Permeability and porosity modelling for resedimented mudrocks – Applications for compaction dominated mudrock systems

Abstract

Mudrocks have complex petrophysical properties due, in part, to their very fine grain size and the associated high specific surface area (m2/g) and cation exchange capacity (CEC, meq/g). The role of the cation exchange capacity and its associated bound water must be included to account for mudrock electrical, transport, and mechanical properties. A widely varying set of samples from the deep water Gulf of Mexico Miocene section, the arctic seafloor, the Pierre Shale, and Brazos River sand (used for varying sand content) were resedimented to study the effects of these variations on compaction and permeability. The mudrocks were characterized in terms of CEC, clay mineral content, bulk mineralogy, and particle size distribution. Resedimentation techniques were employed to measure the mudrock properties under controlled conditions i.e. slurries were made with known clay mineral type, bulk mineralogy, particle size distribution, CEC, and pore fluid salinity, and subsequently compacted in an oedometer cell.The measured permeabilities were found to be strongly dependent on the clay mineral CEC and on initial pore fluid salinity. Two models were developed to explain the variation in measured permeability: a modified Carman-Kozeny model and a staged differential effective medium model (SDEM). The Carman-Kozeny model fits the permeability data to within a factor of three but underestimates the stress dependence. The SDEM model includes the effects of both bound water saturation and porosity on fluid flow. The model was also able to estimate permeability for all the different mudrocks to within a factor of three with an improved match to the stress dependence. These two models were also successfully tested using previously published permeability data for both resedimented and intact Boston Blue clay samples.A compaction model for porosity was also developed and applied to published data for artificial mixtures of smectite and kaolinite. The model predicts the stress dependent porosity of the resedimented clays to within four percent error. It is then extended to incorporate the influence of the inclusion of sand sized material and applied to our resedimentation data and previously published Boston Blue Clay data. The required inputs for the model include the weight fraction of non-clay sized material, and the cation exchange capacity (CEC, meq/gram). No consistent salinity dependence was observed for compaction.