Company
nMAE (normalised Mean Absolute Error) is the average gap between forecast and actual production, expressed as a percentage of the plant's installed capacity: the lower, the better. To compare providers fairly, use the same plants, the same period and the same forecast horizon. In our real day-ahead cases, nMAE is 5.98% for wind and 1.67% for solar; on a 120 MW wind farm this cut imbalance costs by 25–38%, worth around €1 million per year.
Yes, we recalibrate our models weekly using the latest data. As the model continuously improves its understanding of the real plant's behavior, the nMAE improves accordingly.
Renewcast delivers one forecast: the continuous one. It's always the latest, built on the newest weather run, the best-performing model and the most up-to-date digital twin of your asset. It's refreshed every 30–60 minutes, and every export, API call or Portal view gives you the latest version. Use it for both operations and bidding. Day-ahead is not a separate delivery. It's a label we apply after the fact to the last continuous forecast issued before your market closure. We use it only to measure performance. It never changes and can appear a couple of hours after the cutoff, so there's no need to wait for it.
A model can win every benchmark and still cost you money. We work on what actually drives your financial result: local calibration, correction based on each site's historical patterns, up to 25 technical parameters per asset, and forecasting down to the individual turbine.
For wind, an AI-native digital twin of each plant, calibrated for high winds and rapid ramps. For solar, a physical model of the plant corrected by a machine learning layer. For every site, we dynamically select the 3 best performing weather models from a pool of around 15.
Yes. We provide an hourly probabilistic forecast up to 2 days ahead, with 80% average accuracy and 9 out of 10 icing events detected.
A smaller gap between forecast and actual production means lower imbalance costs. In our real day-ahead cases, nMAE is 5.98% for wind and 1.67% for solar; on a 120 MW wind farm this cut imbalance costs by 25–38%, worth around €1 million per year. We will estimate the saving on your own portfolio together with you.
We can model any location in the world. The platform is fully scalable and already supports large, multi-asset portfolios.
We work with single plants as well as large portfolios. There is no minimum size requirement.
We forecast wind and solar assets on the same site, each with its own dedicated model. We also provide support for hybrid sites and co-located storage.
Yes. In addition to plant-level forecasts, we can deliver aggregated forecasts by portfolio.
An annual subscription per plant or per MW, depending on portfolio size. Alternatively, a performance-based model: a base price plus a bonus tied to the results delivered, so that our fee follows the value we generate for you.
Plant technical data, historical energy injected into the grid (ideally 24 months), curtailments and unavailability. For new plants, we start with a physical model that evolves into an AI-powered one as production data accumulates, with accuracy increasing progressively.
We cover intraday, day-ahead, 48 hours and medium term up to 15 days, at 15-minute resolution (down to 5 minutes). Forecasts are updated continuously: the platform checks for weather updates every 15 minutes and recalculates the forecast whenever weather or plant operating conditions change.
We manage the complexity of your data with you: with your support we clean, correct and enrich generators and plant data.
Three weeks: the service goes live within three weeks of receiving and validating all the required data.
Certainly. We will agree with you on the best way to test the service on your plants.
Via API or SFTP, with customizable exports, or through our web platform with reports and data downloads. We integrate with your existing systems.
You can send planned and unplanned unavailability through the API, and the forecast is recalculated as soon as plant operating conditions change.
A dedicated Client Operations team supports you with integration and runs regular performance reviews with you, with support available during business hours and documented SLAs.
Your data remains yours. We use it only to deliver your forecasts and handle it in line with GDPR and all applicable data protection regulations. The platform is a cloud-native solution built on Azure, designed to meet cybersecurity requirements and fully operational on large, multi-asset portfolios.
Book a demo and speak directly with a member of our tech team. They'll look at your specific situation and help you find the right answer.
Get in touch with Renewcast
_10%20(1)%20(1).png)