Denatured MSR Design Efforts

In May 1974, India detonated a nuclear weapon that had been made from plutonium it had extracted from a heavy-water research reactor. Coming several years after the ratification of the Nuclear Non-Proliferation Treaty (which India had not signed) it caused a great deal of political discussion about the role of breeder reactors, plutonium, and chemical reprocessing. During the 1976 election Democratic candidate Jimmy Carter pounded incumbent President Gerald Ford on the topic of nuclear proliferation and chemical processing, leading Ford to announce five days before the election that the United States would suspend plans to chemically process spent nuclear fuel. The presidential election was close, but Jimmy Carter won and continued the newly-announced policy.

In his announcement on October 28, 1976, Ford specifically mentioned that “nuclear power” produced plutonium that could be used to fuel other nuclear reactors or for nuclear weapons. That statement remains highly debatable to this day. Nevertheless, the thorium-fueled breeder reactors that were being designed by ORNL did not produce any appreciable amount of plutonium during their anticipated operations. One might have wondered why this did not lead to their elevation in priority and status in the planning of the ERDA (the successor organization to the AEC). But there is no indication that this happened.

Early in his adminstration, Carter announced that solar energy would be his prime focus and that he would cut back drastically on “the concentration involving the breeder reactor and a plutonium society.” He reiterated his concerns about proliferation and asserted that the US had vast inventories of coal that could be burnt for power. He also announced that there would be further investigations into reactors that did not involve “direct access to materials that can be used for nuclear development.”

A few weeks later, Carter announced to a joint session of Congress that the central effort in ERDA’s fast-breeder reactor program, the Clinch River Breeder Reactor, would be deferred indefinitely. In a fact sheet accompanying his speech, he announced that the US “breeder” program would be redirected toward evaluation of alternate breeders, fuels, and advanced converter reactors with emphasis on nonproliferation and safety concerns.

After the second cancellation of the Molten-Salt Reactor Program in 1976, there was a brief effort during the Carter Administration to evaluate reactors that had enhanced proliferation resistance. This effort was undertaken by a small subset of researchers from the original MSRP, including Dick Engel, Harold Bauman, Warren Grimes, and Herb McCoy. Their results were published in two ORNL reports in August 1978 (ORNL-TM-6413, “Molten-Salt Reactors for Efficient Nuclear Fuel Utilization without Plutonium Separation.”) and March 1979 (ORNL-TM-6415, “Development Status and Potential Program for Development of Proliferation-Resistant Molten-Salt Reactors.”).

Their point-of-departure was the ORNL reference design MSBR, using a single fluid and a sophisticated chemical processing system described previously. Among the modifications considered were elimination of the breeding gain, a reduction in the power density (and specific power) so that protactinium isolation could be avoided, and.several conceptual variations in the fuel processing cycle. Denaturing the reactor with depleted uranium was not initially considered because that would lead to substantial plutonium production and would forego one of the basic advantages of the use of thorium fuel. But their conclusion was that the reference MSBR did not meet the new, more-rigorous standards that Carter had announced.

Chemical Processing Concepts

Denatured MSR chemical processing flowsheet (from ORNL-TM-6415, pg 105).

Figure 1: Denatured MSR chemical processing flowsheet (from ORNL-TM-6415, pg 105).

The chemical processing system for the denatured MSR also used the reference MSBR as the point-of-departure, but it became even more complex, as shown in Figure 1. Denaturing the uranium was the root cause of this additional complexity:

  1. In the reference MSBR, the uranium content was predominantly fissile (233U) whereas in the denatured MSR the uranium content was predominantly fertile (238U). The fissile content had much higher levels of 235U than the reference MSBR, and all of the uranium had to be preserved in the chemical processing steps. This meant that much more uranium would need to be fluorinated and carefully returned to the fuel salt at the end of processing than in the MSBR case.
  2. The presence of 238U led to the generation of 239Pu and other isotopes of plutonium. The thermal-neutron spectrum of the denatured MSR also meant that a significant fraction of this plutonium would not fission on the first pass, leading to the formation of transuranic nuclides like americium and curium. The plutonium was a valuable fissile resource and also needed to be carefully preserved in the chemical processing steps so that it could be returned to the fuel salt. But plutonium was not chemically separable by fluorination and would come out of the salt in the same bismuth extraction step that removed protactinium from the salt.
  3. The goal of a breeding gain in the reference MSBR (~7% surplus fuel generated per annum) was replaced by a less-ambitious goal of “break-even” breeding, where the reactor only supplied enough fissile to compensate for its consumption. This less-ambitious goal was still quite challenging for the denatured MSR because 239Pu was an inferior fissile fuel in the thermal spectrum compared to 233U, generating fewer neutrons per absorption of a thermal neutron.
  4. The core power density of the reactor had been decreased considerably, leading a fissile inventory approximately three times greater than the reference MSBR (for the same electrical power generation). This further increased the amount of fluid that needed to be processed.
  5. The long-term buildup of transplutonic (americium, curium, berkelium, californium) nuclides represented a substantial neutronic loss to the reactor, with approximately two neutrons consumed on transplutonic generation for every neutron released in the fission of a transplutonic nuclide. Substantial uncertainty existed as to whether these transplutonic nuclides should be disposed of at the end of reactor operation or recycled to the next generation of reactor core, where they would continue to exert a neutronic “drag” on the performance of the reactor.

From a chemical processing and neutronic performance standpoint, denaturing the uranium in the molten-salt reactor had no benefits and many disadvantages. Only the political mandate to consider denatured operation pushed research in this direction.

Changes to the processing flowsheet were necessary because of the presence of large quantities of plutonium in the fuel salt. All uranium, protactinium, and plutonium would need to be chemically extracted from the salt before rare-earth fission products could be removed using reductive extraction techniques similar to the reference MSBR. One of the challenges of Pa-Pu removal was that fission-product zirconium would also be removed and later reintroduced to the fuel salt. A careful partial oxidation of the extractive bismuth was proposed to address this concern, followed by hydrofluorination of bismuth containing zirconium and uranium into a waste salt. Improved techniques for zirconium removal were considered highly desirable.

To accommodate the higher inventories in the fuel of uranium, plutonium, and zirconium, more highly-depleted metallic lithium would be used as a reductant, which increased processing costs and also increased the rate at which fuel salt solvent would need to be discarded. An increase in the solvent discard rate also reduced the effective consumption of thorium as an energy resource.

It was also proposed to carry a much higher fraction (~10%) of the uranium in the salt as a trifluoride (UF3) rather than as a tetrafluoride (UF4). It was anticipated that this would lead to immediate reduction of fission product selenium and tellurium and their complete retention in the fuel.

Processing techniques for gaseous fission products (xenon and krypton), noble-metals (molybdenum, technetium, etc.) and tritium were relatively unchanged from the reference concept.

Later Concepts

The final report on the denatured MSR concept was issued in July 1980 (ORNL-TM-7207). In this report, chemical processing techniques that had been proposed in earlier reports (fluorination, reductive extraction) had been eliminated, beyond basic techniques like hydrofluorination to control oxygen contamination. With the abandonment of overt chemical processing came a decrease in the conversion ratio of the reactor, to the point that a break-even conversion ratio was no longer achievable. The DMSR concept now became yet another example of a reactor that burned 235U as its primary fissile fuel, albeit it in a rather exotic way when compared with conventional solid-fueled reactors.

This was summed up as follows in ORNL-TM-7207 as an inevitable consequence of low-enrichment uranium fuel:

The presence of 238U in a DMSR, combined with the effects of flux flattening, sufficiently reduces the nuclear performance so that a net breeding ratio substantially greater than 1.0 probably could not be achieved, even with full-scale fission-product processing.

With the abandonment of breeding, the motivation to continue to consider the use of thorium diminished as well. Since the 1980 report was the last ORNL paper to describe molten-salt reactor designs and development, it is reasonable to assume that this represented the very last effort at ORNL in this field. Perhaps if further funding had been available, molten-salt reactors that used uranium fuel exclusively might have been the next stage of development. Such burner reactors would have retained the physical advantages of molten-salt reactors (high temperature operation at low pressure) but would have lost the fuel cycle advantages of thorium (highly efficient fuel use and minimal transuranic production). However, if such reactors were 235U burners in either incarnation, it would have been desirable to avoid the additional complications that thorium introduced because of its chemical similarity to the lanthanide fission-products.

Chemical Processing Experimental Research

Since the effort to investigate denatured molten-salt reactor concepts took place after the Molten-Salt Reactor Program had been cancelled, there were no chemical processing research activities that took place to accompany the conceptual effort.