Tariff forecasts

This feature is in alpha and not yet generally available.Contact support for early access or more information.

Sometimes you need to know the energy price before it’s available. In most European markets, day-ahead electricity prices are published around 11:00 a.m. — which means your systems may not always have a full 24-hour forecast in time for when you run your planning. Gaps or late arrival of data may also be caused by energy provider system outages. To ensure uninterrupted service, Flatpeak uses machine learning and AI to:

Predict future prices
Populate historical gaps

When this occurs:

The API returns forecasted values with confidence < 1.0
Once the upstream data is available, Flatpeak back-fills with actual data and updates the value to 1.0

This means your system can always operate — even if the market data isn’t there yet. Or there is a connection issue with your customer’s provider systems.

How tariff forecasts work

Flatpeak’s forecasting logic adapts based on the tariff type and the time window in question. Below is a simplified explanation.

1. If there are gaps in data

FIXED / TIME_OF_DAY
Missing periods are interpolated from surrounding values, using the lowest confidence of the adjacent time blocks.
MARKET / DYNAMIC
Gaps are filled using extended historical data for the same period taking into account other factors such as weather.

2. If future data is not yet available

FIXED / TIME_OF_DAY
Flatpeak extrapolates the known pricing pattern into the future. Forecasts start at 0.8 confidence and decay by 0.2 every 90 days, to a floor of 0.2.
MARKET / DYNAMIC
Predictions use a blend of market prices, provider surcharges, weather forecasts and historical patterns.