Mechanical damage and transport properties of synthetic argillaceous rocks

The role of mechanical damage on the evolution of permeability was studied on synthetic argillaceous rocks made of sand, cement and clay. An experimental set-up was developed in order to produce synthetic rock samples made of the above mentioned components in fixed proportions. The mixture is introduced in a tubular mould of internal diameter 20 mm. A uniaxial compression of 10 kN is applied on the mixture to allow better grain contacts and to reduce the residual porosity. In a first step the procedure has been validated on a sand-cement mixture. We measured the uniaxial compressive strength as a function of cement content. The results clearly show a non-linear dependence between both parameters. This can be interpreted as a percolation problem in a medium containing two phases: sand grains and cement matrix. In a second step, the same procedure has been applied on a sand-clay mixture. Again the uniaxial compressive strength was measured as a function of clay content. Like for cement, the results show a non-linear dependence which can be explained as a percolation transition. A second experimental set-up was developed in order to measure the gas (N2) permeability of the synthetic samples. A hydrostatic cell (20 MPa) was designed with water as confining fluid. The sample is isolated from the confining pressure fluid by a Viton jacket clamped on the end pieces connected to the pore circuit. The upstream pore pressure circuit contains the pressurized gas bottle and a pressure transducer, the gas outflow being measured by a volumetric flowmeter in the downstream portion which is kept at atmospheric pressure. Permeability is measured using the stationary flow method. Confining pressure has been kept constant and equal to 3 MPa. Pore pressure is then changed to allow for Forchheimer and Klinkenberg corrections in order to determine the true permeability of the specimen. The procedure has been validated on samples composed of sand and clay pre-compacted at 10 kN. The results show the drastic decrease of permeability (10-12 to 10-15 m2) as clay content is increased from 10 to 40% ) Next we have checked the residual effect of mechanical damage on pre-compacted samples containing sand, cement and clay. Cement proportion has been kept constant and sand and clay content were varied accordingly. The samples were loaded in a uniaxial press up to increasing stress levels, unloaded, and introduced in the hydrostatic cell. Their apparent gas permeability was measured by keeping the pore pressure constant. The preliminary results show that for low clay content a slight permeability decrease precedes a rapid increase (from 3 10-19 to 2.5 10-18 m2 ) as one approaches the peak stress of the deformation curve. This effect of mechanical damage is less pronounced when increasing clay content. (authors)