Optimization of fracturing stages/clusters in horizontal well based on unsupervised clustering of bottomhole mechanical specific energy on the bit

Abstract

Multi-cluster perforation and multi-staged fracturing of horizontal well is one of the main technologies in volumetric fracturing stimulation of unconventional oil and gas reservoirs, but unconventional reservoirs in China are generally of strong heterogeneity, which causes different fracture initiation pressures in different positions of lateral, making it difficult to ensure the balanced fracture initiation and propagation between clusters in multi-cluster perforating. It is in urgent need to precisely evaluate the difference in rock strength in lateral and determine the well section with similar rock strength to deploy fractures, so as to reach the goal of balanced stimulation. Based on the drilling and logging data, this paper establishes an unsupervised clustering model of mechanical specific energy of bit at the bottomhole the lateral. Then, the influence of drill string friction, composite drilling and jet-assisted rock breaking on the mechanical specific energy is analyzed, and the distribution and clustering categories of bottomhole mechanical specific energy with decimeter spatial resolution are obtained. Finally, a fracture deployment optimization method for horizontal well volumetric fracturing aiming balanced stimulation is developed by comprehensively considering inter-fracture interference, casing collar position, plug position, and clustering result of bottomhole mechanical specific energy. The following results are obtained. First, compared with brittleness index, Poisson's ratio and stress difference, perforation erosion area is in a stronger correlation with the mechanical specific energy, and the mechanical specific energy can effectively characterize the difference in the amount of proppant injected into the perforation clusters in the lateral, so it can be served as one of the important indicators for the selection of fracture deployment position. Second, the drilling and logging data cleaning and smoothing and the clustering number selection by the elbow method are the key steps to obtain the clustering results of bottomhole mechanical specific energy, which can tell the difference in the mechanical specific energy with decimeter-level resolution. Third, the interval with mechanical specific energy within 10% of the average value in the section is selected for deploying perforation clusters, and the compiled computer algorithm can automatically determine the optimal position of fracturing section and cluster, so as to realize the differential design of stage spacing and cluster spacing. In conclusion, the research results can further improve the fractures deployment efficiency and balanced stimulation of volumetric fracturing in unconventional oil and gas reservoirs, and this technology is expected to provide ideas and new methods for the fracture deployment optimization of horizontal well volumetric fracturing in unconventional oil and gas reservoirs.