Power Usage Effectiveness - PUE and pPUE

PUE, the abbreviation for Power Usage Effectiveness, was developed back in 2007 by The Green Grid Association (www.thegreengrid.org). Since then it has been adopted by the industry around the world. A detailed explanation of PUE and pPUE has been published by ASHRAE in the book "PUE^TM: A Comprehensive Examination of the Metric".

Today, there seems to be a lot of misuse or misunderstanding of PUE and pPUE in the data center industry. This is my understanding of PUE and pPUE:

PUE

Power Usage Effectiveness is the relation between the total facility energy used to run a data center and the energy used to run the IT equipment as shown in figure 1. The total facility energy can be produced from different forms of energy, not necessarily 100% electricity.

The PUE should be an averaged value over the course of one year in order to consider the influence of the ambient temperature in an appropriate manner. The PUE is intended to be used to document the efficiency of an individual data center over time. The PUE should not be used to compare different data centers.

What PUE considers

The efficiency of the power, cooling, and other infrastructure components and the form of the source energy are the major influencing factors for the PUE.
The total facility energy used to run a data center can consist of electricity, natural gas, fuel (regular generator tests), or water for adiabatic cooling or district chilled water instead of using chillers. The different forms of source energy will then be weighted with defined weighting factors in the PUE formula.
The location of the data center has an influence on the PUE. The colder the climate, the more economizer or free cooling can be used, the lower the energy consumption of the cooling system, and the lower (better) the PUE.
The air temperatures used in the data center have an impact on the PUE as well. The higher the air temperatures, the more efficient the operation of the cooling system, the lower the energy consumption of the cooling system, and finally the lower (better) the PUE.
The servers and other IT components themselves may operate highly efficiently or possibly with a very poor level of efficiency. The useful output, whatever it may be, may be produced very efficiently or inefficiently and this influences the PUE in a "negative" way. The more efficiently the servers work, the lower the energy consumption of the IT, and the higher (worse) the PUE. It may seem counterintuitive, but this is indeed the case.

What PUE does not consider

The electricity used to run a data center might be produced by a nuclear plant, by burning coal, or by using renewable sources such as solar, wind, and water. The PUE does not consider if the electricity has been produced by renewable sources.
Some data centers may reuse heat created during the cooling process. The PUE does not consider any energy reuse in the "energy balance". There are other metrics that consider this issue.

pPUE

The metric pPUE, partial power usage effectiveness, defines a certain portion of the overall PUE of a data center within a clearly defined boundary. In my example I have calculated the PUE and the pPUE for a transformer, UPS/PDU, chiller, CRAH, pump, and others to show how it is calculated and to show the differences in relation to the PUE.

Figure 1 shows the energy distribution of all infrastructure components and the IT equipment. The PUE is calculated as follows: PUE = 11,750 MWh / 8,760 MWh = 1.341

Figure 2. The transformer losses shown as pPUE results in the pPUE(transformer):
pPUE(transformer) = ( 500 MWh + 8,760 MWh ) / 8,760 MWh = 1.057

This article is an extract from the original post (only available in English) that can be found here and includes several examples: www.linkedin.com/pulse/my-understanding-pue-ppue-benjamin-petschke

About the author

Benjamin Petschke was born in 1969 in Germany. After studying physics he joined STULZ in 1996 and worked since then in the R&D, Export and Marketing department on different positions. With 19+ years' experience in the Data Center cooling industry, Mr Petschke is specialised in Data Center cooling design, energy saving and acoustic issues.

He works closely with the Joint Research Centre of the European Commission for the Code of Conduct on Data Centres on the Best Practice section and recently with the German DKE in development of the DIN EN 50600, Information technology – Data Centre facilities and infrastructures.

Mr Petschke authored White Papers on subjects like Best Practice for Data Centre Cooling and Indirect Free Cooling with Dynamic Control Logic.