Banca de DEFESA: CAMILA NASCIMENTO RIBEIRO

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : CAMILA NASCIMENTO RIBEIRO
DATE: 09/08/2021
TIME: 14:00
LOCAL: Auditório Pós-Graduação Engenharia Civil - CTG - UFPE
TITLE:

GRADIENT APPROACH REFINING TECHNIQUES FOR WATERFLOODING OPTIMIZATION

KEY WORDS:

Ensemble-Based Method; Simplex gradient; Optimization of the water injection problem; SQP; Nonlinear Least Squares; Refinements.

PAGES: 127
BIG AREA: Engenharias
AREA: Engenharia Civil
SUMMARY:

In this work, refinement techniques are presented to improve both the performance of the Ensembles Based Method (EnOpt), as well as to improve the performance of the Gradient Simplex (SG), for the optimal management of the water injection problem in constrained reservoirs. The objective function adopted is the Net Present Value (NPV) and the restrictions imposed are at the well and field level. The SQP algorithm was adopted to solve the water injection problem and the Nonlinear Least Squares Method (NLS) to solve the decoupled problem. We propose four techniques to refine the sensitivity matrix: Truncation technique; Damping Technique; Locating Techniques that include the sensitivity of the NPV of the producing wells with respect to the controls of the other producing wells and the sensitivity of the NPV of the producing wells with respect to the controls of the injector wells; and also the Regularization Techniques, namely, the sensitivity of the NPV of the injector wells with respect to the controls of the producing wells and the sensitivity of the NPV of the injector wells with respect to the controls of the other injector wells. These techniques were added to the EnOpt and SG methods, resulting in five distinct refinements, REB, REB*, REB+T, RSG+T and RSG+TD, which were applied to the BCO-Falha and Brugge reservoir models.

BANKING MEMBERS:
Presidente - 1130915 - BERNARDO HOROWITZ
Externo ao Programa - 2695448 - LEONARDO CORREIA DE OLIVEIRA
Interna - 2223613 - SILVANA MARIA BASTOS AFONSO DA SILVA