Information technology — Data centre facilities and infrastructuresPart 4-2: Power Usage Effectiveness BSI Standards Publication... NORME EUROPÉENNE English Version Information technolog
Terms and definitions
For the purposes of this document, the terms and definitions given in EN 50600-1 and the following apply
Information technology equipment energy consumption refers to the energy utilized, measured in kilowatt-hours (kWh), by devices responsible for storing, processing, and transporting data within computer rooms, telecommunication rooms, and control room environments.
Note 1 to entry: Examples are servers, storage equipment and telecommunications equipment
3.1.2 power distribution unit equipment that allocates or partitions power for other energy consuming equipment
Power Usage Effectiveness ratio of the data centre total energy consumption to information technology equipment energy consumption, calculated, measured or assessed across the same period
Note 1 to entry: It is recognized that the term “efficiency” should be employed for PUE but “effectiveness” provides continuity with earlier market recognition of the term
Note 2 to entry: Sometimes the inverse value of PUE, referred to as Data Centre Infrastructure Efficiency (DCiE), is used
3.1.4 partial Power Usage Effectiveness derivative of PUE, which is the ratio of the total energy consumption within a defined boundary to the information technology equipment energy consumption
3.1.5 designed Power Usage Effectiveness derivative of PUE, which is a projected PUE determined by the design targets of the data centre
3.1.6 interim Power Usage Effectiveness derivative of PUE, which is measured over a specified time other than a year
3.1.7 total annual data centre energy consumption total energy consumption for all energy types serving the data centre, measured in kWh at its boundary
Note 1 to entry: Energy is measured with energy metering devices at the boundary of the data centre or points of generation within the boundary
Note 2 to entry: This includes electricity, natural gas and district utilities such as supplied chilled water or condenser water
Note 3 to entry: Total annual energy includes supporting infrastructure.
Abbreviations
For the purposes of this document, the abbreviations given in EN 50600-4-1 and the following apply
CRAC Computer Room Air Conditioner/Conditioning
CRAH Computer Room Air Handler units dPUE designed Power Usage Effectiveness
DX Direct Expansion idPUE interim designed Power Usage Effectiveness iPUE interim Power Usage Effectiveness
PDU Power Distribution Unit pPUE partial Power Usage Effectiveness
PUE Power Usage Effectiveness r.m.s root mean square
Symbols
For the purposes of this document the following symbols apply
E DC total data centre energy consumption (annual) in kWh
E IT IT equipment energy consumption (annual) in kWh
4 Applicable area of the data centre
Power Usage Effectiveness (PUE) is a metric that pertains exclusively to the data center infrastructure within its defined boundaries and measures the energy efficiency of that infrastructure in relation to facilities operating under similar environmental conditions.
IT load characteristics, availability requirements, maintenance, and security requirement; c) illustrates the energy allocation of a data centre.
PUE offers valuable insights and guidance for designing efficient power and cooling systems, optimizing equipment deployment within these systems, and ensuring effective operation of the equipment.
PUE provides a means to determine:
1) opportunities for the improvement of the operational efficiency of a data centre;
2) the improvement of the designs and processes of a data centre over time;
3) a design target or goal for new data centres across the anticipated IT load range.
PUE does not take into account the:
— energy efficiency of the IT load, its utilization or productivity;
— efficiency of on-site electricity generation;
— efficiency of other resources such as human resource, space or water;
— use of renewable energy resources or accounts for re-use of waste by-products (such as heat)
• a standalone, comprehensive resource efficiency metric.
Derivatives of PUE which are useful in certain circumstances are described in Annex C PUE should not be used to compare different data centres
5 Determination of Power Usage Effectiveness
General
E DC = total data centre energy consumption (annual) in kWh;
E IT = IT equipment energy consumption (annual) in kWh
By definition, the calculated PUE is always greater than 1
In scenarios where the sole energy source is the electrical utility, the Energy Data Center (EDC) is influenced by the position of the utility meter Power Usage Effectiveness (PUE) can be utilized in mixed-use buildings to distinguish between energy consumption for the data center and other operational functions Additionally, the partial Power Usage Effectiveness (pPUE) may be implemented for more detailed analysis.
E IT encompasses a range of components, including IT equipment such as storage, processing, and transport devices, as well as supplemental equipment like keyboard/video/mouse (KVM) switches, monitors, and workstations or laptops utilized for monitoring, managing, and controlling data centers.
E DC includes E IT plus all the energy that is consumed to support the following infrastructures:
1) power delivery - including UPS systems, switchgear, generators, power distribution units (PDUs), batteries, and distribution losses external to the IT equipment;
2) cooling system - including chillers, cooling towers, pumps, computer room air handling units (CRAHs), computer room air conditioning units (CRACs), and direct expansion air handler (DX) units;
3) others including data centre lighting, elevator, security system, and fire detection/suppression system.
Total data centre energy consumption
The data centre under consideration shall be viewed at as a system defined by interfaces through which energy flows
The following forms of energy shall be metered at the interfaces: a) electricity; b) gaseous fuel; c) fluid fuel; d) fluids for cooling (comprising water usage when returned fluid and not evaporated).
The following forms of energy are not required be metered at these interfaces:
2) water from natural sources (i.e requiring no energy consumption in its provision).
All electrical energy at interfaces must be measured in kilowatt-hours (kWh) If any required energy forms are missing from the interfaces, the Energy Data Center (EDC) cannot be established, making it impossible to calculate the Power Usage Effectiveness (PUE).
Gaseous or liquid fuels must be measured in kWh or converted into kWh based on their heat of combustion values In cases where combustion values are unavailable, specific standard values should be utilized.
The cooling energy provided by fluids will be assessed through heat meters, which measure flow rate and differential temperature This data will then be multiplied by the appropriate conversion factor for the fluid system in use.
Technical subsystems are classified based on their metering: those with output meters, such as on-site co-generation of heat and electricity, are deemed external to the system, while subsystems with input meters or partial output metering are regarded as internal.
Total data centre energy consumption in mixed-use buildings
The energy consumption of data centres located in mixed-use buildings will be assessed based solely on the data centre's energy use, provided that metering of all shared technical subsystems enables the separation of energy usage.
To accurately assess energy consumption in data centers, it is essential to separate the energy use of shared technical subsystems from the overall building energy usage Implementing the necessary meters for this separation can help mitigate the impact on Power Usage Effectiveness (PUE).
6 Measurement of Power Usage Effectiveness
Measuring energy consumption
General
To calculate Power Usage Effectiveness (PUE), it is essential to measure both Energy Data Center (EDC) and Energy IT (EIT) This process can be complex, particularly in existing data centers, as it often necessitates the installation of specialized instrumentation to gather the required data.
To effectively calculate Power Usage Effectiveness (PUE) for specific equipment and supporting infrastructure, it is essential to measure Energy Data Center (EDC) and Energy Infrastructure Technology (EIT) However, obtaining additional monitoring data from logical subsets is crucial for identifying potential improvement areas and evaluating the impact of these enhancements on PUE throughout the data center.
Measurement period and frequency
The calculation of PUE requires the recording and documenting of EDC and EIT over a coincident period of
The standard allows for a measurement period of 12 months for EDC and EIT, without specifying how often these measurements should occur Since PUE is calculated annually, the frequency of measurements will influence the timing of future PUE calculations on a rolling annual basis.
Meter and measurement requirements
The measurement of Energy Demand Control (EDC) and Energy Information Technology (EIT) can be performed using either watt meters that provide energy usage data or kilowatt-hour (kWh) meters that accurately report true root mean square (r.m.s) energy consumption by simultaneously measuring voltage, current, and power factor over time.
Kilovolt-ampere (kVA) is not a reliable measurement for energy consumption, as it merely represents the product of voltage and current While volts and amperes mathematically yield watts, actual energy usage must account for the power factor, which integrates frequency, phase variance, and load reaction These factors create discrepancies between apparent energy and true energy consumption, particularly in alternating current (AC) systems Although kVA can serve other purposes in data centers, it falls short for accurately measuring efficiency.
Categories of Power Usage Effectiveness
General
The Power Usage Effectiveness (PUE) is classified into three distinct categories: Category 1 (PUE1) offers a basic resolution of energy performance data, Category 2 (PUE2) delivers an intermediate resolution, and Category 3 (PUE3) provides an advanced level of detail in energy performance data.
The higher Categories provide progressively:
1) more accurate measurements of energy usage (as the measurements are made closer to the devices that consume the energy),
2) greater scope for energy efficiency improvements.
Table 1 summarizes the locations where IT equipment energy consumption is measured for each category The total energy consumption of the data centre is assessed from the utility service entrance, which supplies power to all electrical and mechanical equipment necessary for operating, cooling, and conditioning the data centre.
To accurately evaluate Power Usage Effectiveness (PUE), it is essential to consider all systems that support the data center, along with environmental conditions, reliability, security, and availability requirements, regardless of the selected PUE measurement category (refer to EN 50600-4-1:2016, Annex A).
Measuring energy consumption of IT equipment involves three key locations: the UPS output, which reflects the influence of varying IT and cooling loads; the PDU output, which does not account for losses from PDU transformers and static switches; and the IT equipment input, which excludes losses from electrical distribution components and non-IT devices.
Category 1 (PUE1) – basic resolution
The IT load is assessed at the output of the UPS or similar equipment, which can be monitored in several ways: a) directly from the front panel of the UPS, b) via a meter located on the UPS output, or c) in scenarios involving multiple UPS modules, through a single meter connected to the common UPS output bus.
If UPS or an equivalent power failure ride through or conditioning unit is not available, other categories can apply.
Category 2 (PUE2) – intermediate resolution
The IT load in a data center is assessed at the output of the Power Distribution Units (PDUs), typically obtained from the PDU front panel or via a meter on the PDU output, regardless of whether a transformer is used For Category 2, measuring the load at individual branch circuits is also permissible.
Category 3 (PUE3) – advanced resolution
The IT load in a data center is assessed through metered racks, such as plug strips, which track the total power consumption of IT systems, or directly at the receptacle level or from the IT devices themselves It is important to exclude non-IT loads from these measurements to ensure accurate data.
Measurement placement
Each Category enables progressively improved accuracy of measurement of IT equipment energy consumption, as the measurements are taken closer to the IT devices that consume energy
7 Reporting of Power Usage Effectiveness
Requirements
Standard construct for communicating PUE data
To ensure the reported Power Usage Effectiveness (PUE) is meaningful, organizations must disclose specific information, including the data center's boundaries, the PUE value, the category of the data center, and the termination date of the measurement period.
The PUE Category shall be provided as a subscript to the name of the metric, e.g PUE2 for a Category 2 value.
Example of reporting PUE values
Using the construct of 7.1.1, Table 2 provides examples of specific PUE designations and their interpretation
Table 2 — Examples of PUE reporting
Data centre X, PUE1 (2012–12–31) = 2,25 In the year 2012 the PUE value of data centre X was 2,25 It was a category 1 PUE.
Data centre Y, PUE1 (2013–06–30) = 1,75 In the period 2012–07–01 to 2013–06–30 the
PUE value of data centre Y was 1,75 It was a category 1 PUE.
Data centre Z, PUE2 (2013–12–31) = 1,50 In the year 2013 the PUE value of data centre Z was 1,50 It was a category 2 PUE.
Data for public reporting PUE
When publicly reporting Power Usage Effectiveness (PUE) data, organizations must provide specific information: a) only the organization's name or contact should be displayed for public inquiries; b) data center location details, such as address and county or region, should be limited to state or local region information; c) measurement results must include PUE with the correct nomenclature and category designation.
7.1.3.2 Supporting evidence (where required by authorities having jurisdiction)
The data centre information available upon request includes the organization's name, contact details, and a regional environmental description It also encompasses measurement results such as Power Usage Effectiveness (PUE) with the appropriate nomenclature, along with Energy Data Centre (EDC) and Energy Information Technology (EIT) metrics Additionally, the report specifies the start and completion dates of the assessments, the accuracy level based on the EN 62052 and EN 62053 series for electrical energy measurement, and details on the sizes of the computer room, telecom room, and control room Lastly, it provides external environmental conditions, including minimum, maximum, and average temperature, humidity, and altitude.
Recommendations
Use of PUE Category
The PUE Category should be appropriate to the expected value of PUE:
Trend tracking data
To effectively track Power Usage Effectiveness (PUE) trends in a data centre, it is essential to consider several key factors: the size of the facility in square meters, the total design load (e.g., 10.2 MW), the auditor's name and auditing method, and the contact information for the data centre Additionally, understanding the environmental conditions, the mission of the data centre, and the percentages of different archetypes (e.g., 20% web hosting, 80% email) is crucial Other important details include the commissioning date, the number of servers, routers, and storage devices, as well as average and peak CPU utilization rates The percentage of servers utilizing virtualization, the average age of IT equipment and facility equipment (such as cooling and power distribution), and the data centre's availability objectives (referencing EN 50600-4-1:2016, Annex A) should also be documented, along with cooling and air-handling specifics.
NOTE Other KPIs within the EN 50600–4-X series can assist in the recording of the above information.
Measuring energy and calculating Power Usage Effectiveness
Total data centre energy consumption is assessed at or near the utility meter(s) to ensure an accurate representation of the energy entering the facility, reflecting the overall energy usage within the data centre For additional energy sources, refer to section 5.2.
Figure A.1 — Schematic of PUE calculation from measurements
To ensure an accurate Power Usage Effectiveness (PUE) calculation, only the energy consumption relevant to the data centre should be measured Including energy used by non-data centre areas, such as offices in the same building, would lead to a non-compliant PUE Therefore, data centre administrators must measure and deduct the energy consumed by these non-data centre offices from the total facility energy consumption to achieve a precise PUE value.
Measurement locations
Figure A.2 illustrates the measurement points essential for determining Energy Data Center (EDC) metrics EDC is consistently measured at the utility metering point, referred to as point A, across all categories The additional measurement points, labeled B, C, and D, correspond to the three Power Usage Effectiveness (PUE) categories: Category 1 PUE (PUE1), Category 2 PUE (PUE2), and Category 3 PUE (PUE3).
NOTE The measurement points do not correspond to granularity level 1 of EN 50600–2-2:2014
Figure A.2 — Monitoring and measurement points
Monitoring energy consumption in data centers is a complex task that poses challenges for operators The high costs associated with installing measuring instruments at every critical point in the power path can be a significant barrier Additionally, the processes of collecting, processing, and interpreting the vast amounts of data can complicate effective energy management.
There is also some degree of error inherent in each of the meters measuring energy consumption, which can affect results
For a practical and achievable approach to monitoring, data centre operators should identify where it is most beneficial to measure, taking into account associated improvements in PUE accuracy.
Assessment frequencies
Increasing the minimum frequency of the assessment cycle provides a larger and more accurate data set to analyse
Effective management of a data centre's energy performance requires continuous real-time monitoring, allowing for historical trending and statistical analysis to identify potential efficiencies This method also facilitates the early detection of unexpected variations that may signal system problems.
When continuous real-time monitoring is impractical or not cost-effective, it is essential to establish a repeatable process for capturing measurements that allow for the calculation of Power Usage Effectiveness (PUE) as frequently as possible for internal comparisons If automated systems are utilized, assessments should occur at least once a day.
In all cases, the measurement methodology shall be consistent with the Categories and locations defined in 6.2
Calculation of PUE using various energy supplies
Examples of PUE calculation with various energy supplies
Figures B.1 to B.4 show examples of PUE calculation with various energy supplies
Figure B.1 — Example for a data centre purchasing all electricity
Figure B.2 — Example for a data centre purchasing electricity and chilled water
Figure B.3 — Example for a data centre purchasing natural gas
Figure B.4 — Example for a data centre purchasing electricity and natural gas
Examples of PUE calculation with cogeneration using electricity and natural gas
Figures B.5 and B.6 show examples of PUE calculation with cogeneration using electricity and natural gas
Figure B.5 — Method 1: Measured by chilled water flow
Figure B.6 — Method 2: Calculated from energy required to produce chilled water
Examples of PUE calculation with absorption type chiller
Figures B.7 and B.8 show examples of PUE calculation with absorption type refrigerator
Figure B.7 — Method 1: Measured by chilled water flow
Figure B.8 — Method 2: Measured by input gas
General
Purpose of PUE derivatives
PUE derivatives are useful to support an effective energy management process such as those described in
EN 50600-3-1 Each derivative shall be accompanied with specific information that describes the specific situation.
Using PUE derivatives
The PUE derivatives shall be designated and shall be documented as one of the following
Partial Power Usage Effectiveness (pPUE) measures the energy efficiency of specific subsets within a data center's infrastructure It encompasses various elements, including the data center's boundaries and resiliency level, a detailed inventory of shared resources, the methodology for assessing these shared resources, and all relevant supporting evidence for PUE calculations.
Interim PUE (iPUE) describes a PUE measured for a period less than a year (see C.2) iPUE shall include but is not limited to the following supporting data:
— the boundaries of the data centre including resiliency level;
— all other PUE supporting evidence which exists during the defined intervals
Designed Power Usage Effectiveness (dPUE) refers to the anticipated PUE of a data center before it becomes operational or undergoes a specific operational change It encompasses various supporting data essential for accurate predictions.
1) the boundaries of the data centre including resiliency level;
2) a schedule of interim PUE and PUE based on target IT loads and environmental conditions;
3) all other PUE supporting evidence available prior to operation including target commissioning date. iPUE may be used to validate dPUE parameters
Combined use of the terms is permitted to describe specific situations and values An example use of these derivatives is:
• d/i/pPUE (20XX-08-01:20XX-08-31) = 3,1 [ref jjj];
• [jjj]: [boundaries of the data centre, shared cooling, space, physical security];
• 40 % IT load; environmental conditions; etc.
Interim PUE
PUE, or Power Usage Effectiveness, is defined as an annual metric that necessitates ongoing monitoring of both IT energy and total data center energy for a minimum of one year Each reported PUE value must be accompanied by its corresponding category and the measurement period.
For effective energy management, it is beneficial to measure and report shorter timeframes than a full year, referred to as "interim PUE" (iPUE) Each iPUE value should include its category, the measurement period, and additional context and reporting information necessary for annualized PUE.
By decreasing the measurement interval to a minimum, a real-time iPUE may be established.
Partial PUE (pPUE)
General
PUE measures the total energy consumption of a data center, while pPUE focuses on the energy usage of specific subsystems within the data center's infrastructure This metric can be applied to various types of data centers, as it defines the boundaries of these subsystems internally.
Partial PUE (pPUE) is calculated as follows: sub IT sub IT
E sub = energy consumption (annual) of the subsystem in kWh;
E IT = IT equipment energy consumption (annual) in kWh
pPUE, like PUE, pertains to IT energy consumption and is calculated annually, necessitating a full year of data collection To report pPUE, it is essential to provide the same disclosures as for PUE, along with a clear identification of the specific sub-system or zone being analyzed A zone consists of a relevant group of infrastructure components that consume energy, which must be evaluated for energy efficiency.
In energy management for data centers, it is essential to define zones for each subsystem, including electrical distribution (such as UPS), air handling, and cooling These subsystems are crucial for optimizing energy efficiency and are characterized by specific formulas related to electrical IT power.
E electrical = energy consumption (annual) of the electrical systems in kWh
E HVAC = energy consumption (annual) of the heating, ventilation and air conditioning systems in kWh
Ecooling = energy consumption (annual) of the cooling systems in kWh
This standard enables the definition of additional zones to effectively analyze and comprehend the energy contribution of a data center subsystem Its goal is to enhance the energy efficiency of the subsystem being evaluated, ultimately improving overall energy performance.
The pPUE concept (and any reported value) is only applicable to the zones of a data centre under study
It is meaningless to apply a pPUE to a part of the building that is not a zone of the data centre
The pPUE metric is not applicable to areas without IT load, although other KPIs may be relevant It can be used to assess specific sections of a data center where IT equipment is located, even if those sections share resources with other areas However, regions not being evaluated, regardless of whether they contain IT equipment, are excluded from this assessment The evaluation boundaries must be defined according to EN 50600-4-1 standards.
Zoning
The typical application of pPUEs is confined to data centre environments As part of the energy management strategy, it is essential to delineate the zones of infrastructure subsystems within the data centre, which is determined by the facility's technical design.
For most of the data centres in post-commissioning and in operation, the zoning in Figure C.1 applies
Figure C.1 — Zoning for a data centre
Whether or not the zone “other” shall be included depends on the significance of the energy use of that zone
Initially overlooked, the "other" zone in energy management becomes significant later in the process, once the efficiency of the primary zones has improved to a relevant level.
In case the cooling is provided by DX systems, air handling and cooling cannot be separated Therefore, the zoning of Figure C.2 might be a better approach
Figure C.2 — Zoning for a data centre using DX cooling
In case water is used for an additional cooling system and water transportation and treatment uses a significant amount of energy, the zoning of Figure C.3 is a good approach
Figure C.3 — Zoning for a data centre using water
The standard does not outline a specific method for defining a zone; however, it emphasizes that any designated zone must be appropriate for the intended energy management process and, when necessary, should be modified in accordance with the maturity level of that process.
Metering requirements for pPUE
In order to obtain an appropriate measure of Esub it is typically required to install meters at each outlet of the main PDU
Measurements shall be in accordance with Clause 6.
Reporting of pPUE
Use of pPUE in energy management
The primary goal of using pPUE is to analyze and pinpoint potential energy savings by identifying inefficient zones and subsystems within infrastructure Additionally, pPUE serves to verify the effectiveness of implemented improvement measures For instance, Figure C.4 illustrates a data center with designated zones for HVAC and cooling, where arrows mark the times when efficiency-enhancing measures were applied to the relevant infrastructure components.
pPUE is a valuable metric for estimating the potential of improvement measures and calculating the return on investment (ROI) associated with them By understanding the operational conditions and their corresponding pPUE, the impact of a measure can be quantified as a reduction in pPUE Annual savings are derived from this reduction, multiplied by the annual energy costs for IT equipment The ROI is determined by dividing the initial investment by the annual savings, indicating the time required to recoup the investment.
Use of pPUE in mixed use buildings
In mixed-use buildings, the shared infrastructure can complicate the calculation of Power Usage Effectiveness (PUE), as it becomes challenging to attribute energy consumption specifically to either the building or the data centre.
In mixed-use buildings where the cooling infrastructure serves both data center and office spaces, calculating the Power Usage Effectiveness (PUE) can be challenging due to the inability to separately measure energy consumption However, it is still feasible to determine partial PUE (pPUE) for power distribution and HVAC systems Despite this, the usefulness of these pPUE metrics is limited without a comprehensive understanding of the overall PUE Therefore, it is advisable to implement appropriate metering solutions to distinguish energy usage across the primary infrastructure components in mixed-use buildings.
In calculating pPUE for mixed-use buildings, it is important to consider accepted exceptions for ancillary energy loads associated with shared spaces These include offices, laboratories, cubicles, conference rooms, elevators, lobbies, kitchens or break rooms, parking areas, toilets, corridors, stairs, and convenience stores.
Designed PUE
The energy efficiency of a data centre can be anticipated during the design phase by considering the expected growth scenarios and occupancy levels, as well as the projected timeline for fluctuations in energy consumption.
Table C.1 presents predictions for a containerized data centre, illustrating expected loads based on the target occupancy This analysis leads to a designed Power Usage Effectiveness (dPUE) for each stage, culminating in an annualized dPUE value of 1.20.
Table C.1 — Example of dPUE calculation
Month IT equipment Cooling/ventilation/ humidification Power distributi on
UPS Lighting Remaining support Total data centre in
Nr Dura- tion Avera ge load
Energy used a Average load a Energyus ed Energy used Energy used Energy used Energy used Energy used idPUE
# Days kW kWh kW kWh kWh kWh kWh kWh kWh
□ dPUE Σ 365 1 191 240 144 840 20 329 66 214 2 920 8 760 1 434 303 1,20 a Forecasted use or estimate
In the design phase, dPUE serves as a target for optimal operation defined by the designer, factoring in the local climate conditions, including outside air temperature and humidity, relevant to the data center's location.
In the operational phase, dPUE indicates the anticipated PUE value derived from a resource capacity forecast, such as EN 50600-3-1, which considers the expected energy consumption of both installed and planned data center infrastructures and IT equipment The varying demand on supporting systems throughout the forecast period is assessed based on system component characteristics and external factors like weather and system load An example of this capacity forecast over one year is presented in Table C.1, which breaks down the forecast into monthly sub-periods For each sub-period, the expected changes and circumstances are evaluated, with the assumptions for January detailed in Table C.2.
In the capacity planning process, each sub-period generates values for Energy Data Center (EDC) and Energy IT (EIT) based on specific assumptions, as illustrated in Table C.2, which are integral to dPUE reporting The annualized dPUE is then calculated by summing the sub-period values of EDC and EIT.
Where the forecast period exceeds one year, multiple annual dPUE values may be reported
Table C.2 — Example of context description
No Sub- period What Change/external fluctuations
1 January IT equipment Start-up of data centre IT load 50 kW
Cooling ventilation/ humidification Data centre is situated in the northern hemisphere latitude 40 N and uses free cooling
Power distribution installation With a low load the mainly i 2 related power distribution losses are low
UPS The UPS system is on load, efficiency about
90 % Lighting and remaining supporting equipment Constant consumption only varying with the number of days per month
Interpretation of PUE and its derivatives
General
The PUE nomenclature, along with transparent public reporting guidelines and accessible key information about reported results, significantly boosts the credibility and utility of the PUE metric in line with European Standards.
This annex provides guidelines and consideration points for correctly interpreting PUE results
Individuals making claims should be aware of the following issues and ensure they are reporting and interpreting valid numbers prior to making any public claims
Data centres have different: a) characteristics, capabilities, and operational policies (e.g government regulations and policy, climate, location and customer’s requirements), b) primary applications such as:
1) main usage: testing, manufacturing, internal processes, networking, scientific modelling or calculations, database management, communications, etc.,
2) primary business supported by the data centre: financial services, healthcare, telecommunications, research and development, environmental monitoring, industrial manufacturing, etc.,
3) criticality of service: emergency services, civic infrastructure, health and safety, security, and similar,
4) availability objectives: disaster recovery, periodic loss of service, resource backup requirements, auxiliary resource requirements, and similar (see EN 50600-4-1:2016, Annex A); c) capabilities with respect to collecting and analysing energy consumption data.
The performance of a data center is influenced by various factors that must be considered when interpreting any Power Usage Effectiveness (PUE) value Without context regarding the reported results, comparisons of data collected by different organizations, employing diverse methodologies over varying timeframes, can lead to misleading or meaningless interpretations.
PUE, as defined by the European Standard, should primarily be utilized to track trends within a single facility over time and to evaluate the impact of various design and operational choices Direct comparisons of PUE values across different data centers are not recommended, in accordance with the guidance provided in sections D.2 and D.3.
Data centre infrastructure versus IT equipment
In data centres, loads are categorized as IT loads, infrastructure loads, or excluded from calculations Many data centres operate within mixed-use buildings that house significant office spaces or unrelated loads These buildings may share systems like cooling towers and ventilation, necessitating clear PUE reporting on how these loads are factored into calculations Ultimately, for enhancing a specific data centre, the focus should be on maintaining consistency in the PUE calculation rather than the precise allocation of shared loads.
A lower Power Usage Effectiveness (PUE) indicates decreased energy overhead for IT equipment, yet it does not offer insights into the operational efficiency or productivity of that equipment Additionally, modifications in the deployment or operation of IT systems can significantly influence PUE outcomes.
Organizations that adopt virtualization in their data centers may experience a decrease in overall IT load, yet this can lead to an increase in Power Usage Effectiveness (PUE) due to unchanged fixed overhead costs for power distribution and cooling It is essential for PUE users to document the factors contributing to this increase, as they may reveal further opportunities for improvement In contrast, older data centers, which house legacy servers lacking energy-saving technologies, can sometimes achieve better PUE results compared to newer facilities This is because modern data centers utilize "energy proportional" servers that exhibit significant fluctuations in energy consumption based on IT load.
Changes in Power Usage Effectiveness (PUE) are significant as they reflect the data center's adaptation to modifications in infrastructure equipment or operations Research examining the impact of these changes on PUE must accurately consider any variations in IT load throughout the study period.
Comparing PUE results between data centres
As highlighted in D.1, PUE values of different data centres should not be compared directly
However, data centres with similar conditions can learn from the changes made to another data centre provided the measurement guidelines, reporting guidelines, and the additional data attributes are obtained
For a fair comparison of Power Usage Effectiveness (PUE) across data centers, it is essential to consider various factors, including the age, geographic location, capacity loading, resiliency, service availability, facility size, and other load characteristics, as outlined in EN 50600-4-1:2016, Annex A.
In such cases, PUE can be used to improve data centre infrastructure efficiency and provide insight for similar data centres
EN 50600-2-1, Information technology — Data centre facilities and infrastructures — Part 2-1: Building construction
EN 50600-2-2:2014, Information technology — Data centre facilities and infrastructures — Part 2-2: Power distribution
EN 50600-2-3, Information technology — Data centre facilities and infrastructures — Part 2-3: Environmental control
EN 50600-2-4, Information technology — Data centre facilities and infrastructures — Part 2-4:
EN 50600-2-5, Information technology — Data centre facilities and infrastructures — Part 2-5: Security systems
EN 50600-3-1, Information technology — Data centre facilities and infrastructures — Part 3-1: Management and operational information
EN 50600-4-3, Information technology — Data centre facilities and infrastructures — Part 4-3: Renewable
CLC/TR 50600-99-1, Information technology — Data centre facilities and infrastructures — Part 99-1:
Recommended practices for energy management
EN ISO/IEC 13273-2, Energy efficiency and renewable energy sources — Common international terminology
— Part 2: Renewable energy sources (ISO/IEC 13273-2)
ISO 26382, Cogeneration systems — Technical declarations for planning, evaluation and procurement