A Data Driven Approach to Optimizing Gas Reservoir Management Through Machine Learning and Phase Behavior Analysis of Production Data

Abstract

The significance of reservoir management has been recognized by the Oil and Gas industry as a strategy for maximizing resource recovery while minimizing investment. Fowler et al (1996) identify reservoir management as a tool to manage risks while optimizing profitability. Fanchi (2018) points out that successful reservoir management requires an understanding of the reservoir structure, the distribution of fluids within the reservoir, drilling and maintaining wells which can produce fluids from the reservoir, transport and processing of produced fluids, refining and marketing the fluids, safely abandoning the reservoir when it can no longer produce, and mitigating the environmental impact of operations throughout the life cycle of the reservoir. Wiggins et al (1990) underline the three primary components of reservoir management as: Knowledge about the entity being managedManagement environmentAvailable technology Thakur (1996) highlights the increasing role of data in reservoir management activities along with the need for comprehensive, cost-effective surveillance with surveillance management programs and well-planned data-collection with data-collection management programs.