Has anyone actually tested their backup power setup during a real grid outage — not just in the garden?

Last week we had a proper grid outage here in rural Somerset — about 11 hours overnight into the following morning. First time I've genuinely relied on the setup rather than just tinkering with it. Running a 200Ah LiFePO4 bank fed by a 400W roof array, with a Victron MultiPlus-II 3000 handling the inverter/charger duties. Honestly thought I was better prepared than I was.

Biggest surprise was the fridge. I'd mentally budgeted about 60W average for it but it was clearly pulling more during the compressor cycles — by 6am the bank was down to 40% and the solar hadn't kicked in yet because of cloud cover. Nothing dangerous, but it was a reminder that theoretical numbers and real-world behaviour are two different things. I also hadn't pre-charged the bank fully before bed, which was a rookie error on my part.

Curious whether anyone else has been through a genuine unplanned outage and found gaps in their setup they hadn't anticipated. Did anything fail or underperform that you thought was solid? And for those running similar Victron kit — are you using the ESS assistant or just manual switching for grid backup?

KentOffGrid | 847 posts | ⚡ Solar + LiFePO4

@QG_Marine brilliant timing on this thread — there's nothing quite like a real outage to expose the gaps in your planning, is there?

We had a nasty one here in Kent about 18 months back, seven hours during a January cold snap. Honestly, the system performed better than I expected technically, but I was the weak link — kept second-guessing the state of charge readings and being overly conservative with loads out of habit.

The thing that genuinely surprised me was how much mental energy the monitoring consumed. I was checking the BMS display every twenty minutes like a nervous parent.

Afterwards I set proper low-voltage alerts so the system tells me when to worry rather than the other way round.

What was your biggest lesson from it? Curious whether the overnight element caused complications with solar recovery.

T5Life | 1,203 posts | ⚡ Solar + LiFePO4 + Wind

@QG_Marine yes, had a similar baptism of fire last winter — nine hours out in North Yorkshire during a storm. What surprised me most wasn't the power side, it was discovering which loads I'd completely forgotten to account for. The broadband router was fine, but the boiler's controller draws far more than I'd estimated on standby, and that caught me out.

The other thing nobody warns you about is how differently you behave when you know you're genuinely on reserves versus just testing. You become very conscious of every switch you flip. Honestly taught me more about our actual consumption patterns than six months of monitoring software ever did.

How did your battery state look by morning? Curious whether you had headroom left or were watching the percentages nervously.

ValleyTony | 412 posts | ⚡ Solar + Batteries

My shepherd's hut is essentially always off-grid so every night is a real test, but I did have family staying during the Beast from the East follow-up last February — proper stress test with extra bodies and higher consumption than usual.

The weak link turned out to be my inverter behaviour under sustained load rather than battery capacity. Victron MultiPlus handled it fine once I'd tweaked the transfer switch settings, but I'd never have spotted that flaw just pottering about in the garden on a sunny afternoon.

@QG_Marine 11 hours overnight is a decent shakedown — what did your overnight draw actually come to? I'm curious whether heating was in the mix because that's where most people's assumptions fall apart in practice.

CliffSpirit | 634 posts | ⚡ Solar + LiFePO4 + Wind + Genny backup

@QG_Marine great thread — proper real-world testing is worth more than a hundred garden trials. We had a 14-hour outage here in Northumberland back in February and it exposed one thing I hadn't anticipated: my inverter's low-temperature cutoff kicking in because the battery shed dropped below 5°C overnight. Batteries were at 60% but the system effectively locked itself out until things warmed up. Completely missed that during summer testing.

Since then I've added basic insulation around the battery enclosure and a small thermostatically controlled heat mat — cheap fix for an embarrassing oversight.

Worth asking — @QG_Marine did you notice anything surprising about your actual consumption versus what you'd estimated? I found mine was significantly higher than expected once the chest freezer started cycling at 3am.