Maximizing waste heat recovery from a building-integrated edge data center.

Small data centers can be integrated into the energy systems of commercial and tertiary buildings to capture waste heat generated by servers, mitigate environmental impacts and enhance energy efficiency. This study introduces a novel methodology for maximizing waste heat capture from the cooling coils by optimizing workload distribution in an edge data center consisting of air-cooled servers. The maximization of outlet temperatures algorithm (MOTA) was developed using a validated fast thermal evaluation approach. Experimental studies were conducted to characterize parameters of the cooling system such as air and water flow rates and effectiveness of the coil. A multi-region conjugate heat transfer (CHT) numerical model was developed to demonstrate feasibility of the proposed approach. Heat transfer through the cooling coil was numerically modeled using the effectiveness-NTU method. Good agreement was achieved between the simulated and measured water temperatures at the outlet of the cooling coil. Numerical simulations conducted using the validated CHT model show that the MOTA can improve heat recovery by up to 17.1% under various IT loads. Furthermore, optimizing the water flow rate can reduce the cooling load by up to 53.2%. These combined results highlight the potential of the proposed algorithm for energy-efficient management of data centers. Computational cost and real-world applicability of the proposed algorithm were discussed in detail.