SCIENCE

THE PERFORMANCE OF REACTOR CORE MATERIALS

Introduced by Dr Gareth Neighbour, University of Bath, on 22 May 1998

The knowledge and experience on the performance of nuclear core materials is immense and very extensive. This diverse subject covers a vast range of materials, sufficient to employ hundreds, if not thousands, of scientists and engineers world wide.
Commercial nuclear power started in the UK with the opening of Calder Hall Magnox reactors over 40 years ago in 1956. In 1997, the UK had 35 commercial reactors, including fourteen Advanced Gas-cooled Reactors (AGRs) and one Pressurised Water Reactor (PWR) accounting for ~13 GW(e) approximating to 28% of electricity production. Typical electrical output rose from 435 MW for the Oldbury Magnox reactor to 1350 MW for a typical AGR. Although the broad principles remained the same for both Magnox and AGR stations, the design details evolved and were much improved. Globally, approximately 40 new reactors are under construction and a further 60 are in the planning stage. Whether we like it or not, nuclear power is set to rise in the years ahead considering that one third of the world’s population, ~2 billion, do not have access to commercial energy. These facts underpin the need to monitor and improve the use and efficiency of nuclear materials.
A description of the general circuit and components that make up the typical Magnox reactor and AGR was presented, including the reactor pressure vessel, control rods, coolants and the moderator. This was followed by a detailed description on:
the manufacture of fuel pellets and elements and the reasons for the development from pure uranium bars used in the Magnox rea.ctor to oxide fuels used in the AGR.
the fuel cladding and possible deformation mechanisms including embrittlement, cladding collapse and the ratchetting process, deposition on the fuel as a result of extrinsic catalysts and also from high levels of carbon monoxide and methane within the coolant, and the effects of this deposition on heat transfer and fission gas release within the fuel pin. Measurements show that the majority of fission gas releases in an AGR are much less than 1% of that generated, but occasionally, on pins with heavy columnar deposition, 10% can be obtained.
A case study was also presented which illustrated a recent piece of research concerning the behaviour of the moderator graphite with irradiation and its effect upon the coefficient of thermal expansion (CTE). Nuclear graphites differ in several respects from single-crystal graphite in that they have: (a) a complex networks of pores which interlace the microstructure; (b) a wide variety of crystallite sizes dependent upon raw materials and manufacturing processes; (c) two or more ca.rbonaceous species originating as filler, binder or impregnant; and (d) large clustsrs of crystallites (filler particles) that are connected by a binder or impregnant carbon. For these reasons, detailed understanding of the changes in properties of the moderator graphite in reactor life can be difficult. Essentially, there are two factors which affect core life: neutron irradiation which causes dimensional changes and radiolytic oxidation which weakens the structure. One of the advantages of AGR graphite is that when subject to reactor conditions, it will only shrink 3% over the AGR life at a dose of ~125 n/cm2. Neutron irradiation causes an initial increase in CTE of the moderator, peaking at a dose of ~25 n/cm2. At first, this appears to be a paradox since the same crack/pore system is responsible. A second paradox is evident when CTE is shown not to change with oxidation, but the minima in the dimensional change curve is delayed to higher doses. It is surprising that no change in CTE occurs with oxidation since other properties such as elastic modulus, E, decreases with oxidation and other workers have found for a range of polygranular graphites that log E varies linearly with log CTE. A simple model was proposed to provide explanations for these paradoxes. The results indicated that the most plausible explanation is that the porosity required to accommodate dimensional changes is probably in the micron range, but CTE is influenced by pores of width < 50 nm, which are largely unaffected by oxidation and controlled by the materials and processes used in manufacture.
G. Neighbour