Estamos recebendo a visita de Andrés Codas, egresso da ECA e PPGEAS, até 29/02. Andrés está finalizando o doutorado em “Cybernetics Engineering” na NTNU/Noruega. No dia 26/02 (sexta-feira), às 11 h, sala PPGEAS-2 (térreo), Andrés irá apresentar um seminário intulado intitulado “Contributions to Production Optimization of Oil Reservoirs”. Maiores detalhes abaixo.
Presentation title: Contributions to production optimization of oil reservoirs.
This research focuses on efficient optimization algorithms and suitable simulation models for oil production optimization. The emphasis is on the variety of the decision processes for different time-scales. To this end, this research considers three time-scales and proposes tools that lead to the desired integration.
For the long-term time-scale, this presentation introduces the Multiple Shooting (MS) formulation for optimization of reservoir flooding control. In contrast to the traditional Single Shooting (SS) formulation, MS divides the time horizon in several independent time frames, thus it promotes simulation parallelization and broader constraint handling by means of the independent state variables. The independent prediction time frames are synchronized by a tailored Non-linear Programming (NLP) algorithm based on a reduced Sequential Quadratic Programming approach (rSQP). Although, this method was initially proposed for deterministic water-flooding optimization, it will be presented its extension to problems involving fluids in the gas phase and its extension to handle risk measures under uncertainty.
For the middle-term time-scale, we present a framework for integrated production optimization of complex oil fields such as Petrobras’ Urucu field, which has a gathering system with complex routing degree of freedom, limited processing capacity, pressure constraints, and wells with gas-coning behavior. Procedures are developed to obtain a Mixed-Integer Linear Programming (MILP) problem using simplified piecewise-linear approximations which ensure a given accuracy with respect to non-linear models.
For the short-term time-scale, we present a two-layered control structure for the stabilization and optimization of a simple oil gathering network consisting of two wells and a shared riser. The regulatory layer stabilizes the system by cascade control of wellhead pressure measurements whereas the optimization layer instantiates a Nonlinear Model Predictive Control (NMPC) using state feedback. The simplified models are implemented in Modelica and fit to the OLGA model to represent the main dynamics of the system. The proposed two-layer controller was able to stabilize and increase the economical outcome of a gathering network simulated in OLGA.
Biography: Andres Codas is a 4th year PhD. candidate at the Department of Engineering Cybernetics of the Norwegian University of Science and Technology (NTNU). He earned an MS degree in automation and systems engineering from Federal University of Santa Catarina, Brazil. Codas’ research interests are in numerical optimization applied to reservoir management and production optimization.