As data center costs rise, corporate directives to find opportunities for savings are forcing IT managers to find ways of increasing efficiencies without compromising IT capabilities or data center performance. Here are four suggestions for helping companies move toward a high-efficiency infrastructure without compromising availability or scalability.

1. Adapting Cooling to Changes in IT Capacity

As today’s technologies bring higher densities, capacities and availability requirements to data centers, applying yesterday’s cooling strategies becomes prohibitively expensive and inefficient. Historically, to meet changing capacity demands, infrastructure systems have been sized to handle peak or future planned loads, which is an inefficient use of support systems’ energy consumption if they’re running at peak capacity at all times.

A more efficient and flexible approach is to install infrastructure systems that can vary their capacity to the amount of work required, allowing equipment to operate at reduced capacity without paying an efficiency penalty. One example is precision cooling systems that feature variable speed drives or EC plug fans, which improve efficiency by controlling fan speed. A 20 percent reduction in fan speed can result in an energy savings of greater than 50 percent.

2. Moving to a Higher Density

Increasing the density inside the data center is an effective solution for improving efficiency and increasing computing capacity. There are a number of high-density cooling and power distribution technologies that allow data center densities to be increased from the traditional 2 kW to 6 kW per rack to more than 30 kW per rack. This enables data center designers and managers to maximize performance while reducing energy costs and freeing up floor space.

A relatively new approach to cooling higher density environments is the idea of moving cooling closer to the source of heat. Being close to the source of the heat allows for higher cooling system efficiencies and a reduction in the energy used to move air throughout the data center. These cooling technologies can reduce cooling system operating costs by more than 30 percent.

Another cooling option for higher-density environments is cold aisle containment, which uses physical barriers to block off the cold aisle within the data center to most effectively optimize the capacity/efficiency of the cooling system while ensuring air temperature to the servers. Cold aisle containment, which can improve the energy efficiency of existing data centers by more than 30 percent, is often a more practical alternative than hot aisle containment.

Langan Engineering & Environmental Services upgraded its data center in Elmwood Park, N.J., with an uninterruptible power supply (UPS) system that would support efforts to reduce data center energy consumption and provide a clear path for expansion as blade servers and virtualization were introduced. Photo courtesy of Liebert Products, Emerson Network Power.

3. Maintaining High Availability with Efficient Power Systems

In the world of always-on critical systems, availability is a top priority. However, treating the data center as a single space that continuously needs as much power as it can get throughout the day is not efficient. Rather, think of the data center as a dynamic collection of power consumers with varying levels of power needed in different areas and at different times.

There are a number of new power systems on the market that can play an important role in efficiently maintaining high availability. These new uninterruptable power supply (UPS) system technologies are modular and scalable to meet demand and offer flatter load curves to achieve operating efficiencies greater than 93 percent.

4. Taking Advantage of Monitoring Technologies

Determining how much power is being used by any piece of equipment in the data center should not involve guesswork. The more precisely areas and components in the data center can be measured for consumption, the more equipped IT managers are to make real and substantial improvements in the data center’s operating efficiency — and be able to justify from an ROI perspective any incremental funds necessary to implement those improvements.

Branch circuit and rack power distribution unit technologies allow data centers to collect energy consumption information at the branch, rack and server receptacle level. This data can then be analyzed using real-time monitoring software packages to measure overall energy efficiency and optimize the performance of the IT environment. Conclusion

By improving the overall performance of data centers without compromising efficiencies, companies of all sizes can reduce the impact of soaring energy costs on their own bottom lines. Taken together on a global scale, data centers of all sizes — by improving their operating efficiencies — can begin to restrain their growing environmental footprint.