Application of the general microstructural model to erosion phenomena — mechanisms for the chemical–hydrodynamic conversion of bentonite to a pumpable slurry in conjunction with retrieval

One of the requirements of a system for disposal of spent nuclear fuel is that it should be possible and feasible to retrieve the canisters even after they have been put in place and the surrounding bentonite has absorbed water and developed a swelling pressure. This ‘gripping’ of the canister must be released before the canister can be retracted from the deposition hole. One way to achieve this is to convert the bentonite into a slurry by exposing it to a flow of water containing dissolved salts such as sodium or calcium chloride. The potential efficiency of such a process is remarkable in view of the well-known tardiness of the saturation of compacted bentonite with pure water. The present study of the mechanisms involved was prompted by the need to understand the prerequisites and limitations of such a conversion process. Thus, the literature on the molecular structure of montmorillonite (the major constituent of bentonite) was reviewed as well as the literature on the microstructure of montmorillonite–water systems. A review was also made of some chemical literature which led to identification of two rate-limiting factors for montmorillonite microstructure conversion: diffusion over large distances and association–dissociation of primary montmorillonite particles. The knowledge compiled was then used in analyses of the kinetics involved and the following conclusions were made: (1) Exposure of the compacted bentonite with fresh water causes it to swell and to produce free particles by exfoliation. They form gels which cause closure of the pores so that further uptake of water becomes limited by diffusion. (2) Exposure of the compacted bentonite with water containing dissolved salt causes the exfoliated material in the microstructure to shrink (or at least swell less than in the fresh water case). Thus more water can penetrate into the pores and cause differential expansion in the aggregate residues which, in turn, leads to further widening of the pores. (3) The gel formed in the above described process may be removed by the flow of the water thus exposing fresh bentonite surface to continued attack. The paper is based on the general microstructural model but goes beyond it by including also dilute systems. It is concluded that the chemical–hydrodynamical method for removing bentonite from around a deposited canister might be shown to be a robust and efficient one provided that differential expansion of the individual grains in the microstructure is accomplished, that flocculated conditions can be avoided and that the chemically modified material can be removed by flushing. © 1999 Published by Elsevier Science B.V. All rights reserved.