Simultaneous Bayesian inversion for effective anisotropic parameters and microseismic event locations: A physical modelling study

Hongliang Zhang, Jan Dettmer, Joe Wong, Kristopher A. Innanen

To account for anisotropy caused by the presence of a set of aligned vertical fractures in a finely horizontally layered background medium, we present an inversion procedure to simultaneously estimate microseismic event locations and the parameters of an orthorhombic (ORT) anisotropic medium property model. The procedure employs Bayesian inference via Markov-chain Monte Carlo (McMC) sampling with parallel tempering, and principal component diminishing adaptation, to ensure efficient sampling of the parameter space. This approach provides a nonlinear uncertainty quantification, by approximating the posterior probability density with an ensemble of model-parameter sets for effective anisotropic parameters, microseismic event locations and horizontal locations of perforation shots. The noise standard deviation is also treated as an extra unknown in the inversion. To investigate the effects of model simplification, e.g., neglect of horizontal-layering induced vertical transverse isotropy (VTI), we also consider a simpler, horizontal transverse isotropic (HTI), parametrization. The inversion is carried out for simulated data and data from a seismic physical laboratory model. Results suggest that, for field microseismic data processing, neglect of VTI signal caused by horizontal layering in fractured reservoirs may lead to systematic errors in microseismic event locations. Synthetic experiments further demonstrate that the acquisition geometry significantly impacts the resolution of event origin times, event depths, and effective velocity parameters. In addition, resolving these parameters requires an aperture size which may not be practical for field monitoring. Finally, we demonstrate that precise perforation-shot timing information and the incorporation of a vertical downhole array into the small-aperture surface array both reduce the requirement for large array apertures.