While geomechanical analyses targeted to well drilling optimization are well established, full-field reservoir models are challenged to be predictive, supporting field design decisions. In such multiscale scenarios, near-wellbore perturbations are relevant for drilling, but it is pointless to deal with such refinement for subsidence or fault reactivation analyses. We discuss herein our up-to-date understanding of relevant applications for geomechanical models, after 20 years of experience in geomechanical modeling and data assessment. Regardless of their simplicity, single-dimension well models, based on density and image well profiles, provide great control of the operational window. For wellbore models targeting fracture initiation and propagation, further mesh discretization and mapping of several multiphysical phenomena adds complexity to characterize fracture geometry and productivity enhancement either for short-term stimulation jobs or for mapping long-term injection well behavior. For the latter, temperature and poroelasticity play important roles and cannot be neglected and the absence of proppant keeps the fracture less stable. From an even larger perspective, full-field models are challenged to estimate the stress field and provide the boundary conditions for smaller-scale studies. This global view demands numerous mesh elements, in costly parallel computational setups. This work reasons design decisions and tradeoffs, sharing our learnings related to representative scale and modeling strategies, exchange of information among models, mapping of critical boundary conditions, and numerical and computational setups and simulation tools used. Our observations are supported by field examples, discussing how analysis with similar goals may mislead conclusions. Finally, we discuss current technology limitations and challenges for the near future.
Oil fields under drainage have thermal, hydraulic, and mechanical dynamics that can induce undesired outcomes. Mapping risks and understanding the behavior of the rocks along the exploitation lifetime is relevant as the energy demand challenges the industry to develop oil and gas reserves from ultra-deep offshore fields.
Recent accidents related to unforeseen rock displacements and fault reactivation push regulators and companies to invest in prevention. Expensive data acquisition implies scarce availability of information, and geomechanical teams are challenged to build numerical models based on analogs, basin geostructural knowledge, and literature information. The models are then calibrated against the available multiscale and multi-confidence data. Seismic mappings, for example, capture information about large basins with low resolution. Well profile and well fracturing tests provide localized rock attributes and stress states in a few centimeters to a few meters scale. Laboratory experiments, lastly, provide accurate information about micrometric grains, centimetric plugs, and slightly larger whole cores.
