Euradwaste '08 - EU Bookshop - Europa

Shear-tests at increasing normal pressure between highly pre-compacted crushed salt blocks
and host rock (rock salt) including wetting with brine.
The load bearing capacity and dilatancy behaviour of the saltbricks (‘dry’ state) differ significantly
from the behaviour of the compact natural rock salt due to reachable larger deformations and relatively
high load bearing capacities whereby local dilatancy is overlapped by the decrease of porosity
during loading. It is important to note that the strength of the salt bricks is drastically reduced
when moisture is present. Hydrostatic compaction experiments clearly demonstrate that besides the
loading conditions the water content is the key factor for the compaction processes.
Shear tests were performed to investigate contact properties between salt brick surfaces and the
rock salt. Whereas at dry conditions only some friction occurs, significant strengthening is observed
when moisture is present because of activation of cohesion.
The development of a microphysical basis for the constitutive models of the compaction behaviour
and the evolution of transport properties of crushed salt is considered crucial for extrapolating results
from laboratory to in-situ conditions. In NF-Pro special attention was paid to the (controversial)
role of water [4]. An oedometer compaction cell with about 20 mm diameter was used for testing
sieved granular salt samples using the stress relaxation method. This allowed compaction creep
rate vs. axial stress data to be acquired for an individual sample at a range of near-constant porosities.
Stress relaxation data obtained for dried salt samples show that the compaction creep rate
measured at a given porosity is highly
18
sensitive to applied stress ( ε ∝ σ
Figure 3. Compaction creep data for granular salt tested in
presence of saturated NaCl solution as pore fluid.
199
� ) but
insensitive to grain size. If exposed to
the lab air, from which 25-50 ppm of
water is typically adsorbed by dry
granular salt, the samples become
considerably weaker, i.e. faster creep
rates are observed at a given stress and
porosity. When tested in the presence of
saturated NaCl solution as pore fluid,
the samples show much higher
compaction creep strain rates than airdry
samples (Figure 3). The creep rate
also becomes strongly sensitive to grain
size across the whole region of stress
investigated. It is evident from the
experiments, that dislocation
glide/creep and pressure solution both
operate in the samples, with the former being most important in truly dry samples and the latter becoming
important in brine-bearing samples at low stresses and fine grain size. However, an additional
water-sensitive process also seems to operate in lab-dry and wet samples.
5. Conclusions
The studies of the saturation of a bentonite buffer point to the existence of a threshold gradient for
water flow in bentonite. When the effect of a threshold gradient is included in the modelling of the
barrier behaviour, the trend with a very low hydration in hydration tests under thermal gradient can
be well reproduced, but the inclusion of this process makes the model underestimate the hydration