一种用于数据中心脱碳的高效太阳能光伏/热协同制冷与发电系统

Abstract Solar photovoltaic (PV) presents a promising solution for decarbonizing data centers. However, during the hot summer, its temperature rises sharply, significantly deteriorating its power generation efficiency. By utilizing the waste heat from PV modules to drive a chemisorption cold energy storage unit, it is possible to provide cooling capacity for data centers while enhancing PV performance. Nevertheless, the high driving temperature required for chemisorption cold energy storage unit is incompatible with the low waste heat temperature of PV modules. A novel high-efficiency solar photovoltaic/thermal (PV/T) cooling and power synergistic system is proposed in this paper, which deeply integrates PV/T modules with a chemisorption cold energy storage cycle and a vapor compression refrigeration cycle. To actively adapt to the low waste heat temperature, a composite sorbent BaCl 2 /expanded natural graphite /carbon-coated aluminum nanoparticles was developed. Experimental results demonstrated that this composite sorbent achieved effective desorption reaction within the typical thermal output temperature range of PV/T modules (50–70 °C), and smooth sorption reaction under ambient cooling conditions (15–25 °C), making it suitable for the proposed system. The system dynamic performance was evaluated through simulation. By implementing a novel asynchronous start-stop control strategy, the sorption and desorption processes of multiple unit-tube reactors were coordinated to match the daytime thermal output of the PV/T modules, thereby reducing PV temperature, enhancing power generation efficiency, and simultaneously ensuring stable nighttime cooling output. Compared to conventional PV systems, the proposed PV/T system improves daytime power generation efficiency by 13 % and provides 7 h of electricity-free cooling at night, resulting in a 20.2 % improvement in comprehensive solar utilization efficiency. An efficient and low-carbon solar-driven cooling and power solution is provided, demonstrating strong potential for practical application in data centers.