The Pressurized Water Reactor and the LFTR: Some Comparisons

In 1948 exploration of reactor technology was well underway. Most reactors had cores made of solid materials, for example uranium metal clad in aluminum. A second line of reactor development, the which began with the original chain reactor experiment at Cavendish Laboratory and continued with a reactor experiment at Los Alamos, involved the use of uranium compounds dissolved or suspended in water. The reactor was called the Aqueous Homogeneous Reactor. In 1948 reactors were cooled by air, some other gas, or water. Research was underway involving the use of molten sodium metal as a reactor coolant. Alvin Weinberg had proposed the use of water under pressure as a reactor coolant. This concept had the potential to control the heat produced by the reactor and put it to useful work powering ships, or driving electrical turbines. This technology attracted the attention of the United States Navy, and eventually led to the development of the nuclear-powered submarine. Naval reactor technology also had potential for electrical production, and the Navy set up the first project to demonstrate civilian electrical production at Shippingport, Pennsylvania.

Meanwhile the Air Force, which was also interested in its own reactor technology to power bombers, sponsored aircraft reactor research in Oak Ridge. The original aircraft reactor concept explored by engineers at the K-25 facility in Oak Ridge involved the use of liquid sodium as a coolant. The original K-25 aircraft reactor concept had a very significant safety defect, and in 1947 three K-25 engineers, V.P. Calkins, Kermit Anderson, and Ed Bettis began to explore a radical reactor concept involving the use of hot liquid fluoride salts. This was a natural concept, because in 1947 K-25 was the largest industrial facility using fluoride chemistry. The three engineers researched the possibility of using liquid fluoride salts as a reactor moderator, fuel carrier and reactor coolant. The K-25 research lead to the Molten Salt Reactor concept. When the aircraft reactor project was transferred to ORNL in 1950 and assigned to the brilliant chemist Ray C. Briant, Ed Bettis pitched the Molten Salt Reactor to Briant. Briant and Bettis pitched it to Weinberg, and it was agreed that the defective K-25 sodium-cooled aircraft reactor concept should be scrapped, and the promising liquid-salt reactor concept become the focus of ORNL Aircraft Nuclear Propulsion research.

During the next few years a radically different reactor concept was to emerge in Oak Ridge. Conventional reactors are much-evolved versions of Alvin Weinberg's water-cooled reactor. They feature complex cores which contain a ceramic uranium dioxide clad with zirconium metal. This fuel system prevents the escape of radioactive fission products into the cooling water, but creates considerable difficulties for processing the fuel for fuel recycling and the extraction of fission products. The UO2 fuel is also a very poor heat conductor, and the fuel pellets inside conventional reactors become very hot, so much so, that there is a danger if the reactor cooling system fails that the UO2 fuel could melt at 2800°C and create an unholy mess.

The water-cooled reactor is just that, water cooled. A system of pipes carry the water through the core where it extracts heat from the fuel pellets. Water boils at 212 degrees under atmospheric pressure, but scientists had long ago discovered that if water can be kept under high pressure, its boiling point goes up. Engineers had discovered that power conversion becomes more efficient if water is pressurized and prevented from boiling at 212 degrees F. In order to prevent the water from boiling inside the sort of pressurized reactor that the Navy uses, the reactor is placed inside a massive steel pressure vessel, and water is pressurized in the reactor. The pressurized water is superheated, and because it is under pressure it does not turn to steam inside the reactor. A second type of reactor, the boiling water reactor, operates under a little lower pressure, and pressurized water begins to turn into steam in the upper part of the reactor.

Reactors that are cooled with pressurized water are quite complex and can be quite large and pose a number of problems. The presence of pressurized water leads to the danger of a steam explosion. Pressurized water can also leak from pipes outside the reactor, creating a danger that the reactor might not receive coolant water. Coolant failure can lead to core meltdown as it did at Three Mile Island. Core meltdown can lead to containment breach, either by melting through the steel pressure vessel, or by releasing hydrogen gas which in turn can explode with enough force to rupture the pressure vessel.

Pressurized water reactors and their cousins boiling water reactors can be made safe, but at a significant price in terms of complexity and weight. Light water reactors have control issues. Chain reactions may not be uniform throughout the reactor. Operators may need to employ control rods to prevent excess reactivity in parts of the reactor which can lead to local overheating and core damage. This necessitates an elaborate system of internal sensors inside the reactor along with an equally elaborate instrumentation, designed to provide operators detailed information about core conditions. During the 1970's reactor operators could be swamped with information leading to confusion and operator errors. This happened at during the Three Mile Island accident. Computer systems are now in place to manage the flow of information from inside the reactor, and to assist human operators in managing pressurized water reactors. Recent designs of pressurized water reactors have impressive safety features and can be described as demonstrating revolutionary improvements in safety over earlier generations of water-cooled reactors. They are also very expensive, and still use enriched uranium dioxide fuel that is expensive and difficult to reprocess. Pressurized water reactor technology is stuck with once-through fuel technology and the problem of nuclear waste.

When Ray C. Briant and Ed Bettis approached Alvin Weinberg in 1950 to discuss the Molten Salt Reactor concept, Weinberg was already aware of the shortcomings of his invention, the pressurized water reactor. Weinberg's mentor Eugene Wigner believed that the Aqueous Homogeneous Reactor was a better route to low-cost electrical energy than the Pressurized water reactor, and Weinberg was pushing Aqueous Homogeneous Reactor research at Oak Ridge. Ed Bettis' Molten Salt Reactor had many of the attractive features of the homogeneous reactor without some of its drawbacks, but it was to take Weinberg some time before he realized that the MSR represented the preferred route to the pressurized water reactor alternative.

Both the Aqueous Homogeneous Reactor and the Molten Salt Reactor featured a liquid fuel-coolant mixture. The mixture was pumped into and out of the core where moderation and geometry enabled criticality. Eugene Wigner had been attracted to the Aqueous Homogeneous Reactor because its fuel could be continuously run through chemical processors outside the core. This meant that neutron-eating fission products could be removed, making the neutron economy of the Aqueous Homogeneous Reactor so efficient that it could breed Thorium to U-233 advantageously. ORNL reactor designers were to design an Aqueous Homogeneous Reactor with a thorium-containing blanket surrounding core containing a heavy water with a dissolved uranium compound. Before his death Ray C. Briant suggested to Weinberg that a Molten Salt Reactor with a thorium blanket, similar to that designed for the Aqueous Homogeneous Reactor would have superior performance to the latter reactor. Thus Briant can be considered the father of the Liquid Fluoride Thorium Reactor, but in many respects the LFTR had many fathers at ORNL.

Compared to the Light Water Reactor the MSR/LFTR had many safety features, the most outstanding of which was its strongly negative temperature coefficient of reactivity. The liquid salt fuel mixture of the LFTR responds to slow and then stop chain reactions as heat within the reactor increases.

The liquid salt in the LFTR core expands as it heats. As it expands there is less liquid salt in the core, carrying with it fissionable fuel. As fissionable fuel leaves the core, the fission reaction rate slows. At maximum core heat, enough fissionable fuel leaves the core to bring the fissionable mass left in the core down below the amount needed to maintain criticality, The chain reaction stops. Core salts retain heat, and heat is also replenished by the radioactive decay of fission products within the core.

What first attracted Ed Bettis and his associates to the Molten Salt Reactor idea was the way it would respond to a pilot's throttle use.

When the pilot demanded more power for his jet engines, heat is drawn out of the reactor core and transferred into the jet engine where it produces jet power. Heat from the LFTR core can also power powers closed-cycle gas turbines in electrical generating systems. As core temperature decreases, core salts shrink, and more salt is in the core, thus increasing the fission reaction rate. The greater the demand for power for a jet engine or a generator the greater the amount of heat generated by the core, and as a consequence the reaction rate within the core increases. The limitation of power output is determined by the heat removal rate, which in turn is based on the limitations of the turbine generating system.

The reaction rates slow down and then stop as heat withdrawal is decreased, or as temperature increases in Molten Salt Reactors--they basically control themselves. Thus while Pressurized Water Reactors require constant operator monitoring and operator input into its control system, MSRs including the LFTR, basically control themselves. The potential instability of the PWR is simply not present in the LFTR.

Compared to PWR, the LFTR has superior peak load reserve and load-following capacities. Since a LFTR's salts are at maximum heat when a LFTR is on standby, the LFTR can produce maximum power as quickly as its turbines can go to full generating speed under load. Thus the LFTR can not only load follow but can serve as peak demand reserve.

In the case of decreased load demand, less heat is drawn from the core, and the fission reaction rate slows. Thus the same feature that gives the LFTR superior safety over the Pressurized Water Reactor also gives it superior flexibility in generating electricity.