Skip to main content
InvarzTech
Energy & Sustainability

Understanding PUE and Energy Efficiency in Data Centres

What Power Usage Effectiveness (PUE) and DCiE measure, how to calculate them correctly, common pitfalls, and how continuous monitoring supports energy efficiency.

By the InvarzTech team4 min read
Cooling fans inside server equipment lit in blue

Energy is one of the largest operating costs of a data centre, and an increasingly important part of sustainability reporting. To manage it, organisations need a consistent way to describe how efficiently a facility turns electricity into useful IT work. The most widely used measure is Power Usage Effectiveness (PUE).

What PUE measures

PUE was introduced by The Green Grid and has since been standardised internationally as part of the ISO/IEC 30134 series of data centre key performance indicators. It compares the total energy used by a facility with the energy delivered to its IT equipment.

PUE = total facility energy ÷ IT equipment energy

  • Total facility energy includes everything the facility consumes: IT equipment, cooling, power distribution losses, lighting and other support systems.
  • IT equipment energy is the energy consumed by servers, storage, network equipment and similar devices.

A PUE of 1.0 would mean every unit of energy reaches the IT equipment, with nothing used for cooling or lost in distribution. In practice PUE is always above 1.0. The lower the value, the smaller the overhead.

A worked example

Consider a hypothetical facility over one year:

  • Total facility energy: 3,700 MWh
  • IT equipment energy: 2,600 MWh

PUE = 3,700 ÷ 2,600 ≈ 1.42. For every unit of energy used by IT equipment, roughly another 0.42 units are used by cooling, distribution losses and support systems.

DCiE: the same idea, inverted

Data Centre infrastructure Efficiency (DCiE) expresses the same relationship as a percentage:

DCiE = (IT equipment energy ÷ total facility energy) × 100

In the example above, DCiE = 2,600 ÷ 3,700 × 100 ≈ 70.3%. DCiE is simply the reciprocal of PUE, shown as a percentage. Some teams find it more intuitive because a higher number is better.

Measuring PUE well

PUE looks simple, but small decisions about measurement make a big difference.

Use energy over time, not a single power reading

PUE should be calculated from energy (kWh) over a defined period, not from an instantaneous power (kW) snapshot. Cooling demand changes with weather, load and time of day, so a single reading can be misleading. An annual figure captures seasonal variation; monthly figures are useful for tracking trends.

Be clear about where IT energy is measured

IT energy can be metered at different points: the UPS output, the PDU output or the IT equipment itself. Measuring closer to the IT equipment excludes more distribution losses from the IT figure and generally gives a more accurate result. The Green Grid's guidance describes measurement levels based on these points. Whatever level you use, state it, and keep it consistent.

Include everything in total facility energy

Leaving out systems such as lighting, office space within the facility or supporting plant makes PUE look better than it is. Define the boundary carefully.

What PUE does not tell you

PUE is valuable, but it has limits that are worth understanding.

  • It measures overhead, not total consumption. A facility can improve PUE while total energy use rises.
  • It does not measure IT efficiency. Underused servers waste energy, but that waste counts as IT energy and does not worsen PUE.
  • It is affected by climate and load. Comparing PUE between facilities in different climates, or at very different utilisation levels, should be done with care.
  • It is not a carbon metric. Emissions depend on the energy source. Frameworks such as the GHG Protocol treat purchased electricity as Scope 2 emissions, which need separate calculation.

For these reasons, PUE works best alongside other indicators: total energy, IT utilisation, cooling performance and carbon intensity.

Practical ways to improve efficiency

Monitoring identifies where energy is going. Common improvement areas include:

  • Airflow management. Blanking panels, sealed cable openings and aisle containment reduce mixing of hot and cold air.
  • Cooling set points. Operating within recommended inlet temperature ranges, such as those described in ASHRAE TC 9.9 guidance, can reduce cooling energy where equipment allows.
  • Right-sizing. Matching cooling and power capacity more closely to actual load reduces losses.
  • Decommissioning idle equipment. Removing unused servers lowers IT energy directly.
  • Continuous review. Tracking monthly trends shows whether changes are having the intended effect.

How software helps

Accurate PUE depends on consistent metering and reliable data collection. An infrastructure platform can collect facility and IT energy data continuously, calculate PUE and DCiE using a defined method, and present trends on energy dashboards. Organisations running structured energy management programmes, such as those aligned with ISO 50001, can use this data to support measurement and review.

Software supports measurement, data management and reporting. Certification and compliance depend on the organisation's wider processes and independent assessment.

InvarzTech Infrastructure Analytics and DCIM help teams track consumption, PUE, DCiE and equipment performance. Speak to our team about your energy data.

References and further reading

Applying this to your infrastructure?

Talk to the InvarzTech team about your environment.

Talk to Our Team
All articles