As wind farms expand, especially offshore, predicting turbine interactions becomes more complex. Traditional models often fail to account for varying atmospheric conditions — such as atmospheric stability, the boundary layer height, temperature, and turbulence profiles through and above the boundary layer —that all significantly impact turbine wake and blockage effects, and hence wind farm production.

To overcome these challenges, DNV has developed a new innovative tool; WindFarmer CFD.ML v2. This innovative model combines Computational Fluid Dynamics (CFD) with Machine Learning (ML) to create predictions that accurately resolve the impact of atmospheric conditions on turbine interactions at a speed comparable to other rapid models. Trained on a vast amount of high-quality data from high-fidelity simulations, CFD.ML offers a more accurate, efficient, and scalable solution for wind farm developers, operators, and engineers. 

What we cover in this webinar: 

• Understand why existing rapid models fall short in predicting turbine interactions and energy yield accurately.  
• Learn how the CFD.ML V2 model can help you estimate and enhance your wind farm's performance.  
• See how DNV’s WindFarmer CFD.ML leverages machine learning techniques and high-fidelity simulation data of from future wind farm designs to provide a future-ready and reliable modeling.  
• Gain insights from DNV’s collaboration with industry leader RWE, showcasing the real-world performance of CFD.ML against measured wind farm project production data. 

Speakers 

• Tom Levick, WindFarmer Modelling Lead, DNV  
• Karol Mitraszewski, WindFarmer Scientific Developer, DNV 
• Ben Williams, Principal Engineer, DNV 
• Cristoffer Kos, Business Development Manager, DNV