Prologue: The Chairman Returns
On January 20, 1961, John F. Kennedy took the oath of office on the east portico of the Capitol. Robert Frost read a poem. The day was bright and very cold. Among the crowd on the platform, though not prominent in the television coverage, was Glenn Seaborg.
Seaborg had been appointed AEC chairman the previous week. He was forty-eight years old, had won the Nobel Prize in chemistry nine years earlier, and had spent most of the 1950s at Berkeley as associate director of the Lawrence Radiation Laboratory and as chancellor of the university. He had last served in Washington twenty years earlier, as a young chemist at the Chicago Met Lab leading Section C-I, the plutonium chemistry group. It was in that role, in the spring of 1941, that he had presided over the discovery of uranium-233 as a practical source of nuclear energy — the recognition he had recorded in his laboratory journal on April 23 of that year with the sentence about its being “sufficiently long-lived to be a practical source of nuclear energy.”
He had been a graduate student when he made that observation. He was now returning to Washington as the most powerful figure in American nuclear policy, charged with running an agency whose civilian reactor program had, in the twenty intervening years, almost entirely set aside the fuel cycle his own early work had opened.
Seaborg’s appointment was widely welcomed. He was respected in the scientific community, acceptable to the Joint Committee on Atomic Energy, and sufficiently distinguished that his confirmation was a formality. He was also, though few observers noted this at the time, uniquely positioned to understand what had and had not happened to the thorium cycle over the preceding two decades. He had been present at every major decision point. He had written the December 1944 memorandum on uranium-232 contamination. He had sat through the Met Lab Area VII seminars where Wigner taught the eta values that still defined the physics of breeding. He had known Weinberg since 1942.
Whether any of this would translate into action was another question. The AEC chairmanship was a political office, not a technical one. The agency Seaborg inherited was large, institutionally mature, committed to specific reactor programs that reflected fifteen years of accumulated decisions. He could not, as chairman, redirect it merely by preferring different technology. He would have to work within the structure that existed — the Joint Committee, the reactor divisions, the weapons program, the industry relationships that Shippingport and its successors had created.
What happened over the next ten years, during Seaborg’s chairmanship under Kennedy and Johnson and Nixon, is the story of a man who understood better than most what the thorium path had offered, and who nonetheless could not substantially redirect American nuclear policy away from the path that had been chosen.
It was not, in the end, for lack of trying.
The 1962 Report
In March 1962, Kennedy wrote to Seaborg asking for a comprehensive review of the AEC’s civilian nuclear program. The letter was a routine kind of presidential request, the kind that asks an agency to tell the White House what the agency is doing and what it thinks it should do. But Kennedy framed the question with a specific emphasis: he wanted to know whether the current program adequately reflected the long-term energy needs of the nation, and whether the balance between short-term commercial deployment and long-term technology development was right.
Seaborg convened a study group and spent much of 1962 drafting a response. The document that emerged, delivered to Kennedy in November 1962 and published soon afterward as *Civilian Nuclear Power: A Report to the President — 1962*, was a remarkable piece of official writing. It was sober where earlier AEC publications had been promotional. It acknowledged, in the first substantive paragraph of its cover letter, that “our attention had probably for too long remained focused narrowly on short-term objectives.” It argued that the long-term viability of nuclear power depended on the development of breeder reactors capable of using the full energy content of uranium and thorium reserves, rather than the less-than-one-percent utilization of current light-water technology.
It explicitly mentioned the thorium-U233 cycle.
In its technical survey, the report noted that the “spectral shift” reactor and the high-temperature gas-cooled reactor — both being actively developed — had “excellent conversion ratios” and “may eventually be made to breed in the thorium-uranium cycle.” The passage was brief, placed in a list of advanced converter concepts, and described the possibility of thermal breeding on thorium as something that could be achieved “eventually” rather than as an immediate priority. But it was there, in an official document sent to the President of the United States over the AEC chairman’s signature. The cycle Seaborg had identified in 1941 was still being proposed, twenty-one years later, as a long-term possibility.
The report’s principal recommendation, however, was for increased emphasis on the fast plutonium breeder. The language described breeders generally, using the phrase “breeder reactors, which can make use of nearly all of our uranium and thorium reserves.” In the operational sections of the report, though, the near-term resources were directed at the liquid-metal-cooled fast breeder reactor — the LMFBR — descended from Walker Cisler’s Fermi-1 and EBR-I before it. The thorium thermal breeder was mentioned; the plutonium fast breeder was funded.
Why the asymmetry? Partly because the LMFBR program was further developed; Fermi-1 was under construction, the Argonne experimental work had been running for years, and the international community (France, the United Kingdom, and the Soviet Union) was pursuing similar designs. Partly because the fast breeder, producing plutonium, could in principle be integrated with the weapons complex in a way that the thorium cycle could not. And partly because Milton Shaw, whom Seaborg had elevated to director of the AEC’s Division of Reactor Development and Technology in 1964, had no interest in alternatives to the LMFBR.
Shaw, as noted in the previous chapter’s discussion of the MSBR’s demise, had come to the AEC from Rickover’s naval reactors program. He had absorbed the Rickover method — centralized technical control, relentless focus on a single technology, no tolerance for competing concepts — and was applying it to the civilian breeder program. For Shaw, the LMFBR was the one reactor that mattered. Everything else was a distraction, to be minimized or eliminated as resources permitted.
The 1962 Report went to Kennedy. Kennedy approved its recommendations. The civilian nuclear program moved, gradually but decisively, toward increased emphasis on the LMFBR. The thorium thermal breeder remained on paper. Oak Ridge continued its work on the Molten Salt Reactor Experiment, which would reach criticality in 1965, but the program was no longer operating in an environment of official equal consideration with the LMFBR. It was operating in the shadow of a program that was consuming a steadily larger share of the AEC’s reactor development budget.
The Test Ban and the Aircraft
Kennedy’s nuclear priorities during his abbreviated presidency were not primarily about civilian power. They were about weapons. The atmospheric test ban negotiations, which would produce the Limited Test Ban Treaty in August 1963, consumed enormous White House and AEC attention through 1961 and 1962. Seaborg devoted most of the first volume of his memoir — *Kennedy, Khrushchev, and the Test Ban* — to this work. He was deeply involved in the technical dimensions of the negotiation, particularly the question of whether underground tests could be reliably detected by seismic means.
Two of Kennedy’s early decisions, however, had direct consequences for the civilian reactor landscape.
The first was cancellation of the Aircraft Nuclear Propulsion program. The project had been running since the late 1940s, had consumed approximately one billion dollars, and had produced, among other things, the Aircraft Reactor Experiment at Oak Ridge — the world’s first operating molten salt reactor, successfully tested in November 1954. By the time Kennedy took office, intercontinental ballistic missiles had rendered the nuclear bomber strategically unnecessary. Robert McNamara, who had come to the Defense Department from Ford Motor Company with a reputation for eliminating programs whose costs exceeded their benefits, recommended cancellation. Kennedy agreed. The program was terminated in 1961.
The cancellation was, from a weapons-strategic perspective, correct. From the perspective of molten-salt reactor technology, it was significant in a different way. The ANP program had funded the research and development that had made the Aircraft Reactor Experiment possible — the metallurgy of Hastelloy N, the chemistry of molten fluoride salts, the engineering of high-temperature pumps and heat exchangers. When the program ended, the technology it had built did not disappear. It transferred, along with the Oak Ridge engineers who had developed it, to a new project that Weinberg had been proposing for several years: a civilian power reactor using the same fluid-fuel approach.
That project became the Molten Salt Reactor Experiment. It was approved in 1961, construction began in 1962, and it reached criticality on June 1, 1965, running on uranium-235 fluoride fuel.
The second of Kennedy’s early decisions was the continuation and expansion of the Navy’s nuclear propulsion program. Rickover, who had personally known Kennedy’s father and had been a minor presence in Kennedy’s own naval career during World War II, enjoyed uncomplicated support from the new administration. The aircraft carrier *Enterprise*, laid down in 1958, was commissioned on November 25, 1961 — the first nuclear-powered carrier in any navy. The submarine fleet continued to grow. Polaris missile submarines, ballistic missile submarines, attack submarines — all of them used direct descendants of the Shippingport pressurized-water reactor.
The civilian utility industry, meanwhile, was beginning the period that would later be called the “great bandwagon market.” Between 1965 and 1970, American utilities ordered roughly 150 new nuclear power plants. Every one of them was either a Westinghouse pressurized-water reactor or a General Electric boiling-water reactor. Every one of them was solid-fueled, water-cooled, a lineal descendant of the naval reactors program Rickover had built at Bettis in the late 1940s. None of them could breed. None of them used thorium. None of them captured more than a fraction of a percent of the energy content of the uranium that had been mined to fuel them.
The path had been chosen, and the path was now being built, at scale, across the country.
The Molten-Salt Reactor Experiment
The Molten Salt Reactor Experiment achieved initial criticality on June 1, 1965, in Building 7503 at Oak Ridge National Laboratory — the same concrete-and-steel structure in which the Aircraft Reactor Experiment had operated for eleven days in November 1954. The cylindrical reactor cell, originally built for an Aircraft Reactor Test that was never constructed, had been adapted for the MSRE. Inside the cell, the reactor vessel was a cylinder fifty-four inches in diameter and sixty-four inches tall, made of Hastelloy N — the nickel-based alloy that Oak Ridge metallurgists had developed during the aircraft program specifically to resist molten fluoride corrosion. The core consisted of unclad graphite bars moderating the fission reaction. The fuel salt — a mixture of lithium-7 fluoride, beryllium fluoride, zirconium fluoride, and uranium-235 fluoride at 33 percent enrichment — circulated upward through channels in the graphite at 1,175 to 1,225 degrees Fahrenheit, carrying fission heat out of the core to a heat exchanger where it passed the energy to a secondary coolant salt, which in turn dumped the heat to an air-cooled radiator on the roof.
The reactor’s thermal power was limited to eight megawatts. This was not by necessity but by accounting: the Atomic Energy Commission’s rules at the time permitted the laboratory to build a ten-megawatt reactor as an “experiment” using operating funds, without the capital appropriation process that would have been required for a larger machine. MacPherson, who had proposed the MSRE in the fall of 1959 on a budget of $4.18 million, had designed the reactor precisely at the boundary of what the AEC’s accounting rules would allow him to build with minimal bureaucratic interference. When the engineering design was complete, the cost estimate had doubled. By the time the reactor was built, the actual operating power was eight megawatts rather than ten — a minor miscalculation of heat-transfer characteristics.
The purpose of the MSRE was not to produce power and not to breed. The purpose was to demonstrate that a molten-salt reactor, built with materials and methods available in 1960, could operate safely and reliably for sustained periods at temperatures and power densities representative of a commercial molten-salt breeder. The MSRE fuel salt contained no thorium; the reactor did not have a blanket. What it had was a fuel salt chemically similar to what the anticipated breeder would use, a containment made of the same alloys, and a thermal power density low enough to be manageable but high enough to stress the materials realistically. Everything the Oak Ridge team had argued, on paper, since Briant and Weinberg’s 1957 paper in *Nuclear Science and Engineering* — that fluoride salts were stable at high temperature, that Hastelloy N resisted corrosion, that fission gases could be continuously removed, that the fuel chemistry could be managed while the reactor operated — had to be demonstrated in hardware.
Through the second half of 1965 and throughout 1966, the reactor was operated intermittently as the operating crew learned the machine. Sustained full-power operation began in December 1966 with a thirty-day continuous run. Through 1967, the reactor accumulated operating hours. A six-month run ending in March 1968 climaxed the initial demonstration: by that date, the reactor had accumulated approximately nine thousand equivalent full-power hours, and during the last fifteen months of U-235 operation the reactor had been critical eighty percent of the time. This was the operating reliability the Oak Ridge team had hoped to achieve. It was substantially better than the early operating records of most American commercial light-water reactors.
What the MSRE demonstrated in those first three years of operation was, at the level of engineering fact, what Weinberg and MacPherson had been arguing for two decades. Salt handling in an operating reactor was “quite practical,” as Haubenreich and Engel would later write. The salt chemistry was “well behaved.” There was “practically no corrosion.” The nuclear characteristics of the reactor were “very close to predictions.” The system was “dynamically stable” — a consequence of the strongly negative temperature coefficient that the ARE had demonstrated in 1954 and that the MSRE had now confirmed at higher sustained powers. Containment of fission products was “excellent.” Maintenance of radioactive components was accomplished “without unreasonable delay and with very little radiation exposure.”
That last point mattered. The reactor’s piping carried intensely radioactive fluoride salt at 1,200 degrees Fahrenheit. When components failed — pumps, valves, instruments — they had to be removed and repaired. The engineering question underlying every argument for and against fluid-fuel reactors had been whether the maintenance problem could be solved at all. At the MSRE, it could be. The reactor was, by the standards of the industry as a whole, an unusually well-behaved piece of nuclear engineering. Failures occurred, were diagnosed, were repaired. The work continued.
Then the team did something no one had ever done before. They changed the reactor’s fuel.
The logic was straightforward. The MSRE fuel salt contained uranium-235, easy to obtain from the AEC’s gaseous diffusion plants. A commercial molten-salt breeder, however, would run on uranium-233 bred from thorium. At some point the program needed to demonstrate that a molten-salt reactor could operate on U-233. Uranium-233 was available, in small quantities, from irradiated thorium targets at Oak Ridge and a few other sites. Enough of it had been accumulated by 1968 to fuel the MSRE.
But replacing the fuel in a conventional solid-fuel reactor would have meant unloading thousands of fuel rods, physically removing them from the core, and reloading with new rods — a shutdown measured in months, with significant radiation exposure to workers handling the spent fuel. In the MSRE the fuel was a liquid. To change it, the engineers would drain the uranium-235 salt out of the reactor, treat the drained salt with fluorine gas in an on-site processing cell to strip the uranium as gaseous UF₆ (which could then be recovered and stored), add uranium-233 to the stripped carrier salt, pump the new fuel back into the reactor, and restart. The entire operation could be done in weeks rather than months, by a small crew, using equipment that had been built specifically for the MSRE program.
Throughout the summer of 1968, the fluoride volatility facility adjacent to the MSRE building was used to process the fuel salt. The fluorine treatment was successful. The 218 kilograms of uranium-235 that had served as the MSRE’s original fuel were decontaminated from gamma activity by a factor of four billion and stored for later use. Uranium-233 — highly radioactive because of its U-232 contamination — was added to the same carrier salt in quantities sufficient to achieve criticality. The mixture was pumped back into the reactor.
On October 2, 1968, the MSRE achieved criticality for the first time on uranium-233. It was the first reactor in the world to operate on U-233 as its primary fuel. The fuel cycle that Glenn Seaborg and John Gofman had discovered in Room 307 at Berkeley in April 1941, which Wigner and Weinberg had laid out as the basis of a fluid-fuel breeder in May 1945, which Compton had cancelled in November 1943 and which the AEC had treated as a second-line concept through the 1940s and 1950s, had just become, at the scale of engineering reality, the primary fuel of an operating nuclear reactor.
Six days later, on October 8, 1968, Glenn Seaborg flew down to Oak Ridge.
He had been chairman of the Atomic Energy Commission for seven and a half years by then, approaching the end of his tenure. He had spent those years managing weapons tests, negotiating treaties, overseeing the expansion of the civilian reactor industry, fighting for and sometimes losing budget battles within the administration. He had, for the most part, not been in reactor control rooms. On October 8, he wanted to be in one.
The MSRE control room was small by the standards of commercial power reactors — a few consoles, analog instrumentation, a handful of operators. Weinberg was there. So were Paul Haubenreich, who had directed MSRE operations since its first criticality, and Ray Stoughton, who with Seaborg himself had been a co-discoverer of U-233 at Berkeley in 1941. The reactor, which had been running at very low power for the six days since initial U-233 criticality, needed to be taken up to operating power. Seaborg, under the supervision of Haubenreich and the operators, manipulated the controls. The reactor climbed from the watts range up to 100 kilowatts — not full power, but enough for the day’s purpose. The needles on the instruments moved. The reactor was running on the fuel Seaborg had recognized, twenty-seven years earlier, as a potential basis for nuclear energy.
There was no press. There was a photographer. The photograph that survives shows Seaborg at a console, wearing a white shirt and dark tie, gray hair combed back, looking at the instruments with an expression of concentrated attention. Weinberg is visible behind him. The photograph was later used at Oak Ridge as an unofficial symbol of the MSR program — the AEC chairman personally, in physical presence, endorsing the technology that Oak Ridge had spent twenty-two years developing.
Seaborg, who typically spoke in careful bureaucratic cadences, described the moment in language that was, for him, unusually direct. He called it one of the most personally satisfying moments of his chairmanship.
What it was, in the structure of the larger story, was a closing of a circle and a beginning of an end. The circle closed because the man who had first recognized U-233 as a source of nuclear energy in 1941 was in the control room when U-233 first powered a reactor at useful scale in 1968. The end began because the MSRE, having now demonstrated that molten-salt reactors could operate reliably on either U-235 or U-233, had demonstrated everything it was designed to demonstrate. The next step — the step the Oak Ridge program had been planning toward since MacPherson’s early studies in 1956 — was to build a prototype of the commercial molten-salt breeder. That prototype would require funding an order of magnitude larger than the MSRE’s annual budget. It would require construction beginning in the early 1970s and commercial demonstration by the mid-1980s. The case for it was about to be made publicly, in the most comprehensive technical publication the Oak Ridge program would ever produce.
The MSRE itself continued to operate after October 1968. Over the following fourteen months, it accumulated over 2,500 equivalent full-power hours on U-233 fuel. It demonstrated the expected xenon stripping behavior, fission-product deposition characteristics, tritium behavior (the tritium produced by neutron capture on lithium-6 turned out to be a manageable challenge, not a fatal one), and plutonium fuel addition capability. The reactor was finally shut down on December 12, 1969, not because of technical problems but because the Oak Ridge team had run out of the experiments they had planned to do, and the funds were needed elsewhere in the molten-salt program.
The MSRE had operated for slightly more than four and a half years — longer than any other fluid-fuel reactor ever built, before or since.
The Reference Design
On February 2, 1970, the American Nuclear Society’s journal *Nuclear Applications & Technology* published a special issue devoted entirely to molten-salt reactors. Volume 8, Number 2. Eight technical papers by twenty-seven authors from Oak Ridge National Laboratory. A preface by Alvin Weinberg. Together they ran to approximately 180 pages of dense technical writing on reactor physics, fuel-cycle analysis, single-fluid breeder design, graphite behavior, xenon removal, Hastelloy-N metallurgy, fuel salt chemistry, and engineering development of the MSBR fuel recycle. They were the most comprehensive public case for the molten-salt reactor ever assembled. They were also, though almost no one outside Oak Ridge understood it at the time, the program’s engineering high-water mark.
The papers were written in the late summer of 1969, revised through October, and published three months later. Their authors were the working staff of the MSR program at the moment of its greatest technical confidence. M. W. Rosenthal, Paul Kasten, and Beecher Briggs provided the program history and status summary. Paul Haubenreich and Jack Engel — the men who had run the MSRE for four years — wrote the definitive operational record. E. S. Bettis, who had designed the ARE in the early 1950s and had stayed with the program, teamed with Roy Robertson on the reference design of a commercial 1,000-megawatt-electric single-fluid molten-salt breeder reactor. Alvin Perry and Hugh Bauman provided the reactor physics and fuel-cycle analyses that established the breeding performance. Dunlap Scott and W. P. Eatherly analyzed graphite dimensional change and xenon removal — the two technical issues that had driven the program’s 1967-1968 pivot from two-fluid to single-fluid breeder design. Whatley, McNeese, Carter, Ferris, and Nicholson from the Chemical Technology Division wrote up the engineering development of the reductive extraction fuel recycle process that made single-fluid breeding practical.
The papers together described a specific reactor. A cylindrical vessel twenty-two feet in diameter and twenty feet high, constructed of Hastelloy-N, containing a graphite moderator assembly with varying salt volume fractions in the core and blanket regions. The fuel salt would be a mixture of lithium-7 fluoride, beryllium fluoride, thorium fluoride, and uranium-233 fluoride in proportions of approximately 71.7 : 16 : 12 : 0.3 mole percent — nearly identical to the reference salt composition the 1959 Fluid Fuel Reactors Task Force had described a decade earlier. The salt would enter the core at 1,075 degrees Fahrenheit and exit at 1,300 degrees. It would circulate through four primary heat exchangers to a secondary coolant salt loop, which in turn would deliver heat to a conventional supercritical steam cycle producing 1,000 megawatts of electrical power at a thermal efficiency of 44 percent — substantially higher than the 32 percent efficiency typical of pressurized-water reactors.
The reactor would breed. The breeding ratio, at equilibrium, would be approximately 1.06. The specific fissile inventory would be 1.5 kilograms of uranium-233 per megawatt of electrical capacity — lower than any other breeder concept being proposed. The fuel doubling time, compounded, would be roughly twenty years. The total fuel-cycle cost would be approximately 0.7 mills per kilowatt-hour electric. The reactor could be started up on enriched uranium-235 or on plutonium (including plutonium recovered from spent light-water reactor fuel) with essentially the same long-term economics.
The most distinctive feature of the reference design, the feature that made the single-fluid concept possible at all, was the fuel-recycle chemistry laid out in the Whatley et al. paper. Continuous removal of protactinium-233 — the 27-day intermediate step between thorium-232 neutron capture and uranium-233 — was necessary for good breeding performance, because if left in the core Pa-233 would absorb neutrons and spoil the chain. The Chemical Technology Division had developed a reductive extraction process using liquid bismuth at 600 degrees Celsius, in which protactinium and uranium could be selectively transferred from the salt into a bismuth phase, held outside the neutron flux to decay, and then returned to the salt by electrolysis. The process was complex but engineered. Laboratory demonstrations had shown that the chemistry worked. Full-scale equipment would need to be built and tested, but no new physics needed to be discovered.
The other critical technical issue addressed in the special issue was graphite behavior. At the high fast-neutron fluxes the MSBR would generate, unclad graphite moderator bars would change dimensions over time — expanding initially and then contracting. At some point the dimensional change would exceed acceptable limits and the graphite would need to be replaced. Scott and Eatherly’s analysis concluded that, with existing commercially-available graphite, a four-year core life was achievable at an average power density of 22.2 kilowatts per liter. The reactor would be designed for graphite replacement on a four-year cycle — coincident with scheduled turbine maintenance, so the graphite replacement would not impose additional plant downtime. Xenon removal would be accomplished by sparging helium bubbles through the circulating salt and filtering the fission gases at a side stream. The penalty to breeding ratio from xenon was 0.5 percent — negligible, compared to the roughly 3 percent xenon penalty typical of solid-fuel reactors.
The construction cost of a commercial MSBR power station, the Bettis-Robertson paper estimated, would be comparable to the cost of a pressurized-water reactor of the same capacity. The power production cost, including fuel cycle and graphite replacement, would be 0.5 to 1 mill per kilowatt-hour less than a contemporary PWR — largely because of the low fuel-cycle cost and the higher thermal efficiency. The authors were careful to note that these were estimates based on a paper design, and that actual construction experience would be needed to verify them. MacPherson in his 1985 retrospective would observe that “in 1970 a careful detailed estimate was made of the capital cost of an MSR and it turned out to be within 1% of the cost of an LWR” — effectively quoting the same analysis.
What the special issue amounted to, considered as a whole, was a claim that the molten-salt breeder was ready to move from research into engineering development. The technology base was sufficient. The MSRE had demonstrated the reactor engineering at eight megawatts thermal. The chemistry had been developed to laboratory scale. The materials had been qualified for the required radiation exposures. The fuel-cycle economics had been worked out in detail. Every major technical question the program had faced in 1956 had been given, by 1970, either a solved answer or an engineered plan for solving it.
What the program needed next was to build a larger reactor — a demonstration MSBR, perhaps 150 to 300 megawatts electric, at an estimated cost of a few hundred million dollars over roughly a decade of construction and operation. That would be the Molten Salt Breeder Experiment, as it had begun to be called. Proposed formally in 1972, with a projected cost of $350 million over eleven years, it would be the technical next step that everyone at Oak Ridge assumed was coming. The NAT special issue was, in a real sense, the proposal for that next step — laid out publicly, in the peer-reviewed scientific literature, with full technical detail.
The issue was published on February 2, 1970. It reached subscribers in March. By then the decisive budget decision had already been made.
Three months earlier, on November 10, 1969, Budget Director Robert Mayo had eliminated the MSBR line item from the AEC’s fiscal year 1971 budget request. The five-way commissioner vote on which alternative to save — the MSBR or the LWBR — had been held at the AEC on the same day. Seaborg’s compromise appeal to Mayo, which kept the program alive at roughly five million dollars per year, had been worked out that afternoon. Weinberg’s appearance before the Commission pleading for greater support had been on November 25. By the time the *Nuclear Applications & Technology* special issue reached the desks of American nuclear engineers, the federal government had already decided not to fund the program the issue was describing.
The timing was not coincidental. The Oak Ridge team had known since early 1969 that the LMFBR program was consuming an ever-larger share of the AEC’s reactor budget. They had been working since 1967 — since the discovery of the bismuth-based reductive extraction chemistry and the graphite life constraints — on the single-fluid reference design that would be their programmatic proposal. The special issue was scheduled to coincide with the fiscal year 1971 budget cycle precisely so the technical case could be laid out for anyone in Washington who wanted to look at it seriously. What the team had not anticipated was that the decision had been made, and the compromise had been reached, before their case reached print.
The special issue went out anyway. It was cited, in the specialist technical literature, through the early 1970s. MacPherson would cite it in 1985 as the definitive published summary of MSR technology through 1969. A few graduate students in nuclear engineering would read it. A few AEC staff members who had been involved in the November 1969 negotiations would glance at it and file it. The nuclear engineering academic community, which had been orienting itself around light-water reactors for most of a decade, would largely ignore it. The utilities, which had been ordering LWRs at a rate of twenty per year, would not encounter it.
The reference design that had been laid out in those papers — a specific reactor, a specific fuel cycle, specific cost estimates, specific breeding performance — would remain the definitive molten-salt breeder proposal until MacPherson’s 1985 retrospective. It would remain, in essence, the definitive molten-salt breeder proposal for another forty years beyond that. The engineering case that had been assembled in 1969-1970 would not be substantially improved upon — not because later work was impossible, but because later work was not funded.
When a new generation of thorium advocates began consulting the ORNL archives in the 2000s, the documents they would find — the papers they would read, the design they would refer to, the technical foundation they would build on — would be the 1970 special issue of *Nuclear Applications & Technology*. Everything that came afterward, in the way of serious molten-salt engineering, would extend from what had been published in those six weeks of February and March 1970, in a journal that almost nobody outside Oak Ridge read.
The Forced Choice
Two things happened simultaneously in the late 1960s that determined the MSR program’s fate.
The first was the accelerating commitment of the federal government to the Liquid Metal Fast Breeder Reactor. In his September 1967 message to Congress, President Johnson had identified the LMFBR as the “highest priority new technology” in the AEC’s budget. Milton Shaw, who had been running the Division of Reactor Development and Technology since 1964, had built his operation around the LMFBR’s needs. The Clinch River Breeder Reactor project, intended as the first commercial-scale demonstration of the technology, was being planned with the expectation of completion in the early 1980s. By 1968, the LMFBR program was consuming roughly half of the AEC’s reactor development budget, and its share was growing.
The second was the systematic starvation of the alternatives. Shaw’s approach was not subtle. He did not, as a rule, openly attack competing reactor concepts in congressional testimony. He simply ensured that their budgets declined year by year, while the LMFBR’s budget grew. The gas-cooled fast breeder program, the light water breeder reactor program, the heavy water reactor program, the molten salt breeder program — all of these were funded at levels insufficient to make significant progress, while Clinch River and its precursors got everything they asked for.
The decisive institutional moment came in November 1969. Seaborg recounted it later in his memoir of the Nixon years.
On November 10, 1969, Seaborg received the Bureau of the Budget’s markup of the AEC’s fiscal year 1971 budget request. Budget Director Robert Mayo had eliminated all breeder programs other than the LMFBR entirely. The gas-cooled fast breeder, the light water breeder reactor, and the molten salt breeder were all reduced to zero funding. Seaborg called Mayo and protested. Mayo relented slightly: the AEC could keep one of the alternatives alive at a reduced level, but not both the LWBR and the MSBR. He left it to the Commission to choose.
The five commissioners met. Three of them — Johnson, Thompson, and Larson — voted for the MSBR. Two — Ramey and Seaborg himself — voted for the LWBR, which was closer to completion and represented, as Seaborg thought, the less costly budgetary choice. The outcome was made questionable by the fact that Commissioner Larson had a prior connection to Oak Ridge National Laboratory, where the MSR work was being done, and might have been thought to have a conflict of interest. Seaborg went back to Mayo that afternoon with a compromise: carry both, but carry the MSBR “at a low level for a number of years just to keep it alive.” He argued that otherwise Oak Ridge would lose its molten-salt team, along with what Seaborg called “a real chance for an eventual breakthrough on a very inexpensive source of nuclear power.” Mayo reluctantly agreed, conditioned on the AEC finding a compensating cut elsewhere in its budget. They found one.
Three weeks later, on November 25, 1969, Alvin Weinberg appeared before the Commission in Washington. He had come to plead for greater support for the molten salt program. He was unaware that the negotiations on his behalf had taken place two weeks earlier. Learning that the program was still alive, Seaborg later wrote, “seemed a relief” to Weinberg, “but the reduced level was an obvious disappointment.”
Weinberg wrote shortly afterward to Chet Holifield, the chairman of the Joint Committee on Atomic Energy:
> *”I believe the country’s almost single-minded commitment to fast breeder reactors, and its corresponding inability to support alternatives, particularly the molten salt breeder, is an error which I hope the Joint Committee can somehow correct in the not too distant future.”*
Holifield did nothing. The Joint Committee’s institutional commitment to the LMFBR was by then as firm as the administration’s. Holifield himself, a twelve-term congressman from Montebello, California, had been the LMFBR’s most consistent political advocate since the early 1960s. Weinberg’s appeal was, from the committee’s perspective, simply a complaint from a laboratory director whose program was losing its budget to a higher priority.
The MSBR funding continued at the reduced level — roughly five million dollars per year — through 1970 and 1971. It was enough to keep some engineers working. It was not enough to build the next experiment. It was not enough to develop the program toward a demonstration reactor.
Weinberg, for his part, did not confine himself to budget fights. He began speaking publicly about reactor safety — not the safety of molten-salt reactors, which he believed was inherently better than solid-fuel water reactors, but the safety of light-water reactors, which by 1970 were being ordered by American utilities at the rate of twenty per year. Weinberg’s concerns centered on loss-of-coolant accidents, on the adequacy of emergency core cooling systems, on the consequences of core meltdown in large reactors located near population centers. He was not alone in raising these concerns — the early 1970s saw the first serious public debates about reactor safety, with Sierra Club and Union of Concerned Scientists critiques beginning to draw press attention — but Weinberg was unusual in that he was making these arguments from inside the nuclear establishment, as director of its largest national laboratory.
This was not welcomed by Shaw, by Holifield, or by the industry. Weinberg was speaking publicly about problems that the nuclear establishment preferred to address privately. He was, from the establishment’s perspective, giving comfort to the opposition.
In January 1973, the Nixon administration’s reorganization brought a new Commission chairman, James Schlesinger, and a new Oak Ridge director was sought. Weinberg was removed. The official explanation was that Weinberg had served for nineteen years and it was time for new leadership. The unofficial explanation, understood by everyone involved, was that Weinberg had become inconvenient. He had criticized light-water reactor safety. He had advocated for a technology that the AEC had decided not to pursue. He had refused to fall in line.
He left Oak Ridge in April 1973. He would later serve as director of the Institute for Energy Analysis at Oak Ridge Associated Universities, where he continued writing and speaking about reactor safety and alternative reactor concepts until his retirement in 1985. He died in 2006, at the age of ninety-one.
In 1972 ORNL submitted the formal proposal for the next step: a demonstration Molten Salt Breeder Experiment, budgeted at $350 million over eleven years, that would produce a reactor generating modest amounts of electricity and develop the full suite of equipment needed for a commercial plant. MacPherson, writing in 1985 in *Nuclear Science and Engineering*, would observe with restrained bitterness that the 1972 proposal was rejected by men who “were already heavily committed to the LMFBR and guiding a very expensive development program that would be spending about $400 million each year by 1975. It was asking too much of human nature to expect them to believe that a much less expensive program could be effective in developing a competing system.” In January 1973 ORNL was directed to terminate MSR development work.
And then, for reasons MacPherson would say he never understood, the AEC reinstated the program a year later. In 1974 ORNL submitted a revised proposal, adjusted for inflation, with a total projected cost of roughly $720 million over eleven years. That proposal was also rejected. In 1976 — the year of the Ford nuclear-policy statement that would formally close the door on commercial reprocessing — ORNL was “ordered to shut down the MSR program,” in MacPherson’s phrasing, “for budgetary reasons.”
The evaluation document that had supported the 1972 cancellation, prepared internally by the AEC’s Division of Reactor Development and Technology, had identified three major technical concerns: the Hastelloy-N radiation hardening from helium accumulation, the tritium produced by lithium neutron capture, and the tellurium-induced cracking of Hastelloy-N piping. All three problems, as MacPherson would note in 1985, turned out to be solvable — the modified Hastelloy-N alloys with fine carbide precipitates addressed the radiation hardening, the intermediate-loop sodium fluoroborate coolant captured the tritium, and adjustment of the salt chemistry to maintain approximately 2 percent of the uranium as UF₃ rather than UF₄ eliminated the tellurium cracking. Each solution was developed at Oak Ridge in the years after the 1972 evaluation. None of the solutions was incorporated into any subsequent AEC or DOE assessment of the MSR. “I do not believe that such an evaluation was made,” MacPherson wrote in 1985, about whether the Energy Department had ever assessed whether the problems its predecessor had cited as reasons for cancellation had been solved, “or that the Department of Energy is prepared to make one.”
There had also been industry interest that the AEC had chosen not to credit. Two privately funded technical studies of the MSR were conducted at the turn of the 1970s. The first, in 1970, was by a group called the Molten-Salt Breeder Reactor Associates — the Black and Veatch engineering firm, together with five Midwest electric utilities. The second, in 1971, was by The Molten Salt Group, organized by Ebasco Services and including five large industrial firms and fifteen utilities. Both studies reported favorably on the technology. The Molten Salt Group concluded in 1971 that “the existing technology was sufficient to justify construction of a molten salt demonstration plant.” The AEC’s Division of Reactor Development, which had often suggested that evidence of industrial interest would strengthen the MSR’s case, did not engage with either study. MacPherson, writing in 1985 about why the program had ended, would name this pattern specifically: “It was often suggested by the DRD that evidence of industrial support for the MSR was needed to gain AEC enthusiasm. However, the evidence available for this support was ignored.”
The final programmatic document of the MSBR program was ORNL-5018, published in December 1974 — the year between the two cancellations. It described the development plan that would have carried molten-salt breeder technology through engineering demonstration and into commercial deployment over the 1975-1985 period. The budget request it contemplated — about $4 million per year to maintain program viability — was not granted.
MacPherson, looking back from 1985 with the perspective of nine years since termination, would identify two major factors that he believed had caused the program’s end. “The political and technical support for the program in the United States was too thin geographically,” he wrote. “Within the United States, only in Oak Ridge, Tennessee, was the technology really understood and appreciated.” And the second: “The MSR program was in competition with the fast breeder program, which got an early start and had copious government development funds being spent in many parts of the United States. When the MSR development program had progressed far enough to justify a greatly expanded program leading to commercial development, the AEC could not justify the diversion of substantial funds from the LMFBR to a competing program.”
By 1976, with the final termination, the Oak Ridge team that had built and operated the MSRE was disbanded. Some of its members stayed at the laboratory on other assignments. Others left. The files went into storage. The reactor itself, drained of fuel, was placed in caretaker status. The fuel salt, containing uranium-233 and various fission products, was removed to Oak Ridge storage facilities, where it would remain for decades. The documentation of the program — thousands of reports produced by Weinberg’s team over twenty years — was catalogued and archived.
Most of it would not be seriously consulted for another twenty-five years.
The Fast Breeder, in Parallel
While the MSR program was being starved and terminated, the LMFBR program that had consumed its budget was itself running into severe difficulties.
The Clinch River Breeder Reactor Project, authorized by Congress in 1970 and located in eastern Tennessee, had been planned as a four-year, $400 million construction project that would complete the first commercial-scale American fast breeder in the mid-1970s. By 1972, the cost estimate had grown to $700 million. By 1975, it was $1.7 billion. By 1983, when the project was finally cancelled, it had consumed approximately $1.6 billion in federal funds without producing a complete reactor, and total projected costs had exceeded $4 billion.
The Fermi-1 experience had been a predictor. Cisler’s reactor in Michigan, described in the first blog post of this narrative cycle, had been shut down for three and a half years after the 1966 accident, had consumed $143 million, and had produced heat sales revenue of $126,154 before being decommissioned in November 1972. Clinch River was Fermi-1 at ten times the scale, in every respect — ten times the cost overruns, ten times the schedule slippage, ten times the regulatory and political complexity. The LMFBR program, which Shaw and his successors had defended as the American path to energy independence, was in practice a money pit that consumed resources faster than it produced results.
Seaborg, in his later memoir, would describe this period with rueful candor. Writing in the early 1990s, after the LMFBR program had collapsed entirely, he framed the situation this way:
> *”Alvin Weinberg may well have been right. The AEC, with the Joint Committee’s active connivance, may well have erred in putting too many of its breeder eggs in the LMFBR basket. While correctly stating the case for alternative concepts in budget presentations, we gave them only token support compared to the massive emphasis on the LMFBR. When presidential support was sought, it was for the LMFBR only, and when the LMFBR was elevated to the status of a national goal with additional budgetary support, it all but assured that the alternatives would recede further into the shadows.”*
This was written about twenty years after the fact, by a man who had himself presided over the decisions he was describing. It is a remarkable piece of self-indictment — the kind of thing that political memoirs usually avoid in favor of self-justification. Seaborg was telling the truth about what had happened, including his own role in it.
What he did not quite say, but what was implicit in the passage, was that the commitment to the LMFBR had foreclosed not only the MSBR but the entire thorium-U233 path. The LWBR, which had won the November 1969 vote, would eventually achieve a modest demonstration of breeding on thorium at Shippingport in the late 1970s — an ironic coda to the story, with Rickover personally championing a thorium experiment in the one reactor he had built from the start. But the LWBR was not a scalable civilian technology. It was too expensive per kilowatt, required too much enriched uranium, and could not achieve the breeding ratios necessary for a sustainable fuel cycle. It was, in effect, a proof-of-principle experiment rather than a reactor design for commercial deployment.
The MSBR had been the scalable thorium technology. The MSBR had been cancelled.
Oil, India, and the Closing Door
The 1970s were a decade in which American nuclear policy was battered from multiple directions simultaneously.
The OPEC oil embargo of October 1973 imposed on the country its first serious energy crisis since World War II. Gasoline rationing, shuttered gas stations, a stock market crash, and a recession followed. In the immediate aftermath of the embargo, the Nixon administration proposed an ambitious expansion of domestic energy production, including nuclear. Project Independence, announced in November 1973, called for the elimination of American dependence on foreign oil by 1980 through a combination of coal, nuclear, and conservation measures. The civilian reactor industry responded optimistically. Utilities that had placed orders in the late 1960s accelerated their construction schedules. New orders continued through 1974.
Then, in May 1974, India tested a nuclear weapon.
The Indian test — conducted at the Pokhran military testing range, using plutonium produced in a Canadian-supplied research reactor and separated in an Indian reprocessing plant — was the first nuclear explosion by a country outside the five acknowledged nuclear weapons states. It demonstrated, in a way that theoretical arguments about proliferation had not, that the civilian nuclear fuel cycle could be diverted to weapons purposes by a sufficiently determined state. India had not signed the Non-Proliferation Treaty. India had used civilian nuclear assistance to build a weapon. The connection between peaceful and military nuclear technology, which had always existed in principle, had become undeniable in practice.
The Indian test coincided with growing American concern about the expansion of commercial reprocessing. The Allied-General Nuclear Services consortium had been building a large commercial reprocessing plant at Barnwell, South Carolina, since 1970. General Electric had built a similar facility at Morris, Illinois, which was discovered to have serious design flaws and had been placed in cold standby without ever operating commercially. Nuclear Fuel Services had operated a reprocessing plant at West Valley, New York, from 1966 to 1972, but had shut it down for upgrades that proved impossible to complete within the new regulatory environment. The commercial reprocessing industry that the AEC had been promoting since the late 1960s was, by 1974, in serious trouble — its plants either not operating or unable to meet tightening regulatory requirements, its economics increasingly unfavorable, its proliferation implications becoming politically toxic.
For the thorium cycle, the proliferation concerns were particularly damaging.
This requires some explanation. Uranium-233, the fuel produced from thorium, is fissile and can in principle be used to build a nuclear weapon. It is less well suited to weapons than plutonium-239 — it has a higher critical mass, its contamination with uranium-232 (an inevitable consequence of thorium-cycle irradiation) makes it hotter and harder to handle, and it has never been used in a deployed nuclear weapon. But it is a fissile material, and the reprocessing technology required to separate it from thorium reactor fuel is similar in general character to the technology required to separate plutonium from uranium reactor fuel.
From a proliferation-policy perspective, this meant that the thorium cycle shared one of the features that had made the plutonium cycle politically controversial: both required reprocessing. Any effort to deploy thorium reactors at scale would require the deployment of reprocessing infrastructure, and by 1975 the political space for new reprocessing infrastructure in the United States was rapidly closing.
The Nuclear Regulatory Commission, spun off from the AEC in January 1975 as part of the same reorganization that created the Energy Research and Development Administration, took a harder line on reprocessing than the AEC had taken. Its regulatory reviews of the Barnwell plant were protracted and increasingly skeptical. Congressional hearings in 1975 and 1976 featured testimony about the risks of plutonium theft and the inadequacy of existing safeguards. Environmental groups — the Natural Resources Defense Council, the Union of Concerned Scientists, the Sierra Club — filed legal challenges that slowed the licensing process to a crawl. The commercial reprocessing industry that had been the presumed future of civilian nuclear fuel management was grinding to a halt under the weight of its own accumulated problems and the shifting political environment.
Ford’s Decision
Gerald Ford became president on August 9, 1974, three hours after Richard Nixon’s helicopter lifted off from the South Lawn of the White House. He was, as everyone observed, an unexpected president — a Michigan congressman who had been appointed vice president the previous October, after Spiro Agnew’s resignation, and who had never sought national office. He was sixty-one years old. He had been a varsity football player at the University of Michigan in the 1930s. He was well-liked personally, had few enemies on either side of the aisle, and brought to the presidency what the country seemed desperately to need after two years of Watergate: ordinariness.
Ford’s relationship with nuclear policy was generally supportive. He was a Republican from a state — Michigan — with significant nuclear industry interests. He had voted for nuclear programs in Congress. His administration was not hostile to the LMFBR, to commercial power reactors, or to nuclear expansion in general. If Ford had been elected in his own right in November 1976, his administration would most likely have continued the policies it had inherited from Nixon: LMFBR development, commercial reactor expansion, eventual deployment of reprocessing infrastructure.
But Ford was not elected. By the early autumn of 1976, the polls showed him trailing Jimmy Carter, the former governor of Georgia who had won the Democratic nomination on a platform of moral renewal after Watergate. Carter had campaigned, among other themes, on nuclear proliferation concerns. He had served as a submarine officer in Rickover’s program in the late 1940s and early 1950s — the only American president, before or since, with direct nuclear propulsion experience. He understood the technology. He also, increasingly visibly over the course of 1976, understood proliferation as a defining issue of the era. Carter had said, multiple times on the campaign trail, that as president he would ban domestic reprocessing and work internationally to prevent its spread.
This was the political context in which Ford made his October 28, 1976 decision.
Five days before the election — five days before the election he would lose — Ford issued a statement on nuclear policy that effectively ended the commercial reprocessing program in the United States.
The statement was careful. It did not, in legal terms, ban reprocessing; the president did not, under existing statutes, have the authority to do that unilaterally. What it did was direct the executive branch to “delay commercialization of reprocessing activities in the United States until uncertainties are resolved.” It invoked the proliferation concerns that had become salient after the Indian test. It called for international coordination to restrain the spread of reprocessing technology. It directed the Energy Research and Development Administration to change its policies “which heretofore have been based on the assumption that reprocessing would proceed.”
The key passage read:
> *”The United States should no longer regard reprocessing of used nuclear fuel to produce plutonium as a necessary and inevitable step in the nuclear fuel cycle, and we should pursue reprocessing and recycling in the future only if they are found to be consistent with our international objectives.”*
The practical consequences were immediate. The Barnwell plant, on which Allied-General had spent approximately $250 million, was left without a federal licensing pathway. The Morris facility was similarly stranded. The West Valley plant, which had been permanently shut down earlier in 1976 for regulatory reasons, was not reopened. The American commercial reprocessing industry, which had been in serious trouble before Ford’s statement, was effectively terminated by it.
Ford added, in a passage that reflected the continuing importance of the LMFBR to American nuclear planning: *”My decisions today do not affect the U.S. program of research and development on the breeder reactor. That program assumes that no decision on the commercial operations of breeder reactors, which require plutonium fuel, will be made before 1986.”*
But this was a qualification that few observers took at face value. A breeder reactor program requires reprocessing. The LMFBR’s economics depended on the ability to recover plutonium from spent fuel and recycle it as new fuel. If commercial reprocessing was indefinitely delayed, the LMFBR was, in practical terms, on a very long leash. Clinch River was already in trouble for cost reasons; Ford’s statement added policy reasons that would make the project politically vulnerable to any future administration less committed to its continuation.
Ford’s motive for the October 28 decision was the subject of considerable analysis at the time and afterward. The generally accepted interpretation — held by the policy community, by industry observers, and by historians — was that Ford was acting preemptively to define the Republican position on proliferation before Carter could take office and define a Democratic one that would be framed as a break with Republican policy. By suspending reprocessing himself, as a Republican, Ford both addressed the proliferation concern that had become politically salient and denied Carter the opportunity to claim it as a partisan issue.
This interpretation is consistent with the specific timing of the statement — five days before the election, at a moment when Ford’s political options for distinguishing himself from his opponent were narrowing. It is consistent with the private analysis within the Ford administration that Carter would ban reprocessing regardless. And it is consistent with the subsequent trajectory: Carter did indeed ban commercial reprocessing, formally, on April 7, 1977, announcing it as a new policy initiative rather than as a continuation of Ford’s; but the effective policy had been set the preceding October, by a Republican president anticipating what his Democratic successor would do.
For the civilian nuclear industry in the United States, the October 28 statement marked a decisive turning point. It was the moment when the closed fuel cycle — the recovery and recycling of nuclear materials that had been envisioned since the 1940s as the long-term basis of civilian nuclear power — ceased to be a realistic near-term program. The reactors that would be built in America from that point forward would be once-through reactors, using enriched uranium once and then discarding the spent fuel as waste. The breeder programs, deprived of the reprocessing infrastructure that their fuel cycle required, would wither over the following decade. By the mid-1980s, both the LMFBR and its dedicated infrastructure would be cancelled. The thorium cycle, which had always required reprocessing, was by then no longer a live option at all.
Epilogue: Files in Storage
In the spring of 1977, three events occurred within six months of each other.
On January 20, Jimmy Carter was inaugurated. On April 7, he announced his formal ban on commercial reprocessing. In the months that followed, the Clinch River Breeder Reactor Project, which had been the centerpiece of American fast breeder development for a decade, entered the long legislative battle that would eventually end with its cancellation in 1983.
At Oak Ridge National Laboratory, the Molten Salt Reactor Experiment had been in caretaker status for seven years. Its control room was dark. Its fuel had been removed. The Hastelloy N loops in which fluoride salt had circulated for four years were cold and empty, sealed and monitored but no longer part of an active program.
The engineers who had built and operated the MSRE — Paul Haubenreich, Murray Rosenthal, Charles McDougall, the chemistry team, the metallurgy team, the reactor physics group — were by 1977 mostly retired or working on other projects. A few continued producing reports on molten-salt technology under contract to ERDA, producing documents that would eventually be catalogued and archived as ORNL technical reports. ORNL-4812 on chemistry. ORNL-4832 on materials. ORNL-5018, the December 1974 program plan for the MSBR, which had proposed the development path that had been refused funding and then terminated.
These reports went into the laboratory’s technical library and, later, into the Department of Energy’s document management system. They were not classified. They had been prepared as public technical documents. But they were obscure. They were cited occasionally in academic nuclear engineering literature. They were mentioned, sometimes, in advocacy publications produced by small groups that believed the thorium cycle deserved reconsideration. For the most part, however, they were not consulted.
The mainstream of American nuclear engineering education, throughout the late 1970s and the 1980s and the 1990s, focused on pressurized-water reactors and boiling-water reactors — the technology the industry actually operated. A student pursuing a nuclear engineering degree at a major American university during those decades might encounter thorium in a brief mention in a textbook chapter on advanced fuel cycles. Molten-salt reactors might be described in a single paragraph, in the past tense, as an interesting historical experiment. The MSRE itself was known about, within the nuclear engineering community, as a small research reactor that had operated for a few years at Oak Ridge. It was not known about, outside the community, at all.
The public memory of civilian nuclear power in America through the 1980s and 1990s was defined by a different set of events: Three Mile Island in March 1979, Chernobyl in April 1986, the cancellation of the Shoreham plant in 1989, the slow decline of utility orders from the peak of the late 1960s, the persistent waste-management problem, the anti-nuclear movement’s transformation into a kind of default cultural position. The story that was told was not the story of reactor technology choices. It was the story of an industry that had overreached, failed to manage its risks, lost the confidence of the public, and entered a long period of retrenchment.
What had actually happened, at the level of technology, was also a story — but it was a story that almost nobody knew. The story of Wigner and Weinberg in 1944, of the fluid-fuel reactor design that had eliminated xenon poisoning. The story of Cisler’s engineers at Lake Angelus in 1951, picking the thorium breeder as their first choice. The story of Ray Briant’s “bunch of spaghetti” and the Aircraft Reactor Experiment. The story of Seaborg at the MSRE controls on October 8, 1968. The story of the November 1969 Budget Bureau decision. The story of Weinberg’s removal in 1973. The story of ORNL-5018, published in December 1974, proposing a development path that the country chose not to fund. The story of Ford’s October 28, 1976 statement, closing the door on the fuel cycle that all of this technology required.
That story was in the files. The files were in storage.
What happened after 1977 — the long period in which the thorium story was not told, and the shorter period, beginning around 2000, in which it began to be told again — is the story of Part Four.
Notes on Sources
The political and institutional narrative of Part Three is drawn from Glenn Seaborg’s three-volume memoir of his AEC chairmanship: *Kennedy, Khrushchev, and the Test Ban* (1981), *Stemming the Tide: Arms Control in the Johnson Years* (1987), and *The Atomic Energy Commission Under Nixon* (1993). The last of these contains, on pages 177-179, the detailed account of the November 1969 Budget Bureau forced choice between the LWBR and the MSBR, and on page 179 Seaborg’s retrospective self-indictment on the LMFBR commitment.
The 1962 *Report to the President on Civilian Nuclear Power* is cited directly from the document itself; its discussion of breeder reactors and its brief mention of the thorium-uranium cycle are drawn from its text without elaboration.
Chapters Twenty-Three and Twenty-Four, on the MSRE’s operation and the 1970 *Nuclear Applications & Technology* special issue, draw on the special issue itself: Volume 8, Number 2 (February 1970), which collected eight technical papers by twenty-seven Oak Ridge authors together with a preface by Alvin Weinberg. The specific papers used here are: M. W. Rosenthal, P. R. Kasten, and R. B. Briggs, “Molten-Salt Reactors — History, Status, and Potential”; Paul N. Haubenreich and J. R. Engel, “Experience with the Molten-Salt Reactor Experiment” (which provided the operational narrative of the MSRE’s U-235 operation through March 1968, the fluorine stripping of the fuel salt in the summer of 1968, the October 2, 1968 criticality on U-233, and Seaborg’s October 8, 1968 high-power run); E. S. Bettis and Roy C. Robertson, “The Design and Performance Features of a Single-Fluid Molten-Salt Breeder Reactor” (reference design, reactor vessel dimensions, cost comparison with PWR); A. M. Perry and H. F. Bauman, “Reactor Physics and Fuel-Cycle Analyses” (breeding ratio, specific inventory, doubling time, fuel-cycle cost); Dunlap Scott and W. P. Eatherly, “Graphite and Xenon Behavior and Their Influence on Molten-Salt Reactor Design” (four-year graphite life, 0.5 percent xenon penalty); and M. E. Whatley, L. E. McNeese, W. L. Carter, L. M. Ferris, and E. L. Nicholson, “Engineering Development of the MSBR Fuel Recycle” (reductive extraction chemistry using liquid bismuth at 600°C). The Briant and Weinberg 1957 paper in *Nuclear Science and Engineering* Volume 2 (“Molten Fluorides as Power Reactor Fuels,” pp. 797-803) is cited for the earlier theoretical foundation.
The first-person retrospective material in Chapter Twenty-Five is from H. G. MacPherson, “The Molten Salt Reactor Adventure,” *Nuclear Science and Engineering* 90, 374-380 (1985). MacPherson directed the civilian MSR program from 1956 to 1960 and served as deputy director of ORNL through the MSRE program. His 1985 paper is the definitive first-person account of the program. All quoted phrases attributed to MacPherson in Chapter Twenty-Five — including his assessment of why the program ended (“the political and technical support for the program in the United States was too thin geographically”), the AEC evaluation of the tellurium and tritium problems, and the account of the 1972, 1973, 1974, and 1976 program terminations and reinstatements — are drawn from this paper.
The account of the two privately funded industry studies — the 1970 Molten-Salt Breeder Reactor Associates (Black and Veatch plus five Midwest utilities) and the 1971 Molten Salt Group (Ebasco Services plus five industrial firms and fifteen utilities) — is from MacPherson’s 1985 paper. The conclusion of The Molten Salt Group that existing technology justified construction of a demonstration plant is quoted from MacPherson’s paraphrase of that study’s conclusions.
Technical background on the MSRE continues to draw on ORNL-5018 (McNeese et al., December 1974) and Alvin Weinberg’s *The First Nuclear Era* (AIP Press, 1994). The six-day gap between the MSRE’s first U-233 criticality on October 2, 1968 and Seaborg’s high-power operation on October 8, 1968 is established by the Rosenthal-Kasten-Briggs paper and confirmed by Haubenreich and Engel. Previous informal accounts dating the event variously to August 1967 or January 1969 do not align with the contemporaneous operating record.
The account of President Ford’s October 28, 1976 statement is drawn from the statement itself, available from the Gerald R. Ford Presidential Library, and from the Public Papers of the Presidents. Context on the broader American commercial reprocessing situation is drawn from the Congressional Research Service report *Nuclear Fuel Reprocessing: U.S. Policy Development* (CRS Report RS22542, 2008 edition).