Remote status indicators located inside the room (changeover panel and adjacent to main entrance) and the operators’ office, as well as on the generator control panel, confirm whether the generator is running and whether it is on-line.

A mimic panel repeating all the information available on the generator set’s control panel – amps, voltage, fuel, coolant temperature, is usually only provided in the operators’ area.

This is because when a mains failure occurs the operators in the room will engage in a pre-planned procedure of shut down which they will only commence when they believe that the standby source is not going to become available.

In the meantime the facility will be powered solely from the UPS batteries, which will be slowly running down.

The generator engine starter motor is powered from the generator’s own battery. This has sufficient power to attempt to start the engine a finite number of times.

The operators will be informed of the length of time it is likely to take before the generator will start and run up to full power. After that time lapse, they will assume standby power failure and consider an orderly shut down of the systems – depending upon the UPS time resilience.

Alarms
The UPS will usually be equipped with alarms and neon schematics indicating the status of the electrical supply. This will be one of three alternatives:

An alarm is raised when either of the last two conditions occurs.
An alarm will also sound when the UPS senses adverse environmental conditions such as excessive temperature.

All the alarms can be connected to remote mimic panels and/or collected together and connected to a flashing beacon in the room – more likely to be noticed.

Manufacturers’ remote monitoring?
The status of the UPS may also be remotely monitored from a manufacturers facility. This will record the operation of the equipment and provide early warning of likely faults.

Whilst this sounds comforting, a closer inspection of the offer usually reveals that the claimed “live” cover actually refers to the fact that a modem connection ensures constant monitoring. The staffing of the facility however is found to be normal office working hours!

So, should a fault develop during the night or weekend, no report or action will be taken until the next time the office is manned and the reports scrutinised. We have on numerous occasions had to ask for records in order to find out about spikes and power outages, rather than them be volunteered immediately following an incident.

Clearly there is an advantage in having the manufacturer monitor the equipment to look out for developing faults. However they are unlikely to offer an alternative to major fault warning than your own operations staff.

Large rooms
Large modular in room air handling units of any worth, are fitted with a plethora of control settings, information and alarms.

Only a selected few are made available to the operations staff, and there is a convincing argument that none should be.

Others are usually password protected to ensure that only mechanical services engineers are able to adjust the various settings of: timers, off-coil temperatures, set temperatures etc.

The basic alarms of high/low temperature and filter blocked, usually activate a visual alarm.

This is of little use in the noisy and crowded computer room and so it is best connected to another flashing beacon or illuminating warning sign.

Smaller rooms
Wall mounted split system units contain alarm indication of their basic functions.

These can be connected to a room beacon or illuminating warning sign.

Remote monitoring
All the control settings and alarms status can be remotely monitored. Specialist software and a modem provide the mechanical services engineer, not only with access to the controls status, but also the ability to alter those settings.

A simplified version can also be provided to enable the operations staff to similarly monitor or receive alarms.

It is not usual to have live monitoring of air conditioning. This is because we build redundancy into a systems design, such that a whole unit can fail without it affecting the computer room’s environment. This builds in more time for a service response.

Supposing that the non-essential services PDU main breaker fails to reset, when the mains fail and the standby power takes over?

The UPS would carry on supplying all the computer equipment – and eventually the room temperature would rise to a level at which, we hope, the computers would switch themselves off before damage is done.

A high temperature sensor mounted on the underside of the suspended ceiling, in the centre of the room, provides a simple source for a standby alarm indication.

Note: Ensure AC equipment is fitted with auto start relays – or join Pitfalls!

Detection
Water needs to be excluded from server/telecoms rooms because of the large number of electrical connections both inside and outside the equipment they contain.

Detection is via special tape with separate conductors running through. Water bridges those conductors and an alarm condition is created.

Method
Tapes are laid in the floor void adjacent to risk areas – the worst being the in-room units’ condensate drains.

The ceiling void needs cover if there is a roof or there are rooms above it which have a flood risk – toilets, kitchens and canteens, or simply offices having lphw heating pipework.

The room itself is not usually covered, as any flood will flow into the protected floor void. However, where water-carrying pipes occur in the computer room, they should be suitably enclosed and water detected.

Alarm indication
Simple visual alarm indication within the room or in the operations office provides an early warning of leaks in the floor or ceiling void.