A reliable and even power supply is important for the reliable operation of a data center. Not only main power failures, but also smaller voltage fluctuations can disturb or fail delicate IT hardware. The public power supply system in Germany fails approximately one hour per year or 10 seconds a day. However, a server without power supply already fails after 30 milliseconds. A hardware failure through the lack of electricity means not only an interruption of IT business but also loss of not yet stored data and disturbances in software, caused by improper shutdown.

You can find protection through a Static uninterruptable power supply, short UPS. There are varying types of Uninterrupted Power supply systems which are used for different purposes.

Firstly, there is a so called offline device, also called VFD (Voltage and Frequency Dependent). During the normal operation Uninterrupted Power supply systems of this category only pass power from the entrance directly to the exit. Additionally a rectifier is used charging accumulators. If the net supply should break down, the attached consumers are switched over to an inverter supplied by those accumulators. Depending on the module the switchover between net and battery operation is carried out within a delay of up to 10 milliseconds. For delicate equipment this can already be too long.

These systems can only bridge a complete power failure, and not compensate the irregularities of the incoming net voltage. Furthermore these Uninterrupted Power supply systems are only available in a single-phase execution, therefore the performance is not as high as in other, three-phase systems. Those are reasons why offline Uninterrupted Power supply devices are not used in data centers.

Secondly there are the so-called online devices, also called VFI (Voltage and Frequency Independent). They compensate power fluctuations, and they supply IT equipment with power in case of a complete power failure. Therefore this kind of Uninterrupted Power supply consists of two components: The controlling and regulating unit and an energy storage.

The controlling and regulating unit transforms the alternating voltage of the incoming source into direct voltage. This direct voltage is used by the charger to charge the energy storage. In addition, the direct voltage is transformed in turn into an even one or three-phase alternating voltage which then supplies to the downstream consumers. Since an online device is always running, such equipment has no interruption time at power failure.

Such online equipment is particularly fail-safe. It has a so-called by-pass wiring onto which you can switch without any interruption in case of an overloaded rectifier/inverter branch. Moreover, in those systems rectifiers and inverters are designed for continuous operation and are particularly robust and of high quality. This high quality design means relatively high initial costs.

In addition, both the direct and the alternating voltage have inversion losses which reduces the degree of effectiveness. In product descriptions of Uninterrupted Power supply systems, as a rule, the degree of effectiveness is indicated at full capacity. Since most Uninterrupted Power supply, however, are not used at full capacity, the lower degrees of effectiveness at partial capacity are decisive for the correct calculation of the power consumption and the costs. Today the degree of effectiveness of 92% at full capacity can be described as standard for an online Uninterrupted Power supply. Better values will be reached with devices above 7.5 kVA.

Such Uninterrupted Power supply types are also called as "online", "double conversion", "continuous operation", or "double transformer".

Another essential feature of a Uninterrupted Power supply is the maximum bridging time which depends on the capacity of the accumulators. It can measure a few seconds or several hours depending on the requirements. At power failures computers in smaller data centers boot down automatically. In this case, smaller Uninterrupted Power supply systems with a relatively short bridging time are usually installed which allow enough time to shut down all regular programmes. Larger data centers have emergency power systems; the supply from accumulators only serves the purpose to bridge the time until these have started and were taken to nominal power.

The energy storage of a Uninterrupted Power supply is usually a battery with appropriately big storage capacity and high reliability. There are different battery types which vary in consideration of their operating temperature, maintenance requirements, their storage capacity and their initial costs. Other energy storage media are conceivable, for certain applications up to 50 kW super capacitors and hydrogen fuel cells can be used economically.

When using batteries you should care to exchange these regularly before their expiry. The accumulators of a 10 year expiry should be renewed latest after eight years, the accumulators of a 5 year expiry should be renewed completely latest after four years to avoid a failure of the Uninterrupted Power supply system by a faulty battery system. Furthermore the environmental temperature is decisive for the life span of the batteries. Temperatures of approx. 20-22 ° C are ideal, deviations can reduce the life time drastically. In addition the room where the batteries are placed has to be ventilated.

Since there have been changes in technical development within the last few years, operators should consider to invest into a more efficient Uninterrupted Power supply system instead of renewing the batteries of the old system. This means the possibility of saving operating costs with a modular Uninterrupted Power supply system because it operates with lower capacity at the same safety level as a non modular redundant and older system.

Nowadays there are also dynamic Uninterrupted Power supply systems which win required power from kinetic energy. This is assured with an heavy flywheel which is driven by main net voltage and in case of power failure supplies stored energy. However, the temporal bridging of a power failure is limited. To guarantee a timely independent power supply these systems are delivered mostly in connection with a diesel aggregate. It has to be taken into account that the Uninterrupted Power supply system bridges only the time period which the diesel aggregate needs to start running. This kind of the Uninterrupted Power supply system is only economically appropriate as from approx. 1.000 kVA.

Non-modular uninterrupted power supply

Beside the investing costs, For the calculation of the economic value of a Uninterrupted Power supply the costs of a potential failure as well as the operating costs during the complete running time should be considered. To reach desired failure safety at use of non-modular, redundant systems two identical systems are used.

If one of the two breaks down, each of both could supply the necessary capacity. This means, however, different disadvantages: Both systems run permanently even if only one is needed. Moreover, they work with a low degree of effectiveness in turndown operation. And because those systems are difficult to be upgraded at a later stage, they often are planned overdimensional.

Modular uninterrupted power supply

Modular systems with 19” custom Uninterrupted Power supply slide-in units can be designed n+1-redundantly, this means that at failure of a slide-in unit the remaining modules take over the full performance. A120 kVA are not secured conventionally via 2 redundant systems of 120 kVA but with four modules of 40 kVA each. With this concept all modules run at optimal capacity with a very high degree of effectiveness. This has a positive effect on the energy balance of the Uninterrupted Power supply.

At the same time this concept has a positive influence on the system availability. Due to the higher number of modules, it can occur that one module breaks down more often as with a conventional Uninterrupted Power supply, but some manufacturers offer the exchange of a module by the operator himself, even during running operation. With this concept the failure period is reduced considerably in comparison with a non modular system. Hence a modular Uninterrupted Power supply is safer and shows higher availability rates. Additionally faulty modules can often be repaired by the manufacturer and reinstalled during regular operation.

Furthermore modular Uninterrupted Power supplys don’t have to be calculated for the highest expected performance but are able to grow with increasing performance requirements. To avoid single point of failure it is important that every module has its own controller and its own battery.

At times when environmental protection and efficiency play a high role the exchange of an older Uninterrupted Power supply system could be taken into consideration. The investment of a new slide-in unit in modular systems is economically recommendable.

For the reliable operation of a data center a Uninterrupted Power supply system is absolutely necessary, but which type will be the right one? The size and primarily the energy consumption of the data center have to be taken into consideration as well as the building, the maximal necessary bridging times and the size of the budget. We like to advise you at the choice of the right Uninterrupted Power supply system - don’t hesitate to ask us, we’ll take care of IT.

innovIT - CARES ABOUT IT