I have a fully off grid set up, I have wired up the house and a solar company has done the Solar panels, battery and inverter. House is wired the same as you would if it was grid tied, except supplied by a sub board off the inverter.
Although it does pass RCD tests, it does not pass the earth loop impedance test and if I go off the regs, I would need to either change the C/B to B type or fuses. (it currently has c type breakers).
It is my understanding this still satisfies the regs but would it be better to change them anyway? (for piece of mind) Changing out the C/B for b type would be over $1K and fuses around $500
The solar guy says the inverter would shutdown anyway with a fault, which is probably true. I just have some concerns.
Any thoughts/advice would be appreciated.
Best Practice earth loop impedance B type C/B or fuses
Re: Best Practice earth loop impedance B type C/B or fuses
Perhaps you don’t need to verify the EFL on RCD protected circuits anyway?
8.3.9 in 3000 indicates you only need to verify EFL for non RCD protected socket-outlets. I think it’s because the RCD will operate within 300ms, which is less than the 400ms requirement for automatic disconnection/fault protection.
Not quite sure about non RCD protected circuits or P-N short circuits though?
Clause 7.3.5.2.1 is also worth a read especially where it says “A circuit is considered to be protected against prospective short-circuit and earth fault currents when it is supplied from an electricity generation system incapable of delivering a current exceeding the current carrying capacity of the circuit”. Guess it’s saying your cable size etc within the circuit effectively becomes the circuit protection. Not sure if this would apply to a typical off grid house though, I imagine the inverter would be too big.
This topic confuses me a bit too, interested to hear other opinions.
8.3.9 in 3000 indicates you only need to verify EFL for non RCD protected socket-outlets. I think it’s because the RCD will operate within 300ms, which is less than the 400ms requirement for automatic disconnection/fault protection.
Not quite sure about non RCD protected circuits or P-N short circuits though?
Clause 7.3.5.2.1 is also worth a read especially where it says “A circuit is considered to be protected against prospective short-circuit and earth fault currents when it is supplied from an electricity generation system incapable of delivering a current exceeding the current carrying capacity of the circuit”. Guess it’s saying your cable size etc within the circuit effectively becomes the circuit protection. Not sure if this would apply to a typical off grid house though, I imagine the inverter would be too big.
This topic confuses me a bit too, interested to hear other opinions.
Re: Best Practice earth loop impedance B type C/B or fuses
In normal grid-supplied circuits we typically use one over-current device to perform 3 different functions.
First is overload protection of the conductors, , quite straightforward.
Next is short circuit protection, again to protect conductors.
Time requirements as per 2.5.4.5
When using a genset or inverter as the source, it probably won't be capable of providing enough current to trip within this time limit - but in that case it won't heat the cable up in the same way as it would on a grid supply.
Hence deemed to be protected as per 7.3.5.2.
Third aspect is fault protection; which used to be called 'protection against indirect contact', and is about reducing effects of electric shock.
There are several accepted methods, but most common is automatic disconnection of supply as per 2.4.2, 1.5.5.3, & 5.7
This is where EFLI comes in, along with requirement for protective device to operate within 0.4 sec for some circuits and 5 sec for others.
All your circuits supplying lighting or sockets will be RCD-protected; and since the RCD's operating time at rated residual current is 25% under the 0.4 sec limit, there's no need to prove, by testing, that an overcurrent device will operate fast enough.
If using an overcurrent device, we need a suitably low EFLI in order to meet the trip time requirement.
But if using an RCD as the fault protection device, the actual impedance of the EFL simply doesn't matter;
as it's not going to affect operation of the RCD.
Unless it's completely open circuit; in which case we haven't met the requirement for operation "in event of a fault" [1.5.5.3 (a)]
The problem is providing fault protection for non-RCD circuits from a genset / inverter; which may well not be able to provide the high current needed to operate an overcurrent device.
Including for the "mains".
And while 7.3.5.2.1 may cover you for "mains" (provided max output is not greater than CCC of conductors); but it won't cover you for final subcircuits that use smaller cable.
On the other hand, for these circuits the operating time can be up to 5 sec.
Best option is probably just fir RCDs to these circuits as well, not because they are required by 2.6.3; but simply to be sure of providing fault protection
First is overload protection of the conductors, , quite straightforward.
Next is short circuit protection, again to protect conductors.
Time requirements as per 2.5.4.5
When using a genset or inverter as the source, it probably won't be capable of providing enough current to trip within this time limit - but in that case it won't heat the cable up in the same way as it would on a grid supply.
Hence deemed to be protected as per 7.3.5.2.
Third aspect is fault protection; which used to be called 'protection against indirect contact', and is about reducing effects of electric shock.
There are several accepted methods, but most common is automatic disconnection of supply as per 2.4.2, 1.5.5.3, & 5.7
This is where EFLI comes in, along with requirement for protective device to operate within 0.4 sec for some circuits and 5 sec for others.
All your circuits supplying lighting or sockets will be RCD-protected; and since the RCD's operating time at rated residual current is 25% under the 0.4 sec limit, there's no need to prove, by testing, that an overcurrent device will operate fast enough.
If using an overcurrent device, we need a suitably low EFLI in order to meet the trip time requirement.
But if using an RCD as the fault protection device, the actual impedance of the EFL simply doesn't matter;
as it's not going to affect operation of the RCD.
Unless it's completely open circuit; in which case we haven't met the requirement for operation "in event of a fault" [1.5.5.3 (a)]
The problem is providing fault protection for non-RCD circuits from a genset / inverter; which may well not be able to provide the high current needed to operate an overcurrent device.
Including for the "mains".
And while 7.3.5.2.1 may cover you for "mains" (provided max output is not greater than CCC of conductors); but it won't cover you for final subcircuits that use smaller cable.
On the other hand, for these circuits the operating time can be up to 5 sec.
Best option is probably just fir RCDs to these circuits as well, not because they are required by 2.6.3; but simply to be sure of providing fault protection
- Rating: 16.67%
Re: Best Practice earth loop impedance B type C/B or fuses
Thanks AlecK very helpful
I do not however want to protect the mains supply with an RCD as I am going to eventually run a hot water cylinder off the inverter during summer, so I could just install b type circuit breakers on the main cable and hot water circuit. I would probably however just put in fuses as they are less likely to nuisance trip.
I do not however want to protect the mains supply with an RCD as I am going to eventually run a hot water cylinder off the inverter during summer, so I could just install b type circuit breakers on the main cable and hot water circuit. I would probably however just put in fuses as they are less likely to nuisance trip.
Re: Best Practice earth loop impedance B type C/B or fuses
Using B-curve mcbs instead of C-curve lakes a difference only of the supply source us capable of providing the high current necessary for an over-current device.
Yes Tables 8.1 & 8.2 allow for a higher impedance ; but they are based on a normal (ie grid) source.
They are not relevant for low-capacity sources where PFC is inherently limited.
Your inverter is very unlikely to have enough output to meet a 0.4 sec operating time, even with b-curve.
Whether it works OK for a 5 sec trip time will require some calculation, and knowledge of inverter peak output.
You can't just apply the 'normal case' figures from the Tables.
The calculations required are shown in App B
Yes Tables 8.1 & 8.2 allow for a higher impedance ; but they are based on a normal (ie grid) source.
They are not relevant for low-capacity sources where PFC is inherently limited.
Your inverter is very unlikely to have enough output to meet a 0.4 sec operating time, even with b-curve.
Whether it works OK for a 5 sec trip time will require some calculation, and knowledge of inverter peak output.
You can't just apply the 'normal case' figures from the Tables.
The calculations required are shown in App B
Re: Best Practice earth loop impedance B type C/B or fuses
The alternative is to rely on 7.3.5.2.1; and ensure the conductors of the non-RCD circuits have a CCC at least as high as the inverter peak output.
That's peak output, not rated output.
That's peak output, not rated output.
Re: Best Practice earth loop impedance B type C/B or fuses
Peak output on the inverter is 9000 watts, so close to 40 amps, it is a 16mm neutral screen to house so that will be fine but the hot water cylinder is 2.5mm so will fuse it (easier to change size if it becomes and issue).
Thanks very much.
Thanks very much.
Re: Best Practice earth loop impedance B type C/B or fuses
You'll need to check that the fuse (or mcb) will operate within 5 sec with only 40 A available
(compared to the typical 1kA we can expect with a grid supply).
Not sure how easy it will be to get detailed info on operating curves for fuses.
Personally I'd just use an RCD.
For a 30 mA RCD; tripping will not be below 15 mA; and almost certainly will need mire like 25 mA.
Max permitted IR for an element is 10 k ohms which works out to 23 mA.
So if there's enough leakage to trip the RCD, then the w/h is faulty.
For a circuit without elements, min 1 meg-ohm = 0.23 mA
So-called 'nuisance tripping" is mostly a myth;
tripping is generally down to bad decisions during design & installation.
(compared to the typical 1kA we can expect with a grid supply).
Not sure how easy it will be to get detailed info on operating curves for fuses.
Personally I'd just use an RCD.
For a 30 mA RCD; tripping will not be below 15 mA; and almost certainly will need mire like 25 mA.
Max permitted IR for an element is 10 k ohms which works out to 23 mA.
So if there's enough leakage to trip the RCD, then the w/h is faulty.
For a circuit without elements, min 1 meg-ohm = 0.23 mA
So-called 'nuisance tripping" is mostly a myth;
tripping is generally down to bad decisions during design & installation.