A MODEL-BASED MIXED DATA APPROACH FOR OPTIMIZING THE PERFORMANCE OF AN OFFSHORE GAS TURBINE COMPRESSOR

Abstract

Gas turbines play prominent role in aviation, power generation, marine, mining, petrochemicals, onshore and offshore oil and gas industries. Environmental factors result to degradation mechanics of the overall performance of the engine. Industrialist therefore face unplanned equipment and manufacturing downtime which is undesirable because it leads to economic losses and costly repairs. While condition monitoring is used to obtain early warning of impending equipment failure, optimization techniques such as water washing (online and offline) are used to bring the engine/equipment back to lime-light. In this work, several schedule visits were made to a gas turbine plant on industrial duty for electricity generation in an offshore thermal station at Ibeno, near Eket, Akwa Ibom State in Nigeria. Continuous monitoring of the aerothermodynamics and vibration data were taken for a period of twelve months on hourly basis to examine the state of health of the engine due to adverse offshore environmental factors. The diagnostic approach of vibration analysis, trend and performance monitoring were used with a model-based mixed data approach for optimizing the performance of the offshore gas turbine plant.A software code-named VANCANAL written in C++ programming language was used to proactively monitor the gas turbine plant. The software displayed the operational (actual) values of all the parameters of the engine against their designed (reference) values. Once a set limit is exceeded, the software is capable of giving an alarm, signifying the need for maintenance.