A power utility cut its load-forecast error from 18–25% down to 2.75%.
An 18–25% swing between load forecast and actuals meant costly real-time market exposure. The strategic target was under 4%. We delivered forecasting that brought average variance to 2.75%.
Challenge
The utility ran, on average, an 18–25% swing between its load forecasts and actuals — leaving it exposed to high clearing prices in the real-time market every time it missed. Leadership set a strategic KPI to bring that variance under 4%.
Approach
We delivered a forecasting solution built on weatherized load profiles and the drivers that actually move demand — tuned to minimize real-time exposure rather than just fit history — and tracked it against the sub-4% KPI in production.
Technology
Load forecasting on weather-driven features and load profiles, delivered on a modern cloud analytics and BI stack, tracked against a real-time-exposure KPI.
Outcome
Average forecast-to-actual variance came down to 2.75% — inside the sub-4% target — cutting the utility's exposure to high real-time clearing prices and turning load forecasting from a recurring liability into a managed number.
We publish our case studies anonymized as a matter of policy. The engagement and every figure here are real — we describe the work, not the client’s name.
Explore the analytics
Synthetic data — a hands-on view of the kind of dashboard this engagement delivered.
Supply, Demand & Energy Efficiency Analytics
Supply–demand balance and energy-efficiency impact — synthetic data.