Samuel Tumma, Head of Engineering at Aggreko Africa tells us why it’s so important to prioritise reliable power and comprehensive loadbank testing when investing in a robust data centre strategy.
Global data centre spend is expected to reach an impressive total of US$208 billion by the end of 2020. Demand, thanks to expanding business footprints and the move to the cloud and alternative infrastructure solutions, has pushed the market into overdrive with an estimated 628 hyperscalers expected to rise by 2021. The Edge, Artificial Intelligence, cloud IT infrastructure investment, Machine Learning and the rise of emergent technologies are shifting the data centre into high gear and centre stage.
But what if the data centre fails? What if demand overtakes infrastructure? Downtime can incur huge costs, put an organisation’s reputation on the line and can have unexpected knock-on ripple effects throughout the company and its infrastructure.
Therefore, it is essential to tick two particularly important boxes when investing in a robust data centre strategy – reliable power and comprehensive loadbank testing.
As data centre usage continues to rise exponentially alongside technology advances and rapidly changing business requirements, providers must ensure that their facilities can handle this demand. If they miss a crucial step, ignore a potential problem, then the results can be expensive. And damaging. Even if the power outage is not the fault of the service provider, the resultant power and temperature control infrastructure must be prepared for such an eventuality. Especially in Africa.
The crucial task of data centre testing
Effective loadbank testing of facilities is essential. There are plenty of expensive examples of companies that failed to test their systems and lost millions thanks to an unexpected outage. In the UK, leading financial institution TSB suffered a data centre outage that left millions without access to their accounts and cost the company an estimated £370 million. The reason? An untested data centre.
Testing provides insight into any weaknesses inherent in the system, flaws that may only be apparent when pressure is applied to the system. To prevent critical infrastructure from being the source of potential downtime, ensure that every part of the data centre is carefully tested before the facility starts operations. Loadbank testing puts immense pressure on the data centre, determining whether it is capable of withstanding full load demand conditions, or not. This testing process follows five key steps to ensure that maximum capacity is not the point at which a data centre fails, but when it succeeds.
The first step is factory acceptance testing that ensures all equipment has undergone evaluation before and after assembly. This determines whether it has been built according to design specifications and can operate within those parameters. Resistive and reactive loadbank testing can simulate electrical loading for generators and UPS systems to ensure 100% performance at full load. The second step is site acceptance testing. This is carried out by OEMs and installation contractors on-site to ensure that the solution meets specification criteria. At this stage, full load test documentation should be developed to cover quality auditing that includes method statements and schematics. Third place is pre-functional testing that verifies the functionality of the equipment and whether it has been correctly installed, wired, torqued and Megger tested before initial energisation.
The fifth step is individual system testing that takes each item – UPS module, generators, PDUs, bus-bar systems, et al – and individually tests it for full functionality. This includes burn-in tests for hardware using loadbanks or thermal imaging. Finally, integrated system testing wraps the entire system together to test the entire data centre’s performance at full load for all critical infrastructure. This layer of the process should be ensuring that varying loads, staged machinery failures and potential utility problems are applied to the data centre to ensure that it can handle it.
The power conundrum
While testing is always going to be a critical part of data centre maintenance and development, additional power sources and solutions should also be high on the priority list. Power solutions that can hold the data centre upright in the event of an unplanned, and in South Africa, planned, outage. Aggreko not only provides collaborative expertise in testing for data centre capability but is the world’s largest hire company for temporary power and storage alongside precise temperature control.
Aggreko works with data centre providers to ensure that power is always on when needed and to support data centres with anything from seasonal changes in temperature to unplanned outages in power. The company tests all equipment constantly to ensure reliability and works closely with data centre providers to ensure that their systems are ready, no matter what.
In Mombasa, Kenya, for example, Aggreko worked with icolo.io to test all the equipment, including power generators, UPS systems and precision cooling to get their Tier 3 unit in Mombasa online. icolo.io designs, builds and operates state-of-the-art carrier-neutral data centres. Aggreko’s loadbank service helped icolo.io test and commission the data centre and meet the Uptime Institute’s requirement for a Tier 3 facility. With our help, they were able to demonstrate the capability of the data centre and attract global customers with their unique offering.
Extra as a standard
For data centre contractors and operators, it makes sense to enlist testing and commissioning experts, such as those from Aggreko, because they can evaluate exactly what testing should be carried out and provide the correct level of equipment and expertise for carrying it out. For all stakeholders, using an outside consultant and outsourced equipment can generate even more cost savings thanks to the elimination of any capital expenditure on testing equipment.
Aggreko promises extra innovation, reassurance, expertise and flexibility to ensure your data centre is tested properly and eliminating the risk of down-time. We use the latest technology to stress-test your equipment – a task which can be performed remotely with the results sent through to your smart device. We always ensure resilience and consistency by delivering global standards for testing and providing the most extensive range of equipment.
We conduct robust testing of all power and cooling infrastructure – quickly and safely. With our mobile systems mean we can work around you and other trades on-site, reducing the risk of delays and maximising health and safety.
