Every year, in the first weeks of September, the same question comes up in maintenance departments and technical management offices: is the system ready to return to full operation? The answer is usually sought in the wrong place.
Attention goes to the summer heat, to record temperatures, to August's energy consumption reports. But for an uninterruptible power supply (UPS) system, the moment of highest risk isn't the summer heat peak. It's the day of the restart.
Why restart is more critical than summer heat
During a period of reduced operation, a summer shutdown, a production stop, reduced shift schedules, a UPS system doesn't simply stop working. It keeps powering control circuits, but stops being stressed the way it is during the rest of the year. It's precisely in this apparently quiet phase that the most insidious risk factors build up, because they stay silent:
- Batteries sit at partial charge. Without regular discharge and recharge cycles, their actual capacity can drift from nameplate values without triggering any alarm.
- Fans and filters stay idle longer than usual. Dust, humidity and performance degradation accumulate without the airflow that normal operation would provide.
- Temperature sensors don't register drift. A cabinet running at low load doesn't produce the same thermal profile as one at full load: cooling anomalies remain invisible until the load rises again.
None of these factors, on its own, causes a failure. But they add up, and they stay hidden until the moment the system is called on to deliver full load.
What actually happens, technically, at restart
When production resumes, loads rarely switch on in a controlled sequence: production lines, machinery and auxiliary systems tend to restart within a narrow time window. This generates an inrush current peak that can significantly exceed the average nominal load the system was originally sized for.
A correctly sized UPS handles the average operating load without issue across normal months of operation. That doesn't guarantee it will handle a simultaneous restart of every connected load after weeks of downtime — especially when accumulated risk factors (batteries, ventilation, thermal drift) combine with a sizing approach based only on average load rather than transient peaks.
It's the combination of these two elements, latent risk factors and a sudden demand spike, that makes restart day the real stress test for the system.
5 checks before returning to full load
Before bringing a system back to full load, a few hours of targeted verification are worth the investment. Five checks, in particular, catch most problems before they surface as unplanned downtime:
- Cabinet internal temperature. Even a small deviation from expected values, accumulated over the summer months, can significantly shorten battery service life.
- Battery state of charge and internal resistance. Checking that batteries are charged isn't enough: internal resistance is the most reliable indicator of their actual condition, and is typically the origin of most battery-related UPS failures.
- Alarm logs accumulated during the reduced-operation period. These need to be reviewed, not just archived: an intermittent alarm ignored during downtime can be the first sign of a problem that will worsen under full load.
- Transfer test under real load, not just the unit's automatic self-test, which verifies switching logic but doesn't always reproduce the conditions of an actual transfer under load.
- Actual maintenance frequency versus the planned schedule. Extended downtime is often also when scheduled maintenance slips: before restart is the time to catch up on any deferred interventions.
Signs to watch in the following weeks
Even with every preventive check done correctly, some effects of the restart only surface after the system has been running at full load for a while. About two to three weeks after returning to full operation, it's worth paying attention to:
- Intermittent alarms in the log, appearing after the shutdown and often dismissed as "already resolved on their own" — when in fact they warrant investigation, not archiving.
- Batteries that don't return to full capacity after the discharge-recharge cycle triggered by the restart: often an early sign of degradation.
- Thermal drift that reappears under sustained load, even when the initial test, run at partial load, passed without issue.
- Abnormal behavior during a routine transfer test, compared with performance recorded before the summer break.
A targeted check three weeks after restart costs far less, in time and resources, than an unplanned shutdown in autumn — when production is running at full capacity and an interruption has a much greater impact.
A moment to manage, not just get through
A full-load restart isn't a simple switch to flip back on: it's the moment a system reveals whether the checks carried out — or skipped — during downtime actually held up. Treating it with the same attention given to routine maintenance — before, during, and in the weeks that follow — is often what separates a system that handles the restart from one that fails right when it matters most.
Powertronix supports companies and design engineers in verifying and sizing uninterruptible power supply systems, from a single pre-restart check to the design of new installations. To have your system checked before returning to full load, get in touch — replace with the correct link to the website's contact page.