Prediction of formation pressure based on numerical simulation of in-situ stress field: a case study of the Longmaxi formation shale in the Nanchuan area, eastern Chongqing, China

Abstract

The in-situ stress and formation pressure are important parameters in shale gas development. They directly affect the well wall stability, the direction of horizontal well drilling, and the fracturing effect during the shale gas development process. There are abundant shale gas resources in the southeastern Chongqing-Sichuan area, but the structure in the area is complex, and it is difficult to predict the in-situ stress and formation pressure. Therefore, in this paper, a finite element simulation model was established based on a large number of seismic, logging, and experimental rock mechanics data and the prediction accuracy of the stress field simulation was effectively improved. The construction of the stress field was based on the combined spring model, as well as the data related to the measured in-situ stress and the formation pressure obtained during drilling. The coupling relationship between the in-situ stress, the formation strain, and the formation pressure were derived to carry out the prediction of the distributions of the formation pressure and the formation pressure coefficient. The prediction results showed that the present-day maximum principal stress direction in the study area was about NE65°–110°, and the present-day maximum principal stress was 56.12–93.79 MPa. The present-day minimum principal stress direction was about NE335°–20°, and the present-day minimum principal stress was 48.06–71.67 MPa. The formation pressure was 2.8–88.25 MPa, and the formation pressure coefficient was 0.74–1.55. The formation pressure distribution was greatly affected by fault, tectonic location, in-situ stress and rock petrophysical properties, and the overpressure areas of the formation were distributed in the synclines and the deeply buried areas. This study shows that the finite element based formation pressure prediction method is effective.