Imaging source cloud in hydraulic fracturing using wavefield back-propagation methods

Investigation of the source cloud induced by the hydraulicfracturing can provide important information for evaluating and monitoring the unconventional reservoirs. It is challenging to image the source cloud using the conventional arrival-picking-based methods, which appear large uncertainty and error-prone for numerous weak signals induced by the fracking. To improve the imaging quality of the source cloud, we use the wavefield backpropagation methods, including both Arithmetic-mean Reverse-Time Migration (AmRTM) and time-window autocorrelation Reverse-Time Migration (RTM). These methods are based on the full wave equation, which provides a better accuracy over the high-frequency assumption based traveltime approaches. Additionally, there is no requirement of arrival picking in the wavefield back-propagation methods, which helps reduce the uncertainty of source locations. With the generated source cloud imaging, the fractured volume could be estimated. We set up several experiments to indicate the effectiveness of the source cloud imaging using the wavefield backpropagation methods, considering the influence of natural fractures, random noise, and source onset time.