Energy forecasting and analytics provider Amperon has announced the launch of its weather-informed grid demand Mid-Term Forecast (MTF), giving energy market participants a view of electricity demand up to seven months in advance.
Updated daily with hourly granularity, this product enables traders, utilities and operators to better manage risk, hedge positions and prepare for seasonal and extreme weather events that are reshaping power markets.
This past summer, the company also released new Asset-level Renewable Generation Forecasts that allow renewable energy producers to view solar generation forecasts at the individual site-level or aggregate-level for portfolio management.
Traditional mid-term forecasts, which rely upon historic weather simulations, struggle to deliver accuracy in a rapidly changing climate. Amperon’s Mid-Term Forecast leverages sub-seasonal and seasonal weather forecasts from the European Centre for Mid-Range Weather Forecasts (ECMWF), producing 51 unique load predictions for each U.S. grid daily based on its ensemble weather models. Amperon’s data science team has built a machine learning model that trains the ECMWF ensembles on years of historical weather actuals, providing a spatially and temporally accurate hourly curve of each individual weather variable feeding into the demand model — something that hasn’t existed before. The result is a load forecasting solution that helps market participants anticipate demand and price swings well before they arrive.
This shows Amperon’s demand Mid-Term Forecast for ERCOT North load zone. Built using ECMWF’s 46-day and 7-month weather forecast ensembles, this shows the range of outcomes for resulting grid energy demand, as well as the control, mean and min/max values.
Amperon’s new Grid MTF has already proven highly accurate and reliable at forecasting critical demand events up to four weeks in advance. Historical performance analysis covering January 2024 through July 2025 showed that the MTF successfully captured 67 of 68 extreme demand events (the top 1%) across PJM, ERCOT and ISO-NE, delivering forecasting accuracy above 98%.
“Markets have long relied on backward-looking seasonal averages to predict future demand. That’s like driving by looking in the rearview mirror,” said Sean Kelly, CEO of Amperon. “By integrating real weather forecasts into our models, we’re giving customers the clearest view yet of what demand could look like months from now so they can hedge smarter, plan better and reduce costly surprises.”
Available through both Amperon’s API and user interface at launch, the MTF delivers daily updates with hourly granularity, covering all U.S. grids and zones.
News item from Amperon
