Integrated History Matching Workflow to Parameterize and Characterize the Uncertainty Features of a Carbonate Reservoir

Abstract

Abstract This paper details the history matching workflow that was used in calibrating and hence parameterizing the key uncertainties in a carbonate reservoir. Multiple uncertainties in the reservoir model, inability to draw conclusions from data obtained, and how to effectively integrate data from different sources (cores, logs, geological model, well test and production data) were the challenges. Steps were taken to overcome these challenges, to recognize the key uncertainty parameters, using an integrated workflow approach resulted in improved well and reservoir characterization. The details of this workflow are presented in this paper. This paper provides an integrated workflow for dynamic reservoir characterization and improved matching of reservoir behavior. It involves:Analyzing and incorporating dynamic data (well tests, pressure transient interpretation, saturation logs and pressure versus depth),Parameterizing key uncertainties impacting dynamic well and reservoir behavior,Characterizing the key uncertainties and incorporating them into the dynamic model, andValidating the dynamic model with geological information to corroborate the uncertainty findings. This integrated workflow provided an uncertainty parameterization loop by enabling identifying the key uncertainty parameters that have the biggest impact on the dynamic behavior. This then enabled to narrow down the focus on these key parameters from a multitude of different factors. In carbonate reservoirs the dynamic history matching with the aim of understanding and modeling the key parameters through integrating existing dynamic information is not well defined. This workflow enables this integration and identifying the key parameters that can then be integrated into the modeling of the reservoir for improving the model behavior and predictability. An integrated approach and workflow towards reservoir characterization and modeling resulting in improved well and reservoir modeling is presented. Data integration from multiple sources, identification of key uncertain reservoir parameters and their impact on the modeling and production prediction are highlighted.