The gas turbines have been under intense technological development during the past decades. Among the factors that motivate manufacturers to develop differentiated products, it can be cited the rising fossil fuels prices, environmental legislation and increasingly restrictive competition with other alternatives to generate power. Even in the current scenario of global economic crisis, manufacturers and users expect an increase in the number of gas turbines in operation.
In offshore oil & gas exploration, the thermal efficiency of gas turbines has been traditionally at the background compared to the operational continuity. However, nowadays many operators are investing in subsea pipelines to export the associated gas produced in the platforms. In this scenario, the gas, which has been traditionally burned at the flare system, now represents a commercial product.
Performance monitoring of gas turbines for diagnostic of failures has been developed by researchers over the last decades. There are several diagnostic methodologies currently available in the open literature; however, the amount of publications demonstrating successful application in real cases is still reduced. Regarding the application of performance monitoring for prognostic of failures, the scenario is even more premature. Despite of that, both diagnostic and prognostic techniques are seen as very promising from the user point-of-view.
This paper proposes a methodology for performance prognostic based on parameters effectively measured by typical instruments of commercial units. The method makes use of the performance modeling easily found in the literature. The performance parameters measured and also calculated are then corrected to a reference condition in order to eliminate the effects of the variable ambient condition. Future performance parameters are then extrapolated. The methodology was validated through real data of a gas turbine operating in an oil platform. The results showed that the method produces useful information to support the operation and maintenance teams.