We expect the reactor to be producing electricity by the end of 1973. We in the Authority have never proposed that the first commercial fast reactor should not be started until sufficient operating experience of the prototype had been obtained, to be absolutely sure that there were no fundamental problems unresolved. I have, however, always believed in continuity of design and experience and would like to see the next reactor started as soon as the lessons of the first have been fully assimilated by the designers and engineers. Clearly our hopes of a 1974 start are now too optimistic in the light of the commissioning and operating dates for the prototype and the amount of component testing now judged necessary. The design of the CFR is, however, under way…

CFR was the latest acronym, standing for Commercial Fast Reactor.

The February 1973 issue of Atom, the AEA monthly, reported on a meeting the previous November, attended by senior civil servants and nuclear industry management, on "Future Prospects for Energy Supply and Demand," presented by the "New Systems Forum" of the AEA. According to the report of the meeting:

A commercial fast breeder power station programme commencing with a lead station coming on line in 1981 and further stations in the mid — 1980s appears to be a reasonable assumption on the basis that PFR knowhow and experience will be adequate for a first order to be placed for around 1976.

The almost imperceptible note of caution—1976, not 1974, and the "mid-1980s" for subsequent stations—had to be set against the assumption that a station ordered in 1976 could be "on-line in 1981." This allowed only five years for construction and commissioning, compared to the eight plus years already run up at the Dungeness B Advanced Gas-cooled Reactor station and at contemporary fossil fueled stations likewise still unfinished.

Even the faint note of caution in this report was swept aside in an aggressive presentation delivered in the United States in mid-1973 by Tom Marsham, deputy managing director of the AEA’s Reactor Group:

Satisfactory experience with the experimental reactor DFR in the early 1960s led to construction of the 250 MWe power station at Dounreay which will be brought to power this year. Some two or three years from then, we are expecting to start constructing the 1300 MWe lead commercial station with ordering of subsequent commercial plants building up to large scale during the early 1980s. There is nothing adventurous or foolhardy about this plan.

Nothing, perhaps, except its central premise. The end of 1973 arrived and departed with the PFR still awaiting its first criticality, to say nothing of being brought to power. One primary and one secondary sodium pump malfunctioned during tests. Both had to be removed from the reactor for detailed examination. Tests continued with the remaining two primary pumps. On 11-14 March 1974, however, the British Nuclear Energy Society was to play host to a major international conference on "Fast Reactor Power Stations," with delegates from France, the United States, the rest of Europe, Third World countries and even the Soviet Union. The ignominy of welcoming the foreign visitors to the conference with the PFR still cold was too much to contemplate. The week before the conference the AEA pulled out the control rods at Dounreay, and on 3 March 1974 started up their new reactor for the first time. On the opening day of the conference they announced the fact with pride; it was far from coincidental that their French colleagues announced, on the closing day of the conference, that the French Phenix fast breeder had just attained full power.

One paper in particular, by Eric Carpenter, head of reactor physics at the CEGB’s Berkeley Nuclear Laboratories, warned that the CEGB was less enthusiastic than the AEA about a rapid move into fast breeders. Reliability was crucial; together with delays in construction, lack of reliability had "a much bigger deleterious influence on electricity costs than almost any of the advantages claimed in the brochure assessments." The CEGB by this time had all too much firsthand experience of both delays and unreliability of its conventional nuclear stations, and of what the paper scornfully called "brochure assessments." The paper asserted that the putative savings from introducing fast breeders as fast as possible would be no more than 5 percent of total expenditure on a nuclear system and then only in what it called "the unlikely event of capital costs of fast and thermal reactors being equal." The CEGB contributors considered that no order for a fast breeder power station could be placed before 1977 or 1978 at the very earliest.

Throughout much of 1974, staff at Dounreay continued running the PFR at low power. Small leaks appeared in the steam generators, the boilers in which hot molten-sodium passed through thousands of fine tubes to boil the water around the tubes. Such leaks were a particular problem in a sodium-cooled system because of sodium’s reactivity with water. A major leak, like one that had happened at the Soviet fast breeder prototype at Shevchenko in November 1973, would release enough hydrogen and heat to create a serious hazard of explosion. Even a minute leak, invisible to the naked eye, would lead to the formation of hydrogen bubbles in the sodium coolant, presenting at the very least an unwelcome irregularity in the coolant flow, and possibly actual control problems. By the end of October 1974, the most troublesome steam generator was decoupled from the reactor in order to find the leaking tube and plug it.

Six months later, the PFR once again played host to a visiting party. At the end of April the newly formed European Nuclear Society (ENS) held its inaugural conference in Paris. After the conference, one of the side trips took participants from all over Europe to Dounreay. AEA staff were happy to show off their reactor, which was, they said, working fine; a month earlier the plant had generated its first electricity. Unfortunately, however, it had yet to reach a power level above 12 percent of its full thermal capacity. Small but persistent leaks in the sodium water steam generators kept two of the reactor’s three cooling circuits out of operation. PFR staff carried on operating the reactor on its one remaining cooling circuit, but trouble with turbine bearings interrupted even this limited operation. Then, just before the nuclear dignitaries arrived from Paris, more small leaks manifested themselves, this time in a section of the only operative cooling circuit.

The AEA staff at Dounreay put on brave faces, but the ENS visit cannot have been an especially happy occasion. As the editor of Nuclear Engineering International, put it: "Although the reactor itself has been operating very well it has not yet been possible to build up any significant amount of fuel irradiation." Nor, it might be added, to generate any significant amount of electricity. The AEA continued to protest that the reactor itself was working well, and that the stubborn troubles at Dounreay were with the generating set and the steam generators. But the CEGB had already suffered many years of frustration with its own generating sets, and knew what a headache these could be.

Furthermore, to suggest, as the AEA was trying to, that the steam generators were somehow ancillary, not part of the nuclear system, was indefensible special pleading. One of the unique distinguishing characteristics of the fast breeder design selected by the AEA was precisely the choice of molten-sodium as a coolant. If you could not then use the molten-sodium reliably to boil water, you had a basic design problem — one that could not be brushed aside by reference to the satisfactory operation of the reactor core itself.