AmericanChemicalSociety.com

We have conducted a large suite of laboratory experiments to investigate: (1) fault zone hydraulic properties as a function of shear, normal stress, and fault rock composition; and (2) permeability behavior of natural mudstones to develop insights into relationships between permeability, compaction and stress state, and diagenesis. The work has important implications for pore pressure prediction in offshore petroleum development and drilling, transport and flow relevant to water resources and petroleum migration including assessment of reservoir and aquifer compartmentalization and seal integrity, and for tectonic problems involving fluid flow, pressure, and fault mechanics. This project also directly and indirectly supported the work of two post-doctoral fellows at Penn State (Insun Song and Samuel Haines), four graduate students (Rachel Lauer, Matt Ikari, Robert Skarbek, and Enrique Perez), and two undergraduate research students (Nick Adamson and Shaun Sagan).

Our work on this project included measurement of: (1) cross-fault permeability in experimentally sheared synthetic mixtures of chlorite, illite, and smectite clay-rich gouge, at normal stresses up to 60 MPa and engineering shear strains up to ~45; (2) permeability and frictional strength of natural fault zone and wall rocks from a thrust belt setting under the same conditions; and (3) permeability within natural mudstones during uniaxial consolidation at effective axial stresses up to 90 MPa and porosities ranging from 12%-62%. Notably, the natural samples are formed from essentially uniform protolith in a basinal mudstone section, and record the effects of progressive transformation of smectite-illite. Thus, the combination of the synthetic and natural samples have allowed us to address one of the main goals of our planned work – namely to quantify the effects of host rock diagenetic alteration on permeability of both intact mudstone and sheared material (i.e. faults).

We find that permeability of all gouges decreases dramatically with shearing, and to a lesser extent with increasing effective normal stress. Chlorite gouge is consistently more permeable than montmorillonite and illite gouge, and maintains a higher permeability after shearing. Before shearing, the illite gouge is 1-2 orders of magnitude less permeable than montmorillonite gouge, but after shearing the permeabilities of the two materials are comparable. In all three gouge materials, we observe pronounced permeability reduction concurrent with attainment of steady-state (residual) shear strength, suggesting that permeability reduction is linked to fabric development. These results imply that development of excess pore pressure in fault gouge depends on both clay mineralogy and shear strain. In experiments on intact basinal mudstones, our data define a systematic decrease in permeability with progressive consolidation and porosity loss. Values of permeability measured using flow through tests are consistent with those determined using transient methods from consolidation tests, which define permeability continuously as a function of porosity or effective stress. The data are well bounded by the functions: Upper Bound: log(k) = -19.94 + 6.93n; Lower Bound: log(k) = -20.96 + 6.93n; where n is fractional porosity. Notably, we observe no clear trend in the permeability-porosity relationships as a function of in situ sample burial depth, suggesting that, to first order, the mudstone permeability is not significantly affected by clay transformation. We have also conducted several experiments to assess the anisotropy of permeability in the natural mudstones, and find that this effect is negligible (maximum anisotropy of ~2-5 times) over the range of compaction states and stresses we investigated – in contrast to typically assumed values for anisotropy greater than 10-100 times.