Inverse analysis of hydraulic fracture geometry based on data-driven simulation and acoustic emissions

Abstract: Hydraulic fracturing process is invisible in geological media and is difficult to be observed. The acoustic emissions (AEs) or microseismic (MS) technologies are useful approaches to estimate the hydraulic fracturing process. However, the AEs or MS do not provide the geometry of hydraulic fractures directly, but only provides the coordinates of AEs/MS sources. Traditional analytical and statistical approaches of reproducing fracture geometry using AEs/MS sources are relatively empirical. In this work, we monitored the AEs induced by hydraulic fracturing experimentally. Then, data-driven simulation based inverse analysis approach is developed to estimate the fracture geometry according to AEs sources. The difference between hydraulic fractures and AEs sources is defined as objective function. Then, the in-situ stresses are found using the inverse analysis based on data-driven simulation. As shown in Fig. 2, the geometry of hydraulic fractures is reproduced using data-driven simulation and AEs technology.

Keywords: hydraulic fracturing, inverse analysis, data-driven simulation, acoustic emissions technology, AiFrac