Abstract: In this study, an intelligent prediction of the three-dimensional stress field of block Yi 176 of Bonan Oilfield is performed by combining the modeling technology of Petrel and ANSYS and using the construction technology of a threedimensional heterogeneous rock mechanics field. To improve the computational efficiency of the numerical simulation of the stress field, the boundary condition, which is traditionally set manually, is now set automatically using the self-adaptive boundary program developed by the APDL program in ANSYS. The results show that the self-adaptive boundary program can greatly reduce the time taken to set the correct boundary conditions by repetitive manual trials. This approach can improve the efficiency of stress field simulation while ensuring prediction accuracy. The maximum horizontal principal stress of block Yi 176 follows an EW-SEE direction in general. The boundary fault has a significant impact to the stress direction, causing a deflection of around 5°~10°. The in-situ current stress value varies substantially with a high level of discreteness and a belt-like distribution in the plane. Furthermore, the stress is found to be higher in the north region and smaller in the south region. In the current formation, the stress belongs to the type I_a in-situ stress. The heterogeneity of the rock is the main cause of the difference in stress values. In general, a belt-like distribution of "stress dessert" is found in the central part of block Yi 176, which can be used as an advantageous area for well location deployment and fracturing operations.

Key words: three-dimensional stress field, intelligent prediction, self-adaptive boundary program, Block Yi 176, Bonan Oilfield