CASE74 · Data centers · IPOS / Multi-site performance platform
Designing a multi-site data-center platform for continuous PUE optimization
A 2017 IPOS proposal connects metering, common KPI definitions, equipment diagnosis and measurement and verification. Its numerical targets remain proposal-stage expectations.
01 · CASE ANALYSIS
Move from one diagnosis to ongoing management
The proposed platform combines total electricity, IT electricity, cooling electricity and cooling-unit, chiller and tower data across multiple data centers.
Changing IT loads, weather and equipment states mean that a previously identified operating point cannot be assumed to remain optimal.
02 · CASE ANALYSIS
Keep targets separate from results
The first 24-month target was PUE 1.77 → 1.70, with expected annual electricity savings of 6.5 GWh and expected cost savings of approximately KRW 753 million.
These were pre-implementation project targets. The source does not confirm achievement of those quantities.
| Item | Source value |
|---|---|
| First-stage period | 24 months |
| PUE target | 1.77 → 1.70 |
| Expected electricity saving | 6.5 GWh/year |
| Expected cost saving | Approximately KRW 753 million/year |
03 · CASE ANALYSIS
Define the platform layers
Use metering/BMS acquisition, common data definitions and quality checks, KPI/benchmarking, diagnosis/modeling, and optimization/M&V as separately testable layers.
Compare peers with consistent meter boundaries and aggregation periods. Normalize for IT load, weather, cooling architecture and redundancy before ranking sites.
04 · CASE ANALYSIS
Extend automation in stages
Start with monitoring, diagnosis and operator-reviewed recommendations; expand to supervised control only when data quality and service constraints are demonstrated.
Expected-performance models and digital twins are technical extensions. Keep operator override, equipment interlocks and fallback sequences independent of the optimizer.
| Layer | Function |
|---|---|
| Meter / BMS | Acquire timestamped energy and operating states |
| Data quality | Boundary, completeness and meter version |
| KPI / diagnosis | PUE, cooling efficiency and equipment drill-down |
| Optimization / M&V | Review actions and verify adjusted-baseline results |
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