Protecting the Data Center Is Cool

Protecting the Data Center Is Cool

This piece was originally published in the July 2016 issue of electroindustry.

Peter A. Panfil, Vice President of Global Power, Emerson Network Power

enterprise data center emerson
Today’s data centers require high-performance power, thermal, and infrastructure management systems to meet demands for reliability, efficiency, scalability, and productivity. Illustration courtesy of Emerson Network Power.

Once, availability was the singular concern of data center operators. Today, data centers must deliver availability with efficiency, improved productivity, and enhanced scalability. The infrastructure that supports the complex network of servers, storage, and networking equipment in the data center is key to achieving this deliverable.

Best-in-class power systems typically contain switchgear, transfer switches, transient voltage surge suppression, short-term and long-term backup, and electrical distribution systems. The heart of the power system is the uninterruptible power supply (UPS). Application and business considerations will determine the best location for the UPS system, with the centralized, room-scale alternating current (AC) power strategy still predominant. However, organizations that need the flexibility to add capacity quickly are also exploring distributed protection, in which the UPS is deployed in the aisle or equipment rack.

The AC UPS used in data centers provides power conditioning through a double-conversion process that uses a rectifier to convert utility power to direct current (DC) power and an inverter to create a clean AC waveform suitable for electronics. UPS battery systems provide short-term backup power in the event of a utility outage, with longer-term backup power provided by an onsite generator. Continuous battery monitoring is recommended to ensure that batteries work when needed. Redundancy is typically designed into the power path to enable concurrent maintenance and increase fault tolerance.

Downstream from the UPS, a one- or two-stage distribution strategy may be used. In two-stage distribution, power is distributed through power distribution units (PDUs), which may step down the voltage using a transformer, to remote power panels that feed the equipment racks. The most common consumption voltage is 208/120 volts AC (VAC), with some systems going to 240 VAC, to improve the utilization rate.

As an alternate to an AC UPS system, a DC UPS can be employed. The DC UPS system rectifies the incoming AC source to DC and uses batteries for short-term backup power. The most common DC voltage today is 48 volts (DC) VDC, with 380 VDC emerging, to improve the utilization rate.

Cooling Off

To remove the heat generated by dense clusters of electronics, best-in-class facilities employ a thermal management strategy that optimizes protection of the IT equipment and efficiency. Cooling systems with economization are increasingly replacing traditional mechanical cooling, with the use of outside air and water evaporation systems providing the desired supply temperature. These technologies can improve cooling-system efficiency by up to 50 percent.

Perhaps the greatest changes in thermal management are in the use of intelligent thermal controls, which enable machine-to-machine communication so that thermal units across a facility can work as a team to increase efficiency. They also automate cooling system operational routines, such as temperature and airflow management, valve auto-tuning, lead/lag, and others that enhance overall system operation.

Real-time visibility into operating conditions—and interdependencies between systems—is provided by real-time data center infrastructure management (DCIM). The DCIM system uses management gateways and system controllers to collect and filter real-time data and provide data center managers with the centralized visibility required to plan for the future and optimize equipment use, efficiency, and availability. DCIM, along with the emerging Redfish specification (an open industry standard developed by the Distributed Management Task Force) for device communication, will enable the data center to be managed as part of an enterprise-wide Internet of Things.

Today’s data centers are critical to the businesses and users they support. A reliable, efficient infrastructure, including best-in-class power, thermal, and infrastructure management, is essential to their operation.


With more than 30 years of experience in embedded controls and power, Mr. Panfil leads global power sales for Emerson Network Power’s Liebert AC Power business.

Read this month’s issue of electroindustry.

One thought on “Protecting the Data Center Is Cool

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.