Stahlhut, M.; Nefodov, D.; Urbaneck, T.
Utilizing Waste Heat from Data Centers to Supply Residential Quarters: A Simulation Study
The 18th International Symposium on District Heating and Cooling, Peking, 03.09. bis 06.09. 2023, International Energy Agency Technology Collaboration Programme on District Heating and Cooling
https://www.iea-dhc.org/fileadmin/public_documents/DHC2023_Conference_proceedings_CDHA.pdf
Kurzfassung
Data centers (DCs) and their electrical/electronic components require permanent cooling. Most of the cooling demand in DCs is caused by the servers. Direct liquid server cooling enables an efficient heat removal of the server waste heat. With this technology, the cooling medium (water) leaves the rack with a maximum outlet temperature between TIT,out,max = 50...60 °C. Consequently, free cooling with dry coolers is feasible throughout the year. At the same time, waste heat recovery is conceivable for various applications while also necessary for ecological reasons. This paper therefore investigates the heat utilization of DCs with direct liquid server cooling to supply a representative residential quarter with a peak load of 1855 kW and an annual heat load of 5078 MWh. Raising the temperature level of the DC waste heat to the temperature level of the local heating grid Tgrid,s = 75 °C is done with two heat pumps including two compressors each (refrigerant R1234ze(E)). This study uses transient simulation software TRNSYS to analyze the effect of different DC sizes on the coverage of the heating load of a residential quarter as well as the proportion of reused energy of the DC (ERF, Energy Reuse Factor). The results show that for an IT power capacity Pel,IT,max = 2200 kW the heating load of the grid is fully covered with an ERF = 0.38 MWhth/MWhel (web load profile). A further increase in DC size reduces the ERF value. A ratio bigger than Pel,IT,max/Qgrid,a = 0.83 kWel/(MWhth/a) causes an ERF ≤ 0.2 MWhth/MWhel. With regard to discussions on legal obligations for waste heat recovery from DCs (e.g. Energy Efficiency Act in Germany), the ratio of DC size and annual heat load plays a decisive role. Additionally, a simplified economic analysis complements the findings of this study.
Schlagwörter: waste heat recovery, data center, heat pump, district heating, simulation