Trans-dimensional inversion of velocity model in downhole microseismic monitoring

Abstract:

In downhole microseismic monitoring, the velocity model plays a vital role in getting accurate event locations. In the geological environment, the sedimentary rocks are composed chiefly of the mineral grains. Many of them are in the form of fine layers. A reasonable velocity model is needed to describe the real condition. In this research, Bayesian inference is applied to update the velocity structure. Firstly, the Metropolis-Hasting Markov Chain Monte Carlo (MCMC) algorithm is used to solve the fixed dimensional inversion problem and proves that the layer depths should be updated to get better event location accuracy. Then the trans-dimensional inversion theory is introduced and the reversible jump MCMC (rjMCMC) algorithm helps get a more reasonable velocity structure. In the inversion process, the double-difference (DD) method does a lot of help in obtaining exact close event locations. These events are used as fixed sources and they provide full coverage of interest areas to help gain the accurate velocity model. This model can reflect the main and fine velocity structures and greatly reduce the event location errors.