A study on the improvement of velocity model for microseismic event mapping using the modified Fresnel volume approach

Abstract

In multistage hydraulic fracture stimulation, enhancing the connectivity of a fracture network is very important. Hydraulic fractures can be tracked by locating microseismic events, which occur at each stage of the hydraulic fracture stimulation. In previous studies, microseismic monitoring has usually been performed with an initial velocity model, ignoring the alteration of velocity in the fracturing process. However, the velocity structure changes due to hydraulic fractures created during each stage. Thus, to perform accurate monitoring, the velocity alteration that occurred in the previous stage must be accounted for when locating microseismic events in the next stage. In this study, a velocity inversion algorithm was developed to update a velocity model by reflecting the velocity alteration that occurred during the previous hydraulic fracturing stage. To solve the problem of high computational cost when constructing a Jacobian matrix and to improve the efficiency of inversion, the modified Fresnel volume approach was adopted. In this approach, the inversion region is restricted to the zone where velocity alteration is expected due to hydraulic fracturing. The construction of a Jacobian matrix using this method not only significantly reduces the time required for computation but also alleviates the coverage problem of ray-based tomography. However, the altered velocity could not be correctly imaged because the source locations estimated in the previous stage contained errors and the velocity inversion could only update velocities in the Fresnel volume of the ray paths between sources and receivers. To determine the exact locations of microseismic events, e.g., when only P or S waves can be picked due to a low signal-to-noise ratio, a microseismic event mapping algorithm was developed using the station-pair double difference (DD) method. To assess the performance of the monitoring system, which consisted of event mapping and velocity inversion applications, the numerical experiments were conducted for the velocity model to simulate and enhanced/engineered geothermal system (EGS) field. Although the altered velocity was not perfectly inverted due to incorrect source locations, the velocity model updated by the newly developed algorithm well described the true velocity model. The positions of microseismic events relocated with the updated velocity model were closer to the true locations. In addition, aseismic zone, in which microseismic events did not occur but a velocity alteration was evident, were also imaged by the velocity inversion technique. Keywords: Event mapping, hydraulic fracturing, microseismic monitoring, modified Fresnel volume approach, velocity inversion