DCEP key terms

The ratio of total facility energy consumption to the energy delivered to IT equipment over the same period. A value of 1.0 means all energy reaches IT load; higher values indicate overhead from cooling, losses, and other infrastructure.

The reciprocal of PUE, expressed as a percentage, showing what share of total facility energy actually reaches IT equipment. A PUE of 2.0 corresponds to a DCiE of 50 percent.

A PUE calculated for a defined boundary within the facility, such as a single hall or cooling loop, rather than the whole building. Useful for isolating efficiency of specific subsystems or tenants.

The electrical power drawn by IT equipment such as servers, storage, and network gear, typically measured at the output of the UPS or at rack PDUs. It is the numerator basis for PUE calculations.

A metric that credits facilities for exporting recovered heat or other energy by subtracting reused energy from total consumption before dividing by IT energy. Values below PUE indicate beneficial reuse.

Litres of water consumed per kilowatt-hour of IT energy over a reporting period. Captures evaporative cooling and other water use, complementing energy-focused metrics.

Kilograms of carbon dioxide equivalent emitted per kilowatt-hour of IT energy, combining total facility energy with the carbon intensity of its sources. Lower values indicate cleaner operation.

Design PUE is the efficiency predicted at a specified load and condition during engineering; operational PUE is measured from real metered data at actual load. Operational values vary with utilization and weather.

The ratio of average power demand over a period to the peak demand in that period. Low load factors mean equipment runs lightly loaded, often reducing electrical and cooling efficiency.

The electrical power consumed per rack, usually expressed in kilowatts per rack or per square metre of white space. It drives cooling architecture and airflow design choices.

The practice of controlling how cool supply air reaches equipment intakes and how hot exhaust returns to cooling units, minimizing mixing and bypass. Includes containment, blanking, and sealing.

A rack layout where equipment fronts face each other in cold aisles receiving supply air and backs face each other in hot aisles collecting exhaust. Separation prevents supply and return streams from mixing.

Physical barriers such as panels or doors that enclose either the cold aisles or the hot aisles so supply and exhaust air cannot mix. Improves cooling efficiency and raises usable return air temperature.

Cool supply air that reaches the cooling unit return without passing through IT equipment, for example through open rack spaces or cable cutouts. It wastes cooling energy and lowers return temperatures.

Hot exhaust air that flows back into equipment intakes instead of returning to the cooling units. It raises intake temperatures and can create hot spots even when overall cooling capacity is adequate.

Filler plates installed in unused rack units to block air from passing through empty spaces. They force supply air through operating equipment and reduce bypass and recirculation.

A localized area where equipment intake temperatures exceed acceptable limits because of poor airflow distribution. Hot spots constrain load even when aggregate cooling capacity appears sufficient.

The temperature of cooled air delivered from cooling units to the equipment intakes. Raising it within equipment limits typically improves chiller efficiency and economizer hours.

The temperature of air leaving the IT space and returning to cooling units. A higher value relative to supply indicates effective airflow and enables more efficient cooling operation.

The temperature difference between return and supply air across cooling units or equipment. A small delta T often signals excess airflow, mixing, or bypass rather than a cooling capacity problem.

A cooling mode that uses outside air or indirect heat exchange instead of mechanical refrigeration when outdoor conditions allow. Increases free cooling hours and lowers energy use.

Rejecting data center heat without running compressors, using cold outdoor air, water, or a fluid cooler. Availability depends on climate and the chosen supply temperature.

A refrigeration machine that removes heat from a water loop and rejects it outdoors. Chiller efficiency, expressed as kW per ton or COP, strongly influences facility overhead energy.

The ratio of cooling delivered to energy consumed by a cooling machine at given conditions. Higher values mean more cooling per unit of input power.

Cooling output in British thermal units per hour divided by input power in watts, commonly quoted for cooling units. Higher EER indicates a more efficient device at rated conditions.

Cooling that rejects heat by evaporating water, which can be more energy efficient than compressor-based cooling but consumes water. Drives the water usage metric.

A direct-expansion cooling unit that uses a refrigerant cycle to cool air in the IT space. Typically less efficient at scale than chilled water systems.

A cooling unit that circulates air over a coil supplied with chilled water from a central plant. It has no integral compressor and relies on chillers or free cooling.

A system providing conditioned, battery-backed power to IT loads during outages and transients. Its conversion losses are a significant part of facility overhead energy.

The plot of UPS conversion efficiency against percentage of rated load. Efficiency is usually poor at very low load, so right-sizing and modular capacity improve real-world performance.

The ratio of real power to apparent power in an alternating current circuit, between zero and one. Low power factor increases current for the same useful work and can incur utility charges.

Energy dissipated as heat in transformers through core and winding effects, both at no load and under load. Siting and loading practices affect the size of these losses.

Equipment that distributes electrical power to racks or downstream circuits, often with metering. Rack-level metering enables per-device energy measurement and capacity planning.

Providing one more component than the minimum needed to carry the load, so a single failure does not interrupt service. Redundant capacity often runs lightly loaded and less efficiently.

Capacity installed for redundancy or future growth that sits unused because of cooling or electrical limits. It represents capital and efficiency waste that assessment should identify.

Running multiple workloads on fewer physical servers to raise utilization and reduce idle power draw. Consolidation lowers total IT energy and often improves facility efficiency.

The degree to which a server's power draw falls as its workload falls, rather than staying near idle levels. Better proportionality reduces energy waste at low utilization.

Comparing a facility's measured energy metrics against peers, internal baselines, or targets to identify improvement opportunities. Requires consistent boundaries and measurement periods.

A documented process for quantifying actual energy savings from an efficiency measure against an adjusted baseline. Prevents attributing unrelated changes to the project.