At the heart of the energy transition, the efficiency and flexibility of buildings are becoming increasingly central. Through the REEFLEX Horizon project, the aim is to develop solutions that make buildings not only less energy-consuming but also proactive in managing consumption and interacting with the electrical grid. In this context, the REEFLEX project partner, Enerbrain is proud to be a key technological partner, contributing its expertise in predictive control and energy optimisation. Their role in the REEFLEX project focuses on the development, testing, and scalability of an intelligent control system based on Model Predictive Control (MPC) algorithms.
The REEFLEX Project: An Overview
REEFLEX aims to create opportunities for SMEs and start-ups in the demand-side flexibility market by increasing the participation of energy consumers. Through the development of an interoperability platform and a catalog of services, the project aims to maximize the flexibility of distributed energy resources, taking into account the different profiles of end-users and existing infrastructure limitations. Furthermore, REEFLEX plans to create a common operational market model, using artificial intelligence services and automation systems, along with distributed ledger technologies, to improve transparency and trust, reduce barriers to entry and costs, and encourage greater participation of energy consumers.
Predictive Control for Dynamic Buildings
Model Predictive Control is an advanced approach that allows a system to predict how it will behave over time and act accordingly. In the context of buildings, this means optimising the operation of HVAC (heating, ventilation, and air conditioning) systems by considering multiple variables: external weather conditions, occupancy patterns, energy price time slots, comfort constraints, and the thermal behaviour of the building envelope.
The added value of MPC lies in its ability to anticipate needs rather than simply react. Where a traditional control system turns on equipment once a threshold is exceeded, our MPC evaluates hundreds of possible trajectories to decide when, how, and with how much energy to intervene.
Tailored Solution for Each Building
One of the most innovative aspects of Enerbrain’s approach is the system identification phase, where each building is digitally modelled based on real data collected during everyday operation. Thermal characteristics, space usage, internal and external environmental conditions, and occupancy profiles are all taken into account. For example, a classroom requires control strategies that are very different from those of a restaurant or an office building. This is why MPC does not adopt “predefined” logic but adapts its algorithms to the actual needs of the environment.
This customisation involves a longer initial calibration process compared to other systems but ensures reliable and replicable results over the long term. To simplify this step, we are developing automatic model validation tools and alert notifications to facilitate system refinement
Real-Scale Validation
Our MPC algorithm has been tested in real-world scenarios reflecting the variety and complexity of the European building stock:
- The main pilot site in Zaragoza (Spain) is the focal point for REEFLEX validation. Here, the system is fully operational and integrated with the project’s digital infrastructure, demonstrating its ability to manage energy flexibility in real-time.
- Public and private buildings in various European countries, including schools, offices, and commercial activities, have been involved in testing the system under very different climatic, operational, and behavioural conditions.
- In industrial environments, predictive control has successfully addressed challenges related to reducing load peaks and optimising high-energy consumption production cycles.
These tests have provided valuable feedback, confirming the system’s effectiveness and guiding the refinement of control strategies.
Innovations for Smarter Control
Key features introduced in our MPC system include:
- Dynamic startup: This allows the HVAC system to be activated at the most optimal moment, avoiding premature startups and utilising the building’s thermal inertia.
- Temperature range control: This avoids continuous on/off cycles, aiming for stability within predefined comfort ranges.
- Edge mode operation: The system can operate without cloud connectivity, making it resilient even in contexts with limited connectivity or security constraints.
- MIMO (Multiple Input Multiple Output) logic integration: This allows the system to manage multiple HVAC subsystems in a synergistic and coordinated manner, enhancing overall operational efficiency.
A System Ready for Complex Scenarios
Thanks to its flexible architecture, the MPC developed by Enerbrain is suitable for advanced scenarios like Renewable Energy Communities (RECs). In these contexts, the control must consider not only the needs of the individual building but also the energy coordination among multiple users, optimizing self-consumption and promoting shared load strategies.
Our system is also equipped with a consumption forecasting module, which uses historical data and weather forecasts to plan the building’s behaviour 24 hours in advance. This allows for anticipating demand peaks, reducing energy costs, and proactively responding to signals from the electrical grid.
A Vision for the Future
The experience gained from REEFLEX has been a challenging and strategic testing ground for Enerbrain.
As Marco Martellacci, Enerbrain’s Chief Innovation Manager, highlights:
“The implementation of Model Predictive Control within the Reeflex project demonstrates how advanced algorithms can transform energy management in buildings. By combining predictive capabilities with real-time data, we not only optimize consumption but also create a more efficient interaction with the grid and greater energy flexibility. This is a fundamental step towards a more sustainable and intelligent energy ecosystem.”
The journey doesn’t end here. Thanks to the results obtained in heterogeneous scenarios, the system is ready to face new challenges and guide the building sector towards a more sustainable, digital, and interconnected future.
With the REEFLEX project, Enerbrain reaffirms its commitment to promoting technologies that enable more conscious and flexible energy management. A vision that looks to the future, always starting from real data and the needs of the buildings and people who inhabit them.
Contact Information
For more information about REEFLEX, please contact us at: contact@reeflexhe.eu
Follow us on