The PFC will be controlled much more by the inverter than the battery.  LiIon batteries can produce huge currents if you're careless enough to short circuit them.  But inverters are limited to a few kilowatts (3.6kW is common) and they aren't going to produce more current than the inverter's electronics can generate.

I agree - PFC will be 2/3rds of bu**er all from a typical inverter - about 2x rated output would be generous. As long as you have a normal TN-S arrangement the local earth rod won't affect the PEFC (although it will if you have a TT island hanging off the installation - e.g. for a detached garage). As with most embedded generation (domestic sized anyway) you'd be looking at RCDs for ADS (although in practice the output voltage may just collapse anyway in the face of a short circuit).

There is some guidance available - I'm sure Graham can point you to the latest IET book on the subject..

   - Andy.

OK, let's try for an example. A random example (Givenergy) suggest 2600W output (~ 11.3A) with and this particular one, unusually it seems to me, does quote maximum fault current output 40A. It tells us nothing about for how long this fault current might  be able to be produced. - Hence the only safe thing is to presumably not to rely on that being a limiting factor and assume for "as long as it takes" .

Let's suggest we aim to run a lighting system on it using B6 breaker. At 40A this would break in ~ 20 seconds, and by Adiabatic, require in excess of 1.5mm conductors. - Fair enough.

Another random brand (Solis) provides no maximum current at all but just says in the data "output short circuit protection - YES."

Another (Solax this time) 5KW output rated one simply says "Maximum AC output Current" 23.9A. - (Presumably a steady state operating current), again without any figure for short circuit max or short circuit time before internal protection kicks in. 

A similar problem arises with small generators -there is not enough 'oomph' to operate over-current ADS at anything more than a small fraction of the full load output.

As already noted, the trick is to use RCD or RCBO to do the safety of life type ADS with LE amd NE faults, and accept that dead short LN fault is not really going to be detected and  may not trip anything except the inverter internals may collapse after some un-specfied time.

Mike

Thank you mapj1, that was really my point.. I have always wondered this for small generators generally, something till now I have not needed to dirty my hands with. "The inverter internals may collapse after some unspecified time" .. Yes it must do at some point.   

"In all cases, compliance with Chapter 43 SHALL be fulfilled.." so we're good to go, shame there's no data. - Sign it off!

justinneedham said:

Pulling up data sheets for numerous hybrid inverter models, I can't find any data for any of them relating to inverter PFC during "islanding mode" when the emergency loads are supplied directly, during network power failures, or presumably during planned periods. 

As a rule of thumb, inverters current-limit at around 1.2 times their nominal current rating. It's no good trying to measure this with a loop tester.

Also worth considering, as it's an international market, that there are a number of products out there, where the "backup power" is a floating supply (not recommended in the IET Code of Practice, not permitted for domestic installations according to the MCS standard), or otherwise might not comply with switching arrangements necessary to comply with the Electricity Safety, Quality and Continuity Regulations (Regulation 21) as well as BS 7671.

Let's suggest we aim to run a lighting system on it using B6 breaker. At 40A this would break in ~ 20 seconds, and by Adiabatic, require in excess of 1.5mm conductors. - Fair enough.

The adiabatic starts to get silly much beyond 5s (as the heat loss from the cable starts to become significant) - a simpler approach is just to select a device that provides overload protection (e.g. In ≤ Iz) and has sufficient breaking capacity for it's position in the installation (hardly a challenge for small generators) - that's then deemed to provide fault protection in most circumstances (435.1). For MCBs fault protection (between live conductors) is provided by the thermal rather than magnetic element, and the time to disconnect isn't particularly well defined (and doesn't have to be). For L-PE faults (for ADS), RCDs disconnects plenty quick enough to protect the conductors, given the low fault currents.

   - Andy.