All reported experiments were conducted in a triaxial pressure apparatus at the Rock Mechanics Laboratory of the Lamont-Doherty Earth Observatory. The experiments consisted of subjecting a fluid-saturated sand to different pressure and temperature conditions and measuring permeability evolution of the sand over the course of approximately 4 days. All experiments were run at a confining pressure (Pc) of 100 MPa and a pore pressure (Pp) of 50 MPa. Each experiment was run at a constant temperature and deviatoric stress (sd) varying from 25-275°C and 25-90 MPa, respectively. Two series of experiments were conducted; one with varying T and the other with sd varying (see Table 1). This was done to isolate the effect of T and sd on permeability evolution. Pc, Pp and sd were all maintained constant during each experiment by servo-control. To test several hypotheses, other experiments were conducted in which temperature, fluid composition or flow rate were changed during the experiment. The sand used in all but two experiments was 90 weight percent labradorite and 10 weight percent quartz, both with grain sizes of 210-500 µm. Sands were prepared by placing fragments of quartz or labradorite in a chipper and subsequently extracting the 210-500 µm fraction with sieves. Both quartz and labradorite were then run through a magnetic separator to remove most foreign phases. Before each experiment, the sand was soaked with de-ionized water and placed in an ultrasonic bath to remove any fine, adhered particles. In one experiment, 100% quartz was used and in another, 100% labradorite. This was done to determine the effect of sand composition on permeability evolution. The starting fluid used in all experiments was de-ionized water.

The starting sand was jacketed with copper tubing which had been Ni-plated to prevent any corrosion of the jacket (Figure 1). Surrounding the jacketed sample was a coil heater. Porous stainless steel filters were placed at both ends of the sample to disperse the flow of water through the sample and to prevent loss of material into the pore fluid system. Two thermocouples were used to record temperature and provide feedback for the temperature controllers; one on the outside of the jacket and the other within the pore fluid system immediately above the upper filter. The pore fluid thermocouple was used to control temperature.

In all experiments, the sample was first subjected to precompaction after which the heater was re-tightened. This step was necessary because application of Pc resulted in significant radial compaction and decoupling of the sample and heater. Once the sample was re-inserted into the vessel, it was subjected to an axial compression of approximately 1 MPa to affect O-ring seals. This was followed by saturation of the sample and evacuation of any air from the pore fluid system. Pc was then raised to 100 MPa with Pp always maintained at half of Pc so that an effective stress of 50 MPa was never exceeded. sd was then applied at a rate of 1 MPa/sec and time-dependent compaction allowed to stabilize (approx. 45 minutes). Fluid flow through the specimen and temperature were then applied simultaneously. The application of temperature resulted in an additional phase of compaction which will be discussed in greater detail below. Once this final compaction was complete, permeability measurements began and continued until the conclusion of each experiment.

The flow of water through the sand was driven by two pore fluid intensifiers (Figure 2), both of which were connected to pore pressure transducers. The total volume of water in the system was approximately 35 cc, 6 cc of which was driven back and forth through the sample at constant pore pressure differential, much smaller than the ambient pore pressure. The volume of water circulated through the sample was large enough so as to allow mixing with fluid in each intensifier. Once flow had been initiated, the pore pressure difference across the sample was measured from the two pore pressure transducers. Since the cross-sectional area of the sample (A), the length of the sample (DL), pore pressure difference across the sample (DP), viscosity of water (µ) and flow rate (Q) were all known parameters, absolute permeability could be calculated at any point during an experiment by using a modified form of Darcy's Law:

K= (µ DL Q)/(A DP)

In several experiments, tests were conducted with different values of DP, which showed that DP varied linearly with flow rate (Q). This indicated that flow was laminar and that the use of Darcy's Law was justified.

Go to next section

Go back to table of contents