Kam, Dongjae (2015-08). Streamline-Based Transport Tomography and History Matching for Three-Phase Flow. Doctoral Dissertation. Thesis uri icon

abstract

  • Reconciling a geological static model to the available dynamic information, known as history matching, is an essential procedure for the decision-making through predictions of fluid displacement in a reservoir. However, there are several challenges in the history matching workflow because the geologic models are becoming complex and more detailed with a large number of grids. Recently, streamline-based inverse modeling has shown great promise for the high resolution geologic model because of many advantages in terms of computational efficiency and applicability. However, the current approach is primarily focused on handling the water-cut and tracer test data. This dissertation presents a novel streamline-based approach to incorporate a variety of dynamic information into the history matching process for the forecasting of reservoir behavior with increased confidence. We first develop the streamline-based transport tomography by incorporating novel tracer technology. The distributed arrival time made available by a novel tracer provides a significantly improved flow resolution for reservoir characterization. We demonstrate the new approach for streamline-based history matching of distributed water arrival time together with aggregated well production data that clearly shows the benefits of the transport tomography using novel tracers. Second, we propose a new methodology to incorporate bottom-hole pressure data into the geologic model using the streamline-based approach. This approach overcomes the limitation of the sequential process used in previous applications by facilitating the joint inversion, while reproducing reservoir energy during the flow rate matching. The joint inversion with a multiscale approach is suggested to account for the disparity in resolution of different types of data. It leads to capturing of the large- and fine-scale heterogeneity and reproducing the pressure and water-cut responses efficiently. Finally, we extend the streamline-based inverse modeling to the three-phase system by adding gas-oil ratio data simultaneously. We validate that the streamline-based analytical sensitivity of the gas-oil ratio can provide reasonable approximations for the purpose of inverse modeling. The Pareto-front concept is introduced for a multiscale multi-objective approach in combination with the streamline approach to overcome the challenges in the streamline-based three-phase joint inversion. In addition to demonstration of the streamline-based history matching method with a variety of dynamic data, we emphasize the applicability of our approach to the field-scale reservoir model to satisfy the industry demands.
  • Reconciling a geological static model to the available dynamic information, known as history matching, is an essential procedure for the decision-making through predictions of fluid displacement in a reservoir. However, there are several challenges in the history matching workflow because the geologic models are becoming complex and more detailed with a large number of grids. Recently, streamline-based inverse modeling has shown great promise for the high resolution geologic model because of many advantages in terms of computational efficiency and applicability. However, the current approach is primarily focused on handling the water-cut and tracer test data. This dissertation presents a novel streamline-based approach to incorporate a variety of dynamic information into the history matching process for the forecasting of reservoir behavior with increased confidence.

    We first develop the streamline-based transport tomography by incorporating novel tracer technology. The distributed arrival time made available by a novel tracer provides a significantly improved flow resolution for reservoir characterization. We demonstrate the new approach for streamline-based history matching of distributed water arrival time together with aggregated well production data that clearly shows the benefits of the transport tomography using novel tracers.

    Second, we propose a new methodology to incorporate bottom-hole pressure data into the geologic model using the streamline-based approach. This approach overcomes the limitation of the sequential process used in previous applications by facilitating the joint inversion, while reproducing reservoir energy during the flow rate matching. The joint inversion with a multiscale approach is suggested to account for the disparity in resolution of different types of data. It leads to capturing of the large- and fine-scale heterogeneity and reproducing the pressure and water-cut responses efficiently.

    Finally, we extend the streamline-based inverse modeling to the three-phase system by adding gas-oil ratio data simultaneously. We validate that the streamline-based analytical sensitivity of the gas-oil ratio can provide reasonable approximations for the purpose of inverse modeling. The Pareto-front concept is introduced for a multiscale multi-objective approach in combination with the streamline approach to overcome the challenges in the streamline-based three-phase joint inversion.

    In addition to demonstration of the streamline-based history matching method with a variety of dynamic data, we emphasize the applicability of our approach to the field-scale reservoir model to satisfy the industry demands.

publication date

  • August 2015