Coordinating the Energy-Water Nexus in Data Centers: A Synergistic Framework for Cooling Efficiency and Operational Cost Reduction

Abstract

Data Centers (DCs), as an important component of information storage and processing infrastructure, have emerged as a major contributor to global energy consumption. Traditional research often modeled workload optimization, cooling system performance improvements, and waste heat recovery independently, overlooking potential synergies between DCs and adjacent water networks. In this paper, we propose a coordinated optimal scheduling framework that integrates DC operations with water network management. By incorporating models for time-of-use electricity pricing, dynamic server power consumption, cooling water temperature control, and reservoir level regulation, the framework provides a systematic foundation for coordinated scheduling, thereby effectively optimizing both energy consumption and operational cost. The model comprehensively accounts for fluctuations in electricity prices and variations in workloads over different periods. It also incorporates water flow regulation and reservoir management in the cooling system into the scheduling constraints, ensuring efficient circulation of cooling water and effective reuse of thermal energy. Moreover, by introducing temperature target constraints and penalty terms, the server temperature is effectively controlled to ensure the long-term stable operation of the system. Simulation analyses demonstrate that the model significantly reduces overall energy consumption and power costs while exhibiting strong dynamic adaptability, robustness, and improved energy utilization.