Presentations

Abstract

The wind industry is in a pivotal moment. In order to meet 2030 targets, annual wind energy installations must be accelerated. These targets demand larger turbine dimensions and capacities from Original Equipment Manufacturers (OEM), underscoring the need to prioritize turbine reliability. With non-operation times growing more costly, there's a pressing need to prioritize an optimal preventive maintenance, which maximises the availability of the wind turbine. Furthermore, in applications where optimization of operational expenditure (OPEX) is demanded, reliability of the assets is key. In Wind Energy, in particular, the size and therefore output power of the wind turbines has steadily increased in the last two decades, requiring an optimal maintenance management for increasing availability. In response to these challenges, the wind energy sector is increasingly turning to innovative solutions. One such approach is leveraging real-time prognostics to monitor the status of critical components. This proactive strategy enables operators to anticipate potential failures and minimize downtime by optimizing repair times and spare parts management. In this context, Ingeteam has developed a Digital Twin (DT) of the power converter, which models the electro-thermal nature of the Insulate Gate Bipolar Transistors (IGBTs) and allows implementing a Remaining Useful Lifetime (RUL) estimation strategy of the semiconductor modules, driven by supervised AI models and tested against real power stack test-benches. This approach allows for estimating the degradation occurring within the modules, offering critical information about their condition and expected lifespan. The digital model operates in real-time within the Converter Control Unit (CCU), harnessing the versatility and efficiency of edge computing. Additionally, a real-time temperature soft-sensor has been developed in order to monitor the IGBT junction temperature waveforms (hot-spot).