How Natural Cooling Can Cut Data Center Energy Costs by Over 20%

National Policy Overview

In March, China’s Ministry of Industry and Information Technology, the National Administration of Government Affairs, and the National Energy Administration issued a notice on the pilot work plan for national green data centers, confirming 14 pilot regions across six sectors and 103 projects.

The plan aims to create a hundred green data‑center pilots by 2017, improve average PUE by more than 8%, issue four national standards, promote 40 advanced technologies and best‑practice management methods, and publish a construction guide.

Energy Efficiency Basics and Significance

The key metric for data‑center greenness is PUE (Power Usage Effectiveness), the ratio of total facility energy to IT load energy; the closer to 1, the better. U.S. data centers average a PUE of 1.9, with leading facilities below 1.2, while most Chinese centers exceed 2.2.

Many Chinese operators focus on construction rather than operation, resulting in high ongoing energy consumption.

Example: a 200‑rack, self‑owned data center (small‑scale) with 3 kW per rack and a PUE of 2.5 would spend 13.14 million RMB annually on electricity. Reducing PUE to 2.0 cuts the cost to 10.51 million RMB, saving over 2.5 million RMB per year.

For larger facilities with 500–2000 racks and higher rack densities (5–15 kW), even modest PUE improvements translate into massive cost reductions.

Energy Distribution in a Typical Data Center

In a typical PUE 2.5 data center, energy consumption breaks down as follows: IT load 40%, cooling system 40%, and other systems (lighting, fire protection, UPS, distribution losses) 20%.

The cooling subsystem, especially water‑cooled equipment, dominates the cooling energy share.

By reducing cooling‑equipment energy from 30% to 10% of total consumption, overall energy use drops from 40% to 20%, achieving a 50% total saving and lowering PUE to about 2.0.

Natural Cooling Principle

Natural cooling uses outdoor air for heat exchange when ambient temperature meets predefined conditions, allowing compressors to shut off (M = 0%).

Case Study: Dalian Data Center

Analysis of two years of Dalian weather data shows that natural cooling can be applied for more than six months annually, delivering over 50% energy savings with 100% availability.

Key reference values:

M (Cooling System Optimal Condition) : M = 0% indicates natural cooling mode; M > 0% is normal mode.

T1 (Optimal Indoor Temperature) : 23 °C ± 1 (per GB50174‑2008).

T2 (Optimal Outdoor Temperature) : 10 °C, the highest outdoor temperature at which M = 0% can be maintained.

By continuously monitoring indoor temperature, outdoor temperature, and cooling‑unit operating mode, operators can schedule optimal maintenance and switching times, ensuring the cooling equipment operates in natural‑cooling mode whenever T2 conditions are met.

Model Definition and Application

Applying the model to 2014 historical data demonstrates more than six months of natural cooling, 100% system availability, and a total energy‑saving ratio exceeding 50%.

Long‑term analysis shows the first occurrence of T2 is shifting later each year, indicating that climate trends (e.g., El Niño) will affect cooling‑system planning and maintenance schedules.

Energy‑saving and emission‑reduction are responsibilities shared by everyone.