This was first published in the March issue of Data Centre News Magazine.
Richard Clifford, Head of Innovation at Keysource, explains that data centre owners and managers could be using their power infrastructure to generate revenue without sacrificing disaster recovery processes.
The UK energy market has seen significant price increases over recent years - as much as 62.6% between 2006-2016, according to price comparison provider Selectra. The market’s volatility was highlighted in December 2017 when wholesale gas prices hit their highest levels in six years, due to supply disruption in Europe. As such, it’s becoming increasingly important for volume energy users to consider innovative ways to reduce costs and ensure they remain competitive.
Battery storage has been billed as the missing piece of the puzzle in addressing the word’s energy challenges. Companies like Tesla are investing in research to create cutting-edge batteries for homes and businesses that store energy from renewables. But energy storage is nothing new. Similar technology lies within Uninterruptible Power Supply (UPS) systems, which are already used by the vast majority of data centre owners. Using these systems in new ways could allow the sector to guard against the rapid price changes in the energy market.
UPS systems have been used in Data Centres for decades, but it’s only recently that operators have started to consider this infrastructure as a potential source of revenue generation by taking advantage of National Grid’s Firm Frequency Response (FFR) incentive.
On a basic level, UPS draws in energy while infrastructure is running and automatically switches to power the data centre – keeping systems live in the event of a failure. The stored energy is effectively a reserve that rarely gets used. However some operators are now using this reserve to power their IT systems when energy prices are at their highest, switching back to mains supply when prices are lower. In doing so they’re able to take advantage of the best tariffs.
The obvious question is whether doing this will affect disaster recovery. Using UPS in this way means more frequent recharging and should a power failure happen at this time, there could be a gap in uptime. For some data centre operators, this risk has been enough to completely remove any interest in using UPS for battery storage.
There’s an easy fix, however. By using two systems – one for disaster recovery and a second to reduce costs – operators can avoid risk and assure customers and stakeholders that there will always be back-up power available.
The application of this is relatively straight forward. One UPS system sits below the infrastructure and draws in energy which can be routed back to power IT in the event of a failure – purely used for disaster recovery. Meanwhile, a second is connected further upstream, at the transformer. This second UPS simply stores energy directly from the grid as the IT infrastructure is powered from the mains. It then can act as the battery for use when tariffs are high, helping generate savings without the risk.
Operators that have made the change to using UPS in this way have achieved savings in the region of 5 to 10 per cent on their energy bills. And the model doesn’t just allow for savings. Data centre operators can generate revenue by selling stored energy back to the grid too thanks to incentives like the FFR – a framework that allows third parties to feed back to the grid.
Ultimately the viability of this strategy depends on the data centre and the operator. Among the things to consider is an increase to operational cost. If UPS systems are being used more frequently it means that maintenance of these systems may have to ramp up. And there is obviously the capital expenditure that comes with investing in a second UPS system and connecting this to the existing data centre infrastructure.
Yet the case for doing so is compelling. Margins in the sector are tightening – some are nine times less than they were a decade ago. Meanwhile, energy is still among the largest overheads businesses face –anywhere from 25 per cent to 60 per cent of running costs, according to trade association Intellect.
The good news is that data centre demand has never been higher, due to the increasingly business critical nature of IT systems and growing demands for the rapid accessibility of data. But this means that energy costs will likely continue to be a pressure point which restricts some operator’s ability to grow. Using UPS to generate savings is not a panacea for the sector’s challenge, but it is an example of the sort of small changes data centre operates can make to ease some of the pressure.