Data Centre power quality and assurance – key considerations for financial institutions

If the data centre of a financial service provider goes down due to a power glitch, the impact is significant.

May 12, 2014

By Marco da Silva, MD of Jasco Power Solutions

If the data centre of a financial service provider goes down due to a power glitch, the impact is significant – there’s potential damage to equipment, applications and systems. However, the greater impact includes downtime, business loss and damage to customer relationships. Speed and cost of recovery, and limiting damage sustained, will depend on the power management system in place. Clean and constant power is critical. To get it right a number of issues must be taken into consideration.

A data centre typically houses the ICT infrastructure of a business. It is the nerve centre of the business where multiple cabinets of ICT equipment enable communication, transactions and myriad other business processes, and precious business data is stored and retrieved. A power outage, poor or dirty power supply, or a power anomaly such as a surge or lightning strike, can wreak havoc with sensitive systems. Protection is vital, but so is ensuring suitable recovery and business continuity.

Power quality and assurance – the practical considerations

Much will depend on the risk appetite of the business, i.e., how much data can they afford to lose or is instant failover (100 percent up time) required. Depending on the challenges faced in the environment and the sensitivity of equipment, a number of technologies can be applied to drive power quality, from voltage stabilisation, galvanic isolation of the power supply to transient absorption and harmonic filtration.

Practical considerations must also be factored in. Besides system downtime, loss of power may mean cooling equipment is switched off, putting ICT equipment at risk of overheating or suffering damage as a result of rising humidity. The result may be loss of system performance or operational integrity, resulting in unacceptable times to recovery. In addition, when power is restored and systems start up, there is generally a power spike to deal with.

While a branded off-the-shelf solution, such as a UPS, can provide a part of the answer, a customised solution that meets the requirements of a specific environment will deliver better odds. A tailored solution will address failover in terms of speed and capacity, but also ensure power quality. Such a solution may include installation of voltage stabilisers, UPS’, inverters, surge and lightning protection, transformers and/or power supplies. It will certainly require careful assessment of loads, identification of mission critical systems and sensitive equipment, testing and refining of power quality, and ensuring suitable cooling and distribution of power in terms of the feed to the UPS. Heating, ventilation and air conditioning equipment is usually powered separately given the huge inrush current on this equipment. These loads are typically run of a generator.

Smaller, modular, hot swappable

In terms of advancements in system design, power management solutions are getting smaller and modular solutions are providing additional benefits. Space is at a premium in data centres and the traditional requirement for failover specifies replication of a UPS or power management system (N+1). New solutions that allow single modular scalable systems for failover within the same frame, minimise the footprint of solutions.

Modular system designs make almost any configuration possible, while the hot swapability of components within these systems drive mean time to repair, replacement, and ultimately recovery and failover. Good advice is to ensure the solution provider holds sufficient stock to ensure immediate replacement of key components.

Other considerations are energy efficiency, the addition of environmental monitoring to the solution and use of embedded Web servers which will allow the power management solution to be accessed from a remote location. The latter also facilitates monitoring of equipment by the service provider if desired. Maintenance of these systems is crucial. A rule of thumb is to service UPS’ twice a year, testing for electrical or battery integrity.

Depth of experience, engineering capability – a winning combination

While there are a number of proven and recognised UPS brands available to select from, the majority are only available in standard configurations. They cannot be customised. Resellers of these solutions are also usually not capable of putting together or advising on the design of complex, multi-component tailored solutions. A solution provider with the right experience and capabilities will be able to identify challenges and potential complexities and develop a suitable solution.

In selecting a provider, identify companies that have strong engineering capabilities, broad experience and a solid track record in designing and building bespoke solutions. For quality assurance, ensure that a manufactured solution is accompanied by an ISO 9001/2008 rating.