Abstract
In this work we dealt with some problems of operation and maintenance in wind farms. We focused on the main critical aspects of any maintenance strategy which must include identification of the plant elements to inspect as well as the planning of the possible actions aimed at minimise production losses. At the same time any maintenance strategy must take into account the possible costs. In fact, those decisions can be made based on risk-based methods. We designed a Risk-based maintenance approach to plan inspection tasks to be assigned to service robots in Wind power plants. A SCADA system is employed to collect and manage suitable data (Power, Wind Velocity and related machine events), and the Risk is evaluated on a daily basis over the data collected. The evaluation of the risk is strictly related to the healthiness of the power plant itself. Then, tasks are created and scheduled based on a certain priority which is strictly correlated to the evaluated risk. For the analysis of our approach, we used the real data collected on a Wind Power Plant in Greece over 396 days. The power plant is capable to produce an overall power of 7.2 MW, and it is composed of 8 Wind Turbines of 900 KW per each. We observed that, out of 396 days, 50 days presented machine events leading to a related Risk evaluation for which our approach will produce 258 inspection tasks. From this analysis, we concluded that the application of the risk-based methodology paired with the exploitation of permanent robots on field could result in a 225 MWh reduction of the plant’s lost production, in other words an increase of production of 45.6%.

This publication has 14 references indexed in Scilit: