CONTEXT AND ISSUES

Data centers consume a great deal of energy: 259 TWh in the EU in 2020, i.e. 1.7% of world energy consumption (source: European Commission).

Energy-intensive buildings

For example:

  • a typical data center consumes 10 to 100 times more energy per m² than a standard office building.
  • a data center covering 10,000 m² consumes the same amount as a town with 50,000 inhabitants.
  • over 10 years, the operating cost of a data center is the same as its installation cost.
  • the electricity bill represents 10 to 15% of the operating cost.

The building infrastructure currently represents almost half the total energy consumption.

 

it_equipment

 

An increasingly large ecological footprint

The environmental footprint of data centers continues to increase: over the 10 coming years, it is estimated that there will be 30 times more data (including 90% unstructured data) and 1000 times more servers. At this pace, energy needs could double within five years. It is therefore essential to reduce the carbon footprint of data centers and improve their energy efficiency, in order to reduce consumption and costs.

LEGRAND GROUP'S RESPONSE

Objective: Reduce the PUE

The PUE (Power Usage Effectiveness) is an indicator for measuring the energy efficiency of a data center by working out the ratio of the total consumption of the data center to that of the computer and telecoms (IT) equipment.

The ISO/IEC 30134-2:2016 “Information technology -- Data centres -- Key performance indicators -- Part 2: Power usage effectiveness (PUE)” standard defines several PUE categories:

  • Basic PUE (Category 1): This measurement stipulates the conversion of all measurements into kilowatt-hour (kWh). It is a precise method including energy sources other than mains electricity. PUE1 is calculated over a 12-month period.
  • Intermediate PUE (Category 2): This measurement includes the category 1 requirements. However the IT consumption is measured at the level of each PDU (Power Distribution Unit). A clear distinction is therefore made between the infrastructure and the IT equipment and it is easier to measure a pPUE (partial PUE).
  • Advanced PUE (Category 3): This measurement includes the category 2 requirements. It refines them by requiring the IT consumption to be measured at device level.

A data center with optimum efficiency will be PUE 1, whereas the average global PUE of a data center is between 1.8 and 1.89 (source: Uptime Institute survey 2012). Reducing this is therefore a priority in order to ensure that the infrastructure provides ever-higher performance.

 

To reduce the PUE, there are 3 possible actions:

+ Worth noting

ISO/IEC additional indicators

ISO/IEC standardised additional indicators to refine the assessment of the ecological footprint of a data center:

  • the Renewable Energy Factor was standardised as the ISO/IEC 30134-3:2016 Information technology -- Data centres -- Key performance indicators -- Part 3: Renewable energy factor (REF)
  • The Energy Reuse Factor (ERF): this measure of the amount of energy reused outside the data center is in the process of being standardised
  • The Carbon Usage Effectiveness (CUE): this extrapolates a greenhouse gas emission volume based on the electricity consumption of the data center
  • The Water Usage Effectiveness (WUE): it measures the amount of water used in the data center.

These last two KPI are on the list for future standardization.