Probabilistic Approach for Estimating Reservoir Quality and Calculating Water Saturation: Montney Formation, western Canada
Reservoir quality is primarily determined by porosity, permeability, and water saturation (Sw). Accurate determination of water saturation is important both for reserves estimation and because Sw dictates the relative permeabilities of formation fluids. Sw can be measured directly or calculated indirectly using different methods (well-logs analysis or mercury injection data for a single well, with limited ability to propagate results across a production field or identify compositional controls on fluids distribution in the rock. We present a novel method for calculating Sw through matrix and fluids mass-balance using the software GAMLS©. Our model incorporates fluid properties, mineralogical composition and reservoir properties for the Montney Formation in the Pouce Coupe Field of western Canada. The data set includes multiple wells in the Pouce Coupe area, some of which are partially cored through the Montney interval, core analyses (QEMSCAN mineralogy, porosity, permeability measurements, total organic carbon analyses), and full-diameter-core water saturation measurements. One well with a full data set was selected as the model well, while other wells were used for verifying calibration parameters. GAMLS© uses a probabilistic multi-variate cluster analysis of well-logs to identify rock types end-members (electroclasses). During clustering, each digitized sample is probabilistically assigned to one or more end-members, permitting a more refined calibration of the model to core analysis data than is possible in most other available clustering procedures. Next, we used GAMLS© to generate log-scale mineralogical abundances curves for the model well, calibrated to QEMSCAN mineralogical analyses. In addition, we computed bound and free water saturation for this well through matrix-fluid mass-balance. This mass-balance relays on matrix and pore fluids hydrogen content, and the readings of the neutron-porosity well-log. Results show a consistent and correlateable pattern of end-members across the field, allowing for correlation between wells based on physical rock properties. Modeled mineralogy, and computed total and bound water saturation fit core measurements in the model well and the validation wells, and are consistent with previously published data for the Montney Formation. This probabilistic model allows for better estimations of hydrocarbons reserves, and permits more credible decision-making when targeting high reservoir quality zones.
AAPG Datapages/Search and Discovery Article #90350 © 2019 AAPG Annual Convention and Exhibition, San Antonio, Texas, May 19-22, 2019