Hybrid Workflows- A History of Tomorrow

Abstract

This study provides an overview of the evolution of technology in the reservoir modeling field, discussing past, present, and future advancements. It explores how technologies such as central and graphics processing units have enabled engineers to increase the number of simulation grids from 500,000 to 1 trillion cells over the last 20 years. The paper also examines data-driven workflows that have accelerated the reservoir modeling process to a few weeks, highlighting their limitations. To address these limitations, the study introduces hybrid physics-guided data-driven models as a modern digital tool for mapping potential remaining oil in brownfields. These hybrid workflows combine the strengths of both physics-guided and data-driven models, overcoming the weaknesses of each approach. Two case studies, located in the North Sea and the Middle East, are reviewed to demonstrate the effectiveness of these hybrid workflows. The comparison of the water saturation map generated by the hybrid workflow with the contact movement calculated from 4D seismic data in the North Sea field shows a strong correlation, indicating a reliable match between the two maps. In the Middle East case study, the performance of four infill wells calculated by the hybrid workflow is evaluated against the actual outcomes of post-drilling activities. The results show that, except for one of wells, the actual outcomes align with the predicted trends, falling within the P10-P90 band. The paper emphasizes the novelty and benefits of using hybrid workflows, as they enable the generation of potential remaining oil maps in a few days—a task that is challenging with conventional numerical simulation workflows. These hybrid workflows offer significant time and cost savings while maintaining high fidelity. The advanced technologies discussed in this study facilitate the implementation of practical digitalization strategies across all disciplines in the industry, ushering in a new era of data-driven innovation.