Title: |
Derivative-Free Optimization for Generalized Oil Field Development |
Author: |
Obiajulu Joseph Isebor |
Year: |
2013 |
Degree: |
PhD |
Adviser: |
Durlofsky |
File Size: |
10MB |
View File: |
|
Access Count: |
1603 |
Abstract:
Given the substantial costs and potential rewards associated with oil and gas field development and management, it is essential that these operations be performed as close to optimally as possible. In this work, we consider the computational optimization of general oil field development problems. Techniques are devised to simultaneously determine the optimal number and type of new wells, the sequence in which they should be drilled, and their locations and time-varying controls. The general optimization is posed as a mixed-integer nonlinear programming (MINLP) problem and involves categorical, integer-valued, and real-valued variables. Our new formulation handles bound, linear, and nonlinear constraints. The latter are treated using filter-based techniques. Noninvasive, derivative-free and easily-parallelizable approaches are applied for the optimizations. Methods considered include Mesh Adaptive Direct
Search (MADS, a local pattern search method), Particle Swarm Optimization (PSO, a heuristic global search method), a PSO-MADS hybrid, and Branch and Bound (B&B, a rigorous global search procedure that requires relaxation of the categorical variables). The filter-based treatment of nonlinear constraints is extended to PSO and to the PSO-MADS hybrid.
Example cases involving well control optimization, joint well placement and control, and generalized full-field development problems are presented. The well control optimization example highlights the positive features of the filter constraint handling treatments. For the joint well placement and control optimization examples, the PSOMADS hybrid is shown to consistently outperform the standalone MADS and PSO procedures. The joint optimization approach is also observed to provide superior performance relative to the sequential procedure, in which the well placement and well control problems are solved separately. For the field development optimization cases that involve the optimization of well number in addition to well locations and control, there are two examples in which B&B is applied. In these examples, the PSO-MADS hybrid is shown to provide solutions comparable to those from the more established
B&B approach, but at much lower computational cost. This is significant since B&B provides a near-exhaustive search in terms of the categorical variables. A full-field development example is presented where the number, type, drilling schedule, location and control of wells are optimized. This case demonstrates the broad capabilities of the new optimization framework.
We also implement an approach for field development optimization with two conflicting objectives. For this problem, a single-objective product formulation for the biobjective optimization procedure is applied. Three biobjective field development examples, including one that highlights the applicability of biobjective techniques to optimization under geological uncertainty, are presented. Our overall conclusion from this work is that, although they can be demanding in terms of computation, the optimization methodologies presented here appear to be applicable for realistic reservoir development and management problems.
Press the Back button in your browser.
Copyright 2013, Obiajulu Joseph Isebor: Please note that the reports and theses are copyright to their original authors. Authors have given written permission for their work to be made available here. Readers who download reports from this site should honor the copyright of the original authors and may not copy or distribute the work further without the permission of the author, Obiajulu Joseph Isebor.
Accessed by: ec2-54-163-62-42.compute-1.amazonaws.com (54.163.62.42)