Integrated hydro-mechanical and seismic modelling of the Valhall reservoir: A case study of predicting subsidence, AVOA and microseismicity

Abstract

Geomechanical, fluid-flow and seismic modelling have been combined to predict surface subsidence, seismic anisotropy and microseismicity for the Valhall reservoir, North Sea. The constitutive model used in the geomechanical simulation consists primarily of layers having poro-elastic behaviour, but with poro-elasto-plasticity behaviour in the chalk reservoir units. The constitutive model incorporates matrix deformation during simulation, such that areas of compaction and dilation are modelled so that the likely microseismic response of the reservoir can be predicted. In the coupled fluid-flow and geomechanical (hydro-mechanical) workflow, a finite-element geomechanical simulator is coupled to a reservoir fluid-flow simulator and applied to predict seafloor subsidence. Subsequently, the history-matched hydro-mechanical results are transformed into dynamic elastic models suitable for seismic analysis using an empirical static-to-dynamic relationship and stress-dependent rock physics model. The elastic models are then used to predict seismic anisotropy and microseismicity, allowing for an additional assessment of hydro-mechanical simulation via comparison with observed field seismic data. The geomechanical model has been calibrated to reproduce the measured subsidence. Furthermore, the predicted seismic anisotropy extracted from the reflection amplitude variation with offset and azimuth resembles that measured from field seismic data, despite the limited calibration of the rock physics model to the Valhall reservoir rocks. The spatial pattern of modelled microseismicity is consistent with previously published microseismic analyses, where the modelled failure mechanisms are consistent with typical production-induced seismicity. The results of this study indicate that seismic data has the potential to improve the calibration of hydro-mechanical models beyond what is possible from conventional fluid production and surface subsidence data. This is significant as seismic data could provide greater control over the whole field rather than borehole and surface measurements.