Quantitative analysis of distributed acoustic sensing data for time-lapse vertical seismic profiling and multiphase flow monitoring

Abstract

Distributed acoustic sensing (DAS) technology has proven to be extremely valuable in oil and gas industry applications, such as seismic imaging, completion monitoring, and production monitoring. However, quantitative analysis of DAS data is necessary to realize the full potential of the technology. Furthermore, the benefit of DAS technology comes from multiple usages of the same fiber. Hence, this thesis aims to advance the acquisition, processing, and interpretation of DAS data for the following applications in different stages of a well life cycle: time-lapse vertical seismic profiling (VSP) during completion operations, and multiphase flow monitoring during production.

In the first part of the thesis, we developed a processing workflow to quantitatively analyze unique time-lapse DAS VSP datasets acquired after each hydraulic fracturing stage. This workflow inverts properties of transmitted PS-waves observed in the field data to the geometrical and physical characteristics of the stimulated rock volume (SRV). For each VSP dataset, the PS-wave is generated due to the interaction of the incident P-wave from the surface seismic source with the SRV, which can be represented as a temporary low-velocity anomaly. We found that the analysis of the PS-waves can constrain SRV height, SRV width, and fracture closure time. First, we applied the developed workflow in a single well completion case and then validated this workflow with 2D finite-difference time-domain (FDTD) modeling. Next, we extended it to a case of zipper-fracturing stimulation, when multiple wells are completed simultaneously. Finally, we showed and validated with 3D FDTD modeling that the previously developed workflow can be used to invert SRV properties in a fiber well and adjacent wells. Our work provides critical information to optimize hydraulic fracturing operations.

We then used laboratory experiments to advance DAS acquisition and analysis for multiphase flow monitoring, focusing on a slug flow, one of the most common flow types in both wells and surface facilities. The slug flow consists of two phases — a gas phase Taylor bubble and a liquid phase slug. Such flow type can be damaging to facilities; consequently, active flow monitoring is essential and can reduce operational costs through proactive facility management. We built a vertical flow loop facility equipped with air/water inlets, a thermal injection apparatus, and a wrapped single-mode fiber to research slug flow. The flow loop allows the study of acoustic and thermal energy propagation within the slug flow using DAS. After data processing, two moveouts were observed in the data, one associated with the thermal slug, propagating at water velocity, and another with Taylor bubbles, propagating at air velocity. Therefore, by applying our workflow, we demonstrate that velocities of both liquid and gas phases can be measured. Furthermore, the same workflow can be applied to field data.

As a next step, we modified the flow loop and added a special bubble-generating apparatus, which allowed us to study the rising Taylor bubbles in a controlled environment. We explored how DAS response changed with bubble size. We confirmed that no velocity change is observed due to bubble size difference, and DAS can be used as a pressure gradient sensor to estimate the size of Taylor bubbles. Additionally, we demonstrated that DAS is a valuable laboratory tool because it captures spatially distributed processes, such as Taylor bubbles merging, in more detail than is possible with conventional point sensors.

The thesis shows that DAS is a unique tool for reservoir monitoring. Time-lapse VSP allows for characterization of SRV during well stimulation, including the case of multiwell completions. Later in the life of the well, the same fiber cable can be used for flow monitoring, which was studied in a controlled laboratory environment for the slug flow case.