Chris Connors, Project Sales Director, DTGen
The recent rail disruption across the north-west was a reminder of how much modern life depends on power systems that we never see. A short drop in the mains supply took out a Network Rail operations centre in Manchester. The resulting lag meant the knock-on effect cascaded across six train operators, displacing trains and crews and leaving commuters stranded well into the following day.
We don’t know the specifics of that site. But it raises a question I spend a lot of my working life on: if the mains supply fails, does the backup actually do its job?
Most people assume that if there’s a generator on site, they’re covered. But reality is more complicated than that, and the gap between having a backup supply and it performing the moment it’s needed is where organisations can get caught out.
Mind the (start-up) gap
A standby generator can’t predict when it will be called on and obviously isn’t running continuously. If the mains fails, there is always a short break in supply while the set starts, comes up to speed and takes the load – within seconds, but not instantly.
For most sites, that momentary delay isn’t an issue. But for some, it is. Where a load is genuinely critical and even a few seconds of interruption is unacceptable, an uninterruptible power supply (UPS) sits between the mains and the equipment. A UPS is essentially a very fast battery system that steps in the instant the mains drops, holding the critical load while the generator starts and takes over. It’s why data centres invest heavily in UPS protection, because they simply can’t afford to wait for the generators.
When a site with backup installed still loses power for a sustained period, it raises the question of whether the critical equipment was protected in that window at all, or whether the protection was there but didn’t operate as intended.
Available isn’t the same as connected
There’s a second failure point that is often ignored: the changeover switch, or automatic transfer switch.
In a typical standby setup, the generator has no direct link to the mains. The changeover switch is what monitors the incoming supply, and when it sees the mains fail, it signals the generator to start. Once the set is running at the right voltage and frequency, the switch breaks the connection to the failed mains and makes the connection to the generator.
The point here is that the generator can be running perfectly, up to speed and ready to take load, and the site can still go dark. If the changeover switch fails to break from the failed mains and connect the backup source, perhaps through a fault in its mechanism, or simply because it’s never been properly tested, the transfer never happens. The generator is available but not connected. Externally it will look like the backup has failed, but the weak link was the switch, not the generator itself.
Maintenance is the thing that gets forgotten
In my experience, the most overlooked issue in backup power isn’t specification. When a system is designed, the rating, the coverage and the transfer time usually get the right attention. Instead, the problem comes later. Once a generator is sited and commissioned, it’s too easily forgotten, and the maintenance routine slips or is not delivered in a robust way.
This presents a false economy, because a generator is only ever going to be as good as the infrastructure around it – the cables, connections, fuel and fuel tanks – and only as reliable as its upkeep. If starter batteries aren’t maintained, or fuel quality isn’t monitored, or if any of several other elements aren’t carefully reviewed, you can end up with a generator that fails to start.
Good testing means more than a weekly walk-round to check for leaks and obstructions, as valuable as that is. It means running the set on load, regularly, to prove it can deliver stable output under real conditions. Checking that the engine fires is simply not enough. And it means a longer full-load test once a year to confirm the set still meets its rating.
The real cost is the recovery
Sadly, the direct outage is rarely the main event. As the rail disruption showed, a very short interruption can displace complex, interdependent operations and trigger days of knock-on chaos, with the commercial, operational and reputational damage far outweighing the seconds of power lost, or the costs involved with maintaining plant adequately.
The crucial point here is that getting backup right isn’t a box-tick exercise. It’s a question of specification, maintenance and testing, sustained over the life of the asset. The value of standby power is measured in the interruption it avoids.
This article appeared in the September 2026 issue of Energy Manager magazine. Subscribe here.
