Laboratory tests on different backfill types in an oedometer cell
Report of the project:
At BGR, the oedometric compaction of potential backfill materials is a line of research since almost 40 years. It determines the ease of a material to compact, which is of essential importance to know in encapsulating radioactive waste.
“Oedometric compaction” means that a material is subjected to an axial force while the radial volume is fixed. This is similar to lowering a small sized lid onto a material placed in a ridged cooking pot.
The ease of a material to compact is expressed as the so-called “compaction resistance” or “backfill resistance”, which is a measure of the minimal force that is needed for compaction. Backfill resistance is a state variable. It is controlled by the state of compaction that the material already experienced, i.e. by its density. However, it is also controlled by other material intrinsic and extrinsic factors, such as mineralogy, moisture content and temperature. With oedometric testing, we aim to estimate a materials’ final density that would be reached by slowly converging tunnel walls in a salt repository site. This density strongly defines the materials’ permeability and mechanical behaviour. Both parameters being essential in the safety assessment of a radioactive waste repository.
Moreover, oedometric compaction yields cohesive, consolidated blocks that can be used in further measurements, such as triaxial testing or permeability estimation.
We test a large variety of crushed salts. They vary in natural moisture content, grain size distribution, mineralogy and dislocation densities. We investigate crushed salt from flat bedded salt strata as well as from diapirs. We also add different amounts of bentonite as a permeability reducing additive to the crushed salt and we artificially alter temperature, compaction speed and moisture content of the raw material.
Moist crushed salt with more than one weight-% compacts easier, i.e. with less force. It can be artificially pre-compacted during emplacement. However, such moisture contents might cause different viscosities of the final compacted backfill to the less moist natural rock salt.
Bentonite additive lowers the backfills’ permeability, as well as reducing its resistance to compact. However, bentonite might dissipate water when sufficiently heated by the heat-emitting radioactive waste, which is unwanted as it enhances corrosion and promotes fluid pressure build-up when heated.
We extrapolate that the convergence of tunnel walls can compact crushed salt to a density similar to the surrounding natural salt, with similarly low permeability. However, in the beginning of crushed salt compaction, right after emplacement, it is significantly permeable. This early permeability is utile. It hinders the built-up of unwanted overpressures potentially caused by waste canister corrosion. The time required to consolidate the a-priori lose and permeable material to a fluid-tight barrier is a matter of ongoing debate. Therefore, next to compaction tests, we also use microstructure analysis to gain a firm process understanding. For instance in the very slow mechanism of so-called “creep compaction”, which is hard to reproduce in laboratory timescales and yet dominates at low differential stresses when the crushed salt is almost fully compacted.
We strongly collaborate with numerical working groups, both in-house and with external partners by providing and discussing our results.
