Technology to optimise maintenance cost
In the global search for renewable energy sources wind energy continues to experience an increasing market share. This growth drives new technological improvements which aim at increasing capacity, improving reliability and reducing the overall cost of wind turbines and correspondingly of the energy. One of the critical subsystems in this regard is the drive train consisting of a gearbox and/or generator.
There are different cost drivers in wind energy. Focus in this article is on the CAPEX cost and the OPEX cost. Better understanding drive train loading conditions can be leveraged in a dual way. First by developing more cost effective designs by using this loading knowledge in the design process, which is beneficial for the CAPEX by reducing material costs both in drive train and foundation. Secondly the loading information can be used in predictive maintenance strategies for optimizing O&M costs to reduce the need for unscheduled repair jobs and the amount of downtime related to failure. Particularly under offshore conditions this can add significant value given the complex and costly logistics.
The key to an optimal offshore maintenance strategy is optimized logistics, due to the challenging weather conditions in combination with the need for the availability of appropriate vessels. Maintenance planning is essential in this. An optimal approach is the use of a prediction based maintenance strategy. This implies that components are replaced before failure occurs but not earlier than absolutely needed.
The main challenge in this approach is assessing in addition to the current condition of each specific wind turbine also the remaining lifetime. This means that not only occurring failure is detected but also not yet initiated failure is accounted for. This requires the availability of the time series of forces, accelerations,... actually occurred during the life time of the drive train, since loading is site dependent.
The measurement campaign currently being carried out on the drive train of one of the wind turbines at Belwind in the framework of the Offshore Wind Infrastructure Application Lab (OWI-Lab) has this goal. The aim is to develop the instrumentation technology to visualize and track over time the most important drive train parameters needed in the remaining life time calculations and setting the first steps in the online operational remaining life time calculations. By keeping the instrumentation complexity as low as possible it is envisaged to allow the instrumentation of all turbines in a wind park in a cost effective way. By partnering with a gearbox and/or generator supplier it should then be possible to use these generated time series to get a valid remaining life estimate for each of the drive train subcomponents, which the turbine park O&M manager can use to optimize maintenance schedules and logistic plans.
This approach is complementary to the different techniques currently available to assess whether failure is initiating and happening. The most commonly used on wind turbine drive trains are: oil analysis, vibration analysis and acoustic analysis. These techniques assess whether failure is present in the drive train. Nonetheless it is less straightforward to use them in a predictive way.
This project is of special interest to several members of the OWI-Lab user committee, such as ZF Wind Power, LMS International and Belwind, and is carried out in the framework of the IWT Innovation Mandate Post-Doc grant of Dr. Ir. Jan Helsen. The main goal is to develop and do proof of concept of the appropriate instrumentation package for this purpose and set first steps in the development of a software to generate and analyse the time series resulting from the measurement campaign.
This project uses the insights gathered during the IWT O&O Kratos project in which ZF Wind Power together with KULeuven developed simulation model techniques for analysing the dynamic behaviour of wind turbine gearboxes and designed highly accelerated life time tests (HALT tests) which can be carried out on its specifically designed 13.2MW dynamic back to back wind turbine gearbox test-rig.
At the beginning of February the first instrumentation package was installed at the Belwind site. The drive train of one turbine is continuously monitored and data analysed in parallel. The goal for the next months is to extend the analysis software and validate this on the gathered data. Consequent steps include further correlation with other monitoring systems on the same turbine, such as the foundation and tower monitoring (in general named Structural Health Monitoring – SHM), corrosion monitoring, and installing a more extensive monitoring package on the turbine.
