Electrify Europe 2018

Steam Turbine Evolution as a Digital Asset (Room Lehar 3)

Maintenance is the single largest controllable cost in a plant and traditional labor often results in no action until run to failure and lower return on the assets or in time-based inspections and interventions. The modern view of the asset management is based on condition based and predictive maintenance, monitoring the actual health of the equipment with the following benefits: • prompt warning of potential failures reduces the timing for failure reparation and the damages maintaining plant high performances and reliability/availability; • health knowledge permits to optimize the scheduling of the maintenance inspections and/or interventions anticipating or delaying them in most favorable market conditions. The second item involves a continuous check of the equipment health, especially for the critical assets like the Steam Turbine. Ansaldo Energia and ABB are collaborating to mix different approaches in a single fully integrated synergic platform, including: • Process Measurements • Vibration Monitoring • Field Data Treatment with sensor validation and substitution policy • Machine Learning for measurements baseline and to provide the basis for diagnostic indicator calculation • Performance Monitoring • Lifetime Monitoring • Automatic trend-line check The global approach, capable to automatically check-up the equipment and to give a failure diagnosis and relevant remediation policy realizing an effective Predictive Diagnostics, was tested in application in a combined cycle power plant. The paper describes such application, the relevant experimental results and the lesson learned in terms of data mining exploitation.