Domestic consumer unit rating with PV and battery storage.

This discussion is interesting but seems to be a little shy of the actual characteristics of the equipment. Battery inverters cannot supply significant excess current, and have protection built in to prevent this situation which otherwise would rapidly cause electronic failure. Therefore the short circuit condition is not as described, as the inverter MCB can probably never trip, the inverter will simply supply the rated current an very low voltage, or more likely shut down altogether. Therefore the mains MCB will trip too due to the excess current and the supply will not dissipate significant power in the cable, simply rated for the inverter output current.

In a recent discussion, the problem of connections as shown in Andy's second drawing, was discussed but is probably unsatisfactory due to a different problem. In order to provide power back to the mains, although OK in Island mode, a significantly higher voltage is required to overcome any voltage drop in the connecting cables. The solution is that local generation should connect DIRECTLY to the consumer side of the meter via a dedicated circuit, not as drawing 2 on a local final circuit.

The lack of RCD protection at this point is then not a problem either, assuming the cable is installed or chosen to make this unnecessary, such as SWA if buried or perhaps in steel conduit. Note that this arrangement also prevents any kind of overload situation of the connecting cable, so the "difficulties" with 536 etc are avoided. Over-current protection at the inverter is not required and "adding" of the protective device current is not required.

If the Solar PV inverter and battery inverter are mounted 10 m away from the consumer unit and the total load of the installation is being supplied by Solar PV and battery, how is the cable to the consumer unit protected for high impedance fault current. Eg where buried<50mm PVC/PVC cable would require 30mA RCD.

An RCD at the grid-end of the circuit will still detect & trip on earth faults even if all the current is provided by the on-site generator - why? because when running in parallel with the grid the on-site generator must not have its own N-PE link - the only path for any earth fault current is via the grid supply's N-PE link - and so the fault current will pass through the grid-side RCD (if only in it's N pole to return to the on-site generator's N). That won't disconnect the generator from the fault of course, but provided the RCD opens N as well as L(s) (as required 551.7.1(ii)) then the on-site generator becomes isolated from Earth (again because it doesn't have its own N-PE link) in effect becoming a separated or IT circuit, so the shock path is broken - so the danger is removed. No need for a 2nd RCD at the generator end (it wouldn't work anyway for faults (other than on the generator itself) as there's no N-PE link on the generator.

   - Andy.

Island mode?

Thank you.

Having now appreciated that when the inverters are generating in parallel they are operating in separated or IT mode it makes more sense. Your explanation is a great help.

Sparkingchip said:

 RCD and over current protection at both ends of the circuit.

Yes ... and as David says, the arrangement has no load-shedding to prevent large loads being connected.

The diagram is only to illustrate the principles of Reg 551.7.2

David Stone said:

Battery inverters cannot supply significant excess current, and have protection built in to prevent this situation which otherwise would rapidly cause electronic failure.

Whilst this is technically correct, in practice if the inverter manufacturer fails to provide sufficient information to permit compliance with Regulation 419.2 an RCD is often necessary for ADS. This is required at the inverter itself, but it need not be a 30 mA RCD (and perhaps may be time delay depending in disconnection times) unless additional protection is required for other reasons (e.g. cables concealed in building fabric < 50 mm without suitable protection).

Mspark said:

Having now appreciated that when the inverters are generating in parallel they are operating in separated or IT mode

No they are not, operating IT mode, they are either connected to the grid (in which case the grid earthing arrangement applies) or they ought to be connected in TN-S arrangement. IT is not recommended for island mode, and Chapter 82 in AMD2 requires a neutral to be formed by earthing after disconnecting the live conductors to the grid.

One of the vehicles as storage examples in the IET COP Electric Vehicle Charging Equipment Installation 4th edition with RCD protection at both ends of the EVSE and storage batteries circuits. 

shop.theiet.org/code-of-practice-for-electric-vehicle-charging-equipment-installation-4th-edition

Island mode?

Yes in island mode the on-site generator would have to behave in the same way as the grid supply - i.e. provide a N-PE link and any RCDs would have to be positioned after the generator - but that's a different kettle of fish again from 551.7 as it's no longer in parallel with 'other sources'.

I guess if you had multiple on-site generators that could run in various combinations in island mode, things start to get even more complicated and but still you'd need to arrange things to have exactly on N-PE link active at any one time and ensure that earth fault currents do pass through appropriate RCDs. Where an EV charge point as been TT'd and is then generating, in island mode the change point would become TN-S from its local electrode and the main installation then becomes TT (if with fortuitous connections to the grid PME Earth), so it can get much more "interesting".

  - Andy.

Having now appreciated that when the inverters are generating in parallel they are operating in separated or IT mode

Probably more accurate to think of them connected in the same way as an ordinary load (if one with -ve power consumption) - i.e. form part of the same system to which they're connected (usually TN or TT). Strictly speaking they only operate as a separated or IT system in the rare event that the RCD trips (and then just until the inverter senses grid loss and shuts down).

   - Andy.