ML-based Crude Oil Mud Contamination Detection Using EIS Technology
Please login to view abstract download link
Oil-based muds (OBM) are commonly used in drilling operations for their lubrication properties leading to faster operations, and their ability to enhance wellbore stability. OBM can and will invade rock formations during drilling operations to contaminate crude oil reservoir resulting in degraded quality and requiring treatment to separate. Electrochemical impedance spectroscopy (EIS) is a nondestructive technique that applies alternating current across a range of frequencies to measure complex impedance responses. These responses can be analyzed to provide multiple parameters that can be used to differentiate clean crude oil from OBM-contaminated samples. In this study, we conducted laboratory testing using EIS at different frequencies on both clean and contaminated crude oil, under different temperature conditions and with different OBM contaminants. Based on the experimental data, we train a machine learning (ML) model to accurately classify crude oil samples and predict the percentage of OBM contaminant. Our approach aims to enable real-time contamination assessment, thereby improving efficiency and productivity by reducing the time required for formation testing.
