Maintaining high levels of availability and reliability are essential objectives for many industries, especially those that are subject to high costs due to shutdowns of critical systems, e.g. gas turbines. To utilize these systems as effectively as possible, preventive maintenance must be optimized. Determining what is optimal is, however, a multi-variable task requiring detailed knowledge about the components in the system and their different damage mechanisms. These factors have always affected the condition of the gas turbine and maintenance actions, but only recently has it been possible to estimate and measure them correctly for individual components during operation. In the past, it was necessary to construct maintenance intervals from the most critical component (or components), requiring the highest maintenance frequency. An additional worst-case scenario margin was also necessary, taking into account factors such as possible load variation, differences in environment (affecting e.g. power turbine temperatures) and other sources of uncertainty. These uncertainties together have determined traditional maintenance planning, with maintenance packages each containing a set of maintenance activities for a set of components being predetermined and preplanned. With the new CAMP approach, the maintenance strategy is to reach a Retirement For Cause (RFC) strategy, where components are not replaced until a potential failure has been detected. This requires measurement techniques that can monitor how the gas turbine is operated, prognostics capabilities that foresee maintenance needs, and test methods that can determine the state of a component during maintenance events. One important part of CAMP is therefore a prognostic tool which tells us the condition, and therefore the maintenance needs, of individual components within the gas turbine. To handle this information and efficiently make a preventive maintenance plan, software for gas turbine maintenance optimization has been developed. The software can not only calculate the most efficient point in time for a maintenance action, it can also adjust the maintenance plan to any customer’s specific demands. This paper describes the model, gathering and processing of information, risk assessment performance and the result from an optimization which groups maintenance actions as a result of customer prioritized demands. It also describes the software layout and how it is used.

This content is only available via PDF.
You do not currently have access to this content.