Numerical simulation is the main method to predict and manage petroleum reservoirs, in which, traditionally, rock compressibility, that cannot well represent stress-strain behavior of the rock, is the only geomechanical parameter taken into account. Considering this, a more complete representation of rock mechanical behavior is a key point for any hydromechanical numerical analysis, starting with rock mechanics and petrophysical lab experiments, followed by numerical calibration. In this work Modified Drucker-Prager and Mohr-Coulomb constitutive models have been adopted for numerical modeling of rock mechanics experiments on different carbonate types. Hydromechanical behavior has also been numerically assessed for some of the tested samples, aiming for the construction of a pore pressure – porosity – permeability variation table, which allowed the use of pseudo-coupling. In this way, geomechanical effects were taken to reservoir simulation, obtaining a good relation between computational cost and accuracy.
The main method to predict and manage petroleum reservoirs is numerical simulation in which, traditionally, rock compressibility is the only geomechanical parameter taken into account. During exploitation, though, reservoir-rock deforms, causing porosity and permeability variation, phenomenon that cannot be represented by compressibility. Considering this, a more complete representation of rock stress-strain behavior is a key point for any hydromechanical numerical analysis. A good mechanical characterization of a geomaterial based on rock mechanics and petrophysical lab experiments is a starting point, followed by numerical calibration of rock behavior observed in laboratory. This last stage can be done via lab experiments modelling, aiming to obtain mechanical and elastic parameters that best represent the stress-strain rock behavior, based on a proper constitutive model.
This paper intends to describe a methodology based on numerical modeling of lab experiments to take to reservoir simulation the geomechanical effects due to exploitation. For this, three examples of numerical modelling of experiments, and one example of pseudocoupling application are presented, all based on carbonate rock data.