Rock salt provides the world's best barrier, and has thus long been considered as a host rock for nuclear waste repositories. Salt domes have vertical extensions of several kilometres and are consequently well-suited for deep borehole disposal, as an alternative to a mined repository.
The isolation capacity of undisturbed salt rocks is based on the creep behaviour which tends to close any access paths. A number of natural analogues where fluids have been contained in cavities under high pressure for millions of years support this picture. For waste disposal, one has to show that the tightness of the geological barrier is not compromised by the repository excavation and the thermal loading due to the waste, and that suitable geotechnical barriers can be constructed.
We discuss the option of deep borehole disposal in rock salt. In contrast to the mined repositories which are usually considered, this option employs boreholes of several kilometres, drilled into a salt dome. The waste canisters are emplaced in the lower part of the borehole. As backfill material, we propose molten salt mixtures, similar to the ones used in solarthermal plants as heat exchange and storage fluids. The waste-generated heat will keep the salt liquid for a long time, ensuring complete containment without the possibility of ground water reaching the waste. The upper part of the filled boreholes, still inside the salt dome, converges under lithostatic pressure due to salt creep, and the top part can be sealed with asphalt, bentonite and concrete. We present some experimental results on the closure of boreholes under pressure and the properties of liquid salt to support our proposal.
The favourable barrier properties of rock salt have long been exploited for cavern storage of gas and oil, and make it a prime candidate for as waste repository host rock. Usual concepts, both in bedded and domal salt, assume a repository mine in a salt formation in a depth of several hundred to a thousand metres where heat-generating waste can be emplaced in drifts or in boreholes below drifts. After emplacement, drifts can be backfilled, and the repository is closed using suitable shaft seal systems. Over time, salt creep will cause convergence in the mine and enclose the waste, completely isolating it from the biosphere.
Meanwhile, in the context of other host rocks such as claystone or crystalline rocks, the deep borehole disposal concept has been suggested. In this paradigm, boreholes are drilled to a depth of several kilometres, and the waste canisters are emplaced directly from the surface. Details of casing, seals etc. depend on specific concepts and host formations. In the crystalline basement, the waste will be presumably be in contact to ground water, however, due to the large depth and hydrogeological factors, the waste can be effectively decoupled from the biosphere. Technically, this approach can be much simpler than a mined repository.
Here we propose to combine the advantages of these concepts: Salt domes can have thicknesses of several kilometres, and can correspondingly accommodate deep boreholes. In large depth, convergence is much faster, and waste can be completely contained rather quickly, without any groundwater access. A novel backfill material – eutectic molten salt mixtures – can further enhance this idea, offering immediate complete containment of the waste as well as retrievability.
Laboratory studies simulating thermal maturation of source rocks to generate and expel oil/gas differ from natural conditions. Amounts and compositions of products deviate between experiments and those found in natural petroleum source, carrier and reservoir rocks. Experimental data used in numerical models simulating oil/gas generation/expulsion thus seem to require adaption, causing significant uncertainty in present petroleum systems analysis. We designed and tested a procedure to simulate laboratory generation and expulsion of hydrocarbons under conditions most similar to natural conditions.
The critical factors in combined petroleum generation and expulsion simulation studies are: i) the pore and kerogen network of a sediment must remain as closely as possible in its natural state, as destroyed pore-systems are unsuitable for interpretation due to artificial reaction sites and migration avenues created upon experiment; ii) kerogen maturation must occur in the presence of water, as water acts as hydrogen donor for oil/gas formation and via hydrostatic pressure and associated flow stimulates migration of oil/gas; consequently, dry pyrolysis often used in petroleum generation experiments for kinetics calculation will not reflect near-natural oil/gas expulsion; iii) pressure regimes applied upon experiment must mimic differentially the lithostatic and hydrostatic pressures, as lithostatic squeezing will cause collapse of clay mineral aggregates but opposes kerogen swelling in non-lithified rocks; consequently experiments under all-directional identical confining pressures are unsuitable to reflect oil/gas expulsion; iv) the generated and then expelled oil/gas fluids must be allowed to migrate in the given pressure and permeability regime in order to avoid artificial secondary reactions, including condensation as well as cracking, hence rendering closed pyrolysis unsuitable for many petroleum expulsion studies. Exceptional cases may include tight shales, where expulsion is hindered or burial of reservoirs associated with oil to gas cracking.
The review by Professor J. C. Jaeger (Chapter 1) briefly summarizes and discusses a select group of papers by various workers dealing primarily with aspects of fracture in brittle rock materials. Professor Jaeger's treatment is primarily concerned with the correlation of phenomenological and theoretical failure criteria with empirical results from short duration tests at ambient temperatures and generally at low lithostatic and interstitial fluid pressures.
As engineering rock mechanics becomes more concerned with new techniques of mineral exploration and production at increasingly greater depths in the crust, it is well to be aware of the effect of the environment --that of the superposed stress field upon the lithostatic and interstitial fluid pressure, the temperature and the duration of the loading upon the failure of rocks either by fracture or flow. All results to be discussed here are from laboratory tests on small homogeneous samples, thus caution should be used in the direct extrapolation of these results to large inhomogeneous rock masses having quite different boundary conditions in the earth. The purpose of this supplement is to briefly point out the major effects of these parameters as examined singly, with all others held constant, and to cite some recent results which apply to the main theme on the brittle failure of rocks.
The total stress field imposed upon a small volume of porous and permeable rock in the earth's crust may be thought of as the sum of an applied differential stress, plus a lithostatic pressure, plus an interstitial fluid pressure. The differential stress portion may be completely specified by nine components, six shear and three normal. If the small volume element is in rotational equilibrium, by suitably choosing our coordinate axes we may resolve these nine components in to the three principal components. Since neither the lithostatic pressure nor the interstitial fluid pressure* contain shear components, the fluid pressure may be subtracted from the lithostatic to give the effective lithostatic pressure (see Terzaghi 1) and this in turn added to principal components from the differential stress to give the three principal effective stresses.
Brittle failure of rock as evidenced by either extension (tension) or shear fracture may be enhanced or suppressed in all rock types by suitably altering either the differential stress, or the lithostatic or the interstitial fluid pressure. This is reflected in the absolute values of the three principal effective stresses at failure as well. Typically, at the lower values of differential stress, low lithostatic and high interstitial pressures (equaling relatively low principal effective stresses), the rock is weak, being nearly elastic up to its rupture point and failure is by extension fracture normal to the least principal effective stress, s3. As the differential stresses and the lithostatic pressure are increased or as the interstitial pressure is decreased, the rock becomes stronger and more ductile. Failure at these conditions takes place by a combination of the extension fracture (normal to s3) and shear fracture which is inclined at an angle from a few degrees to less than 45 degrees to s1.