小型数据中心气流组织及能耗优化

Optimization on Air Distribution and Energy Consumption of a Small Data Center

数据中心空调系统的首要任务是保证服务器的安全可靠运行.为了达到这个目标,很多数据中心的空调系统在欠负荷下运行,造成能源的浪费.这主要是因为气流组织的不合理造成局部区域报警从而启动了多余的制冷设备.为了节省能源和研究合理的气流组织,以一个存在上述问题的小型数据中心为例进行了研究.首先用数值分析的方法建立了数学模型,然后用实测数据提供了模拟中需要的边界条件,模拟得到其速度场和温度场情况,再用实测数据验证了模型模拟结果,然后根据实测和模拟结果用评价参数评价其能耗水平,最后提出改造方案.结果表明,在热扩散不好的区域增加智能风扇系统有效地改善了气流组织的情况,与一般的通风系统相比,这种系统每年可以降低38%的能耗.

The air conditioning system of data centers gives priority to the normal operation of the servers. However, in many data centers, the air conditioning system runs in the under-load state which is significantly energy-consuming. The main reason for it is that additional equipments are started up because the unreasonable air distributions result in alarms of the monitors. In the paper, a typical data center was selected as an example to investigate the optimal air distributions to save energy. Numerical models were established with the input data obtained from on-site measurements. The measured profiles were also used to validate the air velocity and temperature distributions of simulation results. Then, the energy consumptions were evaluated by indexes based on the measurement and simulation results. Finally, the modification plan was presented. The results show that the air distribution was improved by adding intelligent fan system to the regions with excessive heat. The energy consumption could be reduced by 38% annually compared to normal ventilation systems.