The optimisation of the usage of gas turbine generation sets for oil and gas production using genetic algorithms

Abstract

The privatisations of the energy supply industries world-wide has meant that emphasis is now on how to profitably compete. In this environment the development of effective models for optimisation of power plant are of increasing importance, particularly operational strategies for off-design conditions, and particularly for gas turbine engines. Maximisation of plant profitability necessitates proper and integrated evaluation of many factors, the most important of which are: availability and price of fuel, system efficiency and performance, life cycle costing of plant and machinery, present and future generation of revenue, likely future market dynamics. A major contribution of this work is the application of the proposed method to simultaneously maximise both total profit and usage availability of a typical combination of gas turbines engines used for power generation in oil and gas production. The method allows the user, for example, the opportunity to select locally appropriate daily and seasonal power demands and ambient conditions. Through a genetic algorithm optimisation technique, an additional powerful feature of the method is that it allows the user to choose an optimised operating combination of their existing gas turbine equipment. Both individual engine power setting and number of engines can be varied. Alternatively, the user can apply the code to select the best combination of new and/or replacement equipment to achieve best economic performance and highest availability. The number of variables involved in the optimisation process is, of course, very large. It is, therefore, difficult to find the optimal configuration. To address this problem, the first phase of this study is limited to the analysis of the performance of industrial gas turbine engines. The primary aim is to identify the key parameters in the determination of off-design performance. The second aim for the first phase is to identify those tasks suitable for automation. The Gas Turbine Library (Turbomatch) developed at Cranfield University includes simulation

codes for many different industrial gas turbines and processes. The optimiser developed as part of this research has been linked with that library. The second phase of this project is to develop an economic model for gas turbines analysing off-design performance. The model includes a life cycle cost assessment including: capital cost, maintenance and operating costs, fuel cost, emission and other taxes and disposal cost. By including total revenue it has been possible to develop a model that allows maximisation of total profit under variable operating conditions. The third phase of the project presents an automated optimisation tool based on a listing of the Turbomatch simulation code and a genetic algorithm technique. The tool uses an evaluation of the fitness value of the objective function and takes into account the optimisation constraints. Two case studies considered where real data obtained from oil field in Libya are used to illustrate the use of the new code to maximising the profit.