ASPECTS OF FLUID/ROCK MICROSTRUCTURE THAT MIGHT AFFECT ROCK RHEOLOGY: AN EXPERIMENTALIST'S HISTORICAL PERSPECTIVE

Over the past 15 years, high P-T experimentalists have developed a more-or-less unified view of the microstructures of rocks containing C-O-H fluids. In the mid-1980s it was discovered that, under conditions of chemical and mechanical equilibrium, the geometry and interconnectedness of pores depends upon the mineralogy of the host rock and the composition of the fluid. Long-range connectivity is common for H2O-rich fluids (which are generally "wetting") and nonexistent for CO2-rich compositions ("non-wetting"). Since fluid connectivity is key to the efficacy of diffusion on scales larger than the pores themselves, this finding is potentially important to bulk-rock rheology. Unfortunately, however, the "wetting" data are directly applicable only to monomineralic rock analogs of nearly uniform grain size, and so are of limited usefulness in understanding the behavior of rocks in the Earth.

Within the last 2-3 years, it has been shown that spatially variable mineralogy or grain size in a rock must lead to non-uniform distribution of any fluid present in the system. Both experimental results and theoretical considerations demonstrate that fluids localize around minerals having low crystal/fluid interfacial energies, and also around small grains of any mineral. A complex microstructure characterized by spatially variable fluid abundance may lead to equally complex rheological behavior: regions or units dominated by "wettable" minerals or small grains may weaken due to increased fluid abundance (e.g., drawing of fluid into ductile shear zones due to a local reduction in grain size?).

The foregoing conclusions and suggestions assume that fluids "permeate" rocks on the scale of individual mineral grains. This is a reasonable expectation for "wetting" fluids, given the rapid rates of chemical infiltration recently measured in the lab (millimeters to centimeters per year).