Selection of Optimal Platform Locations

Abstract

Summary A new method for finding, the optimal platform and best well locations in offshore oil fields is presented to minimize the total drilling cost while the productive potential is maximized. The general mathematical model is defined as a nonlinear mixed-integer programming problem in three-dimensional (3D) space that is solved in a two-dimensional (2D) plane. Graph-theory approach and an alternate-location/allocation plane. Graph-theory approach and an alternate-location/allocation algorithm were used sequentially for the solution of this problem. The model has been tested on several actual offshore fields. Some computational results are given for both limited and unlimited platform capacity cases. Introduction The efficient development of offshore oil fields requires that important long-term decisions be made at an early stage. The papers that have dealt with the problem of determining optimal platform siting have not considered the reservoir properties as an effective optimization parameter. It is usually assumed that well targets are known and all will be drilled from the selected platform locations. This assumption certainly is not a realistic one. Because the decision-making in the early life of the project for the first platform location and first potential target is project for the first platform location and first potential target is risky, very careful analysis of the available seismic and geologic information is needed. The model introduced here uses the results, productivity index (PI), and oil-in-place (OIP) maps obtained from the productivity index (PI), and oil-in-place (OIP) maps obtained from the initialization of a 3D, three-phase black-oil simulation model. The map is divided into a line grid of test targets that are assumed to be possible well locations. The general problem of selecting the best platform locations and optimal well assignments is analogous to the standard multiwarehouse and factory location problem. Various proposals have been made for solving such multifacility location problems by linear programming, simulation methods, heuristic methods, a dynamic-programming programming, simulation methods, heuristic methods, a dynamic-programming algorithm, and convex programming. The major drawback to these methods is that each appears to require large amounts of computation time. The solution provided by the model developed in this paper is based on a graph-theory approach. The platform locations are the center of connected subgraphs and the well targets are the nodes connected to each platform. Similar location problems involving graph theory have been platform. Similar location problems involving graph theory have been studied by Donath, Gilbert and Pollak, and Hanam. The advantage of this approach is that nonlinear functional relationships, which require approximation in the models, can be handled better and exactly. The objective of this model is to find the optimal selection of wells and platform locations to maximize the total productive potential while total drilling cost is minimized.