【作者】 刘金；

【导师】 戴源德；

【作者基本信息】 南昌大学 ， 动力工程及工程热物理， 2025， 硕士

【摘要】 数据中心作为支撑算力的基础设施,不仅为信息产业和数字经济高质量发展提供助力,更在人工智能技术与大数据应用的快速迭代中发挥关键作用。然而,传统的数据中心依赖电网供电、冷却方式效率低下,出现高能耗高碳排放的问题,引起业内的极大关注。为实现节能减排,促进数据中心的绿色低碳发展,本文拟研究从设备性能角度提高冷却系统能效、构建适用于数据中心的多能源协同综合能源系统、分析不同气候条件下的综合能源系统表现三个方面开展研究工作,探索数据中心节能减碳的有效途径。首先,拟采用磁悬浮冷水机组提高冷却系统能效。磁悬浮冷水机组具备优秀的性能表现,是一种高效的制冷设备,但是其初始投资较高,难以在节能效率和经济潜力之间做出取舍。针对案例数据中心现有冷却系统构建风冷系统能耗模型,同时提出分别采用变频离心式冷水机组和磁悬浮冷水机组的两种水冷系统改造方案。根据冷水机组制造商提供的数据点拟合高精度的冷水机组性能曲线,构建三种冷却系统能耗模型以分析系统能效和经济性表现。在案例数据中心所在地,磁悬浮冷水机组的节能改造方案自然冷却时长达3813小时,相比原风冷系统节能率达64.54%,PUE降低0.349。并且,磁悬浮冷水机组COP表现优于变频冷水机组,年平均值高出6.7%,使得采用磁悬浮冷水机组的节能改造方案相比变频离心式冷水机组可额外节省3.96%的年能耗。在此基础上,分析了五个典型气候城市的改造方案能效和经济性表现。尽管水冷系统改造方案初始投资较高,但两种改造方案的经济性均表现良好,对比原风冷系统,变频离心式冷水机组的改造方案投资回收期在2.02-5.15年,磁悬浮冷水机组的动态投资回收期在2.16-5.6年。其次,针对数据中心可再生能源利用率低、碳排放高的特点建立多能源协同的综合能源系统,该系统由冷却塔和地下水集成的多冷源冷却子系统、建筑光伏一体化子系统(BIPVS)和余热回收子系统组成。通过各个子系统的热电耦合,所提出的综合能源系统可以高效和低碳运行。以案例数据中心为例,对综合能源系统与数据中心原系统和分立能源系统(水冷精密空调系统和热泵系统独立运行)两者进行对比分析,研究综合能源系统在冷却、制热、系统整体能效和经济性方面的表现。结果表明,在供暖季,综合能源系统中冷水机组的性能系数(COP)比分立系统提高0.4-2.55,而热泵的COP提高1.4-2.1。与原系统和分立系统相比,综合能源系统的年能耗分别降低72.98%和21.3%,并减少68.34%和19.52%的碳排放,其动态投资回收期分别为2.87年和3.54年。最后,本研究基于气候因素对综合能源系统的多维度影响,选取中国五个典型气候区的20个代表性城市作为研究对象。通过构建能源系统能耗模型,对比分析了综合能源系统、分立能源系统以及原系统的能效、环保及经济性等各项关键指标。重点揭示了气候因素对多能源协同利用的影响机制。模拟过程将各气候区域的环境温度、太阳辐照强度、风速、地下水温度等多种参数变化纳入分析。结果表明,所选20个城市中,综合能源系统的能源使用效率显著高于其他两个系统。在降低碳排放方面,与原系统相比,综合能源系统可减少69%-79%的碳排放量,与分立系统相比也可降低17%-31%的碳排放。从经济性角度分析,考虑资源价格差异,综合能源系统的生命周期成本(LCC)比所选20个城市的分立系统低4%-18%。然而,需要特别指出,在气候严寒且资源价格低廉的城市,综合能源系统的经济表现不如原风冷系统。更多还原

【Abstract】 As the infrastructure supporting computing power,data centers not only contribute to the high-quality development of the information industry and the digital economy,but also play a crucial role in the rapid iteration of artificial intelligence technology and big data applications.However,traditional data centers rely on grid power supply and have inefficient cooling methods,resulting in high energy consumption and high carbon emissions,which has drawn significant attention from the industry.To achieve energy conservation and emission reduction and promote the green and low-carbon development of data centers,this paper intends to conduct research from three aspects:improving the energy efficiency of the cooling system from the perspective of equipment performance,constructing a multi-energy collaborative integrated energy system suitable for data centers,and analyzing the performance of the integrated energy system under different climatic conditions,in order to explore effective ways for energy conservation and carbon reduction in data centers.Magnetic bearing chillers have excellent performance,but their initial investment is relatively expensive,making it difficult to balance energy-saving efficiency and economic potential.For the existing cooling system of the case data center,an air-cooled refrigeration system energy model is constructed,and two water-cooled system retrofit plans were proposed,using variable frequency centrifugal chillers and magnetic bearing chillers,respectively.Based on the data provided by the chiller manufacturers,high-precision chiller performance curves were fitted.The energy models of the three cooling systems are constructed to analyze the system energy efficiency and economic performance.By system simulation,the energy efficiency and economic benefits of the three cooling systems in the case data center are analyzed.In the case data center,the energy-saving retrofit plan using the magnetic bearing chiller has annual free cooling period of 3813 hours,with an energy-saving rate of 64.54%compared to the existing air-cooled system and the PUE decreases 0.349.Moreover,the COP performance of the magnetic bearing chiller is superior to that of the variable-frequency chiller,with an average annual increase of 6.7%,enabling the energy-saving retrofit plan with magnetic bearing chiller to save an additional 3.96%of annual energy consumption compared to the variable-frequency centrifugal chiller plan.Furthermore,the energy efficiency and economic performance of the retrofit plans in five typical climate cities are analyzed.Although the initial investment for the water-cooled system retrofit plans is higher,both retrofit plans show good economic performance.Compared to the existing air-cooled system,the investment payback period of the variable-frequency centrifugal chiller retrofit plan are 2.02-5.15 years,and that of the magnetic bearing chiller plan are 2.16-5.6 years.Then,a multi-energy collaborative hybrid energy system is proposed,which consists of a multi-cooling source cooling subsystem,a building-integrated photovoltaic system(BIPVS),and a waste heat recovery subsystem.Through the thermal-electric coupling of each subsystem,the proposed hybrid energy system can operate efficiently and with low carbon emissions.Taking the case data center as an example,a comparative analysis is conducted among the hybrid energy system,the existing system and the isolated energy system(The water-cooled air conditioning system and the heat pump system operate independently)to study the performance of the proposed hybrid energy system in cooling,heating,overall system energy efficiency,and economy.The results show that during the heating season,the coefficient of performance(COP)of the chillers in the hybrid energy system is 0.4-2.55 higher than that of the isolated system,while the COP of the heat pump is 1.4-2.1 higher.Compared with the existing system and the isolated system,the annual energy consumption of the hybrid energy system reduces by 72.98%and 21.3%,the carbon emission reduces by68.34%and 19.52%,respectively.The dynamic investment payback periods are 2.87years and 3.54 years,respectively.Finally,this study,based on the multi-dimensional impact of climatic factors on the integrated energy system,selected 20 representative cities from five typical climatic zones in China as the research objects.By constructing an energy system energy consumption model,it compared and analyzed the key indicators such as energy efficiency,environmental protection and economy of the integrated energy system,isolated energy system and the original system.It focused on revealing the influence mechanism of climatic factors on the multi-energy collaborative utilization.The simulation process incorporated the variations of multiple parameters such as environmental temperature,solar irradiance,wind speed and groundwater temperature in each climatic region into the analysis.The results showed that among the selected 20cities,the energy utilization efficiency of the integrated energy system was significantly higher than that of the other two systems.In terms of reducing carbon emissions,compared with the original system,the integrated energy system could reduce carbon emissions by 69%to 79%,and compared with the isolated system,it could also reduce carbon emissions by 17%to 31%.From an economic perspective,considering the differences in resource prices,the life cycle cost(LCC)of the integrated energy system was 4%to 18%lower than that of the isolated systems in the selected 20 cities.However,it should be particularly noted that in cities with severe cold climates and low resource prices,the economic performance of the integrated energy system was not as good as that of the original air-cooled system.更多还原