UPS users supporting loads from small server rooms to hyperscale data centres mostly view power resilience and efficiency as their two highest priorities. While the balance between these factors varies according to the load’s size and criticality, the power efficiency aspect is assuming an ever-increasing urgency. A few years ago, efficiency improvement was predominantly about reducing operating costs. Today, however, an energy improvement strategy is essential to demonstrate a responsible, green attitude to customers, employees, shareholders and any other stakeholders, as well as to comply with government legislation. This has been highlighted by Theresa May’s recent commitment for the UK to reach net zero carbon emissions by 2050.
In this article, Alex Emms, Operations Director at Kohler Uninterruptible Power, provides some brief background to this shift in priorities, before explaining why the search for ever-improving efficiency can conflict with the Uptime Institute’s Tier lV redundancy and resiliency requirements. He then shows how these issues can be overcome by an approach that uses smart modular UPS technology while providing upfront assistance during the design phase.
Sustainability as the fastest-rising data centre concern
DigiPlex, a pan-Nordic green data centre operator, has released a survey of almost 300 decision-makers in Sweden and Norway, titled ‘A Perfect Storm: Nordic Data Centre Trends 2018’, containing extensive research into changing attitudes and business priorities[i]. Survey results revealed ‘Sustainability’ to be the fastest-rising data centre concern, from 27th place to fourth place in just the two years since the 2016 survey. The topic joins ‘Security’, ‘Operating Costs’ and ‘Energy’ at the top of respondents’ concerns.
Carbon costs are becoming a global issue, with activist pressure from groups like Greenpeace and increased government awareness pointing to potential fines, new taxes or reputation damage down the line.
Tier classification and redundancy
While all data centre loads are to some extent critical, budget realities for most operators mean that the level of resilience provided must be balanced against cost. The Uptime Institute’s four-tier classification system gives users the information they need to make this choice; Fig.1 shows the uptime requirements for the four different levels.
Fig.1: Uptime Institute Tier levels and availabilities
All Tier classifications specify redundancy levels for the power distribution paths and equipment serving the site’s computer hardware. If we compare Tier ll and Tier lV redundancy levels, we can see the differing impacts they have on UPS utilisation and efficiency.
Tier ll calls for redundant capacity components and single non-redundant distribution paths to the site computer equipment. This could be fulfilled by a pair of standalone UPSs on a single path, each rated at 300 kW. The N+1 redundancy requirement is satisfied; however, neither UPS will ever be utilised beyond 50 percent. If the data centre demand drops significantly below normal levels, the UPS loading could fall to a point at which efficiency becomes substantially reduced.
A better solution is to use a modular system as shown in Fig.2.
Fig.2: N+1 modular UPS configuration for a Tier ll installation
This configuration could be fulfilled, for example, by a PowerWAVE 9500DPA modular UPS system. With four 100 kW modules supporting the 300 kW load for N+1 redundancy, utilisation per module improves to 75 percent; more margin for energy-efficient operation.
This energy efficiency issue cannot be so easily solved in Tier lV infrastructures, which call for redundant capacity systems and multiple distribution paths to the site’s computer hardware. The Tier lV requirement can be met with a 2(N+1) UPS system, as shown in Fig.3.
Fig.3: 2(N+1) modular UPS configuration for a Tier lV installation
In Fig. 3’s configuration, even though it takes advantage of modular design, the maximum utilisation per module drops to 37.5 percent. From here, a relatively small drop in the critical load would take the UPSs into an inefficient operating region.
However, it’s possible to overcome this issue using a smart module deployment technique, which we will now describe.
Xtra VFI smart module switching
A load/efficiency profile for a modern, modular UPS system shows that efficiency remains high for most load values, and only falls significantly once loading drops to 25 percent or less. This means that the 2(N+1) nominal loading of 37.5 percent, while low, still allows full-efficiency operation provided it can be maintained.
Xtra VFI can be used to achieve this ‘load level maintenance’. This mode automatically adjusts the number of active modules to match the current load requirements. Modules that are not needed are switched to standby but remain in a state of readiness, primed to transfer to active double conversion operation if the load increases.
Suppose, for example, the load in Fig.3 falls to 200 kW, or 0.66 x its nominal value. The loading per module drops to 25 percent accordingly. However, Xtra VFI could then switch out one 100 kW module in each UPS while still preserving 2(N+1) redundancy for the reduced load. Meanwhile, utilisation improves back to 200kW/6 modules = 33.33 percent, so the UPS continues supporting the reduced load with not only full redundancy but also full efficiency.
Fig.4 shows the load/efficiency curve for a UPS such as the PowerWAVE 9500DPA operating in double conversion mode; it also shows the benefits of the Xtra VFI mode.
Fig.4: The energy efficiency benefits of the Xtra VFI mode
When the system calculates the module switching needed for maximum efficiency, it allows for the redundancy required. Redundancy levels, as well as highest expected load step, can be user-configured. If a mains failure or alarm occurs, the Xtra VFI mode is automatically deactivated while all modules are returned to active status. Modules can also be rotated between active and standby mode to stabilise aging and ensure that the modules are available when needed.
Finding the optimum solution; a consultative approach
From this article we can see that the optimum UPS solution for any installation depends entirely on its circumstances and priorities. Accordingly, when KUP sales engineers discuss new projects with clients, they take a consultative approach. They find that informal, wide-ranging conversations can be the most productive, as these can surface not only the obvious issues, but others that are less so, yet also important.
Some surprising circumstances can be uncovered. In one example, a small server room hidden away in a basement turned out to be fulfilling a highly critical role, supporting all of its organisation’s business transactions worldwide.
Certainly, a starting-point can be to discuss configurations that will help a client to best balance their expectations of budget, resilience and energy efficiency. Resilience discussions should include allowances for battery autonomy; this could be a few minutes if it’s only needed until a generator can come on line, or may be several hours if it must support the load until utility power is restored.
Additionally, it’s well worth explaining the advisability of installing bypass switches, as these allow UPSs to be maintained or taken off line without interrupting power to the load.
UPS monitoring may also be required; KUP can provide a remote monitoring service as part of their maintenance support package.
A review of strategy for future expansion is also important; the planned location for the UPS may be sufficient for the immediate installation, but will it accommodate future growth, in terms of space, access, and cooling capability?
Any relevant legislation should be considered, whether due to region or industry. Healthcare estates and railway installations have industry-specific legislation as well as general electrical requirements, for example.
Conclusion
Whether embarking on a new-build project or an upgrade to an existing system, UPS system users will benefit greatly from engaging with a supplier such as Kohler Uninterruptible Power. Their advanced modular technology, backed by nationwide technical support and long-term business stability, ensures that their clients achieve an optimally-balanced solution while avoiding the many pitfalls that can otherwise occur during design and operation.
