This debut post is a concise summary of the modern approaches to realistic, efficient nuclear power that heed traditional safety concerns and cost effectiveness, which I wish to promote as the clean, modular sources of baseload electricity for the near future.
The molten salt reactor is basically a chamber containing high temperature, unpressurised liquid phase fluoride salts, with a moderation mechanism such as control rods, and inputs and outputs to access the generated heat. In the design built and tested in the 1960s at Oakridge National Laboratory, a mixture of fluorides of lithium, beryllium and zirconium was used as the coolant, containing uranium fluoride as fuel. It was tested for a total of 6000 hours (250 days) without incident.
The striking advantages of this approach to nuclear power would have been realised in the following phase of research, had funding been continued. The MSR fissioned U-235 (and then U-233) to generate heat, but a further layer of subtly different fluoride coolant was intended to “blanket” the main coolant chamber such that it was exposed to the neutrons from the reaction. This blanket would contain thorium fluoride as the fertile fuel.
Thorium exists naturally as a single, ubiquitous radioactive isotope. It is responsible for much of the harmless background radiation in soil, sand and rocks which nobody spends a second’s thought on. Thorium-232 would “breed” uranium-233 after capture of a neutron, and it is this form of uranium which would fission to provide further neutrons to sustain the chain reaction. Used in this way, molten salt reactors would rely on an abundant primary fuel that is currently considered a worthless by-product of rare earth mining, and which, in principle, could be concentrated from soil or rock from nearly anywhere. Moreover, the homogenous nature of the liquid fluoride fuel ensures essentially total conversion: every watt of thermal energy would be produced.
Other actinides, in suitable molten salt form, could also be used to fuel the MSR, hence this technology represents an avenue for permanent disposal of waste and weapons-grade material. In addition, intrinsic passive safety features promise “walk away safety”. For a start, the reactor fuel is already in a high temperature (>650C), molten state, so the concept of “nuclear meltdown” is entirely circumvented. The density of this liquid in the original experiment was observed to oscillate so as to accelerate and decelerate the chain reaction and “load follow” the energy demanded of the reactor. As for emergency shutdown, an outlet pipe at the base of the reactor is cooled by an electric fan which keeps a “plug” of salt frozen within. Failure of the system would cause this to melt and allow the molten salt to drain harmlessly into basement storage tanks. Finally, the near-atmospheric pressure of the reactor means no large, thick concrete containment is necessary.
The reaction heat is exchanged into a separate salt or steam loop to drive a turbine for electricity, but the high temperature is also ideal for chemical and industrial process heat, such as water desalination. Although this reactor concept is being promoted in the U.S. as LFTR (Liquid Fluoride Thorium Reactor), a major Chinese research centre has dedicated a group of about 300 workers to establishing the MSR technology based on the Oakridge results.
The integral fast reactor is envisaged as self-contained reactor, generator and fuel recycling plant. It is specifically based on liquid metal-cooled fast breeder reactor technology, as opposed to traditional water-cooled thermal reactors.The enriched uranium and other fuel derived from spent nuclear or weapons-grade material is fabricated, as oxides, into solid fuel along with sodium metal. At operating temperature the liquid sodium, as well as circulating and transferring the reaction heat, fills the voids left by fissioning material and acts to maintain steady neutron density.
These neutrons interact in the fast spectrum, with much the same energy as they had when they were released. This results in breeding of further Pu-239 from fertile U-238, and thus the eventual consumption of virtually all the nuclear fuel (in comparison, a traditional light water reactor will use less than 1% of the solid fuel material). The resulting spent fuel is recycled in a pyro-processing facility, powered by the reactor, where remaining useful isotopes are extracted and incorporated into new fuel, and actual waste is treated for long term storage.
The most promising IFR is known as PRISM (Power Reactor Innovative Small Module), the result of extensive testing of the liquid sodium-cooled reactor concept in the form of the successful Experimental Breeder Reactor II, which ran from 1965 to 1995. Actual scenarios were demonstrated where coolant flow was shut off at full power, resulting in natural expansion of the reactor liquid and shutdown due to low enough neutron density. Other passive safety features are provided by refuelling and generating mechanisms integrated into the reactor itself under the sodium coolant, which is securely sealed from interaction with oxygen or water. This also implies a modular design philosophy which will enable assembly line production reminiscent of passenger aircraft construction. PRISM is ready for assessment by various countries’ regulatory authorities; China is known to be constructing a similar reactor.
Small Modular Reactors are a broader class of modern reactors which generally offer an output under 3-500 MW, integrated coolant circulation, safety systems and power generation, and rapid assembly line production. Some approaches boast the need for infrequent refuelling. Strictly speaking Generation-III+ technology, among the many designs, the Westinghouse SMR is somewhat like a miniature model of the state-of-the-art AP1000 power plant, many of which are currently being erected in China. The increased economies of scale and standardisation of components which do not require prohibitively large production facilities promise an increased power output per unit area of footprint and per total fabrication costs.
Update December 14th: The SMR design to be prioritised through U.S. federal funding is the NuScale design, an incredibly portable reactor concept with a nominal electrical output of 45 MW. Read about it here.
Here is an animation of the construction of the first Babcock & Willcox mPower plant. The mPower will be rated at 180 MWe per module.