The University of Gdansk has received funding for a new project entitled “HYBRID WIND – Hybrid wind turbine condition monitoring system for different weather conditions” as part of the international Joint Call 2023 competition announced by the Clean Energy Transition Partnership (CETPartnership). The initiative and the competition includes funding of research for energy transition. The project will be implemented in partnership with Gdansk University of Technology, the Institute of Fluid-Flow Machinery of the Polish Academy of Sciences and partners from Belgium, Denmark and Germany.

The submission of the project is the first result of cooperation, which was initiated by the signing of a multilateral letter of intent by the University of Gdansk in June 2022 with EDF Renewables Poland, Gdansk University of Technology and the Institute of Fluid-Flow Machinery of the Polish Academy of Sciences. The main goal of the agreement is to deepen scientific and academic cooperation and implement joint research and development projects.

The project was prepared as part of the work of the all-university substantive team “University of Gdansk for Offshore Wind Energy” (UG for MEW), dedicated to developing cooperation with the socio-economic environment in the offshore wind energy sector.

The project aims to improve the state of the art in wind turbine damage detection and monitoring. Through the use of pioneering new hybrid methods using advanced component and performance damage prediction modeling, the accuracy of wind turbine damage assessment is to be improved.

Wind energy has helped achieve the 20 percent target set by the EU’s Renewable Energy Directive. However, wind turbines (WTs) fail about twice a year, leading to significant economic losses and negative environmental impacts.

Operation and maintenance (O&M) activities have been identified as one of the major costs in total wind farm project expenses, and today almost all turbines are maintained using traditional time-based (preventive) or failure-based (corrective) strategies, which increases the levelized cost of electricity (LCOE),” says dr hab. Michał Suchanek, prof. UG from the Faculty of Economics, project team leader from the University of Gdansk.

The HYBRID WIND project is expected to improve the state of the art in wind turbine damage detection and monitoring by pioneering new hybrid methods using advanced component damage prediction and performance modeling through digital twinning (DT) and modern machine learning (ML).

Applying ideas emerging from the DT field, we will create predictive models so detailed and so closely linked to real materials, components and environmental conditions that many current physical inspections can be transferred to DT simulations, increasing the accuracy of damage assessment, preventing downtime through targeted repairs and predictive maintenance, and lowering the LCOE for wind energy. The resilience and weather sensitivity of damage detection and prediction systems is critical to the cost-effectiveness of wind energy and its subsequent deployment, use and adoption, explains dr hab. Michał Suchanek, prof. UG.

As the university’s team of experts notes, wind turbines and the reliability of their components in operating wind farms are largely determined by the availability of vital business information captured by sensors and monitoring of critical subsystems, while trying to identify emerging major failure modes. However, there is a poor understanding of the links between actual loads, weather conditions and their impact on the components in use, which limits the effectiveness of these measures.

Improving the accuracy of damage detection and monitoring systems therefore requires advances in models, tests, measurements and automated analysis that combine what is detected in the real world with increasingly realistic in-silico modeling in real time.

According to researchers at the University of Gdansk, achieving an optimal hybrid design of a fault detection and assessment system can be achieved by applying new methods such as ML physics, which adopts gray-box models incorporating both data-driven (black-box) and knowledge-driven (white-box) formalisms.

The project team on the part of the University of Gdansk also includes dr hab. Przemysław Borkowski, prof. UG, dr hab. Ernest Czermański, prof. UG, dr Elżbieta Adamowicz, dr hab. Robert Bęben, prof. UG, dr hab. Paweł Antonowicz, prof. UG, dr Mariusz Chmielewski, dr Renata Płoska and mgr Anna Młynkowiak-Stawarz.

The project is expected to last 36 months, with a total cost of €3,202,486.