Monitoring and Characterization of Gas Migration in Oil-Based Mud Using Fiber-Optic DAS and DTS

Abstract

Summary Understanding gas dynamics in mud is essential for planning well control operations, improving the reliability of riser gas handling procedures, and optimizing drilling techniques, such as the pressurized mud cap drilling (PMCD) method. However, gas rise behavior in mud is not fully understood due to the inability to create an experimental setup that approximates gas migration at full-scale annular conditions. As a result, there is a discrepancy between the gas migration velocities observed in the field as compared to analytical estimates. This study bridges this gap by using distributed fiber-optic sensors (DFOS) for in-situ monitoring and analysis of gas dynamics in mud at the well scale. DFOS offers a paradigm shift for monitoring applications by providing real-time measurements along the entire length of the installed fiber at high spatial and temporal resolution. Thus, it can enable in-situ monitoring of the dynamic events in the entire wellbore, which may not be fully captured using discrete gauges. This study is the first well-scale investigation of gas migration dynamics in oil-based mud with solids, using optical fiber-based distributed acoustic sensing (DAS) and distributed temperature sensing (DTS). Four multiphase flow experiments conducted in a 5,163-ft-deep wellbore with oil-based mud and nitrogen at different gas injection rates and bottomhole pressure conditions are analyzed. The presence of solids in the mud increased the background noise in the acquired DFOS measurements, thereby necessitating the development and deployment of novel time- and frequency-domain signal processing techniques to clearly visualize the gas signature and minimize the background noise. Gas rise velocities estimated independently using DAS and DTS showed good agreement with the gas velocity estimated using downhole pressure gauges.

Machine-Learning-Assisted Leak Detection Using Distributed Temperature and Acoustic Sensors

“…Due to the vast differences in accuracies between Bands, frequency analysis should be performed to determine the best Bands prior to leak monitoring in a real application during an initial fiber response calibration process. This can be done using spectrum and spectrogram analysis, as demonstrated in previous studies by the authors [52,53]. The frequency ranges corresponding to the leak and the noise may not be the same between pipelines and fiber deployments, as each pipeline will have different flow rates, types of transported fluids, diameter of pipe, thickness of pipe, type of pipe material, sources of noise, fiber installation method, fiber material, etc.…”