Improved Petrophysical Property Evaluation of Shaly Sand Reservoirs Using Modified Grey Wolf Intelligence Algorithm

• Published in: Computational geosciences Vol. 27; no. 4; pp. 537 - 549
• Main Authors: Jia, Bao, Xian, Chenggang, Jia, Wenfeng, Su, Jianzheng
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 01.08.2023; Springer Nature B.V
• Subjects: Accuracy; Algorithms; Data logging; Earth and Environmental Science; Earth Sciences; Geotechnical Engineering & Applied Earth Sciences; Hydrogeology; Mathematical Modeling and Industrial Mathematics; Oil and gas fields; Oil fields; Oils & fats; Optimization; Original Paper; Phase volume fraction; Programming languages; Reservoirs; Sand; Software; Soil Science & Conservation; Source code; Water salinity; Water saturation; Shaly sand; Multi-objective optimization; Grey wolf optimization; Dielectric
• ISSN: 1420-0597; 1573-1499
• DOI: 10.1007/s10596-023-10217-2

Multi-frequency dielectric scanning logging is an advanced method that plays a critical role in evaluating unconventional oil and gas reserves and residual oil distribution. This method provides higher accuracy compared to conventional logging and can obtain essential formation parameters such as formation water salinity, pore textural index, and dispersive phase volume fraction. Despite its advantages, the inversion of permittivity and conductivity measurements at multiple frequencies into formation properties remains a "black box" problem. This challenge makes it challenging to understand specific implementation methods and inversion techniques without purchasing expensive software and hardware from oil field service companies. To address this issue, this work proposes a publicly available and advanced intelligent optimization algorithm. Our approach reduces calculation complexity and achieves high accuracy in evaluating petrophysical properties, including shaly sand reservoirs, without relying on costly software services from oil field companies. Our method offers distinct advantages over traditional approaches, such as the ability to derive formation properties from dielectric logging information with confidence, without specialized equipment or software from commercial providers. Additionally, the open-source code is readily available in various programming languages, including Python, R, and Matlab, making our approach accessible and easy to implement.