A can of earth worms ? ?

[bernardgreen]

This was a question at a planning meeting for a new project.

"" If all circuits have RCD protection that trip at 35 milliamps why does the CPC have to have an impedance less than 1000 ohms ( yes one thousand ohms ) ?

Reasoning for the proposal was as follows.....

A Live to CPC short with a 1000 ohm CPC will create a fault current of 230 milliamps which will un-balance and trip the RCD. If the CPC impedence is about 1 ohm then the MCB will trip on the the 100 amp plus fault over current. This trip may happen before the RCD trips on the 100 amp unbalance. ( possibly the RCD core may be instantly saturated at that fault current and not trip )

It is 100 amps plus and not 230 amps because I fault = 230 / (Rcpc + Rlive)

A Live to Neutral short will not create an un-balance in the RCD and it will not trip but hopefully the MCB will trip on the over current.

So if all circuits are RCD protected at 35 milliamp fault current then CPCs and earth rods need not be low impedance.

As long as all Live to CPC faults, either direct or via a body or other item, result in an un-balance adequate to ensure fast tripping of the RCD the CPC is effective.

The first reaction was " That is dangerous rubbish ""

The proposer then asked. "" If you were getting an electric shock from a live wire would you expect the RCB to operate and save you ? ""

The answer was " Yes of course. that is what RCDs are intended for. "

The proposer then asked " Then you ensure your body's impedance is always low enough that it complies with the impedance of a CPC for that installation where you got the electric shock ? "

 

[Spark123]

30mA??

Technically there is no reason why we cannot protect against faults (was indirect contact) to the 50v rule with RCD protection which gives a max Zs of 1666 ohms however I'd be looking for the fault if the readings were anywhere near this.
For reliablity I like to see the figures low enough to operate the OCPD in the case of a TN system.
In a TT system you have little choice, however as long as the R1+R2 are low enough then I'd accept it. Ze should be below 200 ohms for reliablity too.

Direct contact or basic protection is a slightly different minefield where an RCD may be used to supplement it, but not as the only means of preventing it.
Put simply - if the current flowing through a body is 30mA or then it isn't fatal. It may pause problems if you are stood on a ladder tho!
If the current is greater than 30mA then the duration of the shock becomes a factor.
Once you get up to the 100mA region then you want the supply to be cut very quickly as it may cause the heart to stop.

 

[ColJack]

there's also the series resistance problem

if you happen to touch something live whilst having another body part in contact with the earthed part of the circuit, then that 1000ohms that you'd have for a direct short then becomes 2000ohms with you in series ( +1000ohms used for clarity, no one has definitively answered "what is the resistance of the human body?" )

which is still enough to operate a 100mA rcd, but not as quickly as it would for a combined resistance of 1200ohms..

 

[Spark123]

100mA RCDs are not allowed to be used for supplementary protection against direct contact (got my 16th edn head on), only for indirect contact protection i.e. metalwork made live through a fault which should be earthed via the CPC.

 

[ColJack]

I thought you used 100mA as incommer on a TT system?
never installed any TT JSYK.

 

[Spark123]

Yep, you used to install a 100mA (preferably type S) as an incomer on a TT system. This is to provide Automatic Disconnection of Supply (protection against indirect contact 16th edn or fault protection 17th edn) in the event of a fault on a protected circuit as the EFLI will be too high to operate a fuse/MCB.
It is not for supplementary protection against direct contact.

It has slightly changed in the 17th edn regs as almost all circuits in a domestic now need protection with 30mA devices anyway and disconnection times have changed.

 

[ban-all-sheds]

And thus, I predict, we'll train a generation of sparks who won't think that Zs matters at all as the RCD will save their sorry a**e...

 

[Spark123]

It has been like that for a long time with TT systems - you have to rely on RCDs for excessive Zs as there isn't any other viable option.
As for TN systems there isn't much excuse for using RCDs for excessive Zs but the 17th edn regs do allow it.

 

[davelx]

And thus, I predict, we'll train a generation of sparks who won't think that Zs matters at all as the RCD will save their sorry a**e...

Which is exactly what the ECA say in their guide to the 17th Ed. C.4.6 5 Irrelevant ELI Specification

 

[bernardgreen]

It is logical to use RCDs to detect earth faults because they work of the difference of current between Live and Neutral. So they will detect and trip on any fault that signifcantly disturbs the equality of currents on the Live and Neutral be it an earth fault or a leak of electricity from the protected wiring to (or from) anywhere else.

 

[securespark]

An RCD is very much a mechanical device: they can and do fail, frequently! I have even had brand new ones straight out of the box that have failed to trip under test.

This is why regular testing makes sense, both of the installation by a qualified person and testing by the householder by pressing the button.

For my money the EFLI is still important, RCD or no RCD. The circuit should still comply.

 

[bernardgreen]

An RCD is very much a mechanical device: they can and do fail, frequently!

Yes.

This is why regular testing makes sense, both of the installation by a qualified person and testing by the householder by pressing the button.

Or by inducing an earth fault at the far end of each protected circuit with a suitable test device ( one that puts a 6 K resistor Live to Earth for a preset test duration )

For my money the EFLI is still important, RCD or no RCD. The circuit should still comply.

Yes. belt and braces are necessary to ensure safety,

 

[electronicsuk]

This is why regular testing makes sense, both of the installation by a qualified person and testing by the householder by pressing the button.

But, unfortunately, nobody (and I feel that I can state that pretty categorically, but no doubt someone out there will prove me wrong...) tests an RCD anywhere near as frequently as instructed in a domestic situation. Nobody wants the hassle of resetting the tens of digital clocks on every conceivable appliance in the modern home.

 

[JohnD]

I do my RCBOs occasionally.

Only got a clock reset on the microwave. the CH has battery backup, the AV stuff picks up a radio signal. The cooker clock is bust

Have a couple of old clock radios, if I could be bothered I'd renew their standby batteries more often.