Looking for opinion on whether this scenario is compliant

[The Ghost]

Went to daughters house to replace a couple of broken sockets and did a Zs and RCD test after replacing. CU was replaced and notice on it said inspected 2015. Plastic dual RCD 30 ma/Main switch 60947-3 isolator. System is TT. (100ma RCD?)
Zs at socket was 11.75 ohms L-PE 2.64 ohms L-N. No bonding present to services. Suggested my daughter might get the council in to check it out. Had a phone call today and electrician said it is all compliant! He also said it disconnects in 0.37ms and I asked to talk to him he would not. He said I was just trying to make money, if only he knew? Apparently he fixed the bonding in a few minutes.
I said to my daughter tell him it should be 0.2 (the property is not cross bonded as in the exception to table 41.1) he said that was the case 20 years ago and I needed to get up to date. Okay!
I must confess to be a little bit taken aback by this and thought I would do a sanity check with the forum members. I do have Amd 3 BS7671 by the way i.e. 2382-15 so I think I am up to date.
Essentially I am of course concerned for my daughter, and what appears to be possibly a cavalier attitude. I told my daughter to get a copy of the report and tests so we await that. Am I being concerned over nothing or just plain stupid??? Tin hat on.

 

[The Ghost]

the Ra

Did not test Ra, just wanted to change sockets and get council to check basically.

 

[Gavin John Hyde]

What is the Ra valie?

If it is anything like the rods I have measured the last few weeks then the readings will be much higher than normal.

 

[telectrix]

That’s for over current protective device disconnection not RCDs.

Wrong..... if you look below table 41.1, it specifically mentions RCDs.

 

[Wilko]

Sorry if I've missed something, but B32 won't trip "instantly" with that L-N impedance. Hopefully appliances are protected by 13A fuses but a short in the cabling would cook for a while, I think.
And stepping out on the edge a bit further, I don't think the ring is intact

 

[The Ghost]

Ok so still, 11.7 ohms Zs on RFC any thoughts? 60898 MCB 32a B

 

[Strima]

Wrong..... if you look below table 41.1, it specifically mentions RCDs.

But if you rely on an RCD for fault protection you must ensure the fault currents are significantly higher than the rated residual current of the RCD.

As RCDs are used for additional protection then the timings for those must comply with table 3A in appendix 3.

Leaving you to ensure that the PSSC of the installation will disconnect the OCPD within the stated times of 41.1.

 

[Strima]

Ok so still, 11.7 ohms Zs on RFC any thoughts? 60898 MCB 32a B

Using a 30mA RCD for earth fault protection then your maximum permitted Zs is 1667 ohms.

But what was the PSSC for the circuit? Only this measurement will tell you if the OCPD will disconnect within 0.2 seconds.

For a 32 amp B type 60898 you need a current of 5x the device rating for a disconnection time between 0.1 and 5 seconds. i.e. 160 amps.

 

[Strima]

Infact looking at one of your earlier posts with you L>N of 2.64 ohms that will give you a fault current of 87 amps which will disconnect in roughly 40 seconds so it will not comply.

How far out in the sticks is her house and what was the voltage reading?

 

[telectrix]

agreed. the RCD will comply with a L-E fault, but not with L-N. for a L-N reading of 2.64, must be a bloody long circuit.

 

[westward10]

Doesn't need to comply across L/N as no values are given for this scenario. We need to know the L/N reading at source.

 

[magnoliafan89]

I tested a tt system yestersay and the zs at db was 9.11 ohms with ipf of 25a, on the socket circuit i got a zs of 10 ohms and an r1+r2 reading of 0.49 ohms and was told this was fine....i always thought and was taught a tt would give higher readings for a zs than as given in appendix table b6. Now im confused

 

[Bellendian]

The higher zs is fine, and dealt with by the application of rcds for fault protection, and in this case since they are 30ma they will also provide additional protection. The curve ball is the 2.64 ohm L-N....generally we assume that if R1+R2 is good then R1+RN would be as well....since the CSA of cpc will be less than or equal to the neutral...but ring finals aside, what specific tests are mandated to check this? Could well be a loose neutral connection, but how are these tested?
Often hear circumstantial evidence that it's always neutrals that burn out or are loose....is it because a loose l or cpc gets picked up during r1+r2 testing, but those pesky neutrals fall by the wayside? Just a thought....

 

[Wilko]

Doesn't need to comply across L/N as no values are given for this scenario. We need to know the L/N reading at source.

A different time to comply perhaps?
Reg 434.5.2 says to interrupt fault current before damage has occurred. So we don't need 200ms but we do need to check dis times with adiabatic (?).

 

[Bellendian]

A different time to comply perhaps?
Reg 434.5.2 says to interrupt fault current before damage has occurred. So we don't need 200ms but we do need to check dis times with adiabatic (?).

Agreed, but this is design stuff, and possibly something that is missed during initial verification? Or am I losing the plot!?

 

[Strima]

Compliance should be checked each time the installation is inspected and tested as supply nature and perameters change.

A circuit that complied when designed may not comply now.

 

[Bellendian]

Yep, but on a typical radial cct, which of the mandated tests would highlight a higher than expected neutral impedance?

 

[johnduffell]

A different time to comply perhaps?
Reg 434.5.2 says to interrupt fault current before damage has occurred. So we don't need 200ms but we do need to check dis times with adiabatic (?).

Good point although adiabatic should only be used up to 5s, go far beyond that and the value would be way too conservative.
Not sure how to work it out, i agree with the rest of your post. I feel like we should be sizing it as an overload protection, so as long as the mcb can protect the cable that way it's ok? There's no special case of worrying about exposed conductive parts being live in this case.

 

[Wilko]

If there isn't a defect in the DNO supply or the RFC, then the RFC must be very long to get L-N of 2.6 Ohms at an outlet. Either way, the voltage drop would slow down my tea and toast, no good at all !

 

[The Ghost]

The rcds are providing fault protection

Are they? This is a dual RCD board and if I understand correctly the RCD are additional protection. As this is a TT system why would one not use an up front 100ma RCD for fault protection. I must admit I very rarely if ever encounter TT and if I do usually get it converted to PME so frankly I am not on top of the implications reg wise of TT.

 

[Ian1981]

Are they? This is a dual RCD board and if I understand correctly the RCD are additional protection. As this is a TT system why would one not use an up front 100ma RCD for fault protection. I must admit I very rarely if ever encounter TT and if I do usually get it converted to PME so frankly I am not on top of the implications reg wise of TT.

30mA rcds offer fault protection for earth faults only and have the added bonus of offering additional protection hence the 2 tests.
1x rating trip within 200ms for circuits upto 32 amps if it’s TT (the standard for bs61008 and 61009 being 300ms) and 5x for additional protection within 40ms.
Your mcbs will be providing short circuit and overload protection only.
It maybe that additional protection via 30mA rcd is required for socket outlets,cables serving a bathroom location and cables buried in walls etc rather than relying on a 100mA rcd which would not comply with the requirements of additional protection and only offers fault protection.
I would not be worrying about your high Zs readings after all it’s a TT arrangement and as your covered for earth faults via the dual rcds in the consumer unit.
what your issue is with the bonding requirements will only be applicable if you have extraneous conductive parts entering the dwelling.

 

[Wilko]

Hi - agree.
But any L-E faults before the CU's RCDs will have to clear via the service fuse. And with TT its likely it won't clear and there's risk the installation E could rise. Case of L tail flapping about inside AMD3 CU for example (very unlikely and why tails glands and good workmanship always needed).

 

[Ian1981]

Hi - agree.
But any L-E faults before the CU's RCDs will have to clear via the service fuse. And with TT its likely it won't clear and there's risk the installation E could rise. Case of L tail flapping about inside AMD3 CU for example (very unlikely and why tails glands and good workmanship always needed).

I like to install an S type rcd rather than the main switch if using a dual rcd board but that’s just my preference on TT arrangements.

 

[westward10]

This is a plastic CU doesn't need a 100ma main switch. The rcds in situ provide additional and fault protection as already stated.

 

[Ian1981]

Yes agreed a plastic consumer unit takes the danger scenario of loose tails touching the enclosure causing a potential hazard, out of the equation

 

[Adam W]

Yes agreed a plastic consumer unit takes the danger scenario of loose tails touching the enclosure causing a potential hazard, out of the equation

... Although of course there's still the risk of the CU catching fire.
I had a quick leaf through the new OSG in the wholesalers the other day - there's bit in it about countering precisely that situation, but I can't remember the exact details.

 

[rompling]

The higher zs is fine, and dealt with by the application of rcds for fault protection, and in this case since they are 30ma they will also provide additional protection. The curve ball is the 2.64 ohm L-N....generally we assume that if R1+R2 is good then R1+RN would be as well....since the CSA of cpc will be less than or equal to the neutral...but ring finals aside, what specific tests are mandated to check this? Could well be a loose neutral connection, but how are these tested?
Often hear circumstantial evidence that it's always neutrals that burn out or are loose....is it because a loose l or cpc gets picked up during r1+r2 testing, but those pesky neutrals fall by the wayside? Just a thought....

Totally agree