End User
SWW
Description
The AI-based Load Forecasting service provides short-term electricity demand forecasts for distribution grid assets and aggregated network areas at 15-minute resolution for horizons up to 48 hours ahead. Using machine learning models (gradient boosting, RNNs, LSTMs) trained on historical consumption, weather, and calendar data, the service supports operational planning, congestion management, flexibility scheduling, and state estimation.
Core Capabilities
Predictive & Prescriptive Analytics
Business Need
Traditional Static Load Profiles (SLPs) inadequately capture the dynamics of modern distribution networks with increasing heat pumps, EVs, and prosumers. Without accurate near-real-time load forecasts, DSOs cannot reliably anticipate demand peaks, assess congestion risks, or validate flexibility dispatch schedules. The service provides the dynamic, adaptive demand predictions needed for modern grid operation.
Key Performance Indicators
Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE)
Minimum 30% improvement over traditional Standard Load Profiles as baseline
Performance at different forecast horizons and aggregation levels
Data Provided
Time-indexed load forecasts with predicted active power demand and optional confidence intervals per asset or grid zone
Inputs: historical smart-meter measurements at 15-min, feeder monitoring data, weather forecasts, calendar information
TEF
TEF DSO