An integrated 3D geomechanical study for accessing the wellbore complications of offshore wells in Mumbai offshore, India

Petroleum geomechanics plays a critical role in analyzing wellbore integrity and identifying the irregularities in subsurface stresses during drilling and production activities. Many oil and gas reservoirs and wells don’t have adequate data to build geomechanical models to study the stress and rock properties of the reservoir. Conventional workflows and approaches will lead to incomplete models and assessments of the geomechanical properties of the reservoir. To address this challenge, we created a comprehensive approach combining seismic inversion, laboratory experiments, and machine learning methods to develop geomechanical models for investigating wellbore stress and rock property anomalies present in carbonate layers with alternative shale zones. In this study, machine learning-assisted lithology classification was carried out to generate electrofacies; later, these lithology types were used as inputs for constructing 1D and 3D geomechanical models for the North-Heera block. We conducted laboratory experiments that included destructive (triaxial loading) and non-destructive (ultrasonic velocity measurements) to support and identify weak layers. Estimated pore pressure and Shmin were validated with field measurements (in-situ Pressure and LOT) for 1D model accuracy. The results of the 1D and 3D models identified in the shallow zone (250–950 m) and alternative layers (mainly shale) in the producing zone show anomalies in stress and rock strength parameters. The Poisson’s ratio (0.25–0.40) and elastic modulus (4–30 Gpa) are lower, indicating that zones of very low strength caused problems in drilling. This combined approach using geomechanical models, rock testing, and machine learning improved an understanding of carbonate and shale behavior, leading to better wellbore stability and prevention of shallow layer issues.