An Edge-Cloud Sensor-Agnostic Framework for Real-Time Microseismic Monitoring and Analysis

Induced seismicity is a critical diagnostic tool for monitoring subsurface processes in a variety of energy applications, including carbon capture and storage (CCS), hydraulic fracturing, geothermal energy, and enhanced oil recovery. In the context of CO₂ injection, microseismic monitoring enables the detection of acoustic emissions generated by rock failure, providing real-time insights into reservoir dynamics, plume migration, and the integrity of both the reservoir and cap rock. These microseismic events serve as sensitive indicators of stress changes near the wellbore and within the reservoir, offering a means to identify and characterize faults or fracture zones that may be activated during injection. This information is essential for assessing containment risks and constraining the maximum expected magnitude of induced seismic events. This paper presents a comprehensive workflow for passive seismic monitoring using Distributed Acoustic Sensing (DAS), a fiber-optic-based technology that is gaining traction in the oil and gas industry due to its deployment flexibility, cost-effectiveness, and ability to transform existing fiber infrastructure into dense seismic arrays. DAS enables high-resolution, large-aperture monitoring systems that can be deployed temporarily via wireline or permanently as part of well completions. Building on extensive experience from 3C sensor-based microseismic campaigns, we propose a DAS-centric monitoring framework designed to enhance event detection, characterization, and long-term monitoring capabilities—particularly in support of Measurement, Monitoring, and Verification (MMV) programs. The workflow begins with the optimization of monitoring geometry through the evaluation of event detectability and location uncertainty across various sensor configurations. A detailed velocity model is developed, incorporating geological, geophysical, and geomechanical parameters, including rock attenuation and ambient noise characteristics. Where available, geomechanical data are used to model expected source mechanisms and avoid nodal planes, improving the reliability of event detection. Waveform simulations for various failure mechanisms are used to benchmark expected seismic signatures. A real-time DAS data simulator is employed to validate the acquisition geometry and processing pipeline under realistic operational conditions. To automate event detection, we introduce a novel real-time processing framework that combines traditional geophysical techniques with image processing algorithms, emulating expert interpretation of DAS-acquired data. The integrated workflow supports both conventional 3C and fiber-optic-based systems, reducing uncertainties in event localization and source mechanism analysis. This workflow is effective in monitoring CO₂ injection wells over extended periods, ensuring safe operations and regulatory compliance. The methodology is equally applicable to hydraulic fracturing, geothermal energy production, and long-term passive monitoring in oil and gas fields, offering a robust, scalable solution for real-time subsurface surveillance.