Clustering-based stress inversion from focal mechanisms in microseismic monitoring of hydrofracturing

In microseismic monitoring of hydrofracturing, the stress state around the wellbore determines the occurrence and extension of the hydraulic fractures when repetitive fracturing is taking place. A number of methods have been proposed to invert stress state. However, on account of the micromesh region scale and complex distribution form of the hydraulic fractures, it is difficult to obtain the ideal stress models that can satisfy a set of focal mechanisms in a research region. For this reason, we sufficiently discussed the mechanism characteristics and tried to make use of them to improve stress inversion results. After evaluating the partition and choosing the optimal number of clusters, a fuzzy clustering algorithm is first applied to classify the microseismic events into several clusters by taking advantage of the directions of P axes, and then the inversion of stress state from focal mechanisms is implemented for each cluster, respectively. As a result, stress models with smaller-fit angles are obtained, proving the validity of this inversion scheme. In addition, the stress regime categories and propagation of hydraulic fractures are analysed using the inverted stress models, and it is more accurate in estimating the stress state of a research region.