WIND FARM ICING LOSS FORECASTS
Minimize Forecasting Imbalance Costs Due to Icing
Generic weather services can’t forecast icing events. IceLossForecast is a state-of-the-art tool tailored for forecasting icing losses, giving BRPs, operators, and TSOs a competitive edge in intraday and day-ahead trading.
✓ Used by 25+ clients and over 70 wind farms across the Nordics, Baltics and Central Europe
✓ Trusted by Fingrid (Finnish TSO) since 2023
✓ Payback time: 3 to 10 weeks
Trusted by leading wind energy companies worldwide
How it works
Step 1: Setup
We configure high-resolution icing forecasts tailored to your wind farm location and turbines
Step 2: Forecasting
Global models (GFS, ECMWF) are downscaled to local conditions using physical and machine learning models
Step 3: Delivery
Forecasts delivered via API or SFTP every 6 hours, with hourly updates available via SCADA tuning
Step 4: Integration
Easily integrate into your existing production forecast system using a simple icing loss factor
Proven Business Case
For a typical Finnish wind farm with 4% annual icing losses, IceLossForecast reduced imbalance costs by an average of 21,000 EUR per year over 8 years, totaling 168,000 EUR in savings.
Payback time is 8 to 10 weeks conservatively, and just 3 to 4 weeks during recent high-icing winters 2022/2023 and 2023/2024.
Benefits
Reduce Imbalance Costs
Accurate icing forecasts for intraday and day-ahead trading reduce costly forecast errors during icing events
Easy Integration
Icing loss provided as a simple factor (0 to 1) that multiplies your existing production forecast. No complex system changes needed
Continuous Improvement
Historical and real-time SCADA tuning keeps forecasts aligned with your turbines’ actual behavior
Advanced Ensemble Forecasting
We combine multiple weather models with a unique ensemble method to capture uncertainties in temperature, cloud placement, and more.
Forecasts include deterministic values, ensemble means, min/max ranges, and probability of icing losses.
Forecast horizon covers every hour for at least 48 hours ahead, with extended 7-day forecasts available. We target a 70%+ correct hit-rate for icing events.
15+
Active Clients
50+
Wind Farms Served
21,000 EUR
Average Annual Savings per Wind Farm
70%+
Icing Event Hit Rate
Get in Touch About Icing Forecasts
Tell us about your wind farm and we will come back to you with a recommended approach for icing loss forecasting.
Mona Kurppa
Christoffer Hallgren
Frequently Asked Questions
How are the forecasts generated?
We combine global models (GFS, ECMWF) and downscale them to local conditions. The system uses both physical modelling (based on our validated IceLoss methodology) and machine learning to estimate icing mass on turbine blades and calculate wind farm icing losses.
How often are forecasts updated?
Every 6 hours by default. With real-time SCADA tuning, updates are available hourly.
What is the forecast horizon?
Every hour for at least 48 hours ahead. Extended 7-day forecasts are also available.
Is it difficult to integrate into our existing systems?
No. The icing loss is provided as a simple factor between 0 and 1. Multiply your existing production forecast by this factor. Delivery via API or SFTP in CSV format.
How accurate are the forecasts?
Accuracy depends on your goals, whether minimizing imbalance costs, reducing MAE, or detecting high-loss events. We offer historical forecast evaluation so you can assess performance before committing. Our target icing event hit-rate is above 70%.
Ready to Reduce Your Icing-Related Imbalance Costs?
With 15+ active clients, nearly 50 wind farms served, and proven savings of 21,000 EUR per year per wind farm, IceLossForecast by Kjeller Vindteknikk is the leading icing forecast tool for wind energy operators in the Nordics.
Norway – Taerudgata 16, 2004 Lillestrom
Sweden – Fleminggatan 7, 112 26 Stockholm
Finland – Tekniikantie 14, 02150 Espoo
Iceland – Hlidasmari 4, 201 Kopavogur