When emergencies arise with UPS systems, the resolution will require being speedy and permanent. In UPS terms you have four common failure scenarios. Here we describe how they can occur – und how to deal with them.
1. in a single-bus UPS system where a loss of redundancy (e.g. when one module trips off-line) has no impact upon the critical load:
In this case the response to the failure can be handled with the investigation of the root cause. If the fault has been caused by operator error then usually it can be fixed by on-site operation staff.
2. in a single-bus system where the entire UPS system (redundant or not) trips off-line and successfully transfers the load to the utility supply:
In this case the failure has not impacted the load immediately but it is at immediate risk from utility-borne interference. A manual transfer to emergency generator supply, with controlled load and restart shutdown, would usually be recommended although this brings additional risk from operator error and it has to have been previously established that the generator can support the ICT load without the UPS in circuit.
3. in a single-bus system where the entire UPS trips off-line and does not transfer the critical load to the utility supply and the load is disconnected:
A loss of data centre load is highly critical in any business and the operators ability to get the UPS bypass connected, probably including a manual starting of the emergency generator system, will not reduce the impact of the failure but only speed up the process of recovery. A service engineer is required in almost every case to diagnose, repair and reinstate the UPS system.
4. in a dual-bus UPS system where one system is negatively impacted but the load remains protected either by being dual-corded or being protected by point-of-use static transfer switches.
This is where the extra investment in a dual-bus power system is rewarded and the provisioned faulttolerance fully utilised. This failure mode is probably the only data centre power event where an immediate intervention is not required.
And what about modular UPS?
The application of modular UPS can reduce the downtime of any individual module drastically. If the on-site staff have the pre-commissioned UPS module and have had the training for a module-swap-out, then the downtime can be limited to less than one hour under most circumstances. The failed module then has to be repaired.