Data Center Energy Efficiency Assessment
What this domain tests
An efficiency assessment starts with a boundary decision. Before any number is calculated, you must know what is inside the fence: IT load only, or IT plus cooling, UPS losses, lighting, and office space. Most assessment questions hinge on whether the candidate keeps the IT and facility boundaries straight.
Core concepts to know:
- Energy audit vs. assessment: an audit reconciles measured consumption against utility bills; an assessment identifies savings opportunities and typically relies on spot measurements and trend data.
- End-use breakdown: IT equipment, cooling (chillers, CRAHs, pumps, fans), UPS/PDU losses, lighting, and miscellaneous loads. Cooling is usually the second-largest end use and the largest non-IT opportunity.
- Measurement tools: power meters, clamp-on CTs, portable data loggers, ultrasonic flow meters, and building management system (BMS) trend exports. Instrument accuracy and logging interval matter.
- Load profiling: average vs. peak vs. part-load behavior. Equipment sized for peak often runs inefficiently at low load.
- Baselining: normalizing for IT load and weather so savings can be attributed to a measure rather than to a change in business activity.
Traps to watch:
- Confusing nameplate rating with actual measured draw; nameplate overstates real load and inflates savings estimates.
- Ignoring part-load performance curves for chillers and fans.
- Comparing sites with different redundancy levels (N vs. 2N) as if they were equivalent.
- Treating a single spot reading as representative of annual energy.
Worked example
A facility measures 500 kW of IT load and a total facility draw of 800 kW. Facility overhead is 800 − 500 = 300 kW, and PUE = 800/500 = 1.60. If a cooling optimization cuts overhead by 60 kW, the new PUE is 740/500 = 1.48 - a 7.5% reduction in total energy for the same work delivered.