A Digital Framework for Fast-track Confidence Mapping and Integration of Multiple Uncertainty Sources in Reservoir Modeling

Reservoir modeling involves multiple uncertainty sources, and their integration is both challenging and critical for subsurface operations. Thus, we propose a method and the corresponding digital framework to rapidly build confidence maps in reservoir models. These confidence maps help to understand how uncertainties are spatially distributed and to appraise risks in specific areas.

Our approach leverages Multiple-Criteria Decision Analysis (MCDA) techniques to combine user-controlled and knowledge-based inputs with a standardized numerical process. This multiple-criteria integration relies on a knowledge graph that formalizes the list of uncertain components and their relationships. Each component is then assessed to generate an uncertainty index and a map, using simple data-driven recipes. Finally, these maps and indices are aggregated in a single confidence map, with a knowledge-based procedure. This aggregation exploits the Analytical Hierarchy Process (AHP), an MCDA technique consisting in pairwise comparisons with a predefined scale.

In a concrete case study, we apply this method to map the confidence associated with the Gross Rock Volume (GRV) of a subsurface reservoir, integrating uncertainties from seismic data, well markers, well-to-seismic tie, velocity models and interpreted horizons. This example illustrates the consistency of the output as well as the efficiency and versality of the digital framework.