Defesa de Tese de Doutorado – Thiago Lima Silva – 19/04/2017
| Defesa de Tese de Doutorado | |
| Aluno | Thiago Lima Silva |
| Orientador
Coorientador |
Prof. Eduardo Camponogara, Dr. – DAS/UFSC
Prof. Bjarne Anton Foss, Dr. – NTNU |
| Data
Local |
19/04/2017 08h00 (quarta-feira)
Sala PPGEAS I (piso superior) |
| Prof. Eduardo Camponogara, Dr. – DAS/UFSC (orientador)
Prof. Denis José Schiozer, Dr. – UNICAMP Prof. Regis Kruel Romeu, Dr. – Petrobrás Prof. Erlon Cristian Finardi, Dr. – EEL/UFSC Prof. Ubirajara Franco Moreno, Dr. – DAS/UFSC Prof. Juliano de Bem Francisco, Dr. – MTM/UFSC Prof. Fábio Luis Baldissera, Dr. – DAS/UFSC |
|
| Título
|
Contributions to Modeling and Optimization of Oil Production Systems |
| Abstract: The traditional approach for the operation of an oil platform has been to make decisions based on sensitivity analysis using simulation software and heuristics. However, this strategy can be rather time-consuming and does not necessarily ensure an optimal production policy. Instead, oil companies are investing towards the development of optimal strategies in order to improve economic margins. An alternative that is gaining acceptance in the industry is model-based optimization, which may be seen as the integration of mathematical models with algorithms into effective optimization tools. This dissertation proposes models and optimization tools for production optimization of oil fields in the short-term and medium-term.
In the short-term, the contributions are a comprehensive study of multidimensional piecewise-linear models for daily production optimization and the modeling of flow splitting in subsea gathering networks. Because much of the literature was scattered, the first contribution consists of a common framework with the existing multidimensional piecewise-linear models for nonlinear function approximations appearing in oil production processes. Such models allow to transform the nonlinear production optimization problem into a mixed-integer linear program, for which off-the-shelf solvers can obtain optimal solutions. Computational and simulation analyses are performed to evaluate the efficiency and approximation quality of these models for a realistic oil field built in a commercial multiphase flow simulator. The second contribution consists of a model for flow splitting in subsea gathering networks. Despite being a common practice in offshore operations, where the wells can be connected to processing facilities by multiple routes, previous optimization models assumed single routes. An automatic routing model which decides upon single or multiple routing was developed and validated against simulation software. The model was further employed in the optimization of a synthetic oil field, where the optimal strategies with flow splitting yielded higher production rates than single-routed ones. In the medium-term, the contribution is a methodology to handle network output constraints in reservoir management problems, particularly in water-flooding processes. Since full-field implicit simulations are prohibitively costly, reservoir management policies are typically developed with standalone reservoir models, while the constraints regarding the network are limited or fully disregarded. We propose to optimize the integrated problem with a multiple shooting formulation, which is a control method suitable for problems with numerous output constraints. The methodology is employed in the optimization of a two-phase black-oil reservoir producing to a gathering network with nonlinear constraints regarding the operation of electrical submersible pumps. The method’s capability to handle network constraints is assesed by constrasting its results against conventional approaches which neglected the gathering network system. These contributions, seen from the integrated operations perspective, may support production and reservoir engineers in decision making processes of real-world fields. |
|
