Locating microseismic events with time reversed acoustics: A synthetic case study

Abstract

We investigated the applicability of using the Time Reversed Acoustics (TRA) technique, and thus the whole waveform of the recorded signal, to locate microseismic events in a reservoir monitoring system (Lu, 2007). Reservoir monitoring has attracted a lot of attention over the last twenty years. One of the challenges in reservoir monitoring is locating the microseismic events that result from fluid injection and fracturing processes. Conventionally, a seismic event is located using the arrival times of P and S phases. The picking of arrival times could potentially be very time‐consuming and difficult in certain situations, especially when the passive seismic events are continuously recorded by the monitoring systems. The retro‐focusing feature of TRA can be applied to such a location problem and might provide an opportunity to locate events using the whole waveforms without picking different arrival phases. The basic concept involved in TRA is the fundamental symmetry of time reversal invariance. In the TRA approach, we first record the full seismograms at an array of stations. The traces are then time‐reversed and numerically sent back into the medium at those station locations assuming an a priori velocity model. The wavefield of the back‐propagation is tracked and, at the appropriate time, energy concentrates at the focal locations of each event. The TRA technique is particularly amenable at reservoir scale, in that a detailed subsurface velocity structure is usually available. We investigate TRA's retrofocusing properties by experimenting with different receiver coverage on a 3D elastic reservoir model using full elastic finite difference simulations. The advancements in modern reservoir networks can provide a wide receiver coverage that can potentially make TRA practical. In this numerical study, we try to address issues involved in the implementation of the TRA technique in reservoir monitoring projects, such as what kind of monitoring network is suitable for good focusing, how sparse can the stations in a network be, and what is the focal resolution in different setups, etc.