Rapid SAGD Simulation Considering Geomechanics for Closed Loop Reservoir Optimization

While numerical modeling and coupling techniques have been continuously studied, analytical solution or proxy modeling for geomechanical coupling of the steam assisted gravity drainage (SAGD) has not been clearly addressed in the literature. Simulations aside, there is no particular study on the use of geomechanics in closed loop reservoir optimization. Past studies have focused mostly on two separate research areas: (a) SAGD and geomechanics and (b) intelligent optimization algorithms for smart fields.;This research has been carried out to cover two major objectives; providing a low order model to work with real-time data, and also investigating high-resolution geomechanical-flow simulation to work with data assimilation algorithms for history matching and reservoir characterization.;As the first step, a physics-based semi-analytical model was proposed based on the original Butler/Reis SAGD theory. The model was proposed for linear steam chamber geometry by modifying the variation of oil saturation in advance of the steam chamber. The model was then verified with the past experimental lab test results and numerical simulation results. Geomechanics was incorporated using the classical limit equilibrium theory. Although the results of the linear geometry model with geomechanical consideration are promising for the available case studies, application of such a theory needs to be further investigated and improved.;The linear geometry model was then replaced by circular geometry model to better simulate the rising and depletion stages of SAGD process. For the circular geometry model, a multiplier coefficient was defined to consider geomechanics called the geomechanical impact factor (GIF). Based on the results from a numerical study, GIF can be effectively employed for the variation in rock properties such as permeability and porosity. The final version of the proposed model was used for history matching two SAGD projects, UTF phases A and B. The results show that the proposed physics-based analytical proxy generally captures the physics of the problem in low order but a very fast fashion procedure.;The application of analytical models in automated history matching and reservoir characterization was further investigated using the extended Kalman filter (EKF). For this case, Butler/Reis theory and the GIF concept were combined with the EKF for history matching the heterogeneous reservoirs with uncertainty. Using synthetic data and stochastic reservoir realizations, it was shown how analytical models are helpful in reservoir characterization.;While the analytical solution is placed at the centre of the optimization process, the second objective of this research was explored by applying the ensemble Kalman filter (EnKF) to link monitoring data to the simulator(s). For this reason, an iterative geomechanical-flow coupling code was developed and assembled with the EnKF. Through numerical simulations using synthetically generated data, the significance of considering geomechanical monitoring data in reservoir surveillance was examined. It was observed that proper coupled simulation mimics larger portion of the real physics of SAGD process, and as a result, geomechanical observations can add value to facilitate the data assimilation algorithms.