Abstract

Various measurement, monitoring and verification (MMV) techniques are used to ensure the safety of operations in petroleum projects. One of these techniques is microseismic (MS) monitoring of the caprock. A challenge in MS monitoring of caprock is that unlike projects such as tunnelling in hard rocks, the number of recorded events are much smaller and also wide spread through the rock mass which makes interpretation of the physical meaning behind the events very difficult.

In this paper, a numerical technique is introduced and validated for coupling a finite difference code, FLAC3D (Fast Lagrangian Analysis of Continua), with a discrete element code, PFC3D (Particle Flow Code). Then, a similar approach is adopted for modeling microseismic events in the caprock. For this purpose, the underburden, reservoir and overburden layers are modeled by FLAC3D. A rather small section of caprock above the reservoir is nulled in the FLAC3D model and substituted by a PFC3D inclusion. The reservoir pressures are then applied to the reservoir section of FLAC3D. Once the FLAC3D cycles, there will be induced pressures in the caprock due to its undrained response to the change in pressures within the reservoir. These induced pressures are transferred to PFC3D as boundary velocities. Then the PFC3D cycles and sends back the boundary forces to FLAC3D. This back and forth process continues until equilibrium is reached. During this process, there will be bond breakages within the PFC3D inclusion that can be regarded as MS events. An algorithm for recording the microseismic energies based on a change in local strains is also implemented. The results are then compared with real data. The effect of injection fluid temperature is also studied. No significant appearance of microseismic events was observed due to injection of CO2 in both cases (i.e. with or without considering temperature). However, this continuum/discontinuum approach seems promising for understanding the microseismic response of caprock in more detail and also using the MS data for validation of geomechanical models.

This content is only available via PDF.
You can access this article if you purchase or spend a download.