A Predictive Maintenance Remote System based on Partial Discharges Measurements on Wind Turbines

The environmental matters fostered the power generation using renewable energies. This led, among the different possible solutions, to the diffusion of wind farms for wind power generation. Different types of sensors capable to detect the Partial discharges (PD) activity in the power generator and in the other electrical apparatuses installed in the nacelle can be useful as a diagnostic tool for a predictive maintenance program. The sensors able to reveal irradiated and/or conducted PD signals can be connected to an automated acquisition system. In order to monitor continuously the electrical apparatuses in the nacelle, this acquisition system can be remotely connected to a control unit by means of an Ethernet or a Wi-Fi connection. After a preliminary measurements campaign to define the environmental noise, four sensors have been installed on board of each of two nacelles: two sensors, antenna type, have been used to detect irradiated PD signals while other two sensors, high frequency current transformers, have been used to reveal conducted PD pulses. All sensors have been optimized and the output signals have been filtered and amplified. The conditioned signals have been acquired by means of a high-speed A/D converter. An ad hoc developed software allows to set the measurement parameters and to store the acquired data on a remote storage unit. In order to obtain information about the actual status of the electrical apparatuses the measurements have been performed on two nacelles for months.