Transdimensional downhole velocity optimization by incremental pseudo-master method

Abstract

The velocity model in downhole microseismic monitoring plays a vital role in accurate mapping of the hydraulic fracturing image, but the conventional velocity optimization method by the perforation shot introduces non-negligible event location errors when the geological formations are complex with inclinations or pinch-outs. We exploit the advantages of the reversible jump Markov Chain Monte Carlo (rjMCMC) approach in generating different dimensions of velocity models and propose a transdimensional velocity optimization approach called incremental pseudo-master (IPM) method to improve microseismic event location accuracy. A master-event location method is utilized to find the accurate adjacent event locations, and then these events are taken as pseudomasters to update the velocity structure by Bayesian inference. With progressive wider coverage of pseudomasters, the velocity model is continuously optimized for improving the whole event location accuracy. The synthetic and field tests show that the IPM method has higher event location accuracy than the conventional method.