CASE47 · Data centers · IPOS hierarchical diagnostics
Turning PUE into a real-time diagnostic and action hierarchy
An IPOS proposal integrates center, floor and cooling-equipment data. PUE is decomposed into cooling electricity per IT load and equipment COP, with investigation and action checklists at each level.
01 · CASE ANALYSIS
Why a single PUE ranking is insufficient
The 2016 portfolio used approximately 176,885 MWh total electricity and 97,429 MWh IT electricity, with average PUE near 1.82. Estimated 2017 PUE was 1.77.
Operators need to know what to inspect when PUE exceeds a target. Seasonal center differences identify questions rather than proving causes.
02 · CASE ANALYSIS
Three-level KPI hierarchy
PUE = total facility electricity / IT electricity. Cooling index = cooling electricity / IT electricity. Cooling-equipment COP = cooling delivered / equipment electricity.
Lower PUE and cooling index are favorable; higher COP is favorable. The hierarchy distinguishes operating/overcooling problems from equipment-performance problems.
| KPI | Definition | Investigation |
|---|---|---|
| PUE | Total facility / IT electricity | Separate IT-load and overhead effects |
| Cooling index | Cooling / IT electricity | Overcooling, setpoints, sensors and staging |
| Cooling COP | Cooling delivered / cooling-equipment electricity | Refrigerant, pressure, water and heat transfer |
03 · CASE ANALYSIS
Action guide and data integration
Investigate overcooling, room versus setpoint temperature, sensor location/calibration and active-unit count when the cooling index deteriorates. If equipment COP deteriorates, investigate refrigerant, pressure, water conditions and heat exchange.
Existing center servers connect DB-to-DB to the IPOS server. Supplement missing power/temperature points, calibrate sensors and retain specifications, control logic and action history.
04 · CASE ANALYSIS
Staged target quantities
The first 24-month target was PUE 1.70, with approximately 6.5 GWh/year and KRW 753 million/year expected benefit, corresponding to 8.81% cooling-energy reduction.
The longer-term target was PUE 1.65, with approximately 11.3 GWh/year and KRW 1.310 billion/year expected benefit, corresponding to 15.3% cooling-energy reduction. These are alternative staged targets, not additive savings.
| Stage | PUE | Expected electricity benefit | Expected annual cost benefit |
|---|---|---|---|
| Estimated 2017 reference | 1.77 | — | — |
| First stage: 24 months | 1.70 | 6.5 GWh/year | KRW 753 million |
| Longer term | 1.65 | 11.3 GWh/year | KRW 1.310 billion |
05 · CASE ANALYSIS
Capacity and verification boundaries
The proposal described approximately 594 cooling units, 14,890 RT/h total capacity and 3,154 RT/h mean load, around 21%. Source unit notation is retained; installed reserve is not automatically waste.
Maintain redundancy and thermal reliability. Decide the savings measurement method before executing actions and adjust for IT load, weather and operating conditions.
APPLY TO YOUR SITE
Connect your operating data to an actionable energy strategy.
Discuss the measurement boundary, data quality, operating constraints and verification approach for your facility.
Discuss your project All English cases