Omission of fault (and overload) protection for inverter

[happysteve]

I don't "do" solar, but I've been asked to comment on an aspect of a new domestic solar installation by others.

There is no overload or fault protection, except for the service cut-out fuse (unknown size, assume 80A). TN-C-S.

Installation is as follows (have only seen a photo of part of it, not seen in person):

80A (assumed) cut-out fuse -> energy meter -> Henley blocks

Henley blocks -> main DB (no problem)

Henley blocks -> solar meter (16mm or 25mm tails) -> 20A 3 phase rotary isolator (L1 and N terminals used only) in ~2.5mm single-insulated cables just underneath -> 2.5mm T&E -> inverter. There is ~10mm G/Y from (presumably) the MET to the rotary isolator, connected to the 1.5mm cpc of the T&E with a terminal block inside the isolator enclosure.

I don't know how far the inverter is from the isolator, or the route the cable takes.

To me, this stinks.

Regarding fault protection, none of the provisions of 410.3.9 apply, so fault protection is needed.

The adiabatic suggests a minimum of about 4mm (slightly higher) for protective conductor, making some assumptions about fault currents, fuse rating etc.

I can't think of a reason why overload protection could be omitted, but maybe this is due to the nature of it being connected to an inverter, which probably has its own internal way of limiting current to <=16A?

In any case, if a fault occurred anywhere between the solar meter and the inverter, then the cable would melt/catch fire long before the service fuse blew, right?

I just can't understand how someone can install this, and when questioned, assure the customer it's safe? Or am I missing something really obvious?

Appreciate your thoughts.

 

[littlespark]

Every PV I’ve seen has had an OCPD. Methinks there’s been a cock up somewhere.

I’m not an installer either, but I’ve seen plenty in houses I’ve worked in.

 

[brianmoooore]

Electricity can flow both ways in a cable! You have 2.5mm2 T&E protected only by the service cut out.

 

[happysteve]

Electricity can flow both ways in a cable! You have 2.5mm2 T&E protected only by the service cut out.

Well, yes, exactly. Hence discussion about lack of fault protection. Not sure if overload protection can be ommitted or not, as per 433.3.1(ii). But even this indent mentions the requirement for fault protection, even if overload protection is not needed (and I don't know enough about inverters to be able to say whether oveload protection is needed or not...)

 

[brianmoooore]

I don't know a lot about inverter either, although I'm fairly sure that recent models have electronic current limiting, and won't self destruct if the PV input exceeds what the inverter is designed for, but the inverter is irrelevant when it comes to fault protection for the cable.
If the cable shorts out, the current is going to come from the grid, not the inverter, unless we're talking about large inverters*, in which case they won't be connected by 2.5mm2 T&E.

*Even these will shut down when they detect the loss of grid voltage.

 

[littlespark]

Actually… I don’t know if things have changed, but my own system, and many I’ve seen domestically on the “just below 4kW” are protected by a 16A, and wired in 4.0mm, not 2.5….

Often wondered why the cable was oversized.

 

[plugsandsparks]

Connecting the inverter directly to the service fuse does appear optimistic, i would not like the inverter located in the house if that is all that is protecting it. When solar is producing, yes the inverter has lots of internal mechanisms to protect itself however if the inverter itself develops a fault which draws current from the grid, it needs that MCB to trip before stuff catches fire. The size of the MCB is in the MIs for the inverter. Unless its a very large inverter i suspect the 80A service fuse may be a little high.

 

[brianmoooore]

The cable needs protecting the same as if a load was connected on the far end of it, rather than a generator, except that it must not share a RCD with other circuits, such as PV connecting via a MCB to a split RCD board. This is because in the event of an earth fault on one of the other circuits, the inverter will continue to provide power for approaching half a second after the RCD has cut the supply.
A PV system might have an apparently oversized cable for voltage drop reasons. 1%for PV?