Narrowboat 12v system keeps browning out when the inverter kicks in — what am I missing?

Been living aboard a 70ft narrowboat for about eight months now and I'm pulling my hair out with this one. Got a 400Ah lithium bank (four 100Ah Winston cells in parallel), a Victron MultiPlus 12/3000 inverter/charger, and a 600W solar array on the roof going through an MPPT 100/30. Most of the time it all works a treat, but whenever the inverter fires up something big — kettle, microwave, that sort of thing — the 12v lighting and 12v fridge flicker or drop out entirely for a second or two.

The voltage doesn't appear to tank massively — I've got a Victron BMV-712 and the lowest I've seen it dip is around 12.1v under that kind of load, which shouldn't be enough to cause issues. My cabling from the battery to the inverter is 70mm² and about 1.2 metres each way, so I don't think it's a resistance problem there. The fuse is a 300A MIDI on the positive run.

What's puzzling me is that the 12v fridge and lights are on a completely separate circuit — they're running off a separate fuse board that tees off the busbars, not directly off the inverter. So I'd expect them to be somewhat isolated from the inverter's demand. Could it be a busbar connection issue, a dodgy cell, or something weird with the BMS cutting in and briefly limiting current?

Has anyone on here dealt with something similar on a boat install? Really keen to understand whether this is a wiring/connection fault or something more fundamental with the battery setup before I start pulling things apart.

SteveGreen74 | 847 posts | ⭐ Trusted Member

@Scouse16 Classic symptom of undersized cabling between your battery bank and the MultiPlus. A 3000W inverter at 12V is pulling 250+ amps at peak, and that's before you factor in startup surge. Even a short run of undersized cable will cause enough voltage drop to trigger the inverter's low voltage cutoff.

What cable size are you running between the bank and the MultiPlus? Victron actually recommend 95mm² minimum for that unit, and many installers go to 120mm². Also worth checking your busbar connections and fuse holder — those can introduce serious resistance that doesn't show up visually.

Secondary question — are your four Winston cells properly balanced? Parallel lithium banks can mask a weak cell until you throw a big load at it. Worth checking individual cell voltages under load.

Kingy | 1,203 posts | ⭐ Trusted Member

Been there, done that, nearly sank the boat (not literally, but it felt that way).

@SteveGreen74 is bang on about cabling, but there's another culprit nobody mentions — your busbar connections. I spent three weeks chasing brownouts on my 58ft trad before realising one of my busbar bolts had worked loose from all the vibration underway. Torque those down properly and apply some copper grease while you're at it.

Also worth checking: what's your battery BMS doing when the MultiPlus ramps up? Those Winston cells can be brilliant but their BMS units sometimes have a nasty habit of momentarily throttling output under surge demand.

Grab a Victron battery monitor if you haven't already — watching the voltage in real-time when the inverter kicks in will tell you exactly where the problem lives.

TerryLewis | 412 posts

@Scouse16 One thing nobody's mentioned yet — check your battery interconnects between those four cells. Parallel connections are only as good as their weakest link, and if even one of those cell-to-cell cables is undersized or has a dodgy crimp, you'll see voltage sag the moment the MultiPlus demands serious current. The 3000VA unit can pull 250A+ at 12v under load, and Winston cells can deliver it fine, but only if every connection in the chain can handle it. Grab a cheap clamp meter and measure voltage at the battery terminals versus at the MultiPlus input while the inverter kicks in. Any meaningful difference there tells you exactly where your problem lives. Worth doing before you rewire anything.