Periodic Reporting for period 1 - HEATWISE (Holistic Energy management And Thermal Waste Integrated System for Energy optimization)
Reporting period: 2024-01-01 to 2025-06-30
In response to key transformative trends shaping the data center market, notably the EU's emphasis on environmental efficiency and renewable energy, the increasing demand for low-latency distributed computing environments, and the heightened power and cooling needs driven by server energy density, the HEATWISE project tackles the integration of waste heat and thermal energy management in tertiary buildings with significant IT loads for the minimization of waste heat, maximization of energy efficiency, and generation of additional value. Through pilots in Denmark, Poland, Switzerland, Sweden and Türkiye HEATWISE aims to pioneer innovative solutions, encompassing hybrid (liquid and air) cooling technology, digital twin-supported holistic data management system, integrated multi-objective building energy management system, and a self-assessment and optimization tool, complemented by a knowledge-sharing platform.
The key activities of the HEATWISE project in the period of 2024 January – 2025 June (M1-M18) are summarized below.
The historical energy data sets have been collected for Empa and AAU. The waste-heat potential and load profile of pilot buildings - Empa, AAU, TOFAS and PSNC - have been collected and analyzed.
A CFD model was developed for the simulation of film boiling validated by numerical data in the literature, and a baffle design was proposed for the heat transfer enhancement in film boiling. A respective paper has been published as open-access publication.
A CFD model for thermal optimisation of data centres were developed and validated with the experimental data for air-cooled pilots Empa, AAU, PSNC and TOFAS and retrofitting alternatives were developed according to the CFD results to enhance thermal and cooling efficiencies.
A KPI-informed computational framework was proposed for the efficiency of the improvement of data centres and results.
A multi-objective thermal-aware algorithm (MOTA) was developed to maximize waste heat recovery from building-integrated edge data centres. Numerical simulations demonstrated that the MOTA improved waste heat recovery by up to 17% and reduced cooling load by 53%, demonstrating strong potential for energy-efficient operation in edge data centres.
The technical and legal feasibility assessment resulted that due to export bans ZutaCore’s cooling solution cannot be implemented at TOFAS pilot and instead it is going to be implemented at RISE.
The hybrid cooled EMPA pilot has been designed, built and passed all the tests by M17.
A data management platform has been developd to which the EMPA and RISE pilots got connected. The concept of the software landscape of pilots has been prepared, and the digital twin devbelopment has been started.
A set of KPIs for the Data Centre-Building ecosystem has been identified.
The development of a self-assessment tool has been started and the first iteration was presented to the consortium.
The historical energy data sets have been collected for Empa and AAU. The waste-heat potential and load profile of pilot buildings - Empa, AAU, TOFAS and PSNC - have been collected and analyzed.
A CFD model was developed for the simulation of film boiling validated by numerical data in the literature, and a baffle design was proposed for the heat transfer enhancement in film boiling. A respective paper has been published as open-access publication.
A CFD model for thermal optimisation of data centres were developed and validated with the experimental data for air-cooled pilots Empa, AAU, PSNC and TOFAS and retrofitting alternatives were developed according to the CFD results to enhance thermal and cooling efficiencies.
A KPI-informed computational framework was proposed for the efficiency of the improvement of data centres and results.
A multi-objective thermal-aware algorithm (MOTA) was developed to maximize waste heat recovery from building-integrated edge data centres. Numerical simulations demonstrated that the MOTA improved waste heat recovery by up to 17% and reduced cooling load by 53%, demonstrating strong potential for energy-efficient operation in edge data centres.
The technical and legal feasibility assessment resulted that due to export bans ZutaCore’s cooling solution cannot be implemented at TOFAS pilot and instead it is going to be implemented at RISE.
The hybrid cooled EMPA pilot has been designed, built and passed all the tests by M17.
A data management platform has been developd to which the EMPA and RISE pilots got connected. The concept of the software landscape of pilots has been prepared, and the digital twin devbelopment has been started.
A set of KPIs for the Data Centre-Building ecosystem has been identified.
The development of a self-assessment tool has been started and the first iteration was presented to the consortium.
KPIs
The project identified seven thermal KPIs for buildings, out of which the waste heat utilisation restrictor (WHRU) was proposed to be used as the main optimisation and decision making KPI in the DC-Building interaction.
CFD simulation tool
A web-based, scalable software has been developed based on open-source libraries for the thermal simulations of air- and liquid-cooled data centres. Accuracy and reliability of the software have been validated through extensive benchmarking against experimental data from three pilot sites demonstrating its capability to optimize thermal performance and energy efficiency in real-world applications. This technology has the potential to significantly improve energy efficiency, reduce operational costs, and contribute to the sustainable development of next-generation data centres.
CFD simulation methodology validation by HEATWISE pilots
With the help of the open-source CFD simulation tool developed by DS Tech, the numerical models of three pilots have been prepared to validate a new systematic methodology to assess and improve the performance of existing data centres. Two pilots, namely AAU and PSNC provided external sensor data of servers’ inlet and outlet tempereatures, while EMPA pilot provided internal sensor data via Intelligent Platform Management Interface (IPMI). The experimental measurements validated the accuracy and robustness of the simulation model. The simulation of different improvement measures e.g. air deflectors or repositioning the racks have shown significant reduction potential in cooling energy use for PSNC, AAU and EMPA pilots reaching 16.4%, 26.92% and 4.21%, respectively.
Waste heat maximisation algorithm (MOTA)
The MOTA (Multi-Objective Thermal-Aware) algorithm was developed to maximize waste heat recovery from a liquid-cooled data centre (Empa). This solution goes beyond the state-of-the-art, as the maximization of waste heat recovery through workload optimization has not been previously addressed, either experimentally or numerically. The numerical simulations showed an improved heat recovery by up to 17.1% under various IT loads, while the optimization of water flow rate a reduction of cooling load by up to 53.2%. The results of this study are submitted to the Case Studies in Thermal Engineering journal, in collaboration with RISE and Empa. To facilitate widespread adoption, MOTA should be integrated into workload management platforms such as Prometheus and OpenStack, ensuring an implementation within existing data centre infrastructures.
Development of a micro-baffle simulation model for two-phase heat transfer
As a beyond state-of-the-art innovation, a micro-baffle is designed and optimized to enhance heat transfer in film boiling, a critical mechanism in two-phase cooling systems. This novel approach improves thermal performance by effectively managing vapor flow and liquid replenishment, leading to higher heat transfer rates (a 57.4% improvement in the Nusselt number and a 66.4% increase in critical heat flux). The results of this study have been validated and published as an open-access paper.
Empa pilot
HEATWISE Empa pilot demonstrates holistic energy management and thermal waste integrated energy management system for energy management and integration of edge data center at Empa’s NEST building. The hardware set-up for the Empa pilot is complete and it is live since the end of May 2025. The Empa team is working on setting up the servers and acquisition and logging of the data in the NEST cloud. It demonstrates hybrid cooling, whereby 70% of the waste heat is going to be captured through Zutacore’s dielectric, direct, phase-changing liquid cooling technology and the other 30% is captured through air cooling. This pilot demonstrates the recovery and reuse of high-grade waste heat from edge data centres, integrated into the NEST high and medium temperature grid.
The project identified seven thermal KPIs for buildings, out of which the waste heat utilisation restrictor (WHRU) was proposed to be used as the main optimisation and decision making KPI in the DC-Building interaction.
CFD simulation tool
A web-based, scalable software has been developed based on open-source libraries for the thermal simulations of air- and liquid-cooled data centres. Accuracy and reliability of the software have been validated through extensive benchmarking against experimental data from three pilot sites demonstrating its capability to optimize thermal performance and energy efficiency in real-world applications. This technology has the potential to significantly improve energy efficiency, reduce operational costs, and contribute to the sustainable development of next-generation data centres.
CFD simulation methodology validation by HEATWISE pilots
With the help of the open-source CFD simulation tool developed by DS Tech, the numerical models of three pilots have been prepared to validate a new systematic methodology to assess and improve the performance of existing data centres. Two pilots, namely AAU and PSNC provided external sensor data of servers’ inlet and outlet tempereatures, while EMPA pilot provided internal sensor data via Intelligent Platform Management Interface (IPMI). The experimental measurements validated the accuracy and robustness of the simulation model. The simulation of different improvement measures e.g. air deflectors or repositioning the racks have shown significant reduction potential in cooling energy use for PSNC, AAU and EMPA pilots reaching 16.4%, 26.92% and 4.21%, respectively.
Waste heat maximisation algorithm (MOTA)
The MOTA (Multi-Objective Thermal-Aware) algorithm was developed to maximize waste heat recovery from a liquid-cooled data centre (Empa). This solution goes beyond the state-of-the-art, as the maximization of waste heat recovery through workload optimization has not been previously addressed, either experimentally or numerically. The numerical simulations showed an improved heat recovery by up to 17.1% under various IT loads, while the optimization of water flow rate a reduction of cooling load by up to 53.2%. The results of this study are submitted to the Case Studies in Thermal Engineering journal, in collaboration with RISE and Empa. To facilitate widespread adoption, MOTA should be integrated into workload management platforms such as Prometheus and OpenStack, ensuring an implementation within existing data centre infrastructures.
Development of a micro-baffle simulation model for two-phase heat transfer
As a beyond state-of-the-art innovation, a micro-baffle is designed and optimized to enhance heat transfer in film boiling, a critical mechanism in two-phase cooling systems. This novel approach improves thermal performance by effectively managing vapor flow and liquid replenishment, leading to higher heat transfer rates (a 57.4% improvement in the Nusselt number and a 66.4% increase in critical heat flux). The results of this study have been validated and published as an open-access paper.
Empa pilot
HEATWISE Empa pilot demonstrates holistic energy management and thermal waste integrated energy management system for energy management and integration of edge data center at Empa’s NEST building. The hardware set-up for the Empa pilot is complete and it is live since the end of May 2025. The Empa team is working on setting up the servers and acquisition and logging of the data in the NEST cloud. It demonstrates hybrid cooling, whereby 70% of the waste heat is going to be captured through Zutacore’s dielectric, direct, phase-changing liquid cooling technology and the other 30% is captured through air cooling. This pilot demonstrates the recovery and reuse of high-grade waste heat from edge data centres, integrated into the NEST high and medium temperature grid.