April 18, 2023
Today’s data centers have a power quality monitoring problem. Addressing it can save time and money down the road.
Data center downtime is extremely expensive - it can cost hundreds of thousands or even millions of dollars. Power issues account for 43 percent of data center outages (according to the Uptime Institute). Surging power demands, increasing power costs, and global initiatives to reduce carbon footprints are driving a need for more innovative and effective ways to monitor power quality along a data center’s entire power chain. As a result, the need for a new generation of reliable, intelligent rack power distribution, monitoring, and control solutions has never been greater.
Electrical reliability and power quality are critical to solving the downtime prevention equation. What is power quality? Power quality is the steadiness and usability of electric current. Good power quality comes from consistent voltage, steady AC frequency, and a smooth waveform that resembles a sine wave.
Monitoring and managing power quality in data operations is critical to operating efficiently and reliably. However, there are still unmet challenges regarding power quality, including visibility into the data, optimization of efficiency and infrastructure, and integration with existing BMS and DCIM systems.
The first power quality challenge comes from having no visibility into granular power quality data.
Power quality assessments are generally completed as one-time occurrences during the bring-up phase or a significant server overhaul of a data center. This means they often do not consider the impact of power quality issues in the facility on an ongoing basis or what power quality issues are caused simply by standard operations.
Within a data center, operating equipment is inherently prone to power quality issues due to carrying non-linear loads. Examples of this equipment include switch-mode power supply units, variable-speed drives, computers, and uninterruptible power supplies (UPS). Non-linear loads can cause harmonic distortion, which results in unwanted current, overheated cables, vibration, false device trips, energy loss, and other equipment malfunction or failure.
A lack of visibility into critical power metrics limits failover planning severely and drives up device failure rates, which exposes facilities to downtime risk.
The second challenge is an inability to optimize efficiency and infrastructure.
Driven by the lack of available power quality metrics, challenges around optimizing for operational efficiency and scaling mission-critical infrastructure arise. Issues like stranded capacity and the data center industry’s “dirty little secret” become obstacles to deploying high-density IT infrastructure due to the inability to maximize space and usage effectively.
The same level of monitoring that paves the way for maximum usage and efficiency in deployments can also be used to identify zombie servers that consume significant power while idle, as well as provide clarity into outlet issues and overcurrent events. All of these events can lead to costly downtime - a problem that could be avoided with better real-time, granular visibility into power quality at the rack and outlet level than what conventional intelligent rack PDU solutions offer.
The third challenge is a lack of integration for power quality monitoring with existing management tools.
While many intelligent PDUs exist on the market today, data center facility managers still need tools and features that can help them further drive operational uptime and efficiency. The risk posed by unchecked power quality issues is real, and the Uptime Institute statistics demonstrate that the threat of downtime is still alive and well. Unfortunately, current intelligent PDU solutions all struggle with integrating seamlessly with any building management systems (BMS) or data center infrastructure management (DCIM) architectures that are in place. This creates a divide between the general monitoring available and the tools that data center managers rely on to operate their facilities, hindering their ability to manage power quality appropriately.
The ideal intelligent PDU used for monitoring and addressing power quality issues in the data center would need to be able to communicate with existing BMS and DCIM systems through universal APIs such as Redfish, JSON-RPC, or SNMP.
As a result, reimagining the intelligent PDU is currently the best place to start when it comes to improving uptime in data center facilities. It’s a tool that lends itself well to providing the real-time visibility facility managers need to protect their facility’s uptime. But as it stands now, many operators are still risking their bottom lines because they are limited by the intelligent PDU technology that is currently available. Improving the solutions for monitoring and addressing power quality issues at the rack and device level will significantly benefit the data center industry. At Legrand, we’re looking to do just that. Stay tuned to our blog because, in the not-too-distant future, we’re going to be sharing more about what exactly we mean in Part 2 and Part 3 of this blog series.
In the meantime, feel free to contact a Legrand team member to discuss how Legrand solutions can aid your power quality management today.