Cities across Europe are rapidly transitioning towards Positive Energy Districts (PEDs), i.e., urban areas that produce more energy than they consume over a year. PEDs require more than renewable installations, they need the ability to predict, coordinate, and optimize energy flows in real time. ExPEDite develops an AI-driven energy forecasting and flexibility optimization tool that helps districts understand their future energy behavior and intelligently shift consumption to maximize renewable use. 

The energy forecasting and flexibility optimization tool, which is developed within Task 4.1 of the ExPEDite project, combines three modules:

Demand forecasting module

It offers short-term energy demand forecasts for district buildings and assets for the next hours or days using deep learning models.

Renewable Energy Sources (RES) forecasting

It provides short-term energy generation forecasts for different district RES assets, such as solar panels, for the next hours or days, using deep learning models that estimate future solar production using weather forecasts and PV system characteristics.

Demand Response (DR) module for flexibility

Using market price signals, as well as forecasted demand and renewable generation, the energy forecasting and flexibility optimization tool calculates the optimal way to leverage the available asset flexibility and shift consumption across the next hours. The goal of this module is to minimize electricity costs while increasing the use of RES.

The energy forecasting and flexibility optimization tool allows PEDs to understand when energy will be needed, when renewable energy will be available, and how consumption can be adapted. The result is a flexible, cost-efficient, and low-carbon district energy system.

The first version of the tool has been demonstrated using real data from the Riga Technical University campus, and will be further expanded in the second release with deeper integration into the ExPEDite Digital Twin and Decision Support System.  Stay tuned, more developments will follow upcoming months!