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

DIY
,
1

Hello,

I’m putting together the compliance file for a Multi RS Solar 48/6000/100-450/100 (PMR482602020), and I’m stuck on one point the datasheet doesn’t cover.

Setup

ESS, zero feed-in. The whole house is backed up through AC-out-1.
In island mode, the internal ground relay bonds N to PE, so the installation runs as TN-S.
Downstream of AC-out-1: a 50 A MCB (C curve), then about 3 m of 16 mm² cable to a distribution board. Every final circuit there is protected by a 30 mA RCD.

The issue

The datasheet gives a 45 A output short-circuit current. That is far below the magnetic trip threshold of any MCB in the installation (a C50 needs 250–500 A). So in island mode, for a phase-to-earth fault on the distribution cable upstream of the RCDs, automatic disconnection relies entirely on the inverter’s own short-circuit protection. IEC 60364-4-41 requires disconnection within 5 s on distribution circuits in TN systems. The datasheet lists “output short circuit” among the protections, but gives no timing.

Questions

Section 6.2.1 of the manual describes the overload behaviour: output voltage is reduced to limit current, then shutdown, a 30 s wait, three restart attempts, then lockout. Does the same sequence apply to a bolted short-circuit (L-N or L-PE) on AC-out-1 in island mode? If so, how long does the current-limiting phase last before the first shutdown?
Does the 30 mA ground fault detection listed in the specs act only on the PV side, or does it also monitor the AC output in island mode?

If the answer to 1 is reliably under 5 s, an extra 300 mA time-delayed RCD at the head of the distribution board isn’t strictly required. Without a figure, I’ll add one as a precaution.

Thanks in advance.

2

What loop impedance was measured at your PE?

3

Peak power duration so less than 3 seconds. (Source - personal experience)

4

That’s why I always use a selective 100 mA RCD directly at the Multis, i.e. where the three cables come together. This also has the advantage that the disconnection conditions are met for any loads connected directly without an RCD. I don’t fit a fuse on the ACout side if, when operating in parallel with the grid, the maximum current does not exceed the current-carrying capacity of the cable on ACout. As you’ve already mentioned, with a C50 the thermal trip (1.45 × In) will never be reached at 45 A either.

For UPS systems, there are further requirements in Appendix D2 of 0100-410. However, Victron doesn’t seem to document these as such.

I also tried to get to the bottom of it, but gave up and now always fit a selective RCD directly at ACout.

So there’s someone else here who’d be interested to hear what your research turns up :wink: