Role of cracks in progressive permeability reduction during flow of heated aqueous fluids through granite

Abstract

Permeability was measured continuously during an experiment in which a heated, aqueous fluid was pumped through a cylindrical sample of Westerly Granite. Fluid flowed radially outward from an axial borehole down a temperature gradient to the outer edge of the sample. The fluid temperature in the borehole was maintained at 300°C, while a pore pressure difference of 0.5 MPa provided the driving force for fluid flow. Permeability decreased by a factor of 25 during the 2‐week experiment. Scanning electron microscope examination of this altered sample indicates that about half of the grain boundaries involving quartz are cracked or were cracked at some time during the experiment and now contain a filling material. Grain boundaries between the two feldspars are closed in the starting material and in all areas of the altered sample that were examined. Intragranular cracks in all three major minerals are open in the starting material. These observations suggest that initially, the quartz forms a more effective fluid pathway than do the feldspars. About half of the intragranular cracks near the borehole are open, whereas the other half contain a massive deposit which is either Si‐rich or Ca‐rich. In most cracks at the outer edge of the altered sample a Si‐rich filling is present. In crack intersections in this area the filling has a platy texture. At several times during the experiment the discharged fluids were sampled and chemically analyzed. The silica content of fluids passing through the sample was calculated as a function of radius, using rate equations for the dissolution and precipitation of silica. The calculated silica concentrations for the discharged fluids are approximately in agreement with measured values. A separate calculation of the reduction of crack porosity due to uniform dissolution and reprecipitation of quartz indicates that these processes caused a maximum of 10% crack porosity change for an experiment with a borehole temperature of 300°C. Assuming that reprecipitation of quartz was homogeneous, this porosity change corresponds to a permeability reduction of 27%, whereas the observed value was 96%. Therefore other processes besides the simple homogeneous precipitation of a layer of quartz must be operating to reduce permeability. Four possible processes are (1) nonhomogeneous precipitation, (2) precipitation in critical narrow places in cracks, (3) precipitation of other minerals in addition to quartz, and (4) crack healing.