Pressure prediction in non-uniaxial settings based on field data and geomechanical modeling: a well example

We apply a new pressure prediction approach that couples sonic velocity data, geomechanical modeling, and a critical state soil model to estimate pore pressure from wellbore data adjacent to a salt body where the stress field is complex. Specifically, we study pressure and stress in front of the Mad Dog salt body, in the Gulf of Mexico. Because of the loading from the salt, stresses are not uniaxial; the horizontal stress is elevated, leading to higher mean and shear stresses. For the Mad Dog field, we develop a relationship between velocity and equivalent effective stress, in order to account for both the mean and shear stress effect on pore pressure. We obtain this equivalent effective stress using a geomechanical model of the Mad Dog field. We show that the new approach improves pressure prediction in areas near salt where mean and shear stress are different than the control well. Our methodology and results show that pore pressure is driven by a combination of mean stress and shear stress, and highlight the importance of shear-induced pore pressures. Furthermore, the impact of our study extends beyond salt bodies; the methodology and gained insights are applicable to geological environments around the world with a complex geologic history, where the stress state is not uniaxial (fault zones, anticlines, synclines, continental margins, etc.).