This talk will focus on “old problems and new solutions” in borehole geophysics with machine learning (ML) from data acquisition, quality control, pre-processing, inversion, interpretation, and multi-well studies. Some of these can be achieved with ML as an automation tool to improve the efficiency and some can be done with ML as a discovery tool to enhance the quality. Data sets for Labels, training and testing are critical in the applications. I will start with a deep review on the basic problems and model-driven methods in well log analysis. Then transfer the processing to machine learning with data-driven methods, and finally will discuss methods driven by both model and data with examples.
Borehole geophysics, or well logging in the oil and gas industry, provides first-hand information for discovery, understanding and recovery of underground oil and gas. Its application arises from the initial stage of oil and gas field exploration to the determination of oil and gas layers and storage parameters, the estimation of reserves, to the formulation of completion programs such as perforation and fracturing, to producibility forecasting, casing testing and cementing evaluation, as well as production dynamic monitoring. From a data scientist’s point of view, well logging is the technology of “collecting, transmitting, processing, interpreting and applying data” in boreholes. All links includingthe entire process are possible to realize and upgrade through the application of theories and methods of machine learning.
Speaker Bio
Prof. Lizhi Xiao
Lizhi Xiao, Ph.D and Professor, has been working for the oil industry for more than 38 years in both the oil fields and the higher education institutions, including worked at Western Atlas International and Halliburton Company. He is currently the Dean of the School of Artificial Intelligence at China University of Petroleum Beijing. He holds visiting position at Harvard Paulson School of Engineering and Applied Sciences, and he has received Technical Achievement Award from the Society of Petrophysicists and Well Logging Analysists and awards from other academic organizations.