- 期刊
資料中心PUE驗證技術之研究
Research on performance evaluation of data center PUE
摘要
資料中心的能源使用效率已成為一個重要議題，而PUE（Power Usage Effectiveness）為衡量資料中心能源使用效率的主要指標之一。國際間已有多項標準針對PUE進行定義和測量方法的制定，但因應各個資料中心的不同設計樣態和環境限制，需定義一個通用性的PUE測量作法，以精確地衡量資料中心的能源使用效率。本研究針對非獨立空調之混合式資料中心能源使用效率測量方案，根據資料中心的設計樣態和環境限制，探討相應的PUE計算式和測量位置之建議方案，以更精確地計算非獨立空調系統對資料中心PUE能源效率的驗證。另外本研究透過能源模擬軟體EnergyPlus，利用能源模擬方法，建立標準的資料中心模型，分別依不同負載率及不同氣候條件下定義各類別資料中心之PUE參考值，以作為未來與現場實際測量之分級認證參考。依所建立之標準模型1A氣候區的大型資料中心高用電密度設備在25%、50%、75%、100%的設備負載下，PUE值分別為1.619、1.366、1.325、1.320。
並列摘要
Energy efficiency in data centers has become an important issue, and Power Usage Effectiveness (PUE) is a key indicator for measuring the energy efficiency of data centers. Various international standards have been developed to define and measure PUE; however, due to the diverse design patterns and environmental constraints of different data centers, a universal measurement approach for PUE needs to be defined to accurately assess the energy usage efficiency of data centers. This research has further improved the TaiSEIA data center energy efficiency measurement scheme. Based on the design patterns and environmental constraints of data centers, different data center design configurations are defined, and corresponding PUE calculation formulas and measurement location definitions are established for each configuration. In addition, in the case of non-independent air conditioning systems, this research conducts classification and analysis to more accurately calculate the contribution of non-independent air conditioning systems to data center energy. Furthermore, this research utilizes the energy simulation software EnergyPlus to create standardized data center models using energy simulation methods. These models define PUE calculation formula standards for different load rates, energy-saving control methods, and climatic conditions, which are used for comparing and certifying measurement results against actual on-site measurements. This ensures the accuracy and reliability of the measurement results and achieves the goal of fair regression correction reflecting the same conditions. According to the established standard model 1A climate zone, the PUE values of high-power density equipment in large data centers under 25%, 50%, 75%, and 100% equipment load are 1.619, 1.366, 1.325, and 1.320 respectively.