Multi RS Solar – disconnection time on a bolted L-PE fault in island mode (TN-S)?

Been wrestling with something similar on my boat install and figured it was worth raising here as I reckon a few others will hit this.

Running a Multi RS Solar in island mode (off-grid, no grid connection at all) and trying to get my head around disconnection times for a bolted L-PE fault. The Victron datasheet is frustratingly vague on this — it'll tell you plenty about the inverter's output characteristics but goes quiet when you actually need the protective disconnection figures for a TN-S arrangement.

On a boat it's slightly different territory to a house, but the core problem's the same: if you've got a fault, how fast does the Multi RS actually disconnect? And does it meet the 0.4s requirement you'd normally expect at 230V in a TN-S system?

My suspicion is the internal protection relies heavily on the upstream MCBs and RCDs doing the work rather than the inverter itself acting as the "source" in any meaningful protective sense. But I've not been able to confirm that with anything official from Victron.

Anyone actually dug into this properly? Particularly interested if someone's done the Ze/Zs calculations treating the inverter output impedance as the source — because that impedance figure is also not easy to find.

Fogstar/Victron combo here on 48V if it matters. Would love to know if anyone's managed to get a proper compliance doc together covering this bit. Victron UK support been helpful to anyone on this?

@LDVWanderer good question, and one I've been quietly digging into myself for the static van setup.

The short answer is: the Multi RS won't clear a bolted L-PE fault quickly in island mode — it's inverter-limited, not grid-backed. Fault current is typically capped around 2× rated output, so you're nowhere near what's needed to trip a standard MCB rapidly.

What I ended up doing:

• Replacing MCBs with RCDs or RCBOs throughout — fault detection shifts from overcurrent to earth leakage, which the inverter can sustain long enough to trip
• Checking Victron's own ESS docs — they're candid that TN-S fault clearance in island mode depends heavily on your protection device choices

Worth looking at Victron's Community forum too; there's a thread where one of their engineers confirms the disconnection time issue directly. Not a criticism of the kit, just a real limitation to design around.

OllieGraham91 | 47 posts | ⚡ Solar Enthusiast

@LDVWanderer interesting one, and relevant to my shed setup too. Worth flagging that in island mode the Multi RS acts as the supply source, so your disconnection time requirements come from BS 7671 Table 41.1 — 0.4s for a 230V TN circuit feeding socket outlets or equipment up to 32A.

The sticking point is that fault loop impedance is dictated by the inverter's output impedance, which is considerably higher than a grid supply. That typically means your Zs figures won't support MCB operation within 0.4s on longer cable runs. RCDs become your practical solution there rather than relying purely on overcurrent devices. Have you measured your actual output impedance under load yet? That'll tell you a lot about what protection scheme is actually achievable.

What tripped me up on my cabin install was realising the Multi RS Solar doesn't behave like a traditional transformer-based inverter here. In island mode with a TN-S arrangement, your L-PE fault loop impedance depends entirely on the inverter's own output impedance — and the Multi RS Solar's high-frequency topology means that impedance can be surprisingly high, potentially limiting fault current enough that a standard MCB won't clear within the 0.4s requirement under BS 7671.

Victron's own documentation is a bit quiet on this, which doesn't help. Worth pulling the actual prospective short-circuit current from the inverter specs and checking your protective device curves against it. I ended up oversizing my busbars and dropping to a lower-rated RCBO as the primary protection layer rather than relying purely on overcurrent disconnection.

@DucatoProject curious what conclusion you reached — did you go the RCD route in the end?

Great point from @ForestJenny about the transformer-based behaviour difference — that caught me out too on my static van.

Worth flagging: in island mode the Multi RS Solar's virtual neutral means your L-PE fault loop impedance is essentially determined by the inverter's own output impedance, not some nice low-Z transformer winding. So your MCB/RCD has to operate against that, and disconnection times can be... optimistic shall we say.

Victron's own docs are a bit vague on guaranteed disconnection times under BS 7671 Table 41.1. I ended up going 30mA RCD as first line of defence rather than relying on overcurrent devices alone. Basically treating it like a TT system in my head even if it's notionally TN-S.

Not saying that's the correct answer — just what let me sleep at night 😅