網路資料中心空調節能設計之模擬分析研究

Simulation Analysis of HVAC Energy Saving Design for Internet Data Center

張銘陽(Ming-Yang Chang)

指導教授 ： 李魁鵬

國立臺北科技大學/機電學院/能源與冷凍空調工程系碩士班/碩士(2009年)

https://doi.org/10.6841/NTUT.2009.00444

摘要

由國內能源用戶各類建築電力消費量(EUI)統計分析研究調查報顯示，IT科技廠、電信機房、網路資料中心(Internet Data Center, IDC)機房分別佔居前三名。美國柏克萊國家實驗室(LBNL)針對資訊機房分析調查平均電腦機房能源效率指標(Power Usage Effectiveness, PUE)在2003年平均為1.95，至2005年平均為1.63，根據LEED之能源評估必要條件，以300RT等級之IDC整體，PUE值至少須達到1.52以下為基本門檻，但反觀國內網路資料中心機房經由實際量測結果PUE平均約為1.9以上。因此若能夠對於IT產業的節能減碳技術提升，將有助於我國溫室氣體減量也更能夠提高我國產業對全球的競爭力。
本研究以國內IDC為案例，依照實際監測之結果，PUE值為1.82，空調用電占整體的38%。針對實際案例之耗能因子分析，與環境整體條件，以改變空調出風方式，規劃冷熱之通道並加入隔板避免冷熱空氣混風，藉由計算流體力學軟體(CFD)模擬預估氣流與溫度分佈之結果，可說明良好的出回風配置可以提高空調之送風溫度及冰水溫度。並將CFD之模擬結果，建立動態建築模擬耗能之模型，運用節能手法預估改善後之節能潛力，結果說明有效之環境規劃，相較實際案例現況可節能了6%，冰水主機部分節能11.8%，泵浦節能26.5%，風扇節省21.8%。最終耗能表現之最佳節能手法為，若結合環境改善之手法，並參照ASHRAE90.1之改善空調設備效率，整體資料中心能源效率可提升13%，冰水主機提升47.99%，泵浦提升35.64%，風扇提升7.45%。

關鍵字

網路資料中心 ； 計算流體力學 ； 動態建築模擬軟體 ； 能源消耗指標

並列摘要

The EUI statistic data surveyed in Taiwan indicates the IT technology industry, telecom and Internet Data Center (IDC) are the top three. The average Power Usage Effectiveness (PUE) of data center surveyed by Lawrence Berkeley National Laboratory (LBNL) was 1.95 in 2003 and 1.63 in 2005. Based on the design per LEED the IDC’s PUE needs to reach at least 1.53 for 300RT HVAC system. However, the average PUE of data centers surveyed in Taiwan is 1.9, which is quite lower energy efficiency than the previous two cases. Therefore, if the IT industry can improve the technology for energy saving and carbon reduction, it will help Taiwan to reduce the greenhouse gas emission and also increase the industry’s worldwide competition..
The case study in this research shows that the PUE is 1.82 after a comprehensive measurement. The HVAC consumed 38% of the total energy in this case. According to the CFD and energy simulation analysis, the energy conservation measures create 6% energy saving of total facility, chiller 11.8%、pump 26.5%、fan 21.8% respectively. The energy conservation measures include improvement of air distribution, creation of hot-aisle/cold-aisle layout to avoid mixing of the cool air with the hot exhaust air. Finally, per ASHRAE Std. 90.1- 2007 and the above two improvements, the total energy efficiency can increase 13%, chiller 47.99%，pump 35.64%, and fan 7.45% respectively.

並列關鍵字

Internet Data Center ； CFD ； eQUEST ； Power Usage Effectiveness

參考文獻

[2] LBNL, Data center website of Lawrence Berkeley National Laboratory, http://datacenert.lbl.gov/, 2003.

[4] ASHRAE,ASHRAE Handbook - HVAC Applications, Data Processing and Electronic Office Areas,2007, pp.17.1-17.16.

[5] Roger Schmidt, Don Beaty, Jay Dietrich, “Increasing Energy Efficiency In Data Centers”, ASHARE Journal, 49, 18-24, 2007.

[6] LBNL, Data center website of Lawrence Berkeley National Laboratory, http://datacenert.lbl.gov/, 2005.

[11] H.S. Sun, S.E. Lee, Case study of data centers’ energy performance, Energy and Building; 38 (2006) 522-533.