Cool-Data: Flexible Cooling of Data Centers

The large energy demand for cooling data centers (DCs) is jeopardizing the effort to reach the deep decarbonization target. The cooling solution includes an innovative phase-change material (PCM) storage and state-of-the-art controller system using artificial intelligence (AI) developed during the project. The AI-based controller integrates the thermal storage to modular cooling equipment, resulting in a more efficient and flexible cooling system. The PCM storage will support the flexible use of cheap and low-carbon electricity and waste heat. The project includes a systemic perspective to assess the impacts of the scaling up of the solution on greenhouse gas emissions and energy transition. We exploit the synergy between ICT and energy sector and develop new business models and regulatory pathways that maximize the benefits of storage in integrated DCs.
Cool-Data will result in the installed and tested integrated cooling systems at the end users' industrial sites, at two of the project partners (GEV Grinsted and Naviair). Cool-Data will create growth and exporting opportunities by developing a ready-to-be-implemented and exportable smart cooling solution.
Cool-Data addresses a critical challenge to the transition identified in Denmark and Europe and addresses the need formulated by the European Commission in its Code of Conduct for Energy Efficiency in DCs to limit their CO2 emissions.

The project is funded by the Innovation Fund Denmark.

A team of researchers from DTU (DTU Compute, DTU Civil Engineering and DTU Management), together with a number of companies including Center Denmark, will develop a new cooling and storage system for small and medium-sized data centers. Enhanced by Artificial Intelligence methods, the researchers behind the Cool Data project expect to be able to reduce the data centers' energy consumption by up to 80 percent and in the same time to reduce CO2 emissions. This will be done by operating the data centers in a flexible way, using different real time inputs.