The intelligent optimization of perforation cluster locations incorporating the fiber optics monitoring results

Abstract

During horizontal well multi-stage fracturing (HWMF), superfractures are often identified. To promote the uniform propagation of multiple fractures, it is necessary to finely optimize the perforation cluster locations based on the geological and engineering parameters. This work proposes an efficient method to design the perforation cluster locations in consideration of the geoengineering sweet spots with similar mechanical properties. Well log data and the precise fiber optics (FO) monitoring results are combined to find the main influencing factors. The principal component is conducted by introducing correlation analysis and Random Forest. Moreover, the K-means++ clustering method is used to evaluate reservoir quality. The fracturing sweet index (FSI) is proposed to measure the fracturing performance of each category quantitatively. The proposed workflow is effectively validated by two production scenarios. Moreover, the workflow can automatically evaluate reservoir quality based on intelligent clustering methods. Compared with the original design, the updated design lowers the gap among multiple fractures within one stage and increases the well production by 20%–50%. This work is beneficial for the on-site treatment of its feasibility and generalizability.