Applying AI and Real-Time computing in the Energy Sector


Real-Time AI Computing in the Energy Sector

Glaze and Beacon Tower are together with Energy Cluster Denmark, Develco, Brønderslev Forsyning A/S, Agerkrantz Controls ApS and AAU Civil Engineering developing a solution for tomorrow’s district heating systems.


Energy consumption in buildings constitutes almost 40% of the total energy consumption in Denmark. The energy is primarily used for heating, ventilation, and lighting. Optimizing the district heating grid is one of the possibilities to obtain energy savings in this field. The impact of lowering the return temperature with just 1gC in the grid are saved energy in 10.000 houses or 9,500 tons CO2. Making energy consumption in buildings more efficient plays a decisive role in order for us to achieve the political goal of being independ-ent of fossil fuels by 2050.


The challenge today in the district heating industry, is that data are used in isolated silos or dedicated systems not prepared for data exchange needed for Machine Learning and data sharing amongst systems. Further, the existing vendors in the district heating industry is using proprietary solutions with vendor lock-in which decreases the flexibility and cross-sector integration that is needed. This is a barrier for obtaining energy efficiency and energy savings in the entire grid as achieving the next level of optimizations re-quires real-time data analysis.

Real-time grid data, cloud computing, Machine Learning and AI Control enable realization of the next generation of district heating systems that have been much anticipated the recent years. Knowing how to apply these technologies in a flexible, secure and thus scalable way will be key to export district heating (as well as cooling solutions ) as modern and future-proofed energy solutions to new markets.

Optimizing district heating in the future requires a combination of advanced technology solutions, flexible integrations and expertise in energy management and systems engineering. To achieve these optimizations, it requires a digitization of the district heating sector by having real-time input from across the grid and a cross-sector integration to different energy and data sources. This increases the amount of data that must be processed, stored, and analyzed and thus imposes requirements of openness and flexibility to the data storage.

Project description

The overall goals of the project are:

  • Develop a flexible and modern district heating platform that is minimising reliance of fossil energy and targeted export to visionary smart city projects using innovative business models.
  • Contribute to the Danish and European AI-community within the energy sector.
  • Develop Machine Learning & AI control systems that utilize cross-sector real-time data to conduct near real-time operational optimization.
  • Demonstrate that the solution optimizes the energy production and distribution by utilizing a newly patented real-time flow-sensor and HydroState Control Box, an open and scalable cloud-based data platform and application of state-of-the-art AI control techniques on multiple real-time as well as external data sources
  • Proof that the solution can save 10% of district heating companies’ yearly energy consumption by reducing power consumption of pumps, reduced heat loss from grid pipes by lower supply and return temperature and higher efficiency at boilers and heat-pumps compared to existing district heating companies that do not have real-time grid data and cross-sector AI-driven control.

In RACE a sensor solution for real-time measuring of pressure, temperature and flow in grid valves is developed. The measured pressure from the grid is used to set accurate pressures on main and booster pumps. We utilise and further develop Beacon Tower and its open and flexible architecture as the data platform for acquiring and structuring machine and time-series data that enables open sharing of data between systems. We develop solutions for decision support for systemwide planning using AI algorithms trained on Machine Learning generated predictive digital twins learned from datasets spanning whole seasons. We utilise general solutions for integration to typical district heating OT-systems, the solutions are based on open and modern industry standards. We develop a solution for controlling district heating valves from SCADA based on AI control technology. All the above is done together by testing on a real district heating company.

This is a project backed by EUDP – The Energy Technology Development and Demonstration Programme in Denmark. The project has a budget of appr. 12,6 M DKK.

More information

Are you interested in hearing more? Then feel free to contact Jakob Appel, Managing Partner at Glaze, at or +45 26 17 18 58 for more information.