Power Usage Effectiveness (PUE) is the headline efficiency metric for any data centre or server room. It captures how much of the electricity drawn from the grid actually reaches the computing equipment versus how much is lost to cooling, power conversion, and other overhead. This calculator turns your two power readings into PUE, the inverse DCiE percentage, projected annual energy waste, and the carbon footprint of that waste.
How it works
The core formula is a simple ratio, but the derived figures make it actionable:
PUE = total facility power (kW) / IT equipment power (kW)
DCiE = (1 / PUE) × 100% (efficiency as a percentage)
overhead power = total - IT (kW spent on non-IT)
annual facility kWh = total kW × 8760 hours
annual overhead kWh = overhead kW × 8760
annual CO2e (kg) = annual facility kWh × carbon g/kWh / 1000PUE can never be below 1.0, because the IT load is always part of the total. The gap above 1.0 is pure overhead. Multiplying power by 8760 (hours in a year) gives annual energy, and applying grid carbon intensity converts that to CO2e.
Benchmarks and tips
The ASHRAE and Uptime Institute bands are a useful yardstick: under 1.2 is excellent (hyperscale class), 1.2 to 1.5 is good and meets the EU Code of Conduct target, 1.5 to 2.0 is typical for enterprise rooms, and above 2.0 signals significant cooling or conversion inefficiency. Remember that PUE measures overhead only — a facility on a clean grid with a mediocre PUE can still emit far less carbon than an ultra-efficient one running on coal power. Track carbon intensity alongside PUE to see your true footprint.
Where PUE overhead actually goes
A PUE of 1.5 means 33% of total electricity is used for non-IT purposes. Where does that power go?
| Overhead category | Typical share of non-IT overhead |
|---|---|
| Computer room air conditioning (CRAC/CRAH) | 35–50% |
| UPS losses (power conversion) | 10–20% |
| Chiller plant and cooling towers | 15–30% |
| Power distribution losses | 5–10% |
| Lighting, security, other | 2–5% |
Cooling dominates, which is why free-air cooling (using outside air when it is cool enough), liquid cooling on high-density racks, and raising the room temperature setpoint are the main levers available to improve PUE in a conventional data centre. Raising the cold-aisle setpoint from 18°C to 24°C can meaningfully reduce chiller run time and cut PUE noticeably.
Improving PUE in practice
The most impactful actions, roughly in order of effect:
- Eliminate hot/cold aisle bypass. Blanking panels, containment curtains, and raised-floor grommets stop hot air recirculating into intakes and reduce cooling load immediately.
- Raise the supply air temperature. Most modern servers tolerate inlets up to 27°C or higher (ASHRAE A1/A2 ratings). Every degree rise reduces chiller energy.
- Implement economiser / free-air cooling. In temperate climates, outside air is cool enough for a large portion of the year to supplement or replace mechanical cooling.
- Replace aging UPS with high-efficiency models. Older UPS systems operate at 80–88% efficiency; modern systems achieve 96%+ at load.
- Consolidate servers. Underutilised servers draw power nearly as fast as loaded ones. Virtualisation improves IT load relative to total facility power.
Regulatory context
The EU Energy Efficiency Directive (EED) and the European Green Deal are increasingly placing PUE and energy use in the spotlight for data centres. Data centres above 500 kW IT load in the EU are now subject to reporting and efficiency requirements, and the EU Code of Conduct for Data Centres sets voluntary best-practice targets, including PUE. Several national governments are tightening planning requirements for new data centre builds, making efficiency documentation a compliance as well as an operational need.