Power players: Lessons other industries should learn from data centres

Following its launch of the Blackout report, an in-depth analysis of how likely the UK is to experience a nationwide electricity failure, Riello UPS General Manager Leo Craig explains why other sectors should follow the data centre industry’s approach to power continuity

It was just a few weeks ago in mid-June when more than 50 million people across South America were plunged into darkness after a massive sudden power failure.

Virtually the whole of Argentina, Paraguay, and Uruguay were left without electricity after a failure to correctly reconfigure an emergency system following the upgrade of a power transmission line triggered a complete grid collapse.

The aftermath of such a huge international incident immediately brings to mind the question “could something similar ever happen here in the UK?”. Coincidence or not, that’s exactly what we’d been investigating in the weeks prior to the South American outage, research which finally culminates now with the launch of our new analysis the Blackout report.

Clocking in at more than 50 pages and nearly 15,000 words, we believe it’s the most comprehensive look yet into whether the UK could experience a complete grid shutdown (spoiler alert: there’s a 1-in-200 chance during the next five years), the threats which are most likely to cause such a failure, and what the catastrophic consequences would be for today’s digitally-dominated society.

As a company which works with a significant number of server room operators and other IT-related businesses, we know just how seriously the data centre industry takes the issue of power continuity.  That’s why we dedicated an entire section of the Blackout report to outlining the painstaking plans data centre operators put in place to mitigate against the risk of power-related downtime.

Resilience Requires Redundancy

Data centres are right at the very heart of today’s ‘always on’ internet-driven world, so anything less than near 100 percent availability for their clients results in damaging downtime that can cost millions of pounds and cause a massive reputational hit.

For entirely understandable reasons, concepts like redundancy are almost second nature throughout the industry. Data centres also install uninterruptible power supplies (UPSs) that provide enough emergency energy for servers and essential equipment to keep running until the backup generators, either diesel or gas-powered, come online and pick up the slack.

Whether it’s a small standalone IT room that meets the ‘basic capacity’ Tier I requirements, or the multi-megawatt-powered bit barns achieving either the Tier III minimum N+1 redundancy or ultimate fault tolerance of Tier IV, it’s evident what procedures and infrastructure is necessary to maximise uptime.

The higher up the Tier Classification, the greater the in-built resilience. For basic Tier I, there are backup power supplies, while Tier II introduces redundancy into the UPS, generators, and pumps.

Looking to Tier III, we see the introduction of multiple paths for both power and cooling, plus enough redundant components to enable equipment maintenance and replacement without the need for system shutdown. While with Tier IV, there’s redundancy for every component across the whole computing and non-computing infrastructure.

For both Tier III and Tier IV, it’s also considered best practice to connect to the grid via a minimum of two transformers to minimise any disruption if one of the elements should fail. Many larger data centres, particularly hyperscale or colocation, will actually have electricity routed from different substations or separate parts of the grid to spread risk even further.

It’s such thoroughness that means Tier III and IV data centres are designed to continue running during a power outage for at least 72 or 96 hours respectively.

Of course, certain other sectors do go to great lengths to minimise and mitigate the risks that could bring their operations crashing down. In healthcare, for example, best practice for facilities like hospitals focuses on something called the “3 Rs”, which comprises:

• Robustness: a system or site should be able to absorb the effects of an event and continue to operate at the required level
• Redundancy: if robustness cannot be guaranteed, it is essential to provide more than one key facility or subsystem
• Reconfigurability: the most devastating risks are often unanticipated. For true resilience, a system or facility should be able to cope with the aftereffects of an unexpected event.

Other mission-critical environments such as manufacturing factories, water treatment works, and utilities adopt similar principles, with vital infrastructure often duplicated.

But even with these types of organisation, and especially in other non-critical industries and with smaller businesses, there are lessons to be learned from the lengths that data centres go to.

Powerless to prevent disruption?

However, even for sites with such safeguards in place, during a sustained national power outage there’s only so much an operator can do to insulate themselves from severe disruption.

Yes, they might well have emergency backup from UPS systems and generators, but what happens when the batteries are drained, or the fuel runs out and there’s no means of replenishing supplies?

Our report also explores whether there is anything else a mission-critical site such as a data centre could do to take their emergency preparedness one step further.

Perhaps larger businesses and organisations above a certain threshold should be compelled – by new legislation where necessary – to draw up worst-case scenario plans for incidents such as a sustained power outage? These could include demands to structure operations to minimise unnecessary power usage during a civil emergency.

Another idea would be for data centres to minimise the electricity they need in order to prolong their most critical services during a sustained power outage. Such a move would alleviate any unnecessary load on the backup generators, conserving precious fuel so it lasts longer.

The big question any data centre operator needs to ask themselves is if, when confronted with just two choices, wouldn’t you rather run just your most critical operations rather than face a complete shutdown?

• The Blackout report by Riello UPS is free to download from www.theblackoutreport.co.uk