Learning Across Wells: Artificial Intelligence for Missing Resistivity Log Reconstruction
Northside Houston
Speaker:
Seminar Date: May 07 2026
Registration Opens: Apr 08 2026 - May 07 2026
Time: 11:30 AM - 01:00 PM (US CDT)
Admission/Registration Link: None
Donation Link: None
Meeting/Webinar Link: None
Contact: Ali Eghbali (VP Northside, SPWLA Houston Chapter)
Corresponding: vpnorthside@spwla-houston.org
Fees: FREENOTES:
Speaker : Dr. Behzad Ghanbarian, Director of iResearchE3 Lab, UT Arlington
Date : Thursday, May 7th, 2026
Time : 11:45 – 1:00 pm (US CDT)
Venue : Virtual
Admission : This event is a free online event.The link to the meeting will be sent to registered individuals on May 6th.
Please ensure you register below using a valid email address.
SPWLA Contact : Ali Eghbali
Corresponding : vpnorthside@spwla-houston.org;
ABSTRACT:
Resistivity logs are widely used in petrophysics to estimate how much oil, gas, or water is present in a reservoir. However, these logs are not always available because they may not have been recorded in some wells due to cost or operational constraints. This study presents an artificial intelligence approach that reconstructs missing resistivity logs by learning from both the available measurements in a well and information from nearby wells. The proposed model analyzes common well logs such as gamma ray and density porosity from the target well and combines them with similar logs and resistivity measurements from neighboring wells. These wells are connected in a data-driven network that allows the model to learn how geological properties vary both vertically within a well and laterally across multiple wells. The method was tested using data from 142 wells in the Groningen gas field. Results show that incorporating information from surrounding wells significantly improves predictions compared with conventional deep learning methods that treat each well independently. In fact, the new approach reduces prediction errors by more than 40%. Overall, the results demonstrate that learning relationships between wells can successfully recover missing resistivity logs, even when no resistivity measurements are available in the target well. This approach provides a practical and scalable tool for improving reservoir characterization when well log data is incomplete.
BIOGRAPHY:
Behzad Ghanbarian is an Associate Professor in the Department of Earth and Environmental Sciences at the University of Texas at Arlington and the Director of the iResearchE3 Lab. His research focuses on applying modern techniques in artificial intelligence and data science to address complex real-world challenges in energy, environmental systems, and industry. Behzad has authored more than 150 peer-reviewed journal articles and three books. He currently serves as President of the SPWLA Dallas Chapter and is an active member of the American Geophysical Union (AGU), Society of Petrophysicists and Well Log Analysts (SPWLA), and Society of Petroleum Engineers (SPE). Behzad received the Donald L. Turcotte Award in Nonlinear Geophysics from the American Geophysical Union in 2015 and has been recognized among the top 2% of scientists worldwide (Stanford/Elsevier ranking) from 2021 to 2025.