You are here

More Efficient Data Centers: Maximizing Airside Cooling

Modular data center

Around the world, unprecedented and exceedingly large volumes digital data are stored in data centers. The data come from online transactions, social networking websites, banks, healthcare facilities, schools, government bodies, and industry, to name a few. It is essential that these data are reliably processed by and securely transmitted to a variety of user/customers. Within the last twenty years, the amount of digital data being generated has greatly increased resulting in ever increasing numbers of data centers.

Data centers house many servers with typically multiple thousands of IT equipment units. The centers require a considerable amount of cooling infrastructure to enable IT equipment to function properly. One way to reduce energy consumption of data centers is to use air-side economization (ASE) and indirect and direct evaporative cooling (I/DEC). These cooling methods do not use compressors, which is good because compressors consume large amount of energy to convert vaporized cooling fluid to liquid cooling fluid. Although not all data centers can be cooled 100% of the time using ASE and I/DEC methods, maximizing the use of ASE and I/DEC (minimizing compressor-based cooling system) can result in significant reductions in energy consumption of the cooling infrastructure. The annual number of hours ASE and I/DEC will be used in a year is expected to increase since IT equipment manufacturers have begun to make IT equipment that can be operated at wider temperature and humidity ranges than was possible few years ago.

Research on ASE and I/DEC for data center cooling has been conducted on a testbed modular data center. Although these methods of cooling are not new to data centers, maximizing their use over compressor based cooling technologies has the advantage of reducing overall cooling costs and carbon footprints. In the specific I/DEC unit that has been studied, a cooling tower is used to cool water that runs through a water-to-air heat exchanger (indirect evaporative cooling). For this system, the cooling tower can be placed at a distance from the main cooling unit. This allows greater ease of maintenance and provides freedom to place the unit at a suitable place in the data center.

Economic Impact:

Electricity consumption by data centers is expected to increase from 91 billion kWh in 2013 to 140 billion kWh by 2020. Reliable operation in data centers of IT equipment requires that the heat generated needs to be continuously and effectively removed; otherwise, the IT equipment may fail to operate resulting in data center downtime. A recent study indicated that about 30% of total data center energy is consumed by the cooling infrastructure of data centers. Reducing energy consumption of the cooling infrastructure of data centers by few percentage points will translate to many thousands of dollars of savings per year in savings for data center owners and significant reductions in carbon footprints of the data centers.

This breakthrough cooling approach avoids the use of compressors or chillers. The approach s 70% less energy than traditional air conditioning systems. Depending on location and environmental conditions, water usage can also be significantly reduced in these systems. This is because water is recirculated and used several times. Continuous R&D testing has helped replace conventional propeller type fans with VFD to a more efficient EC axial fan in the cooling tower. Data centers that implement these cooling technologies will benefit from the electric cost saving. Overall, the cost of purchasing, installing and servicing an EC fan/motor assembly can save significant dollars as opposed to conventional fan/motor selections tied to a VFD.

The nation will benefit from more sustainable growth of energy efficient data centers that maintain smaller carbon footprints that conventional cooling systems.

For more information, contact Dereje Agonafer at the University of Texas, Arlington,, Bio, 817-272-7377.

PDF icon ES2-2016.pdf