The Science of making Torque from wind, commonly known as the Torque conference, is the leading scientific conference worldwide on wind energy. It is a biennial event of the European Academy of Wind Energy , with Torque 2026 being the 11th conference since its inception. The event took place from June 3rd to 5th, 2026, in Bruges, Belgium, and was a great opportunity for researchers and academics to showcase their work.
Researchers from the Technical University of Munich contributed a total of five presentations – three posters and two oral presentations – conducted within the framework of WP2 of the TWAIN project.
- “A holistic framework for site-specific wind power forecasting” by Antonina Vukobrat, Andreas Vad, Abhinav Anand, and Carlo L. Bottasso.
This study addresses the challenge of uncertainty in wind power forecasting caused by highly variable atmospheric conditions. It proposes a hybrid forecasting framework that integrates wake modelling and machine learning to predict key atmospheric variables and resulting power output. Its novelty lies in delivering site- and turbine-specific estimates that account for both physical interactions and operational constraints, such as shutdown strategies. This comprehensive approach enables more accurate 36-hour-ahead power forecasts, which are critical for reliable grid integration and can directly support market participation strategies.
2. “Leveraging directivity of wind farm noise to increase profitability” by Anik Shah, Kenza Amhis, Carlo R. Sucameli, Franck Bertagnolio, Andreas Fischer, and Carlo L. Bottasso.
Wind turbine noise is directional – yet regulations are often applied conservatively. We show how smarter noise-directivity aware operation can boost wind farm profitability while staying compliant. Up to 2% more AEP – simply by listening more carefully!
3. “Wind farm control for inertial and primary frequency support” by Abhinav Anand, Simone Tamaro, and Carlo L. Bottasso.
This study investigates how wind farm control (WFC) can enhance the fast frequency support capability of grid-connected wind power plants (WPPs). The analysis first examines wake steering for power maximization, showing that de-waking downstream turbines increases the inertia provision capability of the plant. Building on this insight, dedicated WFC strategies are developed to maximize WPP synthetic inertia under curtailment. Turbine curtailment strategies are evaluated from a WFC perspective while accounting for converter, drivetrain, aerodynamic, and control constraints. The impact of WFC is quantified relative to conventional greedy turbine operation for representative wind farm layouts, and system-level effects are assessed using a dynamic grid model with varying shares of synchronous and non-synchronous generation. Results indicate that wake-steering WFC can increase plant-level synthetic inertia by up to 30% under high-wake conditions and significantly improve grid frequency response following a disturbance.
4. “Efficient generation of location-agnostic wind turbine load surrogate models using wake slices” by Adrien Guilloré, Abhay Chaudhary, Abhinav Anand, Andi Vad, Anik H Shah, Vasilis Pettas, Tuhfe Göçmen and Carlo L. Bottasso.
The oral presentation took place on Room Forum 6 on June 5th at 9:00.
Wind farm design and control require fast methods to estimate turbine fatigue loads, which are usually expensive to compute because they depend on complex wake interactions and full wind-farm simulations. As a direct outcome of the collaborative work in TWAIN Task 2.1, this work introduces a more efficient surrogate modeling approach that learns fatigue behavior from single-turbine simulations combined with a library of wake inflow slices, making it transferable across turbine types and farm layouts. Tested on multiple turbines and validated against full wind farm simulations including wake steering, the method accurately captures relative fatigue load variations while significantly reducing computational cost, enabling more efficient fatigue-aware wind farm optimization.
5. “Market-aware operation of grid-integrated wind power plants: beyond revenue, towards structural loads and grid services” by Fanning Zheng, Antonina Vukobrat, Abhinav Anand, Simone Tamaro, Tuhfe Göçmen and Carlo L. Bottasso.
The oral presentation took place on the Concertzaal on June 3rd at 14:40.
This paper, developed as part of the TWAIN project (Task 2.3), proposes an Energy Management System for wind power plants that jointly optimizes market participation across day-ahead and intra-day markets. Unlike conventional approaches focused primarily on revenue maximization, the proposed framework incorporates load-dependent costs and operational strategy to maximize overall profit. It integrates generation and price forecasts with a wind farm flow model to accurately track power references under varying wind conditions and derive optimal power bids. Additionally, the framework quantifies the synthetic inertia the plant can provide, accounting for wake interactions and operational constraints. Results demonstrate improved economic performance, reduced turbine fatigue, and reliable inertia-support estimation.
