Uncertainty and Sensitivity of Permeability Models

Abstract

Abstract Quantifying and managing uncertainty leads to improved project and business performance, robust decision making, increased chance of success, fewer surprises and optimised reservoir characterisation outcomes. Informed decisions will be robust to risk and increase the likelihood of an outcome that delivers promised production from a well or development. Petrophysicists have a crucial role in delivering an appropriate product for estimating productivity and its' uncertainty can have a significant impact on exploration, appraisal, and development project economics. This is particularly the case when estimating permeability and permeability thickness (kh) which are fundamental to predict robust productivity rates for wells. When using core to build log-based permeability models, many issues need to be addressed, some of which may be significant. This paper supports an assessment of other factors that may contribute to variability between core, model estimates and well performance, but is primarily focused on a rapid screening of irreducible uncertainty and sensitivity as opposed to rebuilding a new model. Quantifying the sensitivity and uncertainty of model estimates of permeability provides a framework to optimise reservoir management challenges and a workflow to screen for differences that contribute to sensitivity and ranges of permeability in individual cored wells and as a predictor in uncored wells. for setting expectations (irreducible uncertainty), estimating ranges for undrilled reservoir and infill targets identifying other factors that may be significant in well and reservoir performance supporting optimised reservoir management indicating the potential benefits and value of additional data acquisition.