How can a data center operator be sure that a facility’s measured power usage effectiveness (PUE) – the key parameter defining energy efficiency – reflects reality? The profound risks associated with even minor discrepancies in this area can be avoided by implementing ABB’s modular and scalable solutions. Such solutions meet all measurement requirements, ensure the highest level of accuracy, improve energy efficiency, and can translate into as much as a 36 percent reduction in maintenance costs after upgrading electrical systems.
Aleksandar Grbic ABB Electrification – Smart Power Quartino, Switzerland aleksandar.grbic@ch.abb.com
It may not be much of a surprise to hear that Internet traffic has grown by leaps and bounds over the last few years. But what may raise many eyebrows is that dire predictions about an expected explosion in data center CO₂ emissions have not come close to being realized. Since 2010, Internet traffic has grown 12-fold. This trend has been driven by factors such as the rapidly growing number of interconnected devices, the replacement of physical applications with virtual ones, and a doubling of Internet users. Nevertheless, between 2010 and 2019 data-center energy usage remained steady at about one percent of global electricity demand, or about 200 TWh →01 [1]. Furthermore, this energy-stingy trend appears to be continuing. According to the International Energy Agency (IEA), “If current trends in the efficiency of hardware and data center infrastructure can be maintained, global data center energy demand can remain nearly flat through 2022, despite a 60 percent increase in service demand.” [2]
![01 Global trends in internet traffic and data center energy use, 2010-2019, courtesy OMDIA [1].](https://resources.news.e.abb.com/images/2022/7/14/0/Data_center_energy_m2301-01-EN.jpg)
Behind these encouraging figures is the fact that data centers have consistently invested in technologies designed to reduce energy demand and CO₂ emissions. Here, the most important parameter defining data center energy efficiency is Power Usage Effectiveness (PUE), a term developed by the American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) and The Green Grid [3]. In essence, PUE refers to how much energy is used by a data center’s computing equipment in contrast to total data center energy consumption, including cooling, lighting, and other non-IT-related equipment →02.
As essential as PUE is to determining a data center’s energy efficiency, the data upon which PUE is based and the measurement systems that contribute to its determination, are every bit as important. Here, accurate and correct measurement of key electrical parameters, such as voltage, current, power, energy, and power factor depend on the correct use and placement of measurement devices [3]. The EN50600-2-2 standard for data centers requires the measurement of these parameters to an accuracy of 1 percent. Additionally, it recommends the measurement of current and voltage total harmonic distortion (THCD and THVD), stipulates that acquisition of such data must be performed rapidly and simultaneously, and that the resulting data must be analyzed and represented correctly. Only under these circumstances can the owner of a data center be sure that a measured PUE value reflects reality.
Several consequences can follow if there is even a small discrepancy in a facility’s measured PUE value. At risk, for instance, is the ability of a data center to accurately measure its energy efficiency, its ability to allocate power to IT loads, and its ability to plan installation upgrades effectively.
Accurate information is everything
The above-mentioned risks can be avoided by implementing ABB’s unique and flexible solutions. Such solutions meet all measurement requirements, ensure correct class 1 accuracy, and make power monitoring and capacity planning easy, while improving energy efficiency.
In addition, exceedingly efficient ABB devices ensure the highest efficiency for power distribution equipment, including transformers, UPS systems, cables, and protection and switching devices. Thanks to a UPS efficiency of 97.4 percent on system level in double conversion mode, efficient power distribution products, and the right power distribution design, power distribution losses, which usually average 20 percent, can be cut to just five percent.
Additional efficiency increases can be realized through the implementation of ABB’s measurement, monitoring and control solutions. For instance, ABB’s Ekip devices with embedded metering are capable of measuring and controlling all electrical parameters on all distribution levels with a high level of flexibility and class 1 accuracy according to the IEC 61557-12 standard. Having embedded functionalities offers the following advantages:
• No need for additional relays and measurement devices, thus enhancing simplicity and saving time
• High level of flexibility thanks to a choice of several communication protocol modules
• Simple and effective cloud connectivity
• Increased reliability thanks to fewer devices and connections
• Fast design, installation, and integration.
Furthermore, most of the important information from ABB’s embedded metering devices can be easily visualized and monitored from the ABB Ability™ Energy and Asset Manager, which is available as a local or cloud-based solution.
Modular and scalable
Since data centers come in all sizes, ABB offers modular components →03 designed to make it easy to realize three scalable solution levels:
• Essential monitoring is a basic solution that provides the ability to monitor a data center’s PUE. It is ideal for small installations.
• Enhanced monitoring provides a wider and more accurate view of power consumption, enabling analysis of energy efficiency and the possibility of monitoring UPS status.
• Advanced monitoring is a complete package designed for very detailed metering and providing predictive maintenance. It is suitable for larger data centers or data centers with the highest energy efficiency and sustainability requirements.
The above-mentioned scalable solution levels offer many advantages. They reduce project design time by up to 80 percent, cut project risk because digital connectivity is tested by ABB, are easily adapted to different projects, and offer the possibility of being upgraded at any time, without the need to change hardware, by adding advanced functionalities available on the ABB Marketplace™. This can translate into as much as a 70 percent reduction in costs when upgrading electrical systems compared to traditional replacement.
Essential monitoring solution
ABB’s Essential level solution is based on two Emax 2 or Tmax XT circuit breakers. These measure all the electrical parameters (current; voltage; frequency; active, reactive, and apparent power and energy; power factor; peak factor; THVD; THCD) of a UPS output and a facility’s input. The two circuit breakers are connected through the Modbus TCP communication protocol.
Measurements are collected through the ABB Ability™ Edge Industrial Gateway and stored either in the local Gateway or transferred to the cloud. If the latter is chosen, all the information is available from the new ABB Ability™ Energy and Asset Manager cloud platform and is accessible from any location and any device with internet access.
In addition to a cloud connection, it is also possible to connect the measuring devices (circuit breakers) to the local DCIM installed and configured on the premises and to use available information in a customized way.
Although this solution is simple and has a low initial cost, it provides very little information about the energy consumption of the data center, since only two measurement points are installed. Consequently, there is little room for improving the overall efficiency and reliability of data center.
Enhanced monitoring solution
Here, the same considerations regarding measurements, software and communication outlined in the Essential solution apply. However, in the Enhanced solution the measurements are performed by more protection devices and the new System pro M compact® InSite, thus reaching more data.
Information is available in the cloud and/or locally. The safety of the data is ensured thanks to a high level of cybersecurity, developed in collaboration with Microsoft. This allows very flexible, easy and precise computation of the PUE value. Thanks to the grouping of the load feature, it is possible to customize the plant overview in a fast and flexible manner →04. For example, all protection devices protecting the cooling load can be grouped together so that the values of the cooling load consumption can be seen, while keeping the visibility of the individual values as well.
With higher precision, numerous measurement points, and information about equipment status, the equipment that consumes the most energy can be easily identified and corrective action can be taken. In this way, it is easy to make cost-effective changes that improve the overall efficiency of the data center and, thanks to a unique predictive maintenance feature available on air circuit breakers, maintenance can be planned in advance. Furthermore, thanks to UPS Insight, the major UPS parameters can be monitored. These include real-time current and voltage, temperature and battery charge status, as well as alarms and other events.
Advanced monitoring solution
A much more advanced monitoring system is also possible. This solution offers capillary metering architecture that is not limited to IT land mechanical loads and covers a range of additional load types. This advanced solution can be applied to any type of data center type, regardless of size. But it goes without saying that the larger and more complex a data center is, the more important it is to reduce costs without compromising energy efficiency or reliability.
Taking typical costs into account, ABB’s calculations show that the Advanced solution can save up to 36 percent of maintenance costs for some devices such as Emax 2 air circuit breakers. Additionally, with precise information regarding the health of each device and regular maintenance, the reliability of an entire plant increases dramatically, thus reducing the chances of data center outages, which can cost as much as $2.4 million [4] per incident. The probability of encountering an unplanned maintenance situation is further reduced by the fact that all monitored devices continuously perform self-checks. If a device detects any abnormality based on the upper and lower thresholds set by the customer, it issues an alarm.
Cutting distribution losses
In addition to improving a data center’s efficiency through increasingly precise energy use monitoring, operators can turn to technologies for reducing distribution losses. ABB offers equipment that can decrease power distribution losses down to five percent. Furthermore, this approach is ideally suited to installations with large numbers of measurement devices, which can provide insights into the causes of distribution losses, and thus support distribution efficiency optimization. This approach is applicable to data center IT equipment as well. Such equipment can produce power quality issues that cause harmonic distortions in a network. These issues can be overcome by placing suitable filters inside a network. However, to correctly select and locate the filters, operators must identify the sources and levels of the distortions.
ABB equipment can perform measurements up to the 50th harmonic without additional devices, providing the right information and enabling these improvements. Additionally, the same equipment, all of which is outfitted with embedded power quality meters, can monitor other aspects of power quality, such as average voltage, spikes or short interruptions in voltage, voltage imbalances between phases, etc. that, when identified and rigorously managed, can further increase energy efficiency and reliability.
References
[1] IEA. “Tracking Data Centres and Data Transmission Networks 2020,” Tracking Report, June, 2020. Available: https://www.iea.org/reports/tracking-data-centres-and-data-transmission-networks-2020. [Accessed January 15, 2022]
[2] IEA analysis based on Masanet, E. et al. (2020). Recalibrating global data center energy-use estimates, Science, 367(6481), 984-986. Available: https://doi.org/10.1126/science.aba3758. [Accessed January 15, 2022]
[3] Information technology – Data centre facilities and infrastructures – Part 2-2: Power distribution, EN 50600-2-2 Standard, 2014.
[4] Ponemon Institute, Cost of Data Center Outages, Ponemon Institute, 2016.
