Has anyone actually tested their backup power setup during a real outage — not just a drill?

We had a proper grid failure here in rural Shropshire last month — about 11 hours overnight in January. First time our setup has actually been tested under real conditions rather than me just switching the inverter on in the garden and feeling smug about it. Spoiler: it didn't go perfectly.

We're running a 3kWh lithium (Fogstar Drift 200Ah 12V) with a 2kW Victron Multiplus and around 600W of panels on the roof. In summer that would've been fine, but we started the outage at about 60% SOC, it was pitch black by 5pm, and the heating (oil boiler, 12V pump) plus the fridge just quietly drained us. By 4am we were at 12% and I was rationing everything like it was the apocalypse. Ended up running our little Honda EU22i generator for three hours to top up — which I hadn't actually properly tested in about eight months. It started second pull, thankfully.

The whole thing made me realise I'd been planning around best-case scenarios. Has anyone else had their system stress-tested by an actual unplanned outage? What caught you off guard, and what did you change afterwards? Curious whether people keep a petrol generator as a hard backup or whether anyone's gone all-in on battery capacity alone.

@RayHunt — aye, the motorhome lifestyle basically means every night is an unplanned grid outage, so I've had plenty of involuntary stress-testing over the years.

Most revealing moment was a February stopover in the Cairngorms. Minus eight outside, diesel heater running all night, fridge cycling, phone charging. My Fogstar 200Ah lithium and a pair of Victron components handled it without complaint — but the weak link turned out to be a dodgy Anderson connector I'd half-noticed and ignored for months. Thermal stress exposed it immediately.

That's the real lesson I'd take from your Shropshire outage — it's never the main kit that fails, it's the connectors, the fuses, the cable runs you bodged at 11pm on a Tuesday.

Pressure-testing a system in anger reveals the compromises you made during the build. Sounds like yours held up though?

Really interesting thread, this one. We had an unplanned 7-hour outage back in November — transformer fault about two miles away — and honestly it exposed a few gaps I wasn't expecting. The system worked, but I'd never actually tested the automatic changeover relay under load before, and there was about a 4-second delay before it kicked in that I hadn't accounted for. Lost a NAS drive because of it. Also discovered my battery monitoring display needs mains power to run, which felt spectacularly stupid in hindsight.

@EwanMurray is right that real-world conditions teach you things no drill ever will — cold temperatures hit my LiFePO4 capacity noticeably harder than I'd modelled.

The 11 hours overnight in January sounds like a proper stress test @RayHunt — curious how your heating situation held up? That's usually where the watts really start disappearing.

Really reassuring to see others' setups holding up under actual pressure rather than just bench testing! We had a nasty 14-hour outage last February — ice storm took out lines across a good chunk of Hertfordshire.

The thing that caught me out wasn't the batteries or the inverter, it was the heating controls. My Worcester boiler needs mains power to fire up, so even though I had lighting and the fridge sorted, we were relying on the wood burner through the night. Lesson learned — I've since added a small dedicated UPS just for the boiler and thermostat circuit.

@RayHunt curious whether your heating was covered in your setup? That seems to be the weak link most people don't think about until it's too late. @Paddy_2634 same question for you — transformer faults in November must have made warmth a priority!

Really timely thread @RayHunt. We had a 14-hour outage here in mid-Wales last February — storm damage took out a substation. First proper real-world test of our LiFePO4 bank and it threw up something I hadn't anticipated: the inverter kept tripping because our old chest freezer has a massive startup surge. Never showed up in my testing because I'd always manually staged things. Ended up running the freezer directly off a small dedicated AGM I keep as a backup to the backup, which felt embarrassingly bodged at the time.

Lesson learned — stress testing individual loads under controlled conditions doesn't replicate the chaos of everything switching on at once at 2am when you're half asleep. I'd strongly recommend anyone with an inverter setup deliberately tries running their full realistic load simultaneously, not just ticking items off a list individually.