advanced search

What is Power Usage Effectiveness?

Dec 1, 2008 12:00 PM, By Mark Szalkus, GE Digital Energy - Power Quality

A quick lesson on analyzing the energy efficiency level of a data center

A measure of a facility's total power delivered divided by its IT equipment power usage level, power usage effectiveness (PUE) is especially important in data centers. For starters, these energy-intensive buildings are growing at a rapid pace, consuming 10 to 100 times more energy per square foot than a typical office building. In 2006, for example, data centers used more than 60 billion kilowatt-hours, a number that some industry experts expect will increase to 100 billion kilowatt-hours by 2011.

Total facility power is defined as the power measured at the utility meter (i.e., the power dedicated solely to the data center). This is an important factor in “mixed-use” buildings that house data centers, as this is only one portion of the overall load of the facility. If the data center is not wired on a separate utility meter, estimate the amount of power being consumed by the non-data center portion of the building and subtract this value from the equation (PUE = total facility power 4 IT equipment power). Following is a list of total facility power-related components (i.e., typically power distribution-related equipment):

• UPS modules

• Switchgear

• Generators

• Power distribution units (PDUs)

• 

  Fig. 1. Typical components in a data center electrical system.

  Batteries

• Lighting

• Distribution losses external to the IT equipment

Total facility power loads also include cooling system components, such as:

• Chillers

• Computer room air-conditioning (A/C) units (CRACs)

• Direct expansion air handler (DX) units

• Pumps

• Cooling towers

IT equipment power is defined as the devices used to manage, route, store, or process data within a data center. The load components under the IT equipment power heading include:

• Computers (desktop, workstation, and laptop plus servers)

• Storage devices (drives, tape, mass media)

• Network equipment

• Supervisory equipment such as KVM switches and monitors

These power measurements provide a baseline that allows an IT facility manager to compare his or her power usage levels to other data centers. Typical components found in a data center are shown in Fig. 1.

How do you measure IT equipment power? Measure the IT equipment load after power conversion, switching, and conditioning is completed. According to The Green Grid, a global consortium based in Beaverton, Ore., that is dedicated to advancing energy efficiency in data centers and business computing ecosystems, the most useful measurement point is at the output of the computer room PDUs. This measurement should represent the total power delivered to the server racks in the data center.

What is a good PUE rating? The Green Grid has defined a PUE rating of 1.0 equivalent to a 100% efficient facility. A small study by Lawrence Berkeley National Labs, “Best Practices for Data Centers: Lessons Learned from Benchmarking 22 Data Centers,” found that the data centers it studied had a PUE range of 1.3 to 3.0. According to The Uptime Institute, an industry provider of vendor-neutral, research-based information on high-density enterprise computing based in Santa Fe, N.M., the typical data center has an average PUE of 2.5.

What does this ratio truly mean? It means that for every 2.5W of input at the utility meter, only 1W of output is delivered to the connected IT load. The Uptime Institute estimates that most facilities could achieve a PUE rating of 1.6 by using the most efficient-equipment available and following industry best practices.

Is PUE a dynamic value? In reality, PUE is a dynamic number that changes as the load changes in a data center room. How ironic would it be if the best PUE number attainable is when all the servers in the center are running at near capacity, and shutting down servers to save power actually increases the PUE value? Or your energy-efficient cooling system uses large amounts of water in Southern California where it's just a matter of time before water shortages will cause more environmental issues?

The quality of the electrical and mechanical systems and their operations over time are inputs into PUE. As load changes and servers are turned off, the variability of the power and cooling systems influences your PUE. Therefore, PUE can now have a statistical range of operation given the conditions. PUE should be reported as a range of numbers (low to high), and the average calculated over a period of time. For example, Microsoft shows its PUE for one facility in Fig. 2.

Looking at the big picture, UPS systems operating with an energy efficiency rating in the mid 90% range while protecting critical loads are expected to become the norm. Data center facility managers monitor UPS efficiency more closely in the months and years to come to drive their PUE number down closer to 1.0.

Szalkus is a manager with GE Digital Energy - Power Quality, based in Lombard, Ill. He can be reached at mark.szalkus@ge.com.

Want to use this article? Click here for options!

© 2012 Penton Business Media, Inc.

---

Acceptable Use Policy blog comments powered by Disqus

Browse Back Issues Select an Issue EC&M Magazine | April 2012 March 2012 February 2012 January 2012 December 2011 November 2011 October 2011 September 2011 August 2011 July 2011 June 2011 May 2011 April 2011 March 2011 February 2011 January 2011 December 2010 November 2010 October 2010 September 2010 August 2010 July 2010 June 2010 May 2010 April 2010 March 2010 February 2010 January 2010 December 2009 November 2009 October 2009 September 2009 August 2009 July 2009 June 2009 May 2009 April 2009 March 2009 February 2009 January 1, 2009 December 2008 November 2008 October 1, 2008 September 2008 August 2008 July 2008 June 1, 2008 May 2008 April 2008 March 2008 February 2008 January 2008 December 2007 November 2007 October 2007 September 2007 August 2007 July 2007 June 2007 May 2007 April 2007 March 2007 February 2007 January 2007 December 2006 November 2006 October 2006 September 2006 August 2006 July 2006 June 1, 2006 May 1, 2006 April 2006 March 1, 2006 February 2006 January 2006 December 2005 November 2005 October 2005 September 2005 August 2005 July 2005 June 2005 May 2005 April 2005 March 2005 February 2005 January 2005 December 2004 November 2004 October 2004 September 2004 August 2004 July 2004 June 2004 May 2004 April 2004 March 2004 February 1, 2004 January 1, 2004 December 1, 2003 November 1, 2003 October 1, 2003 September 1, 2003 August 1, 2003 July 1, 2003 June 1, 2003 May 1, 2003 April 1, 2003 March 1, 2003 February 1, 2003 January 1, 2003 December 1, 2002 November 1, 2002 October 1, 2002 September 1, 2002 August 1, 2002 July 1, 2002 June 1, 2002 May 1, 2002 April 1, 2002 March 1, 2002 February 1, 2002 January 1, 2002 December 1, 2001 November 1, 2001 October 1, 2001 September 1, 2001 August 1, 2001 July 1, 2001 June 1, 2001 May 1, 2001 April 1, 2001 March 1, 2001 February 1, 2001 February 1, 1995 January 1, 2001 December 1, 2000 November 1, 2000 October 1, 2000 September 1, 2000 August 1, 2000 July 1, 2000 June 1, 2000 May 1, 2000 April 1, 2000 March 1, 2000 February 1, 2000 January 1, 2000 December 1, 1999 November 1, 1999 October 1, 1999 September 1, 1999 August 1, 1999 July 1, 1999 June 1, 1999 May 1, 1999 April 1, 1999 March 1, 1999 February 1, 1999 January 1, 1999 December 1, 1998 November 1, 1998 October 1, 1998 September 1, 1998 August 1, 1998 July 1, 1998 June 1, 1998 May 1, 1998 April 1, 1998 March 1, 1998 February 1, 1998 January 1, 1998 December 1, 1997 November 1, 1997 December 1, 1996 November 1, 1996 October 1, 1996 September 1, 1996 August 1, 1996 July 1, 1996 June 1, 1996 May 1, 1996 April 1, 1996 March 1, 1996 February 1, 1996 January 1, 1996 December 1, 1995 November 30, 1995 November 1, 1995 October 1, 1995 September 1, 1995 August 1, 1995 July 1, 1995 June 1, 1995 May 1, 1995 April 1, 1995 March 1, 1995

browse e-newsletters