June 20, 2018

Data Centres are integral to our daily lives and support virtually every aspect of our modern existence.

Computer performance continues to increase significantly as does the demand on cooling and power supply infrastructure. Data Centres have been identified as significant users of energy and responsible for the associated Carbon emissions. Many data centres use up to around 45% of their energy consumption on support infrastructure services with cooling and power supply system’s the majority users.

Also as Data Centre capacity is increased it is often the performance and capability of the cooling and power infrastructure that is a limiting factor.
Improving the efficiency of Data Centre infrastructure can result in a range of benefits:

• Reduced energy usage costs
• Reduced Greenhouse Gas Emissions and improved NABERS Data Centre and LEED ratings
• Increased cooling and power supply capacity for expansion
• Increased systems operational resilience and reliability

There are a number of best practice approaches that should be considered for improving the efficiency of existing facilities:

Set Point Temperatures – The American Society of Heating Refrigeration and Air-Conditioning Engineers recommend that modern IT equipment (Class A1) can operate reliably at higher temperatures and have broadened their guideline IT operating temperature range out to 15°C to 32°C.

Airflow Management – The efficiency and effectiveness of data centre cooling systems is heavily influenced by the temperature and quantity of air delivered to the IT equipment. There are a number of options to improve the effectiveness of air flow;

• Configure racks in hot/cold aisle arrangements
• Install blanking plates on racks
• Install containment systems or curtains
• Seal floor leakage paths
• Balance floor grilles

Rack Layout – Poorly configured layouts can contribute to inefficient air flow and cooling air distribution and result in cooling units working harder. Aisle containment systems, rack chimney systems and floor pressure management can provide increases in cooling efficiencies and better utilisation of existing equipment.

Server Upgrades – New generation servers use less energy than traditional standalone systems. They utilise variable speed cooling fans, have power saving modes and are able to operate at higher temperatures; all contributing to lower infrastructure energy usage.

Consolidation and virtualisation of servers – The resulting rationalisation of devices can reduce the IT footprint; improve server and storage device utilisation and lower overall energy consumption.

Free Cooling Systems – Improved cooling solutions can incorporate direct or indirect free cooling solutions to leverage cooler outside temperatures to help cool the Data Centre. Higher operating temperature tolerances for IT equipment also facilitates the increased use of free cooling.

Alternate architectures – Shifting the cooling source closer to the heat source with alternate cooling solutions such as In-row coolers or IT cabinet rear door heat exchangers can reduce energy consumption associated with cooling.

Efficiency of Power Supply – Upgrading Uninterruptable Power Supply systems can result in reduced energy usage. Efficiencies of power supplies generally decrease with age, and modern units are more energy efficient.

Lighting Systems – Modern LED lighting technologies and control systems are very efficient, long lasting and reliable, and can result in useful energy usage improvements and maintenance cost reductions.

Computer Room Air Conditioning Units and Chillers – The latest data centre cooling plant is designed for better efficiency, and operating parameters have shifted for better performance. Chilled water flow temperatures are being increased for better chiller performance, and better integration with free cooling systems. Indirect evaporative cooling and free cooling CRAC (DX) units are being incorporated to achieve lower PUE.

In a recent edition of Facility Perspectives magazine Claire Gaynor talked to the A.G. Coombs team of experts responsible for designing, constructing and the facility management of some of Australia’s largest and most critical Data Centres. See the article here.