How Do I Calculate PUE?
Power Usage Effectiveness (PUE) is the most commonly used metric for measuring data center energy efficiency. It was developed by The Green Grid in 2007 and has been widely popular ever since due to growing concerns of the sustainability of data centers and for being able to represent overall efficiency with one easy-to-understand number.
What is the Formula for PUE?
PUE is the ratio of the amount of power needed to drive and cool the data center compared to the power draw from the IT equipment in the data center.
Expressed as a formula, PUE is calculated as such:
PUE = Total Facility Energy / IT Equipment Energy
How is PUE Useful?
PUE indicates how much of a facility’s total energy consumption is being used by IT equipment, expressed as a ratio. The closer to 1.0 your PUE is, the more efficient your data center is. A PUE of 1.0 would be a perfectly efficient data center in which all the power coming into the building makes it to the IT equipment with none used for loads to cool or light the room and none lost in transmission to the IT equipment.
By knowing your data center’s PUE, you can easily understand how energy efficient it is and monitor progress over time. You can spot changes in efficiency at different points in time such as at peak loads or by season. With this information, it’s easier to set goals, compare the efficiency of different sites, and make smarter decisions to improve your energy efficiency.
What is a Good PUE Ratio?
While a PUE of 1.0 indicates a 100% efficient data center, it’s not a realistic target for most. However, some of the greenest data centers in the world are getting close.
According to Uptime Institute, the average PUE is 1.58. If you are above that, your goal should be to get below it.
The target PUE will be different for every data center because of their unique locations, climates, technologies, and equipment. New constructions with state-of-the-art IT equipment and cooling technology should target a PUE between 1.2 and 1.4. Existing data center footprints should try to stay below 1.5.
How Do I Calculate PUE?
Now you know the formula for PUE, what it means, and a rough estimate of what your PUE should be. However, like many data center professionals, you may wonder: “How do I find out what my PUE actually is?”
Here are step-by-step best practices to know what to measure and how to calculate your PUE.
- Create a plan for data collection. To gauge the impact of your energy efficiency initiatives, you need to calculate and trend your PUE per location over time. The intensity of your efficiency initiatives will determine how frequently you should measure your PUE. For a basic program, weekly or monthly measurements will suffice. For an intermediate program, daily measurements are best. For an advanced program, continuous and automatic PUE calculations with data center energy management software is needed.
- Calculate your total facility energy. Total facility power is typically established with a utility meter. If monthly measurements suffice, you can wait for the utility bill each month. However, more aggressive efficiency initiatives require more frequent data collection. You can install a shadow meter on the utility meter to track more granular measurements or use data center management software to collect it automatically.
- Calculate your IT equipment energy. IT equipment energy can be calculated manually by summing up the power readings of your intelligent rack PDUs, floor PDUs, RPPs, or UPSs, depending on which of your devices are instrumented. If you are doing this manually, you will need to account for electrical losses by comparing each device’s input and output values. As with the step above, it is much easier to collect this data automatically over the network and feed it into an energy management or power monitoring solution.
- Calculate your PUE. For manual calculations, simply divide your total facility energy by your IT equipment energy at the frequency you decided, and that is your PUE. Data Center Infrastructure Management (DCIM) software can perform these calculations automatically in real-time so you can focus more on your energy efficiency efforts rather than on manual PUE math.
- Have a goal. At this point, you know how to calculate PUE and are trending it to know the impact of your efficiency programs. It is best to set an aggressive but realistic target PUE for each of your locations and implement the right strategies and tools to help get you there.
How Do I Improve PUE?
Steps you can take to increase the energy efficiency of your data center and reduce your PUE include:
- Instrument your data center with meters and sensors. Power and environment sensors give you critical data on what is happening in your data center. Power meters can feed into your management software to simplify PUE calculations and environment sensors help ensure your equipment is operating within recommended ranges.
- Set thresholds and alerts on your meter and sensor data. Modern data center management tools collect, store, report, and alert on this data so that you are the first to know of potential issues so you can remediate them and maintain high levels of efficiency.
- Safely increase temperatures. Overcooling wastes energy and money. By monitoring your environment, you can know where exactly you are overcooling equipment and by how much. Then, you can raise your temperature setpoints until you are maximizing efficiency without exceed manufacturer or industry guidelines such as ASHRAE.
- Implement hot/cold aisle containment. Avoid mixing cold supply air with hot exhaust air so your cooling system doesn’t waste energy cooling air that is already cold. Containment strategies will dramatically reduce your energy consumption, increase cooling capacity, and allow more power capacity to be used by IT equipment.
- Consolidate and/or virtualize equipment. Reducing the number of physical assets in your data center allows you to also reduce the amount of space, power, and cooling capacity you need. In the average data center, up to 30% of servers are ghost servers that are wasting energy without performing any useful function. By finding and eliminating them, you can more easily consolidate your data center or virtualize the remaining servers.
- Bill back customers for the energy they use. Drive a culture of energy awareness and accountability by charging your internal or external customers for the energy their equipment consumes. As customers become more mindful of energy efficiency because it impacts their bottom line, your PUE will improve.
- Consider innovative cooling technology. Highly efficient methods of cooling high-density racks such as liquid cooling and immersion cooling are becoming more popular. Research your options to see if one of these is right for your data center.
Is There a Downside to PUE?
PUE is a great indicator of overall energy efficiency due to its simplicity and universal applicability. However, there can be a negative side to such a simple metric.
First, PUE does not take redundancy into account. While data center managers are increasing the resiliency of their data centers to protect against climate change and extreme weather, they must deploy more power-consuming equipment which drives up PUE.
Next, focusing solely on PUE can lead to unsustainable water use. To dramatically lower PUE, organizations may deploy water economizers or evaporative cooling and chillers. The trade-off is that water use will rise, and if the data center is in a region where water is becoming scarce, that may cause more harm than the energy consumption of traditional cooling methods.
Last, improvements in PUE can be misleading. According to Uptime Institute, "improving IT equipment utilization efficiency has a significantly higher impact on energy consumption than increasing efficiency of facilities operations, measured by power use effectiveness (PUE)."
While it is important to measure and improve PUE, it does not paint the entire picture of your data center’s efficiency. It is critical that you remain focused on the objective of increasing efficiency rather than optimizing for one metric.
What Other Metrics Should I Track to Measure Efficiency?
PUE is just one of many data center sustainability metrics. By tracking different metrics, you will understand your level of efficiency from many different angles and uncover new ways to drive greener data center operations.
Metrics you should consider tracking include:
- Air Economizer Utilization Factor: How often outside air is used for free cooling
- Airflow Efficiency: How efficiently air moves from the supply to the return
- Cabinets Compliant with ASHRAE Standards: How much progress is made towards ideal environmental conditions
- Carbon Usage Effectiveness: What the overall sustainability of a data center is
- CO2 Savings: How many CO2 emissions were avoided due to data center energy efficiency efforts
- Delta-T Per Cabinet: How effective airflow is at cooling equipment
- Green Energy Coefficient: How much renewable energy is generated onsite
- Water Economizer Utilization Factor: How often indirect water cooling used
- Water Usage Effectiveness: How efficiently water is being used in the data center
Bringing It All Together
Calculating PUE is something that all data centers should be doing, but for most organizations it is a waste of time to do it manually. With modern DCIM software, PUE calculations are done automatically in real-time so you can spend your time on more productive projects that will improve PUE and be able to easily monitor the impact of those energy efficiency initiatives.
Want to see how Sunbird’s second-generation DCIM software makes it easy to automatically calculate, trend, and compare your PUE across all your sites? Get your free test drive today.