Author ORCID Identifier

https://orcid.org/0009-0006-8958-3354

Semester

Fall

Date of Graduation

2023

Document Type

Dissertation

Degree Type

PhD

College

Statler College of Engineering and Mineral Resources

Department

Petroleum and Natural Gas Engineering

Committee Chair

Shahab Mohaghegh

Committee Member

Samuel Ameri

Committee Member

Kashy Aminian

Committee Member

Mohamed El Sgher

Committee Member

Qingqing Huang

Abstract

This dissertation addresses the limitations of conventional numerical reservoir simulation techniques in the context of unconventional shale plays and proposes the use of data-driven artificial intelligence (AI) models as a promising alternative. Traditional methods, while providing valuable insights, often rely on simplifying assumptions and are constrained by time, resources, and data quality. The research leverages AI models to handle the complexities of shale behavior more effectively, facilitating accurate predictions and optimizations with less resource expenditure.

Two specific methodologies are investigated for this purpose: traditional numerical reservoir simulations using Computer Modelling Group's GEM reservoir simulation software, and an AI-based Shale Analytics approach using IMPROVE™ software from Intelligent Solutions, Inc. The investigation covers the impact of key parameters on production prediction, assumptions made, predictive accuracy, data requirements, workflow complexity, and time efficiency.

By comparing these methods, the research aims to offer guidelines for incorporating AI models into reservoir simulation and identify areas for increased efficiency and accuracy. The study concludes by presenting recommendations to advance the field of reservoir simulation and encourage the adoption of innovative methodologies in the energy industry. The results are anticipated to considerably enhance reservoir simulation processes and optimize production strategies for unconventional shale plays.

Download

Share

COinS