heat exchanger in a separate structure where it gives up its heat to a water/steam loop that drives a conventional (Rankine cycle) turbine. This system assures that in the unlikely event of a sodium/water interaction caused by undetected breaching of the double-walled heat exchanger, no radioactive material would be involved and the reactor vessel itself would be unaffected. Such an event, however unlikely, could result in the cessation of flow through the intermediate loop and thus an inability of the system to shed its heat. In a worst-case scenario where such an event happened with the reactor at full power and operators, for whatever reason, failed to insert the control rods to scram the reactor, the passively-safe system would nevertheless shut itself down safely due to inherent properties of the metal fuel (see below), with the large amount of sodium in the reactor vessel then allowing the fission product decay heat from the core to dissipate.
Metal Fuel: The Ultimate Safety Valve
One of the most important of the many superlatives of the IFR is its use of a metal fuel comprised of uranium, plutonium and zirconium, and the ingenious manner in which the Argonne team solved the problems of fuel expansion and fuel fabrication, as well as the potentially dangerous overheating scenario. Unlike the fuel fabrication of oxide-fueled reactors that requires the dimensions of the fuel pellets to be uniform to very exacting tolerances, the metal fuel for the IFR can be simply injected into molds and then cooled and inserted into metal tubes (cladding) with a great deal of dimensional tolerance, with a sodium bond filling any voids. If an accident situation occurs that would cause the core to overheat, such as a loss of coolant flow accident, the metal fuel itself will expand, causing neutron leakage to terminate the chain reaction, relying on nothing but the laws of physics.
The passive safety characteristics of the IFR were tested in EBR-II on April 3, 1986, against two of the most severe accident events postulated for nuclear power plants. The first test (the Loss of Flow Test) simulated a complete station blackout, so that power was lost to all cooling systems. The second test (the Loss of Heat Sink Test) simulated the loss of ability to remove heat from the plant by shutting off power to the secondary cooling system. In both of these tests, the normal safety systems were not allowed to function and the operators did not interfere. The tests were run with the reactor initially at full power.
In both tests, the passive safety features simply shut down the reactor with no damage. The fuel and coolant remained within safe temperature limits as the reactor quickly shut itself down in both cases. Relying only on passive characteristics, EBR-II smoothly returned to a safe condition without activation of any control rods and without action by the reactor operators. The same features responsible for this remarkable performance in EBR-II will be incorporated into the design of future IFR plants, regardless of how large they may be [xi].
While the IFR was under development, a consortium of prominent American companies led by General Electric collaborated with the IFR team to design a commercial-scale reactor based upon the EBR-II research. This design, currently in the hands of GE, is called the PRISM (Power Reactor Innovative Small Module). A somewhat larger version (with a power rating of 380 MWe) is called the S-PRISM. As with all new nuclear reactor designs (and many other potentially hazardous industrial projects), probabilistic risk assessment studies were conducted for the S-PRISM. Among other parameters, the PRA study estimated the frequency with which one could expect a core meltdown. This occurrence was so statistically improbable as to defy imagination. Of course such a number must be divided by the number of reactors in service in order to convey the actual frequency of a hypothetical meltdown. Even so, if one posits that all the energy humanity requires were to be supplies solely by IFRs (an unlikely scenario but one that is entirely possible), the world could expect a core meltdown about once every 435,000 years [xii]. Even if the risk assessment understated the odds by a factor of a thousand, this would still be a reactor design that even the most paranoid could feel good about.
The initial manufacturing and subsequent recycling of the fuel pins themselves is accomplished with a well-understood and widely used electrorefining process, similar to one that is employed every day in aluminum foundries. The simplicity of the system and the small amount of material that would have to be recycled in any power plant—even one containing several reactor modules—is such that factory-built components could be pieced together in a small hot cell at each power plant site. Every 18-24 months, one third of the fuel would be removed from the reactor and replaced by new fuel. The used fuel would be recycled. Approximately 10% of it would be comprised of fission products, which in the recycling process would be entombed in vitrified ceramic and probably stored on-site for the life of the plant. If the reactor core were configured to breed more fissile material than it consumes, then during the recycling process some quantity of plutonium would be removed and fabricated on-site into extra fuel assemblies that could then be used as the primary core load of a new reactor. The long-lived actinides that remain would be incorporated into the new fuel rods, replacing the quantity of fission products removed (and any plutonium that had been extracted for startup fuel for new reactors) with an equal amount of either depleted uranium or reprocessed uranium from LWR spent fuel.
Thus we solve multiple problems at once. The quandary of long-lived nuclear waste is a non-issue since the fission products will decay below the radioactivity level of uranium ore within a few hundred years, (see diagram [xiii]) yet they will be embedded in a stone/glass matrix that won’t leach anything into the environment for thousands of years. The long-lived actinides that cause so much consternation to the public when considering spent nuclear fuel will never leave the site of the IFR power plant (except in the case where new fuel is moved to start up a new IFR), but will instead be recycled back into the reactors, repeatedly, to produce prodigious amounts of clean energy, gradually all being transmuted into either electricity or fission products that pose no troublesome disposal problems. Moreover, all of the spent fuel that has accumulated from operation of past, present and future LWRs can also be consumed as fuel in an IFR; in short, they ‘eat’ nuclear waste.