資料中心PUE量測 (Measuring PUE for Data Centers)

Recommendations for Measuring and Reporting Overall Data Center Efficiency
Version 2 – Measuring PUE for Data Centers 江金海編譯
1

Recommendations for Measuring and Reporting
Overall Data Center Efficiency
Version 2 – Measuring PUE for Data Centers

資料中心總體效率量測和報告建議
第 2 版-資料中心 PUE 量測

17 May 2011

鑕榮工程顧問有限公司
江金海編譯

2

Table of Contents
目錄
1 Introduction
前言 ............................................................................................................................ 3
1.1 Purpose – Recommendations for Measuring and Publishing PUE at All Data Centers
目的 – 對所有資料中心 PUE 量測及發佈的建議 .…………..………………………………….... 3
1.2 Background – Guiding Principles for Data Center Efficiency Metrics
背景 – 對資料中心效率指標的指導原則 ............................................................... 3
1.3 Scope of Recommendations
建議的範圍 ................................................................................................................ 4
2 Definitions Ancillary Services
定義附屬服務 .............................................................................................................. 5
3 PUE Metric Calculation
PUE 指標計算 …………………………………........................................................................... 11
3.1 Data Center Boundaries
資料中心邊界 ……………………………………….................................................................. 11
3.2 PUE Categories
PUE 分類 ................................................................................................................... 12
3.3 Weighting of Energy Types Based on Source Energy
依據能源類型的能源權重 ........................................................................................ 15
3.4 Renewable Energy Sources
可再生能源 ……......................................................................................................... 16
3.5 Combined Heat and Power Plants
供熱及供電組合 ...................................................................................................... 17
3.6 Reused Data Center Energy
資料中心能源再利用 ............................................................................................... 19
4 Publishing Format and Related Resources
發佈格式及相關資源 ………………………………………………………………............................... 21
4.1 Related Resources and Tools
相關資源及工具 ....................................................................................................... 21
5 Next Steps
下一個計畫 .................................................................................................................. 23
Appendix A. Example PUE Calculation Using Source Energy Weighting Factors
附錄 A 使用不同能源權重因素的PUE計算範例 ........................................................... 24

3

1 Introduction 前言
1.1 Purpose – Recommendations for Measuring and Publishing PUE at All Data Centers
目的-對所有資料中心 PUE 量測及發佈的建議
The purpose of this document is to provide recommendations on measuring and
publishing values for PUE at data centers.
本文件目的在於為資料中心的 PUE 量測及發佈提供建議
These recommendations represent the collective work of a task force representing 7x24
Exchange, ASHRAE, The Green Grid, Silicon Valley Leadership Group, U.S. Department of
Energy Save Energy Now Program, U.S. Environmental Protection Agency’s ENERGY STAR
Program, United States Green Building Council, and Uptime Institute.
這些建議代表 7x24 Exchange, 美國冷凍空調協會, The Green Grid, 矽谷領導組織,
美國能源部即刻節能規劃, 美國環保署能源之星規劃, 美國綠建築委員會,及
Uptime 協會.等組織的集團創作
The recommendations are prepared in accordance with the guiding principles, presented
in the following section.
呈現在下列章節的建議, 是為了符合指導原則而準備,
1.2 Background – Guiding Principles for Data Center Efficiency Metrics
背景-對資料中心效率指標的指導原則
As business demands and energy costs for data centers rise, owners and operators have
focused on the energy efficiency of the data center as a whole, frequently using energy
efficiency metrics. However, the metrics are not always applied clearly and consistently.
To address these inconsistencies, a group of leaders from across the industry met on
January 13, 2010 to agree on data center energy efficiency measurements, metrics, and
reporting conventions.
當資料中心商業需求及能源成本提升時,擁有者及營運者將聚焦在資料中心的整體
能源效率, 通常採用能源效率指標.然而這個指標並不總是清楚且一致.為了解決這
個矛盾, 來自整個產業的領導者組織在2010.1.13開會,會議上在資料中心能源效率
量測, 指標及報告方面取得協議
Organizations represented included the 7x24 Exchange, ASHRAE, The Green Grid, Silicon
Valley Leadership Group, U.S. Department of Energy Save Energy Now and Federal Energy
Management Programs, U.S. Environmental Protection Agency’s ENERGY STAR Program,
U.S. Green Building Council, and Uptime Institute.
組織代表包括 7x24 Exchange, 美國冷凍空調協會, The Green Grid, 矽谷領導組織,
美國能源部即刻節能規劃, 美國環保署能源之星規劃, 美國綠建築委員會,及
Uptime 協會
The following guiding principles were agreed to:
下列為同意的指導原則

4

• Power Usage Effectiveness (PUE) using source energy consumption is the
Preferred energy efficiency metric for data centers. PUE is a measurement of the
total energy of the data center divided by the IT energy consumption.
資料中心使用能耗的能源效率指標,優先採用能源使用效率(PUE),PUE 為量測
資料中心總耗能除以 IT 能耗
• When calculating PUE, IT energy consumption should, at a minimum, be
measured at the output of the uninterruptible power supply (UPS). However, the
industry should progressively improve measurement capabilities over time so that
measurement of IT energy consumption directly at the IT load (e.g., servers,
storage, network, etc.) becomes the common practice.
當計算 PUE 時,IT 能耗量測點至少須在 UPS 的輸出.但這個產業的量測能力應
與日俱增到可直接在 IT 負載(例如, 伺服器, 儲存裝置, 網路…等)測量 IT
能耗成為慣例
• For a Dedicated Data Center, the total energy in the PUE equation will include all
energy sources at the point of utility handoff to the data center owner or
operator. For a Mixed‐ Use Data Center, the total energy will be all energy
required to operate the data center, similar to a Dedicated Data Center, and
should include cooling, lighting, and support infrastructure for the data center
operations.
對於專用的資料中心,PUE 公式中的總能耗,必須包含公用設施到擁有者或營
運者切換點的所有能耗.對於混合用途的資料中心,總能耗必須包含資料中心
運行所需之所有能耗.類似專用資料中心,應包含為了資料中心運行之冷卻,
照明及支援之基礎設施.
This guidance is meant to help the industry have a common understanding of energy
efficiency metrics that can generate dialogue to improve data center efficiencies and
reduce energy consumption. Member organizations are committed to applying and
promoting these guidelines to their programs.
A task force was created to further refine these metrics and to identify a roadmap for the
future. The group also aspires to address IT productivity and carbon accounting in the
future.
本指南意在幫助產業有一個共通的能源指標協定,其對增進資料中心效率及降低能
耗可產生對話.會員組織致力於應用及促進這些指南到他們的規劃.
建立一個工作小組來進一步改善這些指標及確定未來路線.在未來,這個小組還有志
於處理 IT 生產力及碳計量
1.3 Scope of Recommendations
建議的範圍
The purpose of this document is to provide recommendations from this task force on

5

how to measure and publish overall data center infrastructure energy efficiency, based
on the agreed upon guiding principles. This document does not address IT efficiency. It
also does not directly address system‐level metrics for cooling or heat rejection, air flow
management, power distribution, lighting, etc.
本文件的目的在於提供來自這個工作小組,如何依據約定的指導原則,量測及發佈總
體資料中心基礎設施能源效率的建議.本文件並不處理IT效率.也不直接處理對冷卻,
排熱,氣流管理,配電,照明…等系統等級指標.
The task force recognizes that many data center operators may not currently have the
capability to measure all energy consuming components within their data center
accurately.
許多資料中心營運者可能目前沒有能力,準確地測量在資料中心內的所有元件能耗
項目類別
The task force therefore recommends four (4) categories of measurement, which
represent a subset of The Green Grid’s (TGG) measurement methods1
.
工作小組因此建議四種量測分類,它代表著 The Green Grid (TGG)的量測方法的一個
子集
The intent is to encourage operators with limited measurement equipment to participate
while also defining a framework that allows operators to add additional measurement
points to increase the accuracy of their measurement program.
其目的在於鼓勵營運者有限的參與設備量測,同時也訂定一個框架,允許營運者增加
額外的量測點以增加它的量測程序準確度
The goal is to recommend a consistent and repeatable measurement strategy that allows
data center operators to monitor and improve the energy efficiency of their data center.
A consistent measurement approach will also facilitate communication of PUE among
data center owners and operators. It should be noted that caution must be exercised
when an organization wishes to use PUE to compare different data centers, as it is
necessary to first conduct appropriate data analyses to ensure that other factors such as
levels of reliability and climate are not impacting the PUE.
目標是推薦一個一致且可重複的量測策略,其允許資料中心營運者監視及增進它們
的資料中心能源效率.一個一致性量測方法,PUE 也將促進擁有者與營運者之間的溝
通.當組織希望使用PUE來比較不同的資料中心時,應指出必要注意事項.因為它有必
要首先進行適當的資料分析,來確認其他因素(如,可靠度及氣候)不會影響 PUE
These categories range from relatively simple measurements that provide a performance
snapshot to more sophisticated measurement means that provide highly detailed
performance data.
這些分類範圍從相對簡單的量測,提供一個性能簡要印象.到更精密的量測,提供更
詳盡的性能資料

6

1
The Green Grid nomenclature and supporting information:
http://www.thegreengrid.org/en/Global/Content/white‐ papers/Usage and Public
Reporting Guidelines for PUE DCiE

These recommendations can also be used in the planning phase for a new data center or
a major renovation of an existing building to assist in the placement and number of
measurement points and monitoring equipment.
這些建議也可被使用在新資訊機房或既有建物重大整修的規劃階段,有助於安裝量
測點數及監視設備
This document provides an overview of how to measure and report PUE per this task
force’s recommendations. This document is not intended to serve as a detailed technical
reference. Further specific details and requirements on how to measure PUE are
available in The Green Grid white papers (http://www.thegreengrid.org/).
本文件提供一個依工作小組的建議,如何量測及報告的概述.本文件並不預期作為詳
細的技術參考.如何有效的量測 PUE,更進一步的具體細節及需求在 The Green Grid
的白皮書內(http://www.thegreengrid.org/).

7

2 Definitions Ancillary Services:
定義附屬服務
This term refers to services that may be part of any data center. While they do not
directly provide infrastructure or power, they are typical building services that will be
present within the data center boundary and should be included in the total energy.
Ancillary services include bathrooms, lobbies, kitchenettes, offices, conference rooms.
本節的服務意指可能是任何資料中心的一部份.雖然他們不直接提供基礎設施或電
力,他們是出現在資料中心邊界內的典型建築服務及應包括在總能耗內.附屬服務包
含廁所,大廳,小型廚房,辦公室,會議室.
Data Center in Mixed‐Use Building (Mixed‐Use Data Center): This refers to a data center
that is located within a larger building that may support other functions. For example, a
data center located on a single floor within a larger office building. In this case the larger
building contains some combination of non‐IT operations, such as general administrative
office space. A Mixed‐ Use Data Center may have its own dedicated infrastructure (UPS
and cooling) or may have shared infrastructure. Refer to the discussion of boundaries in
Section 3.1.
資料中心在混和用途建築內(混合用途資料中心),指的是資料中心位在多用途功能
的大型建築內.例如,資料中心位在大型辦公大樓的單一樓層.在這個案例,大型建築
包含了一些非IT運行組合,像是一般的管理辦公室空間.一個混合用途資料中心可能
有自己的專用基礎設施(UPS 及冷卻)或可能有分享的基礎設施.參考 Section 3.1 中
邊界的討論
The Mixed‐Use Data Center recommendations covered herein apply to data centers that
have dedicated infrastructure (power and cooling).
混和用途資料中心的建議適用於具有專用基礎設施(電力和冷卻)的資料中心.
Dedicated Data Center: In this context, a Dedicated Data Center is a free standing
building in which all the spaces and supporting infrastructure (HVAC, lighting, electrical)
are directly associated with the operation of the data center. Dedicated Data Centers are
standalone buildings whose exclusive purpose is IT operations and their support
functions.
專用資料中心,指的是資料中心是一個獨立的建築,其所有空間及支援基礎設施(空
調,照明,電力)是與資料中心的運行有直接關係.專用資料中心是一個獨立的建築,
只有一個目的,就是 IT 的運行及 IT 的支援功用.
IT annual energy is calculated as annual electrical consumption of the IT. IT includes all IT
equipment, servers, networking and storage as well as telecom equipment typically
installed in telecom data centers.
IT 年度耗能計算如同 IT 的年度用電.IT 包含所有 IT 設備,伺服器,網路及儲存裝置,
像電信設備一樣,通常安裝在電信機房內
Peak IT electric demand (kW) is the annual peak electric demand for the IT load.

8

尖峰 IT 電力需量(kW)是 IT 負載年度尖峰電力需量.
Peak Total electric demand (kW) is the annual peak electric demand for the entire data
center including IT and supporting infrastructure measured at the data center boundary
(e.g. point of electric feed for Mixed‐Use Data Centers or utility meters for Dedicated
Data Centers).
尖峰總電力需量(kW)是整個資料中心的年度尖峰電力需量,包含 IT 及在資料中心邊
界內支援基礎設施量測(如,混合用途資料中心的電力供應點或專用資料中心的公用
事業電錶)
PDU Output is defined as the output of the PDU that serves IT loads. PDU output does
not IT equipment input is defined as the IT load as measured at the point of connection
of the IT device to the electrical power system, e.g., the power receptacle(s).
IT equipment input captures the actual power load of the IT device exclusive of any
power distribution losses and non‐IT loads such as rack mounted fans. Include efficiency
losses of any transformation that occurs within the PDU but may include downstream
non‐IT ancillary devices installed in IT racks such as fans. If there is non‐IT equipment
supported by the PDU system, (e.g., CRAC, CRAH, In‐row coolers, etc.) it must be metered
and subtracted from the PDU Output reading (the metering approach should be
consistent with the metering required for the PUE category e.g. continuous consumption
metering for PUE categories 1,2 and 3).
PDU 輸出定義為供應 IT 負載之 PDU 的輸出.PDU 輸出並不是 IT 設備輸入,IT 設備輸入
定義為 IT 負載的量測在電力系統到 IT 裝置的連接點,像是電力插座.
IT 設備輸入抓取實際 IT 裝置電力負載,排除任何配電損失及非 IT 負載,像是安裝在
機櫃上的風扇.包括任何發生在 PDU 內的變壓器損失,也可能包刮 PDU 下游裝在 IT 機
櫃內的非 IT 輔助裝置,像是風車.假如有非 IT 設備由 PDU 系統供
電.(如,CRAC,CRAH,In-row cooler 等),其必須設置儀表並從 PDU 輸出讀值(儀表設置
位置須符合 PDU 類別要求,例如 PDU 分類 1,2,3 要求之連續耗能計量)扣除.
Supporting infrastructure includes the following:
支援基礎設施包含下列
• Power systems: Transfer switch, UPS, DC batteries/rectifiers (non UPS – telco
nodes), generator, transformer (step down), power distribution unit (PDU), rack
distribution unit (RDU), breaker panels, distribution wiring, lighting.
電力系統:切換開關,UPS,直流電池/整流器(無 UPS-電信節點),發電機,變壓器
(降壓),PDU,RDU,斷路器盤,配線,照明.
• HVAC systems: Systems, equipment and control strategies including cooling towers,
condenser water pumps, chillers, chilled water pumps, computer room air
conditioners (CRAC’s), computer room air handlers (CRAH’s), dry cooler, supply fans
, return fans, direct and indirect air economizer, direct and indirect evaporative
cooling, water‐side economizer, humidifier, in‐row, in‐rack, & in‐chassis cooling

9

solutions.
空調系統:系統,設備及控制,包含冷卻水塔,冷卻水泵,冰水主機,冰水泵, CRAC
,CRAH,乾盤管,送風風車,回風風車,直接及間接空氣節能器,直接及間接蒸發
冷卻,水側節能器,加濕器,in-row,in-rack,in-chassis 冷卻解決方案
• Physical security: Fire suppression, water detection, physical security
servers/devices.
物理安全措施:消防滅火,漏水偵測,物理安全伺服器/裝置
• Building management systems: computer‐based control system that control and
monitor the building’s mechanical and electrical equipment such as ventilation,
lighting, power systems, fire systems, and security systems.
建築物管理系統:電腦控制系統,控制及監視建築物的機械和電力設備,像是通
風,照明,電力,消防及安全監視系統
Total annual energy is calculated as the weighted sum of the annual energy consumption
for all UPS Output is defined as the output of the UPS that serves IT equipment loads.
UPS output does energy types serving the data center at the data center boundary. In a
Dedicated Data Center building, this will include all energy captured on utility bills; for a
Mixed‐Use Data Center, all the same fuels must be sub‐metered if they cross into the
data center boundary. This includes electricity, natural gas, fuel oil, and district utilities
such as supplied chilled water or condenser water. All energy types must be converted
into the same units before they are summed. Total annual energy must include
supporting infrastructure and dedicated ancillary services (see definitions above). Section
3.3 describes how to weight different energy types. (Note that weighting only applies for
data centers that have energy types in addition to electricity.)
總年度耗能計算為所有能源消耗年度加權總和,所有 UPS 輸出定義為供應 IT 設備負
載之 UPS 輸出,UPS 輸出僅是供應資料中心的能源種類之一.在一個專用資料中心建
築,將包含所有能源費用,在混合用途資料中心,所有能源在跨入資料中心邊界應設
置分錶,這些包含電力,天然氣,燃油及地區公用事業供應之冰水或冷卻水.總年度耗
能必須包含支援之基礎設施及專用之輔助服務(詳前面定義).Section 3.3 說明不同
能源種類的權重.(注意,此權重僅適用於資料中心,電力除外)
UPS Output is defined as the output of the UPS that serves IT equipment loads. UPS
output does not include efficiency losses from the UPS system but does include losses
from downstream electrical distribution components such as PDU's. It may include non‐IT
ancillary devices installed in IT racks such as fans. If there is non‐IT equipment supported
by the UPS system, (e.g., CRAC, CRAH, In‐row coolers, etc.) it must be metered and
subtracted from the UPS Output reading (the metering approach should be consistent
with the metering required for the PUE category e.g. continuous consumption metering
for PUE categories 1,2 and 3).
UPS 輸出定義為供應 IT 設備負載之 UPS 輸出.UPS 輸出不包含 UPS 系統的效率損失,

10

但包含下游配電元件損失,像是 PDU 的損失.它可能包含裝在 IT 機櫃內的非 IT 輔助
裝置,像是風扇.假如有非 IT 設備由 UPS 系統供電.(如,CRAC,CRAH,In-row cooler
等),其必須設置儀表並從 UPS 輸出讀值(儀表設置位置須符合 PDU 類別要求,例如 PDU
分類 1,2,3 要求之連續耗能計量)扣除.

11

3 PUE Metric Calculation
PUE 指標計算
Power Usage Effectiveness (PUE) is the recommended metric for characterizing and
reporting overall data center infrastructure efficiency. The task force strongly
recommends annual energy consumption (kWh) for all energy types as the unit of
measure for PUE calculation. However an entry level measurement category has been
included in the recommendations to allow operators that do not have consumption
measurement capability to utilize demand based power readings.
能源使用效率(PUE)是推薦的指標,用來表現及報告總體資料中心基礎設施效率.工
作小組強烈建議量測 PUE 計算單位,所有能源皆採用能源年度消耗(kWh).然而,一個
入門級量測類別以包含在建議之中,允許沒有耗能量測能力的營運商,採用電力使用
需量讀值.
PUE is defined by the following formula:
PUE = (Total data center energy consumption or power / IT energy consumption or power)
The following sections detail boundary considerations for both Dedicated Data Centers
and Mixed‐Use Data Centers, describe the four recommended PUE categories for data
centers, and address calculation details associated with proper accounting of all fuels and
on‐site systems.
PUE 定義公式如下
PUE=(總資料中心能源消耗或用電/IT 能源消耗或用電)
下節將詳細說明專用資料中心及混和用途資料中心邊界考量,說明資料中心建議的
四種 PUE 分類,及詳細計算相關所有燃料適當用量及現場系統
3.1 Data Center Boundaries
資料中心邊界
The PUE calculation guidelines presented in this document address two data center
configurations:
本文件中提出兩種資料中心規劃 PUE 計算指南
• Dedicated Data Center i.e. the data center is a free‐standing building. (see definition)
專用資料中心即資料中心為獨立建築.(詳定義)
• Mixed‐Use Data Center with dedicated infrastructure i.e. the data center is within a
larger building that has other uses, but the data center has dedicated HVAC and
electrical systems. (see definition)
有專用基礎設施之混合用途資料中心即資料中心在有其他用途之大型建築內,
但資料中心有專用空調及電力系統.(詳定義)
The PUE calculation approach presented below is the same for both these configurations.
However, there are some minor differences in how the boundary for total energy use is
defined for each configuration, as shown in Figure 1:

12

下列為兩種類型資料中心PUE的計算方法,然而,在總能源使用的邊界有些許的不同.
如圖 1 所示.
Figure 1: Energy use boundaries for Dedicated Data Centers and
Mixed‐Use Data Centers.

• In Dedicated Data Centers, the boundary is the building boundary i.e. all energy uses
within in the building are included.
在專用資料中心,其邊界為建築物邊界,即建築物所有能源使用皆包含在內
• In Mixed‐Use Data Centers, shared ancillary services such as common lobbies,
common bathrooms and elevators may be excluded from the energy use boundary.
However, ancillary services that are dedicated to the data center must be included
(e.g. lobby, bathrooms, office spaces that are dedicated to the data center operation).
在混合用途資料中心,分享輔助服務,如公用大廳,公用廁所及電梯,可能排除在
能源使用邊界之外.然而,屬於資料中心專用的輔助服務則必須包含在內.(如資
料中心營運專用之大廳,廁所,辦公空間)

3.2 PUE Categories
PUE 分類
The task force recommends four (4) measurement categories for the measurement of
PUE within a data center:
工作小組對資料中 PUE 的量測,建議四種量測分類.

13

PUE Category 0
PUE 分類 0
This is a demand based calculation representing the peak load during a 12‐month
measurement period. IT power is represented by the demand (kW) reading of the UPS
system output (or sum of outputs if more than one UPS system is installed) as measured
during peak IT equipment utilization. Total data center power is measured at the data
center boundary (e.g. point of electric feed for Mixed‐Use Data Centers or utility meters
for Dedicated Data Centers) and is typically reported as demand kW. As this is a snapshot
measurement, the true impact of fluctuating IT or mechanical loads can be missed.
However consistent measurement can still provide valuable data that can assist in
managing energy efficiency.
這是一個在 12 個月量測周期期間,代表尖峰負載需量基準的計算.IT 用電代表為系
統 UPS 在 IT 設備使用尖峰期間,輸出需量讀值(kW)(假如 UPS 系統安裝超過 1 個以上
時,則為系統輸出總和).總資料中心用電量測在資料中心邊界(如,混和用途資料中
心在電力饋線點或專用資料中心在公用事業電錶),通常報告需量為 kW.像這種簡要
印象式量測,可能錯失 IT 或機械負載變動的真正影響.然而,一致性的量測仍可提供
有用的數據資料,有助於管理能源效率
PUE category 0 may only be used for all‐ electric data centers i.e. it cannot be used for
data centers that also use other types of energy (e.g. natural gas, district chilled water,
etc.).
PUE 分類 0 可能僅適用於全電力式資料中心即它不適用於有使用其他類型能源(如,
天然氣,區域冰水,等)的資料中心

PUE Category 1
PUE 分類 1
This is a consumption based calculation. The IT load is represented by a 12‐month total
kWh reading of the UPS system output (or sum of outputs if more than one UPS system is
installed). This is a cumulative measurement and requires the use of kWh consumption
meters at all measurement points. The total energy must include all fuel types that enter
the data center boundary (electricity, natural gas, chilled water, etc).
這是一個耗能基準的計算.IT 負載代表為系統 UPS 輸出的 12 個月讀值總和(kWh)(假
如 UPS 系統安裝超過 1 個以上時,則為系統輸出總和).這是一種累計量測並且需要在
所有量測點安裝瓦時錶.總能源需包含資料中心邊界內的所有能源總類(電力,天然
氣,冰水等.)
In a Dedicated Data Center building, this will include all energy captured on utility bills;
for a Mixed‐Use Data Center, all the same fuels must be sub‐metered if they cross into
the data center boundary.
在一個專用資料中心建築,將包含所有能源費用,在混合用途資料中心,所有能源在

14

跨入資料中心邊界應設置分錶.
Annual reading should reflect 12 consecutive months of energy data. This measurement
method captures the impact of fluctuating IT and cooling loads and therefore provides a
more accurate overall performance picture then PUE Category 0.
年度讀值應該反應連續 12 個月的能源資料.本量測方法可捕捉到 IT 及冷卻負載變動
的影響,因此可提供一個比 PUE 0 更精確的總體性能圖.
PUE Category 2
PUE 分類 2
This is a consumption based calculation. The IT load is represented by a 12‐month total
kWh reading taken at the output of the PDU's supporting IT loads (or sum of outputs if
more than one PDU is installed). This is a cumulative measurement and requires the use
of kWh consumption meters at all measurement points. The total energy is determined in
the same way as Category 1.
這是一個耗能基準的計算.IT 負載代表為 12 個月的讀值總和(kWh)發生在供應 IT 負
載之 PDU 輸出(或,假如 PDU 安裝超過 1 個以上時,則為輸出的總和).這是一種累計量
測並且需要在所有量測點安裝瓦時錶.總能源的決定與 PUE 1 相同
This measurement method provides additional accuracy of the IT load reading by
removing the impact of losses associated with PDU transformers and static switches.
本量測方法提供更精確的 IT 負載讀值,移除了與 PDU 變壓器及靜態開關的損失影響
PUE Category 3
PUE 分類 3
This is a consumption based calculation. The IT load is represented by a 12 month total
kWh reading taken at the point of connection of the IT devices to the electrical system.
This is a cumulative measurement and requires the use of kWh consumption meters at all
measurement points. The total energy is determined in the same way as Category 1.
這是一個耗能基準的計算.IT 負載代表為 12 個月的讀值總和(kWh)發生在 IT 裝置到
電力系統的連接點.這是一種累計量測並且需要在所有量測點安裝瓦時錶.總能源的
決定與 PUE 1 相同
This measurement method provides the highest level of accuracy for measurement of the
IT load reading by removing all impact of losses associated with electrical distribution
components and non‐IT related devices, e.g., rack mounted fans, etc. Table 1 summarizes
the four categories.
本量測方法提供最精確的 IT 負載讀值,移除了所有配電元件及非 IT 負載相關裝置的
損失影響,如,安裝在機櫃上的風扇等.表 1 總結四種分類

15

Table 1: PUE measurement categories recommended by this task force
PUE Category 0* Category 1 Category 2 Category 3
IT energy
measurement
location
UPS output UPS output PDU output IT equipment
input
Definition of
IT energy
Peak IT electric
demand
IT annual
energy
IT annual
energy
IT annual
energy
Definition of
Total energy
Peak Total
electric demand
Total annual
energy
Total annual
energy
Total annual
energy
*For PUE Category 0 the measurements are electric demand (kW).

3.3 Weighting of Energy Types Based on Source Energy
依據能源類型的能源權重
For data centers that have electricity as well as other energy types, the different energy
types must be weighted according to their source energy. Source energy represents the
total amount of raw fuel that is required to operate the building. It incorporates all
transmission, delivery, and production losses, thereby enabling a complete assessment of
energy efficiency in a building.
對於有電力及其他類型能源供應的資料中心,不同能源類型必須有符合其能源來源
的權重.能源來源代表建築物運行需求原燃料總量.它包含了所有傳輸,交付,及生產
損失.因此能夠在一棟建築內完整的能源效率評估
Table 2 provides the weighting factors for each energy type, normalized to electricity.
These weighting factors are based on national average source factors used by EPA in their
building energy benchmarking.
表 2 提供每一種能源的權重係數,通常與電力做比較.這些權重係數是依據國家平均
能源係數,使用美國環保署的建築能源基準
Weighted energy for each energy type = (Annual energy use X source energy weighting
factor).
每一種能源的能源權重=(年度能源使用*能源來源權重係數)
Note that all energy types must be converted into the same units before they are
summed2

注意,所有能源類型在加總前需轉換成同單位
. For example, if electricity is in kWh and natural gas is in kBtu, both must be converted to
a common unit.
. 例如,假設電力使用 kWh 及天然氣使用 kBtu,兩者必須轉換成同樣單位

16

Table 2: Source energy weighting factors3

Energy Type Weighting Factor Electricity
Electricity 1.0
Natural gas 0.31
Fuel oil 0.30
Other fuels 0.30
District chilled water 0.31
District hot water 0.40
District steam 0.43
Condenser water 0.03
Appendix A has an example of PUE calculation using source energy Weighting Factors.

3.4 Renewable Energy Sources
可再生能源
Electricity from renewable energy (RE) sources should be included in the total energy or
power and assigned the same source factor as grid electricity. In other words, RE sources
are outside the PUE calculation boundary (Figure 2). Therefore, the installation of RE
sources does not change the PUE (i.e. an all‐grid building and all‐RE building would have
the same PUE). The rationale for this is that the purpose of the PUE is to evaluate how
efficiently energy is used in the data center, regardless of the energy supply. Renewable
energy in this context includes solar and wind power.
電力來自可再生能源應包含在總能源或電力內,並且指定相同的來源係數,如同電網
一樣.換句話說,可再生能源不在 PUE 計算的邊界內(圖 2).因此,再生能源的設置並
不會改變 PUE(即,全電網建築及全再生能源建築,其 PUE 仍是一樣).這樣闡述 PUE 的
目的,在於評估如何在資料中心有效率的使用能源,不論能源供給.在本文所說的可
再生能源包含太陽能及風力.
2
For conversion factors, see:
http://www.energystar.gov/ia/business/tools_resources/target_finder/help/Energy_Unit
s_Conversion_Table.htm

3
The EPA source energy factor methodology for all energy types except condenser water
is available at:
http://www.energystar.gov/ia/business/evaluate_performance/site_source.pdf. The
weighting factors presented in this recommendation document are obtained by dividing
each EPA source factor by the reference source factor for electricity, 3.34. The factor for
condenser water was calculated assuming 20 gpm/hp, 3 gpm/ton and 1% line losses.

17

Figure 2: Data center PUE calculation boundary with renewable energy

3.5 Combined Heat and Power Plants
供熱及供電組合
If the data center has a dedicated combined heat and power (CHP) plant, the inputs to
the CHP should be included in the total energy or power (Figure 3) and assigned the
same source factors as those shown in table 1. The outputs from the CHP should not be
included in the total energy or power. The use of CHP (like the use of more efficient
cooling or other equipment) should lower the PUE. The rationale for this is that CHP is
considered to be an efficiency measure rather than an alternative source of energy.
假如資料中心有專用的供熱及供電站(CHP),則CHP的輸入應包含在總能源或電力(圖
3)並指定與列在表 1 之相同能源係數.來自 CHP 的輸出不應包含在總能源或電
力.CHP(像是使用更有效率的冷卻或其他設備)的使用,應有較低的PUE.這樣闡述CHP
是考慮效率量測,而不是能源來源的替代

Figure 3: Data center PUE calculation boundary with CHP.

18

To enable equitable PUE calculations in data center designs that operate on‐site electric
generation two topics need to be explained:
為了能夠合理的計算,現場設計有運轉發電的資料中心 PUE,將有兩個主題需要解釋
• IT Source Energy Conversion Factor
IT 能源轉換係數
o To ensure the PUE calculations for a design including on‐site electric generation
and a design excluding on‐site electric generation are equitable, the taskforce
uses an IT Source Energy Factor (Figure 4). The IT Source Energy Factor enables the
data center owner / operator to include the efficiencies of generating chilled water
(as in all calculations for PUE), but exclude the efficiencies of generating electricity
(as in all calculations for PUE). PUE is not a metric to reward or penalize the
efficiencies of how electricity is brought to the data center, it is a metric to identify
how efficient the electricity is used from the data center boundary to the IT
Equipment.
對於一個設計包含現場發電及設計排除現場發電,確認PUE計算是合理的.工作
小組使用一個 IT 能源來源因素(圖 4).IT 能源來源因素能使資料中心擁有者/
營運者對包含冰水產生的效率(像是對PUE的所有計算),但排除發電的效率(像
是對 PUE 的所有計算).PUE 不是資料中心如何用電效率的一個獎勵或懲罰指標
,PUE 指標是識別從資料中心邊界到 IT 設備如何使用用電效率.

Figure 4: Data center with electric and natural gas (Co‐generation) purchase

19

• Cogeneration Input Fuel Assumption
汽電共生燃料假設
o This taskforce has adopted the assumption that 67% of a cogeneration plants’
input fuel is allocated to the primary form of generation (electricity in the provided
example) and respectively 33% of the primary input fuel is allocated to the
secondary form of generation (chilled water in the provided example).
工作小組已接受汽電共生廠輸入燃料假設 67%分配到主要產生的形式(在提供
的範例-電力)及主要輸入燃料的 33%分配給產生的第二類型(在提供的範例-冰
水)
o This assumption is based on a similar steam based cogeneration study from the
U.S. Environmental Protection Agency’s Energy Star Program. Differences do exist
between the two scenarios, but to solidify the methodology of calculating PUE and
to drive consistent and equitable calculations for all design types, this taskforce is
proceeding with the recommendation to adopt this approach. As further data
specific to cogeneration examples are provided, the data will be taken into
consideration. Provided there is a material change in calculations, the taskforce will
consider updating the recommendations specific to this assumption.
這個假設基準是依據美國環保署能源之星計畫中簡單的蒸氣汽電共生研究.兩
種情境間確實存在著不同,但鞏固的PUE 的計算方法及對所有設計類型驅於一致
及公平計算,工作小組繼續推薦採用此方法.為了提供汽電共生範例更進一步的
數據,此資料將被帶入考慮.在計算中,有提供一個材料的改變,工作小組將考慮
更新建議規定到假設中.

3.6 Reused Data Center Energy
資料中心能源再利用
The concept of reusing waste energy (generally in the form of heat) is receiving a lot of
study as part of an overall effort to improve energy efficiency.
廢能源(通常以熱的型式)的再利用概念是接收了大量的研究,像是為了增進能源效
率總體努力的一部份.
The value of this energy may or may not affect PUE, depending on the location of the
re‐use.
這些能源值可能或不可能影響 PUE,端看再使用的地方.
• If the energy is reused within the datacenter, the value of the reused energy will
already be contained within the PUE calculation and needs no special consideration.
An example will be reusing waste heat to warm a battery space in a cool climate.
The PUE would generally be lower because the HVAC system does not have to add as
much energy to heat that battery space, thereby reducing the PUE when energy is
measured and PUE calculated.

20

假如能源再使用是在資料中心內,再利用能源值將包含在 PUE 的計算內並且無需
特別考慮.例如,再利用的廢熱被使用到寒冷氣候時的電池室加熱.PUE 將會變得
較低,因為空調系統不須要更多的熱能到電池空間,當能源量測及 PUE 被計算時,
因此降低 PUE.
• If the energy is reused outside the data center (e.g. re‐using heat to warm a lab that is
not part of the data center), the PUE of the data center is not affected. While the
effort to conserve energy is laudable, it cannot be accounted for in the calculation of
the data center PUE due to the strict definition of PUE used in these
recommendations. Currently there are on‐going industry efforts to define a metric
that could be used to account for this beneficial use, but it is specifically excluded
from PUE.
假如能源再使用是在資料中心以外(如,再利用廢熱使用在非資料中心的實驗室
加熱),則資料中心PUE不受到影響.然而為節能而努力是值得的.由於被使用的這
些建議中,對 PUE 有嚴格的定義,因此它無法納入資料中心的 PUE 計算內.目前有
對持續產業努力定義一個指標,其被用來解釋這些有益的利用,但是它還是排除
在 PUE 之外.

21

4 Publishing Format and Related Resources
發佈格式及相關資源
When publishing PUE, the category must clearly be indicated using a subscript e.g. PUE0,
PUE1, PUE2, PUE3. A PUE reported without the subscript is not considered to be in
compliance with these recommendations.
當發佈 PUE 時,類別必須清楚的標示下標,如 PUE0,PUE1,PUE2,PUE3.一個沒有 PUE 下
標的報告是不認為符合這些建議的.
Note that PUE0 may only be used for all‐electric data centers. It cannot be used for data
centers that also use other types of energy (e.g. natural gas, district chilled water, etc.).
An all‐electric data center can utilize any of the PUE categories.
注意,PUE 0 可能僅適用於全電力式資料中心.它不適用於有使用其他類型能源(如,
天然氣,區域冰水,等)
一個全電力式資料中心可適用於任一個 PUE 分類
When publishing PUE calculated in accordance with these guidelines, users may state
that “PUE was calculated in accordance with the recommendations of the Data Center
Metrics Coordination Taskforce, sponsored by 7x24 Exchange, ASHRAE, The Green Grid,
Silicon Valley Leadership Group, U.S. Department of Energy Save Energy Now Program,
U.S. Environmental Protection Agency’s ENREGY STAR Program, United States Green
Building Council, and Uptime Institute.”
當發佈符合這些指南的 PUE 計算時,使用者可說 PUE 的計算符合資料中心指標協調工
作小組的建議,主辦單位有 7x24 Exchange, 美國冷凍空調協會, The Green Grid, 矽
谷領導組織, 美國能源部即刻節能規劃, 美國環保署能源之星規劃, 美國綠建築委
員會,及 Uptime 協會

4.1 Related Resources and Tools
相關資源及工具
The Green Grid (TGG) white paper #224
provides a range of options for calculating PUE,
as well as an annotation system to indicate how PUE was calculated. The
recommendations of this task force are essentially a subset of the options in TGG
framework, as shown in Table 3, and are therefore consistent with TGG framework.
The Green Grid (TGG)# 224 白皮書提供了一系列可供選擇的 PUE 計算,也提供一個
注解系統來說明如何計算 PUE.工作小組的建議,本質上是在 TGG 框架內的選項的一
個子集,如表 3 所示,因此其與 TGG 框架一致

22

Table 3: Mapping of PUE categories to TGG framework
PUE category 分類 TGG annotation 註解
PUE0 PUEL1,Y‐‐
PUE1 PUEL1,YC
PUE2 PUEL2,YC
PUE3 PUEL3,YC

The U.S. EPA Energy Star Portfolio Manager tool will be incorporating PUE calculations
that are consistent with PUE1.
美國 EPA 能源之星 Portfolio Manager 工具將把 PUE 的計算符合 PUE1
The U.S. DOE’s DC Pro tool will be incorporating PUE calculations that are consistent with
PUE2.
美國 DOE 的 Data center Pro 工具將把 PUE 的計算符合 PUE2
The US Green Building Council’s forthcoming LEED credit adaptations for data centers has
information related to the amount and type of monitoring equipment that are either
prerequisites or for achieving additional points in the monitoring and verification credit
category. The PUE definition used is consistent with PUE2.
美國綠建築委員會即將推出的 LEED Data center credit 與監測設備類型及數量相
關的資訊,在監視及驗證得分分類上,要麼是必要條件或要麼是得分條件.PUE 定義上
是使用符合 PUE2。

4
The Green Grid nomenclature and supporting information:
http://www.thegreengrid.org/en/Global/Content/white‐papers/Usage and Public
Reporting Guidelines for PUE DCiE

23

5 Next Steps
下一個計畫
This document (Version 2) provides guidance for PUE measurement and publication for
Dedicated Data Centers and Mixed‐Use Data Centers with dedicated infrastructure.
Following the release of this document each participating organization will work to
disseminate this information and promote measurement and reporting of PUE across the
IT industry.
本文件(第 2 版)提供專用資料中心及具專用基礎設施的混合用途資料中心之 PUE 量
測及發佈指南. 本文件發佈後,每個參與的組織將致力於宣傳此一資訊,促進整個 IT
產業對 PUE 的量測與報告
At the same time, all participating organizations will reconvene to discuss the important
next steps to continue to move the industry towards effective energy efficiency metrics,
including metrics of IT productivity. The organizations will continue to work together, and
may establish new task forces as necessary to help provide consistent guidance to the
industry.
同一時間,所有參與的組織將再召開會議討論下一個計畫重點,持續推動產業前進到
實際的能源效率指標

24

Appendix A. Example PUE Calculation Using Source Energy Weighting Factors
附錄 A.使用能源來源權重係數 PUE 計算範例
PUE should be based upon Source Energy as described in Section 2.2 of this document as
well as the EPA source energy factor methodology as described in:
http://www.energystar.gov/ia/business/evaluate_performance/site_source.pdf.
PUE 應依據本文件 Section 2.2 所描述的能源來源,像是 EPA 在
http://www.energystar.gov/ia/business/evaluate_performance/site_source.pdf.
所描述的能源來源係數方法論一樣
Table 2 from the main document is reproduced here for convenience
Energy Type Weighting Factor Electricity
Electricity 1.0
Natural gas 0.31
Fuel oil 0.30
Other fuels 0.30
District chilled water 0.31
District hot water 0.40
District steam 0.43
Condenser water 0.03

Each component of the PUE calculation needs to be multiplied by the appropriate
Weighting Factor.
每一個 PUE 計算的元件需乘以其適用的權重

Recall the definition of PUE: Total Energy divided by IT Energy.
PUE = Total Energy / IT Energy
召回 PUE 的定義 : 總能耗除以 IT 能耗
PUE = 總能耗 / IT 能耗
Total energy should include all fuel sources. For an example data center that purchases
electricity and district chilled water, the equation can be expressed as:
PUE = (Electricity + District chilled water) / IT Energy
總能耗應包含所有燃料來源.例如一個具有購買電力及購買地區冰水的資料中心,則
方程式表現為
PUE = (電力 + 地區冰水)耗能 / IT 耗能
To compute a PUE in accordance with these recommendations it is critical to include all
fuels. For illustrative purposes, consider two fictitious data centers: Case A is an all
electric data center, while Case B purchases chilled water from a local utility. The
following tables show the input fuels, the end uses, and the approximate energy use for

25

each end use. In order to compute the PUE it is not necessary to have the energy for
each end use separately. The total energy is simply summed across all energy inputs
to the data center (at the data center boundary). The IT energy is the only end use
that must be sub‐metered within the building.
計算一個符合這些建議的 PUE,對包含所有燃料是非常重要的.為了解說目的,考慮兩
種假設的資料中心:案例 A,為全電力資料中心,而案例 B 則由地區公用事業購買冰水.
下表顯示輸入燃料,終端使用及對每個終端使用的近似能源使用.為了計算 PUE,不須
分開每個終端使用能耗.總能耗僅僅是所有輸入到資料中心(在資料中心邊界內)的
能耗加總.IT 能耗也只是終端使用,其必須是在建築物內的分錶.

Case A – All Electric Data Center
Energy Input to Data Center End‐use Energy Use
Electricity
(1,705,000 kWh total)
IT Load 1,000,000 kWh
Power delivery loss 250,000 kWh
Lighting 50,000 kWh
Cooling 400,000 kWh
Other 5,000 kWh
The PUE for Case A would be computed as:
案例 A, PUE 計算
PUE = (1,705,000*1.0) / (1,000,000*1.0) = 1.70
Case B – Data Center with District Chilled Water
Energy Input to Data Center End‐use Energy Use
Electricity
(1,305,000 kWh total)
IT Load 1,000,000 kWh
Power delivery loss 250,000 kWh
Lighting 50,000 kWh
Other 5,000 kWh
District Chilled Water
(1,300,000 kWh total)
Cooling 1,300,000 kWh
(4.44 M BTUs)
In Case B, district chilled water is used to provide cooling (remove excess heat).
The PUE for Case B would be computed as:
案例 B,地區冰水被使用來供應冷卻(移除過多的熱)
PUE 計算
PUE =(1,305,000*1.0 + 1,300,000*0.31) / (1,000,000*1.0) = 1.70

It is informative to consider that the PUE for each case is 1.70. The implication of this is
that the efficiency of the cooling system in the all electric data center is identical to the
efficiency with which the District Chilled Water is produced. If the Case A’s cooling
system was more efficient, then it would have required less than the 400,000 kWh in the

26

example and the PUE would be lower. If Case A’s cooling system is less efficient then it
would have used more than the 400,000 kWh to cool the data center (including
distribution losses and lighting) and the PUE would be higher.
它是廣泛的考慮了.兩個案例的 PUE 皆為 1.70,這暗示了在全電力資料中心的冷卻系
統效率與地區產出的冰水系統效率是相同的.假如案例 A 的冷卻系統更有效率,那麼
它的耗能應小於 400,000 kWh,及 PUE 應較低. 假如案例 A 的冷卻系統效率較差,那麼
它冷卻資料中心(包含配電損失及照明)的耗能應大於 400,000 kWh,及 PUE 應較高.

資料中心PUE量測 (Measuring PUE for Data Centers)

Recomendados

Recomendados

Mais conteúdo relacionado

Mais procurados

Mais procurados (20)

Destaque

Destaque (15)

Semelhante a 資料中心PUE量測 (Measuring PUE for Data Centers)

Semelhante a 資料中心PUE量測 (Measuring PUE for Data Centers) (20)

Último

Último (20)

資料中心PUE量測 (Measuring PUE for Data Centers)