Maintaining a best-in-class reputation

To provide this, Ark needed to understand how engineering and operational decisions affected the performance and Total Cost of Ownership (TCO) of the data center. With this knowledge it could plan its data center capacity more effectively and understand the exact impact of investment decisions and customer activity on performance and costs.

“We needed to clearly see what was using energy and how: while we believed our UPS and chillers generated the most energy use, we couldn’t back that up with undeniable evidence. At the same time, capacity planning was crucial in us offering world-class services in an efficient manner. For example, a data center could be built for 800KW total capacity. Yet if we could only fill 350KW of that, we would essentially be paying for 450KW-worth of unused infrastructure.”

Initially, Ark attempted to predict and plan its performance, costs and capacity in-house using complex spreadsheets. However, this was impractical, time-consuming and could not guarantee verifiable results.

“In-house tools and resources were not going to be practical in the long term, and would impact on our ability to deliver products and services. Romonet’s ability to model the designed, built and operational environment, rather than simply metering it, could give us a deeper understanding of our data centers: allowing us to accurately analyze and understand future and actual performance.”

Ark initially chose Romonet Software Suite to help design and commission new data centers. However, it soon found that the suite could be used to calculate, rate and improve the efficiency and accountability of its existing facilities as well.

Using accurate prediction to improve confidence, accountability and performance
“We are already much more confident with our capacity planning. We have been able to defer hundreds of thousands of pounds worth of investment, as we don’t yet need the extra capacity. Taken over the lifecycle of a data center this represents a huge saving on the annual interest alone, and better control over our finances. Looking ahead, we will also be able to plan our energy use more effectively: only paying for the appropriate amount of energy for a data center, instead of having power on reserve that will never be used.”

Ark can also take faster design and planning decisions when investing in its data centers. For example, when replacing its UPS Ark could model competing systems to see what would be most efficient in its environment.

Greater confidence in its TCO predictions means Ark can act with greater authority. For instance, the control sequencer for one of Ark’s chiller units was not providing optimal efficiency. Using Romonet, Ark could identify the increased energy use and prove to the manufacturer that its product was using unnecessary energy. Once adjusted, Ark could also verify that energy performance improved as predicted. Ark expects that this will reduce the chillers’ operating time by an average of 20% over their lifespan; producing a significant net saving year-on-year.

“We now have much greater accountability with our clients: we can quickly and accurately predict the operational costs that a client will incur and ensure we are providing the best service at the best value. The same applies to individual client projects: by predicting the exact effect it will have on IT infrastructure and energy use, we can know a project’s cost and the capacity required.”

Ark can also advise the client on best practices to make individual projects more efficient and alert the client if performance doesn’t meet expectations. One client project required the use of more server space: simply by advising on the best timing for removing blanking plates and bringing these servers online, Ark could save the client 5-15% in energy costs.

“With Romonet, we can give our clients even more confidence in our ability not just to understand their requirements, but to meet them and even suggest ways to further improve performance.”