Unconventional Measurements Bootstrap Reservoir Simulation

Abstract

Abstract Production history matching data is an important step in any study that seeks to optimize unconventional completions and well development criteria. Understanding the reservoir mechanisms during production allows for better optimization of the hydraulic fracture system. Generating a model that fits historical data can be easy but honoring the true petrophysics and fluid dynamics of the reservoir is often challenging. Some of the major challenges during reservoir simulation are uncertainties in water saturation, permeability, phase behavior, and effective fracture surface area during production. This paper discusses how fit-for-purpose core measurements help reduce the uncertainty in these parameters, ultimately requiring a multiple porosity reservoir simulation model to account for these improved measurements and understandings. Industry accepted core analysis techniques under-estimate reservoir water saturation due to loss of water from evaporation and core handling techniques (preservation, crushing, time). Proper evaluation of the void space will be shown and how this is better calibrated to field data. A review of how steady-state liquid permeability testing provides better estimates for reservoir permeability and deliverability in shale reservoirs will be discussed. Coupling these measurements with imbibition effects from hydraulic fracturing fluids and lab studies showing oil-wet and water-wet pore systems acting independently of each other, a slightly "outside of the box" reservoir simulation model was needed to mimic these physics. The proposed reservoir simulation methodology consists of multiple porosities and was developed to incorporate near-wellbore hydraulic fracture effects that are observed during lab testing. Combining this methodology with other lab measurements and a fully three-dimensional (3D) hydraulic fracture model, the number of "knobs" that need to be turned to get a good history match are reduced. Two examples will be presented in this paper showing how the proposed model better honors the physics of lab measurements and provides the user more flexibility during reservoir simulation, especially when buildup data is available. Reducing the uncertainty in these parameters has provided a workflow that helps minimize the multiple non-unique realizations during the history match process and provides a more reliable model for the engineer while reducing the amount of time needed to obtain a match.