Data Centres
Infrastructure Management
Power & Cooling
Data centre outages are costing more, with power failure the culprit
By Paul Brickman, Commercial Director at Crestchic Loadbanks
One of the more consistent problems that the data centre industry faces is the issue of outages. While it is well known that data centre outages can cause critical work problems for an enterprise of any size, what is becoming more and more apparent is that these outages are becoming increasingly costly too.
Recent findings from Uptime Institute’s 2022 Global Data Centre Survey revealed that the data centre industry is growing immensely, becoming more and more dynamic and resilient. Despite persistent staffing shortages, supply chain delays and other obstacles, there is a renewed focus on being more sustainable. The report indicated, however, that light was not at the end of every tunnel, as it additionally highlighted the fact that downtime in the data centre industry is becoming ever more expensive. Indeed, power failures have been identified as the main cause of this increase in cost.
The Global Data Centre Survey focuses on responses from more than 800 owners and operators of data centres. This included those responsible for managing infrastructure at the world’s largest IT organisations. While the aforementioned notes of sustainability, efficiency gains, staff shortages and supply chain issues also dominated the report, the issue of power resiliency remained a persistent and dominant theme throughout.
Back up power failure - a growing concern
Further analysis in related research from the Uptime Institute identifies that the biggest cause of power-related outages is the failure of uninterruptible power supplies, followed by transfer switch and generator failures.
Although this data shows a trend towards improved outage rates, the frequency of these outages is much too high and, with costs also on the rise, the consequences of an outage are getting much more severe.
Data centre operators are well aware of the impact that a power outage can have, and many have put measures in place to mitigate these risks. However, with back up power failures identified as the primary cause of power outages, as well as external issues around grid reliability, energy shortfalls, and the transition to more sustainable power sources, it has never been more important that operators test their back up power systems.
£1m failures are becoming increasingly common
The data highlighted in the report indicates that the costs of outages are on the rise. This is likely down to several factors, such as industry changes, the cost per minute of downtime increasing, and the prevalence of technology that is susceptible to outages.
In fact, a quarter of the respondents that were interviewed reported that their most recent outage cost them more than £1 million in not only direct costs, but in indirect costs also. This 25% is a significant percentage increase from 2021, which showcases a continuing upward trend over the last five years.
The report states, ‘Uptime’s 2022 annual survey findings are remarkably consistent with previous years. They show that on-site power problems remain the single biggest cause of significant site outages by a large margin.’
Considering that data centre equipment vendors are caught between high demand and lingering supply chain problems, and that attracting, but, moreover, retaining qualified staff remains highly problematic for many operators, it is becoming increasingly clear that using a load bank is an essential cost saving tool.
Using a load bank to commission or regularly test the back up power system not only tests the prime movers and the batteries (UPS), but also ensures that the more critical components of the system, such as the alternator and crucially the transfer switches are tested as well. These load bank tests not only prove that the UPS or generators will start, operate, and run efficiently in the case of a power outage, but also that the sets can be safely turned off with no interruptions when mains power is restored.
Put simply, in a data centre environment, the business case for using a load bank is clear cut - not testing is an extremely costly risk to take.
Beatrice - 14 November 2022