With world energy demand projected to more than double by 2050, nuclear energy is fast becoming a prime option for the future of global power generation; indeed, the proportion of the world’s energy produced by nuclear means would need to grow by 15 times to stabilize global CO2 emissions. The reactors that will be commissioned to replace and augment the currently operating crop of generation II and III designs (mostly light-water reactors, or LWRs) will need to be safe, proliferation resistant, sustainable, and competitive with other power generation options in order to be politically and economically viable. One possible option for this fourth generation of nuclear reactors is the lead-cooled fast reactor, or LFR; a design which combines favorable features of some other potential Gen IV reactor-types with significantly better behavior in severe accidents.
It is important to understand what motivates the move from conventional LWRs to the more exotic Generation IV proposals. Perhaps primary among these motivations is the desire to move to reactors that use fission induced by fast neutrons rather than the slower moving “thermal” neutrons used in LWRs. Even in LWRs, some of the U-238 is transmuted into isotopes of Plutonium, two of which (Pu-239 and 241) can fission to produce heat. Fast-neutron reactors (FNRs) are designed to capitalize on this process, using the Uranium up to 60 times more efficiently than a traditional reactor. Additionally, more neutrons are produced than are required to sustain the chain reaction. These excess fast neutrons can be used to “breed” additional fuel, or to transmute nuclear wastes (primarily long-lived actinides) into less hazardous forms.
Because of their reliance on fast rather than thermal neutrons, FNR designs are usually quite different from those of LWRs. For one thing, the presence of a moderator to slow fast neutrons into thermal neutrons is unnecessary in FNR designs. This presents a challenge because the coolant in LWRs, water, does double duty as the moderator. For this reason, FNRs must be cooled by something other than water, something that is either transparent or reflective to neutrons but doesn’t slow them down. Commonly, the coolant chosen is a liquid metal such as sodium or lead (or lead-bismuth eutectic in lead-alloy reactors), although designs that are cooled by gasses like helium or by liquid salts have also been suggested.
There has been considerable experience over the last half century with liquid-metal-cooled fast reactors, totaling around 390 reactor-years of operation, 70 of them with the compact Pb-Bi cooled reactors aboard Soviet Alfa-Class attack submarines. Most of the reactor-years of experience have been with sodium-cooled fast-reactors (SFRs); so what could motivate us to pick the less tried lead-cooled designs over the sodium-cooled ones?
The main reason is that while Sodium has an admirably low melting point, it also has a low boiling point of 1156 K. Additionally, Sodium is highly reactive with air and water. This creates an obvious safety risk should the coolant-loop be breached in some way, as well as increasing design complication due to the necessity of an intermediate coolant loop. A sodium leak and accompanying fire was the reason for the discontinuation of operations at the Japanese MONJU reactor in 1995.
Lead and lead-alloys, on the other hand, are nonreactive with air, water, steam, or CO2, which means that they do not require an intermediate coolant loop with all its inherent inefficiencies and complications. Lead also provides better gamma-ray shielding and neutron reflection properties, the latter resulting in better neutron economy. Because lead has a much higher boiling point and lower vapor pressure than sodium, it is attractive to contemplate increasing the temperature of the working fluid to take advantage of high-efficiency heat-transfer cycles; either steam-Rankine or helium-Brayton cycles have been suggested.
Lead or lead-alloy reactors also show better behavior during severe accidents than do sodium-cooled reactors. Simulations show that in an “unprotected loss-of-flow (LOF) accident” in which the coolant flow through the core rapidly drops off, the sodium in an SFR rapidly begins to boil causing an increase in reactivity that boils yet more sodium. This positive feedback cycle increases the core temperature until the fuel melts, and ends only when the core is roughly half-molten and enough fuel has been swept out to sufficiently reduce the reactivity coefficient. In contrast, the simulations show that the lead or lead-bismuth alloy in an LFR never gets within a thousand degrees of its boiling point, and no increase in reactivity leading to a positive-feedback cycle occurs.
This huge difference in LOF accident behavior was determined to be due to the much greater coolant pressure drop across the core of an SFR than in that of an LFR. However, in the case of loss of power (LOP) accidents, the traditional approach of externally cooling the reactor vessel to remove excess heat after a LOF accident would appear to be inadequate for both LFRs and SFRs; as within 5 days the simulated LFR approached the stress limits for its reactor vessel. Thus, an IRACS (In-Vessel Reactor Auxiliary Cooling System) may be needed. In the case of a fuel-pin failure, the LFR still has the safety advantage due to the higher probability of the released fuel being swept out. This is because of the much higher inertia of lead and the low pressurization during fuel-coolant interaction.
LFRs also appear to have an economic advantage over SFRs. Since it is unnecessary to construct an intermediate coolant-loop and several safety systems found in LWRs, LFRs can have significantly lower capital cost. Reduced maintenance because of lower mechanical complexity and a longer fuel-cycle also helps to bring down operating costs, and Russian publications claim that their lead-alloy-cooled SVBR-75/100 reactor-type will generate electricity at a cost even lower than that of gas-fired powerplants. It is possible, though, that some of the economic advantages of the LFR could be realized by SFRs if their intermediate cooling-loop were replaced by a supercritical CO2 Brayton cycle. In order for such a system to be viable, more research must be done to find a solution for possible clogging due to reaction products of the sodium with the CO2.
There are, however, several challenges that must be met if LFRs are to live up to their potential. Possibly the biggest problem endemic to LFRs is the fact that at their temperatures of operation, the reactor structural materials and the fuel cladding may corrode in the liquid lead or lead-bismuth alloy. There are several possible solutions to this problem, most of them suggested by the long Russian experience with LFRs. Firstly, steels with high concentrations of vanadium and silicon but low concentrations of nickel seem somewhat resistant to the corrosive effects of molten lead- alloys at temperatures of up to 873 K. This is no help to LFRs which use pure lead as coolant, as pure lead is just beginning to melt at that temperature. The use of oxidation inhibitors (zirconium, tungsten, chromium) dissolved in the coolant has also been shown to reduce the corrosion rate of steel by lead-alloys, but these increase the cost of the reactor as their concentrations must be very strictly controlled to avoid depletion. A protective coating of aluminum, molybdenum, zirconium, or carbide on the reactor’s internal surfaces can also help, but is very difficult to get them to deposit and adhere evenly in addition to increasing the cost of the reactor. Active and careful control of the presence of oxygen, hydrogen, and water concentrations in the reactor can also help prevent corrosion.
The choice between lead and lead-bismuth alloy is not necessarily easy. Lead-alloy has the advantage of a much lower melting temperature that lets it take advantage of some of the cheaper options for making the reactor components corrosion resistant, and it experiences much less volume-change with solidification than does pure lead. Also, because of its lower melting temperature, LFRs using lead-alloy are not as susceptible to blockages or clogs due to solidification of the coolant. On the other hand pure lead is ten times cheaper than lead-bismuth alloy, and the bismuth in lead-bismuth alloy can become activated, producing small amounts of radioactive polonium which can be carried into the energy-conversion system.
Despite the difficulties associated with their construction, it appears that LFRs may form a very attractive solution to the energy generation problems facing our civilization. Their relatively compact and simple construction means that they can be designed for modular installation in remote areas, and their lower capital/operation costs could make them very competitive in the energy market. These factors, combined with the ability of all fast neutron reactors to transmute wastes produced by other nuclear powerplants, may make them an increasingly important part of the worlds energy-production mix in the near future.
“Fast Neutron Reactors: WNA.” World Nuclear Association. Web. 28 Oct. 2009. <http://www.world-nuclear.org/info/inf98.html>.
Loewen, Eric P., and Akira T. Tokuhiro. “Status of Research and Development of the Lead-Alloy-Cooled Fast Reactor.” Journal of Nuclear Science and Technology 40.8 (2003): 614-27. Google Scholar. Web. 20 Oct. 2009. <http://wwwsoc.nii.ac.jp/aesj/publication/JNST2003/No.8/40_614-627.pdf>.
Tucek, Kamil, Hartmut Wider, and Johan Carlsson. “Comparison of Sodium and Lead-Cooled Fast Reactors Regarding Severe Safety and Economical Issues.” Proc. of 13th International Conference on Nuclear Engineering, Beijing, China. Environmentalists for Nuclear. Web. 25 Oct. 2009. <http://www.ecolo.org/documents/documents_in_english/SFRvsLFR-05.pdf>.