Clay formations are present in reservoirs and earthquake faults, but questions remain on their mechanical behavior, as they can vary from ductile (aseismic) to brittle (seismic). An experiment, at a scale of 10 m, aims to reactivate a natural fault by fluid pressure in shale materials. The injection area was surrounded by a dense monitoring network comprising pressure, deformation, and seismicity sensors, in a well‐characterized geological setting. Thirty‐two microseismic events were recorded during several injection phases in five different locations within the fault zone. Their computed magnitude ranged between −4.3 and −3.7. Their spatiotemporal distribution, compared with the measured displacement at the injection points, shows that most of the deformation induced by the injection is aseismic. Whether the seismicity is controlled by the fault architecture, mineralogy of fracture filling, fluid, and/or stress state is then discussed. The fault damage zone architecture and mineralogy are of crucial importance, as seismic slip mainly localizes on the sealed‐with‐calcite fractures which predominate in the fault damage zone. As no seismicity is observed in the close vicinity of the injection areas, the presence of fluid seems to prevent seismic slips. The fault core acts as an impermeable hydraulic barrier that favors fluid confinement and pressurization. Therefore, the seismic behavior seems to be strongly sensitive to the structural heterogeneity (including permeability) of the fault zone, which leads to a heterogeneous stress response to the pressurized volume.

The first microseismic monitoring operations of hydraulically stimulated wells were run in Bahrah and Sabriyah oilfields, Northeastern Kuwait. The main objective was to evaluate the capacity of the microseismic in optimising the fracturing process and consequently improving the production of these reservoirs. The major phases of such monitoring projects are sensors network design, deployment, acquisition, data processing, results delivery, and interpretation. Fit-for-purpose monitoring networks were designed by modelling the expected sensitivity and location accuracy of various sensors geometry scenarios, considering local reservoir properties. Geophones were deployed in observation wells nearby treatment wells to record the seismic waves emitted by the microearthquakes induced by the rock fracturing process. This seismicity was located and characterised to image the fracture networks growth under the effect of pumping. From this, fracture geometry parameters were assessed, stress and hazard characterised, unexpected behaviours were monitored and analysed. By providing information in real-time during rock stimulation operations, microseismic monitoring successfully helped improving production while maintaining a focus on the risk assessment indicators. In Bahrah, seismic response to the treatment was assessed for the target carbonate formation Mauddud, evaluating stimulation effectiveness while characterising unexpected and unwanted behaviours. In Sabriyah, fracture geometry estimates helped calibrating injection models and fine-tuning stimulation plans. Furthermore, a strong focus was also placed on monitoring hazard and anomalies in the Tuba carbonate formation being stimulated near a natural fault. Monitoring procedure, results and lessons learned from these projects can be transferred to other existing or upcoming wells to be drilled in the same formations, adding value to these reservoirs by optimising the fracture design, and making hydrocarbon recovery safer and more efficient. This paper reports on the first usage of microseismic monitoring in Bahrah and Sabriyah oilfields in Kuwait. Monitoring met the initial objectives and both the approach as well as results are now a baseline for the effective development of hydraulic stimulation in these reservoirs and others with similar characteristics.

No abstract