Studying the direction of hydraulic fracture in carbonate reservoirs: Using machine learning to determine reservoir pressure

Abstract

Hydraulic fracturing (HF) is an effective way to intensify oil production, which is currently widely used in various conditions, including complex carbonate reservoirs. In the conditions of the field under consideration, the hydraulic fracturing leads to a significant differentiation of technological efficiency indicators, which makes it expedient to study the patterns of crack formation in detail. Studies were carried out for all wells, which were considered as the objects of impact, to assess the spatial orientation of the cracks formed. The developed indirect method was used for this purpose, the reliability of which was confirmed by geophysical methods. During the analysis, it was found that in all cases, the crack is oriented in the direction of the section of the development system element characterized by the maximum reservoir pressure. At the same time, the reservoir pressure values for all wells were determined at one point in time (at the beginning of HF) using machine learning methods. The reliability of the machine learning methods used is confirmed by the high convergence with the actual (historical) reservoir pressures obtained during hydrodynamic studies of wells. The obtained conclusion about the influence of the reservoir pressure on the patterns of fracture formation should be taken into account when planning hydraulic fracturing under the conditions studied.

Keywords

• Hydraulic fracturing direction
• Machine learning
• Actual field data
• Correlation coefficient
• Well flow rate
• Well geophysical studies