Why wind engineering models fall short as offshore projects scale - and how DNV's new model closes the gap
Join us for an insightful webinar on how next-generation turbine interaction modelling can strengthen early-stage design decisions, improve certainty at financial close and reduce discrepancies during IE reviews for the larger scale projects of tomorrow.
| Date: | Thursday, 26 March 2026 |
| Time: | 15:00 - 16:00 CET/10:00 - 11:00 EDT |
| Location: | Online |
Over and underprediction of energy yield remains a major risk in offshore wind financing - especially as wind projects scale into multi‑GW clusters and neighbouring-project effects become harder to quantify.
DNV benchmarking of energy predictions for 30 future offshore wind farms shows a widening gap between engineering models and high-fidelity CFD simulations - with differences of up to 5% on annual energy yield at the largest project scales. The gap grows as turbine interaction losses increase, meaning the risk of overestimated AEP compounds as projects scale. Because high-fidelity CFD resolves more of the controlling physics, it generalises more reliably across project scales than engineering models alone - a gap that translates directly into overestimated yield, IE review discrepancies, and financial risk at project close.
Join DNV experts for a practical, data‑driven webinar exploring how DNV’s new CFD.ML model - a machine‑learning surrogate trained on DNV’s high-fidelity WRF‑to‑CFD simulations - delivers improved accuracy in minutes.
What you’ll learn:
- How different turbine interaction models compare to DNV’s high-fidelity CFD for turbine interaction loss predictions, and why a gap widens between current engineering models and CFD at the larger project scales of tomorrow
- Why patched engineering approaches are unlikely to accurately predict wake and blockage losses for future wind farm clusters - and how CFD.ML models both effects together
- How you can apply DNV‑backed methods to capture long‑range wake impacts from neighbouring and future wind farms that engineering models overlook
- How to align your internal yield assessments with DNV's independent engineering methodology before financial close, reducing discrepancies during external review
- How a multi-fidelity approach - running CFD.ML iteratively through development, and full CFD where it adds most value closer to financial close - reduces uncertainty across the project lifecycle and gets you to a more defensible number sooner
- How CFD.ML enables you to run iterative analysis with improved accuracy using CFD-trained models, for layout optimization, scenario comparisons and repowering or extension assessments.
Agenda:
- The growing gap between engineering models and high‑fidelity CFD
Results from DNV analysis and validation of the performance gap between CFD, CFD trained models, and engineering models, and the practical implications. - How CFD.ML works and what it captures that engineering models miss
An explanation of the CFD.ML approach, including validation results, site‑specific atmospheric conditions, cluster‑to‑cluster wake effects, and unified wake and blockage modelling, with real project examples. - CFD.ML in practice
How the wind advisory team uses CFD.ML as standard in offshore energy assessments and the resulting commercial impacts. - How to use CFD.ML the way DNV does
Practical guidance on applying CFD.ML effectively within project workflows. - VindAI integration
How CFD.ML integrates with VindAI, explained by the VindAI team. - Q&A
Questions or support? Contact our team
