UPS Series/Parallel redundancy

UPS Configurations

Series Redundancy

If redundancy is required then the two UPS's must both be capable of supporting the entire load. They are then wired with one feeding the bypass of the other. If first unit fails it drops back to bypass transferring the load to the second unit. No additional boards are required. In the event of both units failing load is dropped back to raw mains.

Parallel working

Used to increase the load capability of a UPS system therefore if you parallel connect two 60kVA units they will support a load of 120kVA. In practice you get slightly more power out of two units in parallel than a single unit of the same size. Each UPS requires a 'parallel' interface board fitting. If one of the units fail the load will be dropped back to bypass as they are in effect working as a single unit.

Parallel Redundancy

Parallel redundancy tends to used mostly in dc systems where you only require the addition of a blocking diode to achieve isolation of the two or more units and there is no bypass. Again both units must be capable of supporting the full load but they are wired in parallel and if one unit fails the other takes the full load. This is achieved by either a master/slave configuration where the one unit is in standby mode and is only switched in when the master unit fails or by current sharing which requires expensive electronics or transformers.

(MasterElite will operate in parallel up to a maximum of 6 units but not work in a parallel redundant configuration.)

Battery sharing

Where two or more UPS share the same battery bank to reduce cost or because customer has large dc supply of the right voltage available. In a series redundant configuration the same cost reduction can be achieved by each UPS having batteries for half the required run time. As long as both UPS are working the full run time is achieved when the first unit switches to bypass on flat battery transferring the load two the second unit. The only disadvantage with this configuration is that if one units has failed and there is a power failure there will only be half the required run time available. If both units have batteries capable of the full required run time and both units are working you will actually get twice the required run time. MasterElite will only work in a battery sharing configuration when fitted with parallel boards.

N+1

This is a form of redundancy where you use the number of units to achieve full load capability plus one extra so that a single unit can fail and be removed while the remaining still support the full load. If the required load is 60kVA being supplied by three 20kVA units you actually use four units. Series redundancy can also be achieved as long as the load requirement is supported by one of the UPS. Parallel redundancy/N+1 are more common in dc systems as it is easier and cheaper to achieve with dc units and they do not usually have a bypass arrangement to fall back on. The simplest form of N+1 is using two units in a load sharing capacity where the total load is less than the maximum load of an individual unit - this is called Dual redundancy.