1. Early History
In terms of use in the nuclear industry the complex properties of graphite are yet to be fully understood. The British nuclear industry has heavily invested in graphite reactor technology, with every reactor now constructed having a graphite core.
In 1934 Enrico Fermi found that neutrons activated target materials more efficiently if they were moderated: he did so by passing neutrons through materials with relatively light atoms to dissipate the high initial kinetic energy of fast neutrons. At this time there was an abundance of experience in graphite manufacture in the US, so Fermi used graphite to produce the first graphite moderator reactor (known as Chicago Pile Number 1). All subsequent US military reactors were graphite moderated.
During the war British scientists were essentially separated from US research, so continued to develop graphite as a moderator material (the US on the other hand worked closely with Canada to develop CANDU reactors using Canada’s ample supply of heavy water). Initially water was to be used as a coolant for the graphite cores (as was done in the USSR), however the UK later decided to use air cooling on safety grounds.
By 1952 a CO2 cooled graphite-moderated power reactor design was evolved from the Windscale Pile, with the British Electricity Authority – the forerunner of the CEGB – collaborating on the conventional side and establishing its interest in nuclear power. This quickly lead to the construction of the Calder Hall and Chaplecross Plant.
Here the core lattice, pieced together by vertical fuel channels, consisted of vertical blocks stacked in columns with pairs of tiles in between. Adjacent columns came into contact only at the tiles, and even then only at those tile faces whose position was determined by graphite behaviour parallel to extrusion (the direction of minimum growth). Space was left for growth in the perpendicular direction, and to prevent channels from buckling, the core periphery was clamped by spring loaded restraints into a continuous horizontal arch – otherwise the graphite played no real structural role. When operating and hot the core expanded as graphite, and each column was therefore mounted on a ball-bearing to avoid mismatch with the steel support structure. The whole core was enclosed in a steel pressure vessel, so none of this graphite could be replaced.
2. Magnox Stations
In 1955, before Calder Hall had operated, the first purely civil programme was announced: the CEGB’s Magnox Stations (Bradwell, Berkeley and Hinkley Point). The designs were closely based on Calder Hall: block-and-tile columns on ball bearings, encircled by a graphite arch compressed by carefully temperature compensated steel restraints.
As these were being built however the AES found that instead of remaining stable, PGA graphite actually shrank parallel to extrusion. It was feared that when this occurred gaps would open between adjacent columns, permitting distortions which might endanger fuel cooling or control rod insertion. Zirconium pins were successfully inserted into the blocks at a late stage in core construction to maintain lateral alignment, and the reactors are still well within distortion limits.
Nevertheless, after there the complicated block-and-tile concept was dismissed for radial keying – the principals of which are shown in Figure 11 and Figure 45. Here keys are used to allow graphite block and core expansion without distorting the core lattice, or more importantly the channel alignment. Furthermore the channel spacing remains uniform as the core periphery expands and contracts; the lattice just opens out or closes up. However in this configuration the graphite plays a structural role.
The progressive increase in coolant inlet temperature (from 160 – 180°C at Berkeley and Bradwell to 245°C at Oldbury and Wylfa) reduced the rates of dimensional change. However, power densities were increased with consequential increases in core loadings (due to pressure, thermal and dose rate gradients), and also in radiolytic corrosion with its consequential increased rate of weight loss and the concomitant decreases in graphite strength.
3. Advanced Gas Reactor
By the late 1950s space enrichment capacity was becoming available, and the AGR was evolved as a potentially more economic and secure system than Magnox. It retained the well tested carbon dioxide / graphite combination, but used enriched uranium oxide fuel clad in stainless steel to achieve the high gas temperatures, and hence the advanced steam conditions and thermodynamic efficiencies of conventional power boilers. Power densities were also increased in an attempt to reduce overall size and hence capital costs. However the radiation dose rates are many times those in Magnox reactors.
It was determined that the new strength and density requirements could be met with a new fine-grained multi-impregnated isotropic graphite. This was made possible by using gilsonite coke as the raw material, which is found in Uinta Basin in Utah, USA. The improvement in dimensional stability is shown in Figure 44.
Work on the first prototype AGR was commenced two years after Calder Hall began first operation, known as the Windscale AGR, or WAGR, which maintained many of the technologies of the older Magnox Stations. By the 1960s commercial AGRs were being designed with the cores being designed to take advantage of the new graphites and engineering advances.
Very high priority has been attached to graphite quality – for example to get maximum consistency and purity the Gilsonite mines were specially re-opened and all veins surveyed to get the best possible material. Major investments have been made in graphite machining to close tolerances under ultra-clean conditions.
