Reduce imbalance costs with breakthrough Large-Eddy Simulation (LES) weather modelling technology

In renewable energy trading, weather unpredictability and inaccurate forecasting can lead to substantial financial risks and missed opportunities. Addressing these challenges, Whiffle has developed an advanced weather model that combines atmospheric modelling, high-performance computing, and machine learning to deliver forecasts of unmatched precision. This breakthrough model provides forecasts of any meteorological parameter at resolutions of 100m or finer, empowering traders and asset owners with the critical data they need to make informed decisions that ensure reduced imbalance volumes. 

Large-Eddy Simulation (LES): A game changer in weather forecasting 

At the core of Whiffle’s solution, is the world's first operational weather model based on Large-Eddy Simulation (LES). This model employs a sophisticated pre-processing engine to transform large-scale weather data into customized input fields. The data is then processed through a multi-core CPU/GPU, which enables the production of short-term, hyper-local forecasts for variables such as wind speed and solar irradiance, at a resolution of 100m or finer. It captures complex phenomena such as wakes, low-level jets, stability, fog, and cloud formations, even in the most challenging terrains. 

Whiffle's model output is not only a unique weather signal but is also further enhanced by state-of-the-art machine learning algorithms and ensembles for meteo and power production forecasts. This results in exceptionally accurate wind and solar forecasts, that reduce imbalance volumes by 10%. 

“Whiffle's advanced Large-Eddy Simulation weather modelling technology marks a new era in forecasting. It combines real-world physics with unparalleled computing power, offering a level of clarity and precision that is essential for energy traders. In a market heavily influenced by the weather, our technology provides traders with the detailed insights needed to make informed decisions, linking meteorological factors to quantifiable financial impacts," - Remco Verzijlbergh , Co-founder and CTO of Whiffle. 

Key features: 

  • Forecast the weather on a resolution < 100m 
  • Captures wakes, low-level jets, stability, fog, and cloud formations, even in complex terrain 
  • Short-term forecasts for meteo and power production 
  • Intraday updates  
  • Probabilistic and deterministic forecasts  
  • Asset, park, or portfolio level 
  • Real-time observation ingestion  

In a sector where accuracy determines risks or rewards, the precision of Whiffle’s meteo and power production forecasting solutions help its clients (energy traders, asset operators or utilities) to minimize forecasting errors, reduce imbalance volumes and maximize their net revenues. 

Boilerplate 

Whiffle B.V. was founded in 2015, starting as a spin out of the Dutch Delft University of Technology. With its roots in science, the company continues to use cutting-edge R&D to further develop Large Eddy Simulation (LES) models and its unique implementation on high performance computing systems. This resulted in the world’s first Large Eddy Simulation (LES) based operational weather model that produces highly accurate and ultra high-resolution weather forecasts. Application areas of Whiffle’s model include short-term wind and solar power forecasting and wind resource assessments 

Press contact 

Tasneem Hooghart 

Marketing & Communications Manager 

Tasneem.hooghart@whiffle.nl 

Neueste Nachrichten

Press Releases
Press Releases

E-world 2025: Four Forums Offer an Innovative Program on the Energy Transition

The high-profile expert forums are accessible free of charge with a trade fair ticket

Press Releases

E-world Kick-off 2025 Focuses on the Transformation of the Energy Sector

Exchange with top-class speakers from November 19 to 21, 2024

Press Releases

E-world 2025: Get your tickets now

Ticket launch is accompanied by the relaunch of the trade fair logo

Newsletter

Register for the official E-world Newsletter and stay updated!