Evaluation of an Advanced Multi-Fidelity Asset Modeling Workflow to Enhance Dry Gas Reservoir Management

Abstract The goal of this paper is to demonstrate the effectiveness of multi-fidelity integration using a new-generation high-resolution reservoir simulator and multiphase network flow simulators for faster and accurate production forecasts. Low-fidelity models, using decline curve representations of well behavior from the reservoir model and hydraulic tables-based modelling for the flowline network, are employed to quickly evaluate development concepts. These models are validated against full physics integrated models, ensuring enhanced accuracy and reliability in reservoir management. The workflow described in the next sections begins with constructing full-physics reservoir and network models. The reservoir model generates proxy decline-curves for all active wells, while the multiphase flow simulator is used to create hydraulic tables correlating flow parameters and pressure drops across the network. These low-fidelity models form the basis of a much faster integrated model, which, once validated against the full-physics model, can be used to rapidly assess concepts and asset development strategies. This method was implemented on a dry gas reservoir and led to an improved understanding of the reservoir dynamics and a better capture of the full asset physics while adhering to the operating constraints, as well as tremendous improvement in simulation performance for concepts validation using the integrated asset modelling multi-fidelity framework. It requires a one-time deployment, after which various scenarios can be rapidly evaluated and subsequently validated using a model that integrates the original high-fidelity framework. The proxies offer significant versatility, enabling their use in uncertainty and optimization workflows that span from seismic to network analysis. For illustrative purposes, in this case study, the impact of the compressor’s startup time, the rerouting of the pipeline to low-pressure, and various other what-if scenarios were assessed using this method. By streamlining the evaluation process and maintaining accuracy, this approach facilitates more efficient and comprehensive decision-making, ultimately enhancing the ability to optimize asset performance under varying conditions.