ISO/IEC 30134-2:2016

DRAFT INTERNATIONAL STANDARD
ISO/IEC DIS 30134-2.2
ISO/IEC JTC 1/SC 39 Secretariat: ANSI
Voting begins on: Voting terminates on:
2015-08-03 2015-10-03
Information Technology — Data Centres — Key
performance indicators —
Part 2:
Power usage effectiveness (PUE)
Titre manque
ICS: 35.020
THIS DOCUMENT IS A DRAFT CIRCULATED
FOR COMMENT AND APPROVAL. IT IS
THEREFORE SUBJECT TO CHANGE AND MAY
NOT BE REFERRED TO AS AN INTERNATIONAL
STANDARD UNTIL PUBLISHED AS SUCH.
IN ADDITION TO THEIR EVALUATION AS
BEING ACCEPTABLE FOR INDUSTRIAL,
TECHNOLOGICAL, COMMERCIAL AND
USER PURPOSES, DRAFT INTERNATIONAL
STANDARDS MAY ON OCCASION HAVE TO
BE CONSIDERED IN THE LIGHT OF THEIR
POTENTIAL TO BECOME STANDARDS TO
WHICH REFERENCE MAY BE MADE IN
Reference number
NATIONAL REGULATIONS.
ISO/IEC DIS 30134-2.2:2015(E)
RECIPIENTS OF THIS DRAFT ARE INVITED
TO SUBMIT, WITH THEIR COMMENTS,
NOTIFICATION OF ANY RELEVANT PATENT
RIGHTS OF WHICH THEY ARE AWARE AND TO
©
PROVIDE SUPPORTING DOCUMENTATION. ISO/IEC 2015

ISO/IEC DIS 30134-2.2:2015(E)
© ISO/IEC 2015, Published in Switzerland
All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized otherwise in any form
or by any means, electronic or mechanical, including photocopying, or posting on the internet or an intranet, without prior
written permission. Permission can be requested from either ISO at the address below or ISO’s member body in the country of
the requester.
ISO copyright office
Ch. de Blandonnet 8 • CP 401
CH-1214 Vernier, Geneva, Switzerland
Tel. +41 22 749 01 11
Fax +41 22 749 09 47
copyright@iso.org
www.iso.org
ii © ISO/IEC 2015 – All rights reserved

ISO/IEC DIS 30134-2.2:2015-07-14
34 Contents Page
36 Introduction . 4
37 1 Scope . 5
38 2 Normative references . 5
39 3 Terms, definitions, abbreviations and symbols . 5
40 4 Applicable area of the Data Centre . 6
41 5 Determination of Power Usage Effectiveness (PUE) . 7
42 6 Measurement of Power Usage Effectiveness . 8
43 6.1 Measuring energy consumption . 8
44 6.2 Categories of Power Usage Effectiveness . 8
45 7 Reporting of Power Usage Effectiveness . 9
46 7.1 Requirements . 9
47 7.2 Supporting data for publicly reported PUE data . 10
48 Annex A (normative) Energy Measurements . 12
49 A.1 Measuring energy and calculating Power Usage Effectiveness . 12
50 A.2 Measurement locations. 12
51 A.3 Assessment frequencies . 13
52 Annex B (normative) Calculation of using various energy supplies . 14
53 B.1 Usage of different energy sources . 14
54 B.2 Examples of PUE calculation with various energy supplies . 15
55 B.3 Examples of PUE calculation with cogeneration using electricity and natural gas . 16
56 Annex C (informative) Interpretation of PUE . 18
57 C.1 General . 18
58 C.2 Data centre infrastructure versus IT equipment . 18
59 C.3 Interpretation in terms of energy . 18
60 C.4 Type of primary applications of data centre . 19
61 C.5 Better-than-perfect PUE measurements . 19
62 C.6 Comparing PUE results between data centres . 19
63 Annex D (normative)PUE derivatives . 21
64 D.1 General . 21
65 D.2 Partial PUE (pPUE) . 21
66 D.3 Design PUE (dPUE) . 25
67 D.4 Interim PUE (iPUE) . 27
68 Bibliography . 28
ISO/IEC DIS 30134-2.2:2015-07-14
71 Information Technology ‑ Data Centres ‑ Key Performance
72 Indicators - Part 2: Power Usage Effectivenss (PUE)
73 Introduction
74 The global economy is now reliant on information and communication technologies and the associated
75 generation, transmission, dissemination, computation and storage of digital data. All markets have
76 experienced exponential growth in that data, for social, educational and business sectors and, whilst the
77 internet backbone carries the traffic there are a wide variety of data centres at nodes and hubs within both
78 private enterprise and shared/collocation.
80 The historical data generation growth rate exceeds the capacity growth rate of the information and
81 communications technology hardware and with less than half (in 2014) of the world's population having
82 access to an internet connection, that growth in data can only accelerate. In addition, with many governments
83 having ‘digital agendas‘ to provide both citizens and businesses with ever faster broadband access, the very
84 increase in network speed and capacity will, by itself, generate ever more usage (Javons Paradox). Data
85 generation and the consequential increase in data manipulation and storage, is directly linked to increasing
86 power consumption.
88 With this background it is clear that data centre growth, and power consumption in particular, is an inevitable
89 consequence and that growth will demand increasing power consumption despite the most stringent energy
90 efficiency strategies. This makes the need for Key Performance Indicators (KPIs) that cover the effective use
91 of resources (including but not limited to energy) and the reduction of carbon emissions essential.
92 Within the ISO/IEC 30134 series, the term “resource usage effectiveness” is more generally used for KPIs in
93 preference to “resource usage efficiency”, which is restricted to situations where the input and output
94 parameters used to define the KPI have the same units.
95 In order to determine the overall resource effectiveness or efficiency of a data centre, a holistic suite of
96 metrics is required. This International Standard specifies Power Usage Effectiveness (PUE), which has
97 become a popular metric to determine the efficient utilisation and distribution of energy resources within a
98 data centre.
100 NOTE: It is recognised that the term “efficiency” should be employed for PUE but “effectiveness” provides continuity with
101 earlier market recognition of the term.
103 This International Standard is one of a series of standards for such KPIs and has been produced in
104 accordance with ISO/IEC 30134-1, which defines common requirements for a holistic suite of KPIs for data
105 centre resource usage effectiveness or efficiency.
106 At the time of publication of this International Standard the ISO/IEC 30134 series comprises the following
107 • ISO/IEC 30134-1,Information Technology - Data Centres - Key Performance Indicators - Part 1: Overview and
108 General Requirements,
109 • ISO/IEC 30134-2,Information Technology - Data Centres - Key Performance Indicators - Part 2: Power Usage
110 Effectiveness (PUE),
111 • ISO/IEC 30134-3,Information Technology - Data Centres - Key Performance Indicators - Part 3: Renewable
112 Energy Factor (REF),
113 • ISO/IEC 30134-4,Information Technology - Data Centres - Key Performance Indicators - Part 4: Equipment
114 Energy Efficiency for Servers (ITEE ),
SV
115 • ISO/IEC 30134-5,Information Technology - Data Centres - Key Performance Indicators - Part 5: IT Equipment
116 Utilization for Servers (ITEU ).
SV
117 Additional standards in the series ISO/IEC 30134 will be developed, each describing a specific KPI for
118 resource usage effectiveness or efficiency.
119 These International Standards do not specify limits or targets for any KPI and do not describe or imply, unless
120 specifically stated, any form of aggregation of individual KPIs into a combined nor an overall KPI for data
121 centre resource usage effectiveness or efficiency.
ISO/IEC DIS 30134-2.2:2015-07-14
122 1 Scope
123 This International Standard specifies the Power Usage Effectiveness (PUE) as a Key Performance Indicator
124 (KPI) to quantify the efficient use of energy in the form of electricity.
125 NOTE: It is recognised that the term “efficiency” should be employed for PUE but “effectiveness” provides continuity with
126 earlier market recognition of the term.
127 This international Standard
128 a) defines the Power Usage Effectiveness (PUE) of a data centre,
129 b) introduces PUE measurement categories,
130 c) describes the relationship of this KPI to a data centre’s infrastructure, information technology equipment
131 and information technology operations,
132 d) defines the measurement, the calculation and the reporting of the parameter,
133 e) provides information on the correct interpretation of the PUE.
134 PUE derivatives are described in Annex D.
135 2 Normative references
136 The following documents, in whole or in part, are normatively referenced in this document and are
137 indispensable for its application. For dated references, only the edition cited applies. For undated references,
138 the latest edition of the referenced document (including any amendments) applies.
139 ISO/IEC 30134-1, Information Technology - Data Centres - Key Performance Indicators - Part 1: Overview
140 and General Requirements
141 3 Terms, definitions, abbreviations and symbols
142 3.1 Terms and definitions
143 For the purposes of this document the definitions of ISO/IEC 30134-1 and the following apply.
144 3.1.1
145 information technology (IT) equipment energy consumption
146 energy consumed, measured in kilowatt hours (kWh), by equipment that is used to store, process, and
147 transport data within the computer room, telecommunication room and control room spaces
148 Note 1 to entry: Examples are servers, storage equipment, and telecommunications equipment.
149 3.1.2
150 power distribution unit (PDU)
151 equipment that allocates or partitions power or other energy consuming equipment
152 3.1.3
153 Power Usage Effectiveness (PUE)
154 ratio of the data centre total energy consumption to information technology equipment energy consumption,
155 calculated, measured or assessed across the same period.
156 Note 1 to entry: Sometimes the inverse value of PUE, referred to as Data Centre Infrastructure Efficiency
157 (DCiE), is used.
158 3.1.4
159 partial Power Usage Effectiveness (pPUE)
160 derivative of PUE, which is the ratio of the total energy consumption within a defined boundary to the
161 information technology equipment energy consumption
ISO/IEC DIS 30134-2.2:2015-07-14
164 3.1.5
165 designed Power Usage Effectiveness (dPUE)
166 derivative of PUE, which is the a projected PUE determined by the design targets of the data centre
167 3.1.6
168 Interim Power Usage Effectiveness (iPUE)
169 derivative of PUE, which is measured over a specified time other than a year
170 3.1.7
171 total data centre energy consumption
172 total energy consumption for all energy types serving the data centre, measured in kWh at its boundary
173 Note 1 to entry: Energy measured with energy metering devices at the boundary of the data centre.
174 Note 2 to entry: This includes electricity, natural gas and district utilities such as supplied chilled water or condenser
175 water.
176 3.2 Abbreviations
177 For the purposes of this document the abbreviations of ISO/IEC 30134-1 and the following apply:
178 CRAC Computer Room Air Conditioner units
179 CRAH Computer Room Air Handler units
180 dPUE designed Power Usage Effectiveness
181 DX Direct Expansion
182 iPUE interimPower Usage Effectiveness
183 IT information technology
184 PDU Power Distribution Unit
185 pPUE partial Power Usage Effectiveness
186 PUE Power Usage Effectiveness
187 r.m.s root mean square
188 ROI return on investment
189 UPS Uninterruptible Power Supply
190 3.3 Symbols
191 For the purposes of this document the following symbols apply:
192 E total data centre energy consumption (annual)
DC
193 E IT equipment energy consumption (annual)
IT
194 4 Applicable area of the data centre
195 Power Usage Effectiveness (PUE) as specified in this standard
196 a) is associated with the data centre infrastructure within its boundaries only,
197 b) describes the infrastructures energy efficiency relative to facilities with given environmental conditions, IT
198 load characteristics, availability requirements, maintenance, and security requirement,
199 c) illustrates the energy allocation of a data centre,
200 d) measures the relationship between the total data centre energy consumed and the IT equipment energy
201 consumed.
ISO/IEC DIS 30134-2.2:2015-07-14
202 When viewed in the proper context, PUE provides effective guidance and useful insight into the design of
203 efficient power and cooling architectures, the deployment of equipment within those architectures, and the
204 operation of that equipment.
205 PUE provides a means to determine
206 1) opportunities for the improvement of the operational efficiency of a data centre,
207 2) the improvement of the designs and processes of a data centre over time,
208 3) a design target or goal for new data centres, which includes the anticipated IT load range.
209 PUE does not take into account the
210 - energy efficiency of the IT load, its utilization, or productivity,
211 - impact of on-site electricity generation,
212 - efficiency of other resources such as human resource, space or water,
213 - use of renewable energy resources or accounts for re-use of waste by-products, such as heat.
214 PUE is not a
215  data centre productivity metric,
216  a standalone, comprehensive resource efficiency metric.
217 Derivatives of PUE which are useful in certain circumstances as described in Annex D.PUE should not be
218 used to compare different data centres.
219 5 Determination of Power Usage Effectiveness (PUE)
220 PUE is defined as:
E
DC
221 PUE (1)
E
IT
222 where
223 E = total data centre energy consumption (annual)
DC
224 E = IT equipment energy consumption (annual)
IT
225 By definition, the calculated PUE is always greater than 1.
226 Where the only energy source is from the electrical utility, then E is determined by the energy measured at
DC
227 the utility meter. PUE may be applied in mixed use buildings that allow of the differentiation between the
228 energy used for the data centre and that for other functions. Alternatively, the derivative partial PUE (pPUE)
229 may be applied (see Annex D).
230 E includes but is not limited to
IT
231 a) IT equipment (e.g. storage, processing and transport equipment),
232 b) supplemental equipment (e.g. keyboard/video/mouse (KVM) switches, monitors, and work-
233 stations/laptops used to monitor, manage, and/or control the data centre).
234 E includes E plus all the energy that is consumed to support the following infrastructures
DC IT
235 1) power delivery - including UPS systems, switchgear, generators, power distribution units (PDUs),
236 batteries, and distribution losses external to the IT equipment,
237 2) cooling system - including chillers, cooling towers, pumps, computer room air handling units (CRAHs),
238 computer room air conditioning units (CRACs), and direct expansion air handler (DX) units;
239 3) others including data centre lighting, elevator, security system, and fire suppression system.
ISO/IEC DIS 30134-2.2:2015-07-14
240 6 Measurement of Power Usage Effectiveness
241 6.1 Measuring energy consumption
242 6.1.1 General
243 In order to calculate PUE, it is necessary to measure E and E . This is not a trivial task, especially within
DC IT
244 existing data centres which may require the installation of instrumentation to collect the data.
245 NOTE: Although measurement of E and E are adequate to calculate PUE for the defined equipment and supporting
DC IT
246 infrastructure, more monitoring data of logical subsets is necessary to assess areas for potential improvements and to
247 evaluate the resulting improvements to PUE across the data centre.
248 6.1.2 Measurement period and frequency
249 The calculation of PUE requires the recording and documenting of E and E over a coincident period of
DC IT
250 twelve months. This standard does not specify the frequency of measurements of E and E , since PUE is
DC IT
251 calculated on an annual timeframe. However, the frequency of measurement employed will define the timing
252 of subsequent PUE calculations on a rolling annual basis.
253 6.1.3 Meter and measurement requirements
254 Measurement of E and E shall be undertaken using either
DC IT
255 a) watt meters with the capability to report energy usage
256 or
257 b) kilowatt-hour (kWh) meters that report the actual energy usage (true r.m.s), via the simultaneous
258 measurement of the voltage, current, and power factor over time.
259 NOTE: Kilovolt-ampere (kVA), the product of voltage and current, is not an acceptable measurement. Though the product
260 of volts and amperes mathematically results in watts, the actual energy consumption is determined by integrating a power
261 factor corrected value of volts and amperes. The frequency, phase variance, and load reaction causes energy calculation
262 difference between apparent energy and actual energy consumption. The error is inherently significant when power
263 delivery includes alternating current (AC). Kilovolt-ampere (kVA) measurements may be used for other functions in the
264 data centre, however, kVA is insufficient for efficiency measurements.
265 6.2 Categories of Power Usage Effectiveness
266 6.2.1 General
267 Three categories of PUE are defined:
268 a) Category 1 (PUE ), which provides a basic level of resolution of energy performance data;
269 b) Category 2 (PUE ), which provides an intermediate level of resolution of energy performance data;
270 c) Category 3 (PUE ), which provides an advanced level of resolution of energy performance data.
271 The higher categories provide progressively
272 1) more accurate measurements of energy usage (as the measurements are made closer to the devices
273 that consume the energy),
274 2) greater scope for energy efficiency improvements.
275 Table 1 provides a summary of the locations for the measurement of IT equipment energy consumption
276 associated with each category. In all cases, the total data centre energy consumption is measured from the
277 utility service entrance that feeds all of the electrical and mechanical equipment used to power cool and
278 condition the data centre.
279 To properly assess PUE, it is critical to account for all systems that support the data centre, in addition to the
280 environmental conditions, reliability, security and availability requirements independent of which PUE
281 measurement Category is chosen (see ISO/IEC 30134-1:201X, Annex B).
ISO/IEC DIS 30134-2.2:2015-07-14
284 Table 1 - PUE categories
PUE PUE PUE
1 2 3
a b c
Location of IT equipment energy UPS output PDU output IT equipment input
consumption measurement
a
includes impact of fluctuating IT and cooling loads.
b
excludes impact of losses associated with PDU transformers and static switches.
c
excludes impact of losses associated with electrical distribution components and non-IT related devices.
286 6.2.2 Category 1 (PUE ) – basic resolution
287 The IT load is measured at the output of the UPS (or equivalent) equipment and may be read
288 a) from the UPS front panel,
289 b) through a meter on the UPS output,
290 c) in cases of multiple UPS modules through a single meter on the common UPS output bus.
291 The incoming energy is measured from the utility service entrance that feeds all of the electrical and
292 mechanical equipment used to power, cool, and condition the data centre.
293 If UPS or an equivalent power failure ride through or conditioning unit is not available, other categories may
294 apply.
295 6.2.3 Category 2 (PUE ) – intermediate resolution
296 The IT load is measured at the output of the PDUs within the data centre and is typically read from the PDU
297 front panel or through a meter on the secondary of the PDU transformer. Individual branch circuit
298 measurement is also acceptable for Category 2.
299 6.2.4 Category 3 (PUE ) – advanced resolution
300 The IT load is measured at the IT equipment within the data centre. This can be achieved either by metered
301 rack (e.g. plug strips) that monitors aggregate set of IT systems or at the receptacle level or by the IT device
302 itself. Note that non-IT loads shall be excluded from these measurements.
303 6.2.5 Measurement placement
304 Each Category enables progressively improved accuracy of measurement of energy usage, as the
305 measurements are taken closer to the IT devises that consume energy.
306 7 Reporting of Power Usage Effectiveness
307 7.1 Requirements
308 7.1.1 Standard construct for communicating PUE data
309 In order for a reported PUE to be meaningful, the reporting organization shall provide the following
310 information
311 a) the data centre (including the boundaries of the structure) under inspection,
312 b) the PUE value,
313 c) the category,
314 d) the termination date of the period of measurement.
315 The PUE category shall be provided as a subscript to the name of the metric, e.g. PUE for a category 2
316 value.
317 The PUE category should be appropriate to the expected value of PUE:
ISO/IEC DIS 30134-2.2:2015-07-14
318 1) PUE > 1,5: Category 1 to Category 3;
319 2) 1,5 > PUE > 1,2: Category 2 or Category 3;
320 3) 1,2 > PUE: Category 3.
321 7.1.2 Example of reporting PUE values
322 Given the construct above, Table 2 provides examples of specific PUE designations and their interpretation.
323 Table 2 – Examples of PUE reporting
Sample PUE Designations Interpretation
In the year 2012 the PUE value of data centre X was
Data centre X, PUE (2012-12-31) = 2,25
2,25. It was a category 1 PUE.
In the period 2012-07-01to 2013-06-30 the PUE value
Data centre Y, PUE (2013-06-30) = 1,75
of data centre Y was 1,75. It was a category 1 PUE.
In the year 2013 the PUE value of data centre Z was
Data centre Z, PUE (2013-12-31) = 1,5
1,5. It was a category 2 PUE.
325 7.2 Supporting data for publicly reported PUE data
326 7.2.1 Required information
327 The following data shall be provided, when publicly reporting PUE data
328 a) contact information;
329 NOTE: Only the organization’s name or contact should be displayed in public inquiries.
330 b) data centre location information (address, county or region),
331 NOTE: Only state or local region information should be displayed in public inquiries.
332 c) measurement results: PUE with appropriate nomenclature including category designation,
333 d) measurement completion date: completion date identifies the annual period of the assessment.
334 7.2.2 Required supporting evidence
335 Information on the data centre which shall be available upon request includes
337 a) organization’s name, contact information and regional environmental description,
338 b) measurement results: PUE with appropriate nomenclature,
339 c) E and E ,
DC IT
340 d) start and measurement(s) dates the assessments were completed,
341 e) the accuracy level (IEC 62052 series and IEC 62053 series provide a reference for measurement of
342 electrical energy),
343 f) report on the size of computer room, telecom room and control room spaces,
344 g) external environmental conditions consisting of minimum, maximum and average temperature, humidity
345 and altitude.
346 7.2.3 Optional data
347 The following information may be useful in tracking the PUE trends within a data centre
348 a) data centre size (facility square meters),
349 b) total data centre design load for the facility (e.g. 10,2 MW),
350 c) name of the possible auditor and method used for auditing,
ISO/IEC DIS 30134-2.2:2015-07-14
351 d) data centre contact information,
352 e) data centre environmental conditions,
353 f) data centre’s mission,
354 g) data centre archetype percentages (e.g. 20% web hosting, 80% email),
355 h) data centre commissioned date,
356 i) numbers of servers, routers, and storage devices,
357 j) average and peak server CPU utilization,
358 k) percentage of servers using virtualization,
359 l) average age of IT equipment by type,
360 m) average age of facility equipment by type (cooling and power distribution equipment),
361 n) data centre category of reliability (see ISO/IEC 30134-1:201X, Annex B),
362 o) cooling and air-handling details.
363 NOTE: Other KPI’s within the ISO/IEC 30134 series can assist in the recording of the above information.
SSupupplplyy trans transffer er
sswiwitctchghgeaearr
ISO/IEC DIS 30134-2.2:2015-07-14
365 Annex A
366 (normative)
368 Energy measurements
369 A.1 Measuring energy and calculating Power Usage Effectiveness
370 In Figure A.1, total data centre energy consumption is measured at or near the data centre’s utility meter(s) to
371 accurately reflect the energy entering the data centre (see Annex B for other energy sources) in order that the
372 measurement represents the total energy consumed in the data centre.
BBuuiillddiinngg l looaadd -- ddeemmaanndd ffrroomm g grriidd
PPowoweerr
•• sswiwitctchhggeeaarr
ITIT e eqquuiippmmeenntt
•• UPUPSS ITIT l loadoad
eenneerrggyy
•• bbaatttteerryy b baacckkuupp •• sseerrvveerrss
TTootatall d daatata cceenntre tre ccoonnssuummpptitioonn
•• ooththeerr •• sstotorraaggee
eenneerrggyy
EE CooCoolilingng EE •• nneetwtwoorrkk t teelleeccoommss eeqquuiippmmeenntt
DCDC ITIT
ccoonnssuummpptitioonn
•• cchhiilllleerrss •• ooththeerr
•• CRACRACsCs
•• ooththeerr
374 Figure A.1 - Schematicof PUE calculation from measurements
375 Only the data centre portion of a facility’s utility consumption meter relevant to the data centre shall be
376 measured, since including in the calculation any energy that is not intended to be consumed in the data
377 centre would result in a non-compliant PUE calculation. For example, if a data centre resides in an office
378 building, the total energy drawn from the utility will be the sum of the total facility energy consumption for the
379 data centre and the total energy consumed by the non-data centre offices in the building. In this case, the
380 data centre administrator shall measure and subtract the amount of energy being consumed by the non-data
381 centre offices in order to calculate an accurate PUE.
382 A.2 Measurement locations
383 Figure A.2 shows measurement points to support the determination of E . E is measured after the utility
DC DC
384 feed at the utility metering point (i.e. point A). This measurement is consistent across all Categories. The
385 additional measurement points shown in Figure A.2 relate to the three PUE Categories. The measurement
386 points to allow determination of Category 1 PUE (PUE ), Category 2 PUE (PUE ) and Category 3 PUE
1 2
387 (PUE ), are indicated by B, C and D respectively.
PPrriimmararyy/s/sececonondadarryy
ssupupplplyy
NonNon--ITIT l loaoadd
ITIT eq equiuipmpmeentnt
PPrriimmararyy PPDUDU
ccononnenecctitiononss
AA
UPUPSS
LoLoccalal P PDUDU
AAddddiititiononaall
ssupupplplyy
UPUPSS
ITIT eq equiuipmpmeentnt
AA
BB
ccononnenecctitiononss
CC
DD
MeMeasasuremuremeent nt MeMeasasuremuremeent nt MeMeasasuremuremeent nt MeMeasasuremuremeent nt
llococatiationonss f for or llococatiationonss f for or llococatiationonss f for or llococatiationonss f for or
EE PPUEUE PPUEUE PPUEUE
DCDC 11 22 33
389 Figure A.2 - Monitoring and measurement points
ISO/IEC DIS 30134-2.2:2015-07-14
390 Monitoring energy consumption involves many aspects that can prevent it from being easy and
391 straightforward for the data centre operator. Costs can be quite high to install measuring instruments at every
392 point in the critical power path. Collecting, processing, and interpreting all the data also can be complex.
393 There is also some degree of error inherent in each of the meters measuring energy consumption, which can
394 affect results.
395 For a practical and achievable approach to monitoring, data centre operators should identify where it is most
396 beneficial to measure, taking into account associated improvements in PUE accuracy.
397 A.3 Assessment frequencies
398 Daily, weekly, monthly, and seasonal variances affect PUE. Increasing the minimum frequency of the
399 assessment cycle provides a larger and more accurate data set to analyze.
400 To fully understand and successfully manage the energy performance of a data centre, continuous real-time
401 monitoring should be used so that historical trending and statistical analysis can be done to determine where
402 efficiencies can be gained. This approach also enables early detection of unexpected variations that could
403 indicate system issues.
404 In cases where continuous real-time monitoring is not practical or economically justifiable, some form of
405 repeatable, defined process should be in place to capture measurements enabling the calculation of PUE as
406 often as possible for internal comparison purposes. If automated systems are employed the minimum
407 assessment frequency should be daily.
409 In all cases, the measurement methodology shall be consistent with the categories and locations defined in
410 6.3.
ISO/IEC DIS 30134-2.2:2015-07-14
411 Annex B
412 (normative)
414 Calculation of PUE using various energy supplies
415 B.1 Usage of different energy sources
416 E includes all energy consumed for data centre operation. The energy includes not only grid electricity but
DC
417 also on-site generated electricity and non-electricity forms of energy such as purchased district chilled water.
418 Energy measurements for PUE shall be captured in the form of kWh. To ensure PUE calculations for a data
419 centre utilizing on-site electric generation, the generated electricity which is measured by kWh meter should
420 be accounted for.
421 For a data centre utilizing non-electric energy, such as district chilled water, an energy conversion factor
422 enables the data centre owner/operator to include the energy of the chilled water.
423 Energy conversion factors are region-dependent because the amount of fuel needed to produce one unit of
424 chilled water will depend on the predominant method of chilled water generation in a given region. Some
425 countries have published values for these conversion factors. If regional energy conversion factors are
426 unavailable, the Standard Energy Conversion Factors of Table B.1 shall be used. These represent average
427 values of factors from different regions of the world. Naturally occurring cold water from the ground and air of
428 direct free cooling is not accounted for.
429 Table B.1 - Standard Energy Conversion Factors
Standard Energy
Energy type
Conversion Factor
District chilled water 0,4
District hot water 0,4
District steam 0,4
Fuel (for absorption type refrigerator) 0,35
430 .
431 The following formula shows an example of how a data centre that purchases electricity and district chilled
432 water can be expressed:
E
EE1,0 0,4
DC electrical chilled water
433 PUE (B.1)
E E
IT
IT
434 where
435 E = electrical energy consumed
electrical
436 E = district chilled water energy consumed
chilled water
ISO/IEC DIS 30134-2.2:2015-07-14
437 B.2 Examples of PUE calculation with various energy supplies
438 Figures B.1 to B.4 show examples of PUE calculation with various energy supplies.
DaDatata c ceenntre sitre sitete
ITIT E Enneerrggyy
UPUPSS PPDUDU
11 0 00000 000000 kkWWhh
11 0 00000 000000 kkWWhh
110000 0 00000 kkWWhh
EElleecctritriccaall
EElleecctritricciittyy p puurrcchhaassee
ddiisstritribbuutitioonn lloosssseess
11 6 63333 3 33333 kkWWhh
ChChiilllleerr ppllaanntt ChChiilllleedd wwaateterr
553333 3 33333 kkWWhh
PPUEUE == T Tootatall DC DC E Enneerrggyy/I/ITT E Enneerrggyy
PPUEUE == (1 6 (1 63333 3 33333))/(1 0/(1 00000 000000))
PPUEUE == 1 1,6,633
440 Figure B.1 - Example for a data centre purchasing all electricity
DaDatata c ceenntre sitre sitete
ITIT E Enneerrggyy
EElleecctritricciittyy p puurrcchhaassee
UPUPSS PPDUDU
11 1 10000 0 00000 kkWWhh
11 0 00000 000000 kkWWhh
110000 0 00000 kkWWhh
EElleecctritriccaall
ddiisstritribbuutitioonn lloosssseess
ChChiilllleedd wwaateterr
ChChiilllleedd wwaateterr
ppuurrcchhaassee
11 6 60000 0 00000 kkWWhh
11 6 60000 0 00000 kkWWhh
PPUEUE == T Tootatall DC DC E Enneerrggyy/I/ITT E Enneerrggyy
PPUEUE == (1 1 (1 10000 0 00000 x x 1 1 + + 1 1 6 60000 0 00000 x x 00,4,4))/(1 0/(1 00000 000000))
PPUEUE == 1 1,5,588
442 Figure B.2 - Example for a data centre purchasing electricity and chilled water
DataData c cenentrtre se siitete
ITIT E Enenerrgygy
3 03 00000 kkWWhh
EEllecectrtriicciittyy
NatuNaturralal ga gass GGeneneratorerator
5 05 00000 kkWWhh
NonNon--ITIT E Enenerrggyy
2 02 00000 kkWWhh
PPUEUE == T Totaotall DC E DC Enenerrggyy/IT/IT E Enenerrggyy
PPUEUE == ( (5 05 00000))/(/(3 03 00000))
PPUEUE == 1, 1,6767
444 Figure B.3 - Example for a data centre purchasing natural gas
ISO/IEC DIS 30134-2.2:2015-07-14
DaDaDaDatatatata c c c ceeeennnntre sitre sitre sitre sitetetete
ITIT E Enneerrggyy
33 0 00000 kkWWhh
EElleecctritricciittyy p puurrcchhaassee
EElleecctritricciittyy a at t ssiitete
55 0 00000 kkWWhh
22 5 50000 kkWWhh
22 5 50000 kkWWhh
NoNonn--ITIT E Enneerrggyy
pprroodduucceedd
22 0 00000 kkWWhh
NaNatuturraall g gaass GGeenneerraatotorr
PPUEUE == T Tootatall DC DC E Enneerrggyy/I/ITT E Enneerrggyy
PPUEUE == (2 5 (2 50000 + + 2 2 5 50000))/(3 0/(3 00000))
PPUEUE == 1 1,6,677
446 Figure B.4 - Example for a data centre purchasing electricity and natural gas
447 B.3 Examples of PUE calculation with cogeneration using electricity and natural
448 gas
449 Figures B.5 and B.6 show examples of PUE calculation with cogeneration using electricity and natural gas.
DaDatata c ceenntre sitre sitete
11 4 45500 0 00000 kkWWhh
11 5 50000 0 00000 kkWWhh
EElleecctritricciittyy ITIT
EElleecctritricciittyy
ffaacciilliittyy eeqquuiippmmeenntt
445500 0 00000 kkWWhh
NaNatuturraall g gaass CoCoggeenneerraatitioonn
CoCooolliinngg, , aaiirr h haannddlliinngg, , lliigghhtitinngg a anndd
ccoonntroltrollleerrss
440000 0 00000 kkWWhh
mmeeaassuurreedd bbyy TT a anndd ffllooww
EEEllleeeccctritritriccciiitttyyy
HeHeHeaaat t t ooorrr Ch Ch Chiiilllllleeeddd
MMMeeeaaasssuuurrreeemmmeeennnt t t pppoooiiinnnttt
PPUEUE == T Tootatall DC DC E Enneerrggyy/I/ITT E Enneerrggyy
PPUEUE == (1 4 (1 45500 0 00000 + + 4 45500 0 00000 + + 4 40000 0 00000 x x 0 0,4,4))/(1 /(1 550000 0 00000))
PPUEUE == 1 1,6,677
451 Figure B.5 - Method 1: Measured by chilled water flow
452 If the energy consumption of the chiller or the refrigerator cannot be measured, the heat energy of chilled
453 water may be multiplied by 0,4. Note that in this case the efficiency of the chiller is not taken into account for
454 PUE calculation. Since the performance of the chiller is regarded to be the same as district chilled water, the
455 actual PUE value can increase or decrease.
DataData c cenentrtre se siitete
1 450 000 1 450 000 kkWWhh
1 500 000 1 500 000 kkWWhh
EEllecectrtriicciittyy ITIT
EEllecectrtriicciittyy
ffaacciilliittyy eqequiuipmpmenentt
450 000 450 000 kkWWhh
mmeeasasured ured bbyy WWHMHM
1 000 1 000 000000 kkWWhh
NatuNaturralal ga gass CogeCogenenerratiationon
CoolCooliingng, A, Aiirr ha handndlliingng, L, Liighghtitingng an andd
cconontrtrololllersers
EEElllececectrtrtriiiccciiitttyyy
HeatHeatHeat or Chi or Chi or Chillllllededed
MeMeMeasasasuremuremuremeeent nt nt popopoiiintntnt
PPUEUE == T Totaotall DC E DC Enenerrggyy/IT/IT E Enenerrggyy
PPUEUE == ( (1 41 45050 00 000 +0 + 45 450 00 0000 +0 + 0, 0,3535 x x 1 1 00000 0 000000 xx 0 0,47,47))/(/(1 51 50000 00 000)0)
PPUEUE == 1, 1,3838
457 Figure B.6 - Method 2: Calculated from energy required to produce chilled water
ISO/IEC DIS 30134-2.2:2015-07-14
458 In the example of method 2, the input energy ratio for chilled water and electricity should be measured for this
459 co-generation system, based on the benchmark method defined in ISO 26382. The input energy ratio is
460 assumed to be 47 % for chilled water and 53 % for electricity.
461 B.4 Examples of PUE calculation with absorption type refrigerator
462 Figures B.7and B.8 show examples of PUE calculation with absorption type refrigerator.
DataData c cenentrtre se siitete
1 200 000 1 200 000 kkWWhh
1 000 000 1 000 000 kkWWhh
EEllecectrtriicciittyy ITIT
EEllecectrtriicciittyy
ffaacciilliittyy eqequiuipmpmenentt
AAbsbsorptiorptionon
NatuNaturralal ga gass
rrefefrriigegerratoatorr
CoolCooliingng, a, aiirr ha handndlliingng, l, liighghtitinng ag andnd
cconontrtrololllersers
350 000 350 000 kkWWhh
EEElllececectrtrtriiiccciiitttyyy
HeatHeatHeat or Chi or Chi or Chillllllededed
MeMeMeasasasuremuremuremeeent nt nt popopoiiintntnt
PPUEUE == T Totaotall DC E DC Enenerrggyy/IT/IT E Enenerrggyy
PPUEUE == ( (1 21 20000 00 000 +0 + 35 350 00 0000 x0 x 0, 0,4)/(4)/(1 01 00000 00 000)0)
PPUEUE == 1, 1,3434
464 Figure B.7 - Method 1: Measured by chilled water flow
465 If the energy consumption of the chiller or the refrigerator cannot be measured, the heat energy of chilled
466 water may be multiplied by 0,4. Note that in this case the efficiency of the chiller is not taken into account for
467 PUE calculation. Since the performance of the chiller is regarded to be the same as district chilled water, the
468 actual PUE value can increase or decrease.
DataData c cenentrtre se siitete
1 200 000 1 200 000 kkWWhh
1 000 000 1 000 000 kkWWhh
EEllecectrtriicciittyy ITIT
EEllecectrtriicciittyy
ffaacciilliittyy eqequiuipmpmenentt
1 000 000 1 000 000 kkWWhh
AAbsbsorptiorptionon
NatuNaturralal ga gass
rrefefrriigegerratoatorr
CoolCooliingng, a, aiirr ha handndlliingng, l, liighghtitinng ag andnd
cconontrtrololllersers
EEElllececectrtrtriiiccciiitttyyy
HeatHeatHeat or Chi or Chi or Chillllllededed
MeMeMeasasasuremuremuremeeent nt nt popopoiiintntnt
PPUEUE == T Totaotall DC E DC Enenerrggyy/IT/IT E Enenerrggyy
PPUEUE == ( (1 21 20000 00 000 +0 + 0, 0,3535 x x 1 01 00000 00 000)/(0)/(1 1 00000 00 0000)0)
PPUEUE == 1, 1,5555
470 Figure B.8 - Method 2: Measured by input gas
ISO/IEC DIS 30134-2.2:2015-07-14
471 Annex C
472 (informative)
474 Interpretation of PUE
475 C.1 General
476 Each data centre has individual characteristics, capabilities, and operational policies (e.g. government
477 regulations and policy for data centres, climate and locations, and customer’s requirements) that affect its
478 performance and PUE value. Interpretation of the PUE value has to understand the context with these
479 conditions. With such varying conditions and aspects, the most beneficial use of PUE is for self improvement.
480 Relative to these conditions, the reduction of PUE implies a reduction in the energy overhead needed to
481 house the data centre equipment. Each data centre also has different capabilities with respect to collecting
482 and analyzing energy consumption data. Without additional information about the reported results,
483 interpretations of data collected by different organizations using different approaches over different
484 timeframes may be meaningless or misleading.
485 There has been a number of common misunderstandings with regard to the calculation and interpretation of
486 PUE results. Professionals making claims should be aware of the following issues and ensure they are
487 reporting and interpreting valid numbers prior to making any public claims. This Annex provides guidelines
488 and consideration points for interpreting PUE results properly.
489 C.2 Data centre infrastructure versus IT equipment
490 Each load in a data centre is designated as either an IT load, an infrastructure load, or not included in the
491 calculation. Many data centres are in mixed use buildings where there are significant offices or other loads
492 that are not related to the data centre function. Mixed use buildings may also have shared systems such as
493 cooling towers, switchgear, or ventilation systems. In these cases, PUE reporting shall explicitly describe how
494 the loads have been incorporated into the calculation. For the purpose of improving a single, specific data
495 centre, what remains important is not the exact allocation of shared loads to the PUE calculation, but rather
496 that the calculations be performed in a consistent manner.
497 Issues such as availability of equipment and the productivity of the data centre require different metrics and
498 different analyses. Data centres are complex systems. Frequently, changing certain aspects of the facility
499 (e.g., commissioning a new class of server) can produce apparent changes in PUE results for
...