It is no secret that data centers serve as the backbone of the fast-evolving global digital transformation, actively supporting the growing adoption of advanced technologies and high-performance computing, the introduction of 5G, as well as the shift of numerous organizations to a cloud environment to facilitate remote work and communications.
To cope with these ever-increasing demands of both industries and individual users, the operations of data centers have become much more complex. Based on a study conducted by AFCOM, power densities in data centers averaged between 4 to 5 KW roughly ten years ago. With the proliferation of data-intensive processes and applications, however, a number of data centers are now operating at least 10 KW per rack, and is projected to increase another 5 to 10 KW by 2025.
With demand for data center services showing no signs of slowing down, data center operators and managers must be able to create a solid foundation of uptime and reliability within their respective facilities, including having the appropriate power and cooling redundancy measures in place in case of power interruptions.
In the face of growing reliance on their services and the increasing financial and reputational cost of downtime, it is imperative for data centers to keep uptime up and outages out.
Understanding and overcoming threats to uptime
The Uptime Institute released its 2022 Global Data Center Survey recently, which discloses that the cost of service disruptions continues to climb, despite having a lower number of serious power outages.
Twenty-five percent of the survey respondents have reported incurring over $1 million in direct, opportunity, and reputational costs, because of downtime. The 2022 survey also reports on-site power issues as the leading cause of data center outages.
These findings were echoed by Dr. Moises Levy, senior principal analyst for OMDIA, when he, together with Suresh Pichai of Equinix, participated in a panel discussion on ‘How data centers are negotiating threats to uptime’, stating that the lack of reliable power and optimization in data centers considerably affect their uptime and reliability.
Due to ever-growing demand and reliance on data center services, coupled with the ongoing delay in energizing new facilities due to supply chain difficulties, it is possible that some data centers may be operating beyond their ideal capacity, increasing stress on critical equipment and the risk of system failure. Consequently, this heightened demand necessitates data centers to be ready to store large amounts of data to meet industry requirements.
Thankfully, technological advancements in UPS systems are available to help ensure continuous and reliable power in mission-critical facilities such as data centers. However, choosing a UPS that suits current and future operational requirements is also crucial in protecting the facility against power disruptions.
Data center operators, especially those that are mindful of their overhead costs, traditionally make use of low-voltage UPS systems. But with operations and processes becoming more complex, shifting to the use of medium-voltage UPS systems can be a sound decision, as they offer the benefits of increased reliability and reduced costs, because of the lesser number of units required to sustain robust operations.
Apart from offering higher energy efficiency levels and reducing electrical losses by operating at lower currents, medium-voltage UPS systems, especially those with modular architecture, are scalable, and allow data center operators to prepare for future growth in increments, doing away with heavy one-time expenditures.
In addition, medium-voltage UPS systems are flexible enough to support varying data center requirements, and have the capacity to protect mission-critical loads, supplying continuous clean power while providing much-needed protection from a broad spectrum of utility voltage events that can damage critical infrastructure.
Speaking of the not so distant future, data center operators and managers must also find ways to integrate the generation and use of renewable energy within their facilities, especially as the industry faces the enactment of stricter regulations on energy efficiency and sustainability.
Given this, data centers must be able to accelerate the adoption and use of renewable energy in their respective facilities, as clean energy can help in increasing operational productivity while reducing carbon footprint.
Energy storage solutions are seen as one of the key elements in the important shift to clean power, as they enable the storage and use of energy generated from both traditional and renewable sources when it is most needed.
With the increase of power providers, energy costs become more reasonable, and help the grid operate more efficiently and sustainably, due to the reduced demand on power generation equipment. Similarly, energy storage solutions can help data centers increase their profitability, by allowing facilities to sell excess or unused energy back to the grid during peak demand.
Looking after our human resources
Suresh Pichai, who is currently the director for Innovation and Development at Equinix, also shared how the organization views their people as one of the contributors to the efficient operations of their data centers. At present, the industry still faces a shortage in skilled labor, in light of the construction of more facilities and the introduction and use of advanced technologies.
Suresh discussed the importance of having proper procedures and processes to boost productivity while addressing occupational health and safety. As data center operations become more complex, the risk for injuries and accidents in the workplace also increases, especially for those that work with electrical systems.
This is where real-time insights and visibility, facilitated by intelligent digital solutions, can be fully utilized. These smart devices gather and transmit critical data in real-time to an energy and asset manager, which processes, analyzes, and provides information on equipment health and efficiency and signals when equipment and systems are not performing optimally.
This does away with the need for manual condition monitoring, which can pose a safety hazard for data center workers, and helps facility managers anticipate and prepare for potential system failures or implement methods to improve energy use. With operations as complex and intricate as data centers, it is paramount to monitor and measure equipment and operations to effectively optimize processes and manage issues.
Similarly, immersive AI-based tools can help facilitate remote service activities. Instead of waiting for service personnel to be on site to address equipment failure, AR-based tools and applications save up on much-needed time and resources by providing on-site data center employees with professional support in the form of troubleshooting guides and live tutorials to correctly identify problem areas and recommend the necessary remedial measures.
Digital tools and applications can help reduce the possibility of human error that may result in further system or equipment downtime and prolonged exposure to electrical equipment. Leveraging on the power of digitalization can help data center operators managers gain a deeper understanding of their current and future operations while providing existing employees with the required support and resources to ensure that they perform their tasks safely.
Working together to achieve operational reliability
For data centers to achieve a higher degree of uptime and reliability, they need to have support from advanced technologies, as well as trusted industry partners. Achieving optimal data center performance requires teamwork, regardless of their size and complexity, along with the provision and use of products and solutions that can provide real-time information and expansive visibility.
Without the appropriate measurements and indicators, data center operators and managers will not be able to manage operations. Always remember – you cannot manage what you do not measure.