Modelling silicitation of a clay buffer subject to opposing temperature and hydration gradients

Following emplacement and closure of an engineered barrier system (EBS), it is important to understand what kind of phenomena may take place as a cons equence of the coupling between various physical and chemical processes: th e buffer resaturation profile, temperature gradient, material/mineral alter ation (through heating or precipitation), corrosion and gas generation, and aqueous chemistry. These governing processes take place together, and focu sing on any one of them in isolation is difficult to justify, and may invok e unrealistic assumptions. Just as for other well-known coupled process models, such as geometrical fi ngering when flow is fully coupled to flux driven mineral dissolution, it i s expected that the resulting phenomena are robust with respect to the prec ise details of the submodel terms (fingering models are functionally the sa me for all types of mineral dissolution), and the important point is to ide ntify which parameters can most effectively control or limit various effect s. Hence the impact of coupling can be understood qualitatively by coupling relatively simplified submodel terms, before stepping on to a quantitative understanding. Here we take a phenomenological view and consider the problem of a resatura ting compacted clay buffer in which there remains a relatively dry zone clo se to the hot waste canister due to compression of air (and possibly vapour initially until pore pressures exceed 1.5 MPa), displaced by incoming pore water. The 'dry', low thermal conductivity, region is 'baked' by the preva iling temperature, resulting in the possible breakdown of the clay structur e. Throughout the remaining partially and fully saturated clay pore space: aqueous silica complexes are formed (enhanced by temperature) and redistrib uted from the heated zone by diffusive transport. On cooling these result i n the precipitation of amorphous silica. causing strong cementation and the formation of a clay stone layer adjacent to the metal overpack. There is a lso significant silicitation as far as the outer clay boundary. These pheno mena were observed in published experiments, and are shown to arise robustl y within a suitable coupled model of thermal, hydration and chemical altera tion Their relevance to barrier performance and potential impact upon perfo rmance assessment will be discussed. (C) 1999 Elsevier Science B.V. All rig hts reserved.