An uninterruptible power supply (UPS) is one of the most important security components in any business, especially IT. Most business processes will come to a standstill without power and hardly any administration can continue to function properly without it. Power failure problems in computer centres bring additional consequences such as data being damaged or lost. Even more devastating is the interruption of the power supply in hospitals, where human life and limb can be endangered.
A reliable UPS, in addition to bridging power failures, has the additional positive property of improvement in the quality of the power supply. For example, voltage dips, over-voltages and harmonics are kept away from consumers attached to the UPS. Good UPS systems supply their consumers with an absolutely stable frequency, despite possible frequency fluctuations on the public mains network. Often imperceptible mains faults lead to data failures in computer centres.
More than 47% of all data failures are caused by voltage fluctuations; just 16% of the data failures are caused by the IT systems themselves and the remaining 37% are caused by various external effects. These include storm damage, fires, explosions, flooding, water damage, earthquakes (Source: Contingency Planning, USA). If the voltage fluctuations that cause data failures are further investigated, it becomes apparent that almost two thirds of the voltage fluctuations last less than 0,5 seconds.
UPS terms and definitions
In accordance with the UPS product standard EN 62040-3, there exist 10 different network faults that should be caught by the use of a UPS system. They are: power failure; voltage fluctuations; voltage peaks; under-voltage; over-voltage; voltage transients; voltage surges; frequency fluctuations; voltage distortions; and harmonics.
Historically, the terms used to differentiate the various UPS technologies were not defined precisely. However, the terms will be assigned to the classifications specified in EN 62040-3 in accordance to their most frequent use. The terms standby, passive parallel operation or off-line are often used for VFD systems. The terms single conversion, active parallel operation or line-interactive UPS plants are used for VI systems while continuous operation, continuous converter or on-line are also known for UPS plants classified in the VFI category.
Almost all UPS systems available in the marketplace consist of the following five components. To better understand the functions of each specific component, they will be described individually.
1. The rectifier of a UPS system is responsible for converting the three-phase or single-phase alternating current (AC) supplied by the public power mains supply into direct current (DC).
2. The direct current link circuit carries a direct voltage. If a battery is used as a storage medium for the UPS system, it will be charged from the direct current link circuit. In normal operation, the direct current link circuit is supplied from the rectifier. Should a power failure occur, the direct current link circuit will be supplied with energy from the battery.
3. The energy storage is responsible for the energy supply in the case of a power failure. The most frequently used storage medium for UPS systems is the lead battery. In isolated cases, Ni Cad batteries are also used. If the UPS system only needs to bridge very short power failures, capacitors can also be used as energy storage. Another type of energy storage is the fuel cell. The fuel cell produces electrical energy by combining hydrogen and oxygen. This energy can then be used to operate a UPS system.
4. The inverter is responsible for converting the energy provided by the direct current link circuit into a sinusoidal alternating current with the amplitude and the frequency of the mains voltage.
5. The static bypass is a current path from the public power supply mains network directly at the output of the UPS system and thus to the critical load. If the static bypass is activated, the inverter is then non-operational. The static bypass is switched on for an excessive loading of the inverter or for a defect of the inverter, rectifier or battery. Most UPS systems can also be manually switched to the static bypass.
For UPS systems, a differentiation is made between normal operation, battery operation and bypass operation.
In normal operation, the rectifier is fed with energy from the public power supply mains and the battery will be kept fully charged from the direct current link circuit.
If the power from the public power supply mains fails, the UPS switches to battery operation. In battery operation, the inverter will be fed with energy from the battery. The UPS system continues to operate in battery operation until the energy of the batteries becomes exhausted or the mains supply is restored.
The third operating mode is bypass operation. Here, the critical load is supplied directly via the bypass from the public power supply mains. The UPS switches, for example, into bypass operation when too many consumers are connected so that the inverter is overloaded. Defects with the rectifier, inverter or battery also cause switching to the bypass operation.
Transformer or transformerless
Traditionally, UPS systems have a transformer switched in front of the alternating voltage output. The transformer is responsible for stepping up the alternating voltage of the inverter output. This transformation is required because the voltage in the direct current link circuit is not sufficiently high to create an alternating voltage at the required level.
Progress made in semiconductor technology means there are nowadays no problems in building UPS systems without transformer. A DC booster is used for the transformerless technology. The booster is a DC/DC converter that converts the voltage behind the rectifier to a significantly higher direct voltage; the increased direct voltage then allows the inverter to create a higher alternating voltage without needing a downstream transformer.
Transformerless technology provides several advantages. Firstly, UPS systems without transformers can be built with very compact dimensions. Secondly, the transformer is not exactly a ‘lightweight.’ This means transformerless systems are significantly lighter than conventional systems with transformer. Furthermore, the noise level of a UPS system without transformer is significantly less than a comparable system with transformer. The efficiency over the complete load range is better for the transformerless technology than for UPS systems with transformer.
The rectifier is also different for the two technologies. An uncontrolled rectifier suffices for transformerless systems. Systems with transformer, in contrast, require a controlled rectifier in order to provide the voltage in the direct current link circuit at the correct level. However, the controlled inverter also has other disadvantages – it is responsible for the contamination of the feeding mains and a larger reactive power.
For UPS systems with transformer, the battery is connected directly to the direct current link circuit. This is undesirable for the battery, because a certain residual ripple is always present in the direct current link circuit and is responsible for reduced battery lifetime. The battery is charged from a separate charging unit for transformerless systems. This charging unit supplies a smooth voltage and so does not place any load on the battery.
A frequent objection to transformerless UPS systems is that they do not have any electrical isolation between the input and the output. Although this is true, UPS systems with output transformer also do not have any electrical isolation. The electrical isolation produced by the transformer is cancelled by looping the neutral conductor through to the alternating voltage output. This means that the input and the output for UPS systems with transformer are not electrically isolated.
Redundancy, modularity and scalability
To make the power supply as reliable as possible, it is desirable to install redundant UPS systems. This means that a second UPS is available should the first one fail. Thus, if the load is to be protected redundantly, each of the two UPS systems must be capable of supplying the complete critical load with power. It is also possible to achieve redundancy through the use of a modular UPS system. This requires that at least one module more than actually required must be used. The use of the additional module means that the UPS is capable of supplying the complete critical load even if one module fails.
Scalability is another important item for the selection of the correct UPS. Scalability means that the UPS system does not consist of one large system, but rather consists of several modules. This offers several advantages. For example, a scalable system can grow with the associated power. If a non-scalable UPS system is used, it must already be taken into account at the time of system purchase what consumer load the UPS system must supply sometime in the future. This power with the appropriate batteries must then be installed right from the beginning in the computer centre. In contrast, for a scalable UPS, only so many UPS modules as currently required need to be installed when the computer centre is established. The number of battery modules can also be selected to handle the currently installed power and provide the required bridging time for independent operation.
The modular construction means that the term ‘modular UPS system’ is used instead of ‘scalable UPS system’. This means that when initially installed, a modular UPS system can be significantly smaller than a non-modular UPS system that must already handle all future power expansions. In addition, the UPS batteries have only a limited lifetime. This means that non-modular UPS systems must install a large amount of battery capacity at the beginning which may possibly not be required and so costs unnecessary money. In addition, redundancy for a modular UPS can be realised with less power than for a conventional UPS. For a modular UPS, the redundant power is the power still available when a UPS module fails. The following example illustrates this.
A computer centre equipped with a UPS system has a total connected load of 115 kVA. If we want to equip this computer centre with a redundant conventional UPS system, we need to install two UPS systems each with 120 kVA. This means that a complete UPS power of 240 kVA is installed. The power of the batteries attached to the UPS systems would then also suffice to provide the required standalone time with 240 kVA load.
If, in contrast, the 115 kVA load installed in the computer centre is to be handled by a modular UPS system constructed from 40 kVA modules, we proceed as follows. The system is built from four 40 kVA modules. This provides a total UPS power of 160 kVA. If a 40 kVA module now fails, the system with three 40 kVA modules is still capable of providing the installed power of 115 kVA. The UPS system consisting of four 40 kVA modules used here thus has a total power of 160 kVA or a redundant power of 120 kVA. This clearly shows the advantage of the modular UPS.
The complete installed power of the modular UPS system lies significantly below the power of the redundant conventional UPS system. This also means that not so much excessive battery power needs to be provided than only required for the redundancy.
Gone are the days when a UPS was just there to support the load in the event of a mains failure. Scalability, modularity, uptime and now more importantly efficiency, are crucial factors to not only provide resilience within an IT system but also long-term financial gains in operating costs.