You want one how big?

Oct 24, 2012 | UPS & Standby Power

As Paul Norgate of Eaton Power Quality explains, deciding on the right UPS, and in particular the right size of UPS for a specific application, can pose a challenge and is by no means an exact science. However, here are some useful guidelines

To answer the question of UPS sizing, there are a number of factors that need to be considered including the configuration of the UPS installation, the size of the load, the required run time, the possibility of future expansion, and resilience. Let’s take these factors in order, while noting that they are, to some extent, interdependent.

Configuration

There are two main options for the way that UPS systems are configured – a decentralised system (where each server rack is fitted with its own individual UPS), or a centralised system (where one large UPS is used to supply multiple racks). The first approach has the benefit of easy expandability, but the UPSs do occupy potentially valuable space in the racks.

The second approach typically allows the UPS to be located in less valuable grey space, such as an electrical switchroom, but careful consideration must be given to how easy it will be to expand the UPS system should the need arise.

The load

The next step is to assess the size of the load which in principle, is a relatively straightforward matter of taking the loading information from the rating plate on each server fed from a particular UPS, adding the loads together and then allowing a safety factor.

In practice things aren’t quite so simple, as the server rating plates will show the loading in watts (W) whereas the UPS will be rated in VA, which means that the power factor of the UPS must be taken into account. For example, some line interactive UPSs have a power factor as low as 0.8, which means that a 1,000VA UPS of this type can only support a load of 800W (1,000VA x 0.8). With modern double conversion UPSs, the power factor is more usually around 0.9, so the difference between the VA rating of the UPS and the power rating in watts of the load is less of an issue, but it should still be taken into account.

The UPS should ideally be selected so that it will be operating at about 75% of its rated capacity under normal circumstances, in order to achieve the highest efficiency while allowing scope for expansion. It is nevertheless worth noting that modern UPSs lose very little efficiency with loads down to around 40%, so if future increases in load are likely, the installation of an over sized UPS should not necessarily be ruled out. Over sizing shouldn’t, however, be taken too far – even the best UPS operating at only 20% or 30% of its rated load will be inefficient.

Run time

In non-IT applications, run time requirements vary greatly, and should always be considered on a case by case basis. As a rule of thumb for IT systems, however, a run time of 30 minutes is typically considered appropriate but, in sizing the UPS to provide this, the details of the application must still be taken into account. For example, if there is an on-site standby generator, a shorter run time may be adequate.

Also, in some applications, it may be possible to shut down the power to some servers almost immediately, while for others it will be essential to maintain power long enough for complex data back-up and shutdown procedures to be executed. In reality, most data centres have servers that fall into both of these categories. In such cases, an effective and flexible power management system will allow selective control over which servers remain powered and which do not.

Scalability

Provision for future expansion, often referred to as scalability, is something that is frequently neglected. As has already been suggested, this depends very much on the configuration of the UPS installation. Where each rack has its own UPS, expansion is very straightforward – simply add more racks with more UPSs.

Where a centralised UPS system is used, the situation is rather different. The decision then has to be made whether to buy an over sized UPS at the outset, bearing in mind the points made earlier about minimum loads, or to invest in a modular UPS system where extra modules can be added at a future date. With this last approach however, it’s important to be sure that sufficient space will be available to accommodate the new modules and that the electrical infrastructure is in place to deal with the expansion.

The final point to consider is resilience. What happens if there’s a UPS failure or the UPS has to be taken out of service for maintenance? For the most critical systems, it may be necessary to have a redundant standby UPS to cover such eventualities.

Sizing a UPS system isn’t an intrinsically difficult process, but there are many factors that have to be taken into account. Careful attention to these factors, ensuring particularly that none are overlooked, will allow the best and most economical choice to be made.

Eaton Power Quality

www.eaton.com

T: 01753 608 700