Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
IEC 62018: Power Consumption of Information Technology Equipment — Measurement Methods and Energy Efficiency Guidelines
In an era where data centers consume approximately 1–2% of global electricity and ICT infrastructure continues to expand exponentially, the need for standardized power consumption measurement has never been more critical. IEC 62018 provides a comprehensive framework for measuring, reporting, and optimizing the power consumption of information technology equipment (ITE). This standard serves as a foundational reference for engineers designing energy-efficient ICT systems, enabling consistent cross-vendor comparisons and supporting global sustainability initiatives.
Tip IEC 62018 focuses specifically on the measurement methodology for IT equipment power consumption — it does not prescribe absolute energy limits but rather defines how to measure so that reported values are reproducible and comparable.
Scope and Core Principles
IEC 62018 applies to all information technology equipment connected to mains power, including servers, storage systems, networking devices, and peripherals. The standard addresses both AC and DC power input configurations and covers operational states from idle to full load. Its primary objective is to establish a uniform testing framework that eliminates the measurement variability that previously plagued energy efficiency comparisons.
The standard defines several key operating states that must be characterized during testing: Off (standby with minimum draw), Sleep (low-power state with preserved context), Idle (powered on but not processing workload), and Active (processing typical workloads). Each state requires specific measurement durations and sampling rates to ensure statistical validity.
| Operating State | Description | Typical Power (% of Rated) | Measurement Duration |
|---|---|---|---|
| Off | Connected to mains, equipment not functional | 0.5–2% | 5 minutes |
| Sleep | Low-power state, quick resume capability | 5–15% | 10 minutes |
| Idle | OS loaded, no application workload | 30–50% | 15 minutes |
| Active | Processing benchmark workloads | 60–100% | 30 minutes or more |
Measurement Methodology and Test Setup
IEC 62018 mandates strict control of environmental conditions during power measurement testing. Ambient temperature must be maintained at 23 °C ± 2 °C, relative humidity between 25% and 75%, and atmospheric pressure within 86–106 kPa. The power supply voltage must be held at the nominal rated value within ±1%, and frequency within ±0.5%. These constraints ensure that measured power variations reflect equipment behavior rather than environmental drift.
The standard specifies that power measurements must be conducted using a wattmeter with an accuracy class of at least 0.5 (per IEC 61557-2) for active power measurements. For equipment with power factor below 0.5, a wide-bandwidth power analyzer is recommended to accurately capture harmonic content and crest factors. The measurement uncertainty budget must be documented, with combined uncertainty below 2% for all reported values.
Warning Many engineers overlook the impact of power supply harmonics on measurement accuracy. IEC 62018 requires that the measurement instrument have a bandwidth of at least 3 kHz to capture significant harmonic content up to the 50th order (at 50/60 Hz fundamentals).
A critical aspect of the methodology is the stabilization period. Before any measurement begins, the equipment must reach thermal equilibrium, defined as less than 1% change in power consumption over a 10-minute interval. For high-power density equipment such as blade servers, this stabilization period can extend to 60–90 minutes due to thermal mass and fan speed control hysteresis.
Engineering Design Insights for Energy Optimization
Understanding IEC 62018 measurement principles enables engineers to implement targeted energy optimization strategies. The standard’s structured approach to measuring power across operating states reveals that the most significant energy-saving opportunities often lie in the idle-to-active transition efficiency rather than peak load performance.
Design techniques that improve measured efficiency include: adaptive voltage scaling (AVS) that reduces core voltage during light loads, dynamic frequency scaling coordinated with workload prediction algorithms, and high-efficiency power supply units (PSUs) achieving 80 PLUS Titanium certification (≥96% efficiency at 50% load). Power gating at the subsystem level — selectively disabling unused memory ranks, I/O controllers, and accelerator blocks — can reduce idle power by 40–60% without compromising responsiveness.
For system architects, the standard highlights the importance of measuring power at the point of common coupling (PCC) rather than relying on component-level estimates. In practice, this means instrumenting the AC input to the PSU and accounting for PSU losses, fan power, and auxiliary system overhead that component datasheets often omit.
| Optimization Technique | Typical Saving | Implementation Complexity | Applicable States |
|---|---|---|---|
| Adaptive Voltage Scaling | 15–30% | High (HW/SW co-design) | Idle, Light Load |
| Subsystem Power Gating | 40–60% | Medium (RTL design) | Idle, Sleep |
| High-Efficiency PSU | 8–12% | Low (component selection) | All states |
| Fan Speed Optimization | 5–15% | Low (firmware tuning) | Idle, Active |
Frequently Asked Questions
Q1: How does IEC 62018 differ from ENERGY STAR requirements for IT equipment?
IEC 62018 defines the measurement methodology for power consumption, while ENERGY STAR sets specific efficiency thresholds that equipment must meet for certification. The two standards are complementary — IEC 62018 provides the testing framework, and ENERGY STAR references these methods for compliance verification.
Q2: Does IEC 62018 apply to DC-powered equipment in telecommunications environments?
Yes, the standard includes provisions for DC input power measurement (typically -48 VDC for telecom applications). The methodology adapts the same principles, with specific guidance on DC current sensing using shunt resistors or Hall-effect sensors and accounting for DC distribution losses.
Q3: What is the recommended sampling rate for accurate power measurements under dynamic workloads?
For quasi-static states (Off, Sleep, Idle), a sampling interval of 1 second is sufficient. For Active state with dynamic workloads, the standard recommends sampling at 10 Hz (100 ms intervals) to capture transient peaks that affect average power calculations. Some server workloads with burst characteristics may require 100 Hz sampling.
Q4: How should power data be reported to ensure compliance with IEC 62018?
The standard specifies a reporting template that includes: equipment identification, test conditions (temperature, humidity, voltage), instrumentation details (model, calibration date, accuracy class), measured power values for each operating state with confidence intervals, power factor, total harmonic distortion, and measurement uncertainty budget. Reports must clearly distinguish between average, peak, and RMS values.
