Stress-Dependent Permeability: Characterization and Modeling

Abstract During the production lifecycle of a reservoir, absolute permeability at any given location may change in response to an increase in the net effective stress and a concomitant decrease in the value of in-situ permeability. This paper focuses stress dependent permeability in unconsolidated, high porosity sand reservoirs (offshore turbidites) and consolidated reservoirs (tight gas sands). Specifically we address: i) fundamental controls on stress dependent permeability, as identified through analysis of core samples, ii) rock-based log modeling of stress dependent permeability in cored and non-cored wells, and iii) implications for production based on data from reservoir simulation. This work reveals a fundamental difference between the stress dependent permeability behavior of unconsolidated and consolidated sand reservoirs. In unconsolidated sand reservoirs, the greatest permeability reduction with stress occurs in the sands with the highest values of porosity and permeability. In cemented sandstone reservoirs, the opposite is the case: most of the reduction in permeability occurs in sandstones with the lowest values of porosity and permeability. This difference in behavior between unconsolidated and consolidated reservoir sands is controlled by pore geometry. We present a practical, rapid and cost efficient methodology to improve evaluation and enhance the productivity and management of stress-dependent reservoirs. The method is based fundamentally on the identification of ΑRock Types≅ (intervals of rock with unique pore geometry). Thin section evaluation, together with integrated nuclear magnetic resonance and SEM-based image analysis of core material is used to quantitatively identify various Rock Types. Rock Type data is integrated with measurements of permeability at various levels of stress. Results demonstrate that, within a particular field, some Rock Types lose <10% of original permeability while others lose >90% of original permeability as a function of increasing stress. Rock Types are then identified using routine suites of wireline logs, allowing for field-wide determination of the net footage and distribution of each Rock Type in all wells and the foot-by-foot calculation of permeability at any value of net effective stress. Based on geological input, the reservoirs are divided into flow units (hydrodynamically continuous layers) and grid blocks for simulation. Several cases are presented of a conceptual, single well model of an overpressured, tight gas sandstone reservoir that include stress dependent permeability. Results of simulation analyses for varying conditions of reservoir stress demonstrate the importance of stress dependent permeability in more accurate forecasting of reserves and predicting optimum well bore producing conditions.