Dual Victron MultiPlus Setup with Separate Batteries on a Shared AC Bus

Running two Victron MultiPlus 48/5000 units on a shared AC bus without parallel mode — genuinely surprised this doesn't come up more on here, because I nearly went down this exact rabbit hole with my van-to-static conversion project last year.

The idea being each inverter feeds its own battery bank but they're both chucking power onto the same AC ring, right? Sounds elegant on paper until you start thinking about what happens when one bank is gasping and the other's full — do they start arguing like two blokes over the last sausage roll?

My concern would be load balancing. Without proper coordination (and without VE.Bus linking them in parallel), you're essentially hoping the loads distribute themselves sensibly. They won't. One MultiPlus will thrash itself silly while the other has a nice sit-down.

Has anyone actually done this with separate Fogstar or similar LiFePO4 banks? I'm curious whether using a Victron Cerbo GX with some clever Node-RED automation could manage the handoff — basically scripting priority switching based on state of charge rather than letting the AC bus sort itself out.

Also wondering if the 12 kW PV side complicates things further — are you running separate MPPT controllers per inverter, or one big array feeding both?

Seems like the kind of setup that works brilliantly right up until it absolutely doesn't at 2am in February.

Would love to hear if anyone's actually commissioned something like this in the UK — especially with grid-assist considerations under G98/G99 thrown into the mix.

@AnneButler this is basically what I ended up doing almost by accident with my static caravan setup. Had one MultiPlus already running a 48V Fogstar Drift bank, then added a second unit months later when I expanded. Rather than go full parallel (which terrified me honestly — the synchronisation requirements are brutal if you get it wrong), I just kept them on the same AC bus feeding different circuits via a distribution board.

The key thing nobody warns you about: make sure your AC bus can handle both units potentially dumping full output simultaneously. I learned that the slightly awkward way when I tripped a busbar connection during a particularly stormy week in Wales.

VictronConnect still monitors both independently, which is actually rather elegant — you get granular data without the complexity of ESS parallel configuration.

Done this in my shepherd's hut — two MultiPlus units, separate battery banks, shared AC bus, and the key thing nobody mentions is AC input priority sequencing or you'll have them fighting each other like two blokes over the last pasty.

Each unit needs its own VE.Bus config treated as a standalone, not parallel — the moment you let VEConfigure assume they're aware of each other, chaos ensues.

The shared AC bus works fine provided you're disciplined about which unit owns grid input — give that job to exactly one of them, full stop.

Really glad this thread exists — I've been running a similar arrangement in my static caravan for about two years now and the bit that caught me out initially was AC input priority conflicts when both units are connected to grid simultaneously.

Worth being explicit: each MultiPlus needs its own VE.Bus configuration profile treating it as a standalone unit, not set to parallel or three-phase. The shared AC bus then becomes your load distribution point.

The critical thing I'd add beyond what @SaltyTrekker mentioned — set your AC output voltage tolerances tightly (within ±2V ideally) before connecting them to a shared bus, otherwise you'll get circulating currents that'll confuse both units and potentially trip protection.

I used Victron's VEConfigure tool to fine-tune this. Also worth logging both units via Cerbo GX separately so you can actually see which bank is doing the heavy lifting.

Running this exact setup in my garden office — two MultiPlus 48/3000 units, separate Fogstar Drift 48V batteries, shared AC bus. Took me an embarrassingly long time to convince myself Victron hadn't made a terrible mistake allowing this.

The moment of panic I haven't seen mentioned: first time one unit went into absorption and the other