Should a string of recent electrical outages lead to a rethink in our approach to data centre power continuity?
The events of the last few months should be a cause for concern for everyone working in the power continuity industry.
Across the globe, a spate of major power outages has caused chaos and mass disruption. In June, virtually the whole of Argentina, Paraguay, and Uruguay lost power plunging 50 million-plus people into darkness.
September saw a failure in the Central American Electrical System. A blackout hit most of Honduras, Nicaragua, El Salvador and Guatemala.
Throw in mass power outages in New York and Tenerife. Then the unusual case of California in October, where power was deliberately cut to around one million customers to try and avoid a repeat of the devastating wildfires of previous years.
And who can forget Friday 9 August here in the UK? A lightning strike kick-started a string of events that saw two power plants disconnect from the grid, causing frequency to drop below the safe level.
As a result, nearly one million people lost power to help ensure the entire grid didn’t collapse.
In November’s edition of Data Centre Review (DCR) magazine, we explore whether all these incidents in such a short period of time should leave us questioning the seemingly tried and tested approach to data centre power continuity.
Thinking The Unthinkable – the Blackout report
Earlier this summer, Riello UPS published the Blackout report, an in-depth investigation about the possibility of a prolonged nationwide power cut taking place in the UK.
As we make clear in the DCR article, our document doesn’t predict events such as the August incident, it merely attempts to shine a light on the UK’s power grid, the growing threats we face to maintaining power continuity, and what the consequences might be for business and wider society.
However, in the aftermath of August’s outage, several of the key questions highlighted in the Blackout report were raised, including whether National Grid needs to up the amount of reserve power it holds.
The very idea that the electricity system could be brought to its knees just because a couple of power stations dropped off at short notice should send alarm bells ringing
– Professor Dieter Helm, government advisor on energy policy
Debate For Data Centres
Compared to many other industries, the data centre sector already adopts a more serious approach to maintaining power. Operators frown on downtime, with the Uptime Institute’s Tier Classification System the go-to guidelines for guaranteeing availability in mission-critical facilities.
At the top end of the scale, Tiers III (N+1 minimum) and IV (2N minimum) provide enough backup to run off mains for 72 and 96 hours respectively.
But what if we take a wider view? Take August’s outage. National Grid deployed 1,000 MW of reserve power, one of the reasons grid frequency and supplies were back to normal within the hour. Nearly half of this backup (475 MW) came from large-scale battery storage.
Isn’t it time data centres took more responsibility and started using their UPS batteries in a similar way? Most operators to date have been cautious, perhaps even dismissive, of the concept of energy storage. They think why put additional strain on their vital emergency backup.
But with battery storage systems now at a mature phase of development, isn’t this viewpoint slightly outdated?
We explain the benefits to data centres of participating in demand side response (i.e. increased not diminished system resilience, reduced electricity bills, lower network charges, additional revenues from selling surplus back into the network).
In addition, should energy-intensive organisations including data centres be forced to minimise their electricity use during a major power outage or civil emergency?