Feature UPS & Standby Power What are the implications of big data?
The explosion in big data (data generated from many sources), will present opportunities as well as challenges, and data centre operators need a strategy for this large and sometimes unpredictable growth. Here, Kenny Green of Uninterruptible Power Supplies, looks at a UPS topology that supports data centres in meeting the challenges of big data
here are a number of factors that are fuelling and will continue to fuel the exponential growth in data - known as big data - from the volume and detail of information cap- tured by enterprises to the rise of multi- media, social media and the internet of things. While this creates a challenging environment, with the correct tools organisations can use this data to pro- vide deeper insights into their opera- tions, allowing better and more accurate business decisions.
T
One thing that is required is an increase in computing power and its associated infrastructure, so that the growth in data can be handled correctly. A report commissioned by Cable & Wireless Worldwide shows that compa- nies studying big data applications have discovered that capacity requirements will grow by 40% to 50%. Yet most organisations have not considered this within their forward capacity planning. As data centre managers seek to accommodate this fast, sometimes unpredictable growth, they must ensure that their infrastructure expan- sion keeps pace with that of their data processing equipment, without com- promising that equipment’s reliability or availability. The UPS is one critical component of this infrastructure, and it’s an area in which Uninterruptible Power Supplies (UPSL), can help.
Scalable UPS topology Advances in semiconductor technology have allowed the development of trans- formerless UPSs, which have several
Above: Kenny Green, technical support manager at Uninterruptible Power Supplies
Right: handling the growth in data calls for a significant increase in
computing power and its associated infrastructure
Below: there are good reasons why modular UPS topology offers an attractive way forward for data centre operators needing to
accommodate or plan for big data
important advantages over transformer- based implementations. Efficiency is improved by about five percent, sub- stantially reducing heat loss and elec- tricity running costs. However, the key benefit is a substantial reduction in the UPS’s size and weight. For example, a transformerless 120kVA UPS would weigh 310kg and occupy 750 x 850mm, compared with 1,200kg and 800mm x 1,650mm for transformer-based units. These reductions have had a pro- found impact on UPS design, because it means that UPS systems (even 3- phase types) can realistically be built using rack mounting modules rather than monolithic floor standing units. Great new opportunities in flexibility, space saving and scalability become possible - UPSL’s PowerWAVE 9000DPA UPS, comprises a vertical racking frame which accepts up to five modules, rated from 10kVA to 50kVA. These can be added or removed incre- mentally to efficiently track data centre load changes over time. For instance, a data centre load could initially be 50kVA, serviced by a single 50kVA rack mounted module. Another 50kVA module could be added to provide N+1 redundancy, and more as the load grows, until it reaches 200kVA - five 50kVA modules would service this with N+1 redun- dancy maintained.
Modular UPS topology’s flexibility and easy scalability are valuable to data centre operators managing big data. By using vertical scaling within a UPS rack as described, they make effi- cient use of their data centre room, adding UPS capacity without increas- ing floor space. The PowerWAVE 9000DPA can achieve up to 342kW/m2 power density in this way. Equally as important, UPS capacity purchase can be deferred until needed, eliminating unnecessary hardware expenditure. When needs do arise, new modules can be simply mounted into the frame without even interrupting power to the load - there is no disruption, build- ing or engineering work involved. If the capacity demand forecast has led to sufficient power cabling and floor space allocation, expansion beyond the original frame’s capacity is also fast and
28
simple. More frames can be provided, with each containing all of the input protection, input and output isolation required by the UPS modules. Therefore, any associated switchgear panels can be smaller, more easily installed and lower cost than those essential for stand alone UPS systems. It is also why no additional electrical installation work is required when more capacity is added.
High availability With modular UPS’ easy vertical and horizontal scalability, data centre oper- ators can effectively plan for expected future growth in the critical load, and accommodate it as it happens. Availability, which depends on Mean Time Between Failures (MTBF) and Mean Time To Repair (MTTR), is opti- mised by modular topology. MTBF is improved in the PowerWAVE 9000DPA because the modules operate in a truly redundant mode, with no single points of failure. Modular UPS topology also reduces MTTR, because a faulty module can be quickly and easily replaced, cutting repair times from six hours to about half an hour. The overall result is a UPS system that can achieve 99.9999% availability. To summarise, there are good reasons why modular UPS topology offers an attractive way forward for data centre operators needing to accommodate or plan for big data and its associated impact on capacity demand. Provision can be made for large scale and unpre- dictable growth, without having to pur- chase hardware unless or until it is needed. When extra capacity does become essential, it can be installed quickly, easily and without disruption to the critical load.
Uninterruptible Power Supplies
www.upspower.co.uk T: 0800 171 2320
Enter 222 JUNE 2013 Electrical Engineering
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56