Skip to content Skip to footer

Meeting UPS operator priorities in micro and medium data centres

Electrical Review Logo

A survey of data centres decision-makers has found that their key priorities for introducing new technologies and solutions are availability, energy efficiency, reduced operating costs and future-proofing for expected IT growth. Here, Kenny Green,technical product manager at UPS Ltd, a Kohler Company looks at how these priorities can be met for data centre UPS systems up to 1 MVA by specifying modular solutions such as UPSL’s PowerWAVE 9000DPA Series

Data centres are continuously being called on to support ever larger and more critical loads, while being more efficient in doing so. Operators are introducing new technologies and solutions to meet these challenges, but what are their key metrics for success? To answer this, a survey has been conducted across over 4000 decision makers involved in the design, construction and operation of critical data centres across North America and EMEA to identify their views on these key priorities.

Availability

 

The survey shows availability as one of the users’ highest priorities. Because power problems are the largest single source of downtime, increasing power availability is the most effective way to increase overall systems availability.

Availability is a measure of how much time per year a system is operational and available for use. It is formally defined as

 

Availability =     MTBF

(MTBF-MTTR)

 

Figure.1

 

where MTBF is the Mean Time Between Failures, and MTTR means Mean Time To Repair. We can see from this equation that Availability can be increased either by increasing MTBF or, importantly, by reducing MTTR.

 

To show how reducing MTTR can significantly improve availability, consider first a free standing UPS system that has an MTBF of 600,000 hours and an MTTR of 6 hours. Applying these figures to equation (1) yields an Availability of 99.9990% (‘5 Nines’). By comparison a UPS configured using rack mounting modules may have an MTBF of 400,000 hours, because it uses more components in a more flexible and efficient topology. However it has a dramatically reduced MTTR of just 0.5 hrs, because it can be repaired without being taken off line simply by ‘hot swapping’ a faulty module, rather than requiring specialist repair effort in situ after being taken off line. With these figures a modular system offers an availability of 99.9999% (‘6 Nines’) which is an improvement by a factor of 10 over the free standing figure.

Increasing energy efficiency and other cost reduction factors

Data centre operators today are highly motivated to improve energy efficiency wherever possible, partly due to sharply rising fuel costs, but also because of increasing social and political pressure to be seen as an environmentally responsible, ‘green’ organisation.

Table 1 below shows how modular UPS systems such as the UPSL PowerWAVE 9000DPA can increase energy efficiency and reduce operating costs compared with free standing UPSs. The Table shows how each system would be configured to support a 120 kVA load with N+1 redundancy. The 9000DPA system’s incremental, modular topology allows a system capacity more closely matched to the 120 kVA load than possible with the free standing solution. The result is more efficient operation with reduced power and cooling costs, as the Table shows. The savings are significant for the scale of operation, amounting to more than £25,000 over five years, and CO2 reductions of 140 Tonnes or 217 Trees, according to the Carbon Neutral Company.

Table 1: Comparative running costs of a parallel redundant UPS installation to support a load of 96kW ([email protected]) using both legacy ‘free-standing’ and modular systems

 

*Costs based on 7.84 p/KWh. **Source: Carbon Neutral Company

 

The PowerWAVE 9000DPA modular approach brings other operating cost savings in addition to those arising from its energy efficiency. Maintenance of bulky free standing units with a large population of relatively expensive individual components can be time-consuming and costly. By contrast, modular system components are smaller, and easier to manage and replace. In any case, a modular UPS system can be restored to its full redundancy and capacity within 30 minutes, without being taken off line or interrupting its supply to the critical load, by simply hot swapping a faulty module. This can then be repaired or returned to the supplier as convenient.

 

Forecasting maintenance needs and buying the right spares set for an entire standalone UPS unit is difficult and expensive, with no guarantee that the right part for any possible failure will always be available. There is also a time overhead for stock management and logistics. These costs, difficulties and uncertainties are eliminated by modular topology, where the only spares requirement is a single replacement module. This approach can save up to 50% on logistics and stock management costs.

 

Further savings possibilities exist, related to training, as in depth training for different UPS types and sizes can be reduced to module replacement training for a single UPS architecture used for all load sizes across the entire enterprise. By eliminating the need for systems specialists, up to 67% can be saved on service personnel training.

 

Flexibility

 

A UPS system’s efficiency relates to how closely its capacity is matched to its critical load size. However this load can both increase and decrease by considerable amounts during a data centre’s operating life. A sharp reduction can come from a virtualisation exercise, which creates more efficient use of data centre server hardware. However there is typically also an ongoing trend of growth as new users and applications demand more data centre capacity.

 

With its modular architecture, the PowerWAVE 9000DPA can be configured to closely match a data centre load, and then to track it accurately as it grows or shrinks over time. Table 1’s example shows a 120 kVA load. Achieving N+1 redundancy for this with legacy technology calls for two 120 kVA units, each of which is only 50% loaded. Efficiency is only 90% as a result. The alternative arrangement shows N+1 redundancy being achieved with four 40 kVA rack modules, providing a total 160 kVA capacity. This configuration is 75% loaded, so operates at 95% efficiency. Modules can then be simply added to or removed from the PowerWAVE rack in response to load changes over time, keeping the UPS loading and efficiency at the high levels established in the original setup.

 

This review has shown why modular topology is so well-suited to the data centre operators’ priorities as outlined above, meeting their criteria for Availability, energy efficiency and reduction of operating costs, with agility and flexibility in responding to changes in IT loads. The PowerWAVE 9000DPA Series in particular is ideal for micro and medium data centres, as its capacity can be incremented from 10 kVA with N+1 redundancy in one cabinet, all the way up to 1 MVA N+1 in multiple cabinets.

 

Uninterruptible Power Supplies ltd

Tel: 01256 386700

www.upspower.co.uk

email: [email protected]

You may also like

Stay In The Know

Get the Electrical Review Newsletter direct to your inbox, and don't miss a thing.