Supply Isolator required to be metal on TT

[DaveOLW]

Does the Supply Isolator require to be metal on TT, is so metal could become live because before RCD

 

[westward10]

I assume you are talking about a separate domestic double pole isolator if so, no.

 

[DaveOLW]

I assume you are talking about a separate domestic double pole isolator if so, no.

Regulation 421.1.20 of BS 7671:2018 states, for consumer units and similar switchgear assemblies in domestic premises, to be manufactured from non-combustible materials or be enclosed in a cabinet constructed from non-combustible materials.

 

[oscar21]

Regulation 421.1.20 of BS 7671:2018 states, for consumer units and similar switchgear assemblies in domestic premises, to be manufactured from non-combustible materials or be enclosed in a cabinet constructed from non-combustible materials.

It does but that doesn't seem to bother the DNO or meter installers, I see lots of plastic DP isolators in plastic enclosures just after the meter.

 

[westward10]

The fact they are still being produced unlike plastic consumer units suggests they are not the same thing.

 

[James]

I look at it as a poor way of saying consumer unit or similar i.e. a collection of switches, fuses, circuit breakers or similar protective devices as an assembly.

I do not personally consider a single isolator as falling into the regulation listed above.

 

[oscar21]

I look at it as a poor way of saying consumer unit or similar i.e. a collection of switches, fuses, circuit breakers or similar protective devices as an assembly.

I do not personally consider a single isolator as falling into the regulation listed above.

In that case how do yo feel about a "S" RCD upfront in a plastic unit before the main CU.

 

[mainline]

In that case how do yo feel about a "S" RCD upfront in a plastic unit before the main CU.

Rcds are often installed ahead of a consumer unit on a TT supply.
Requiring a metal enclosure would introduce new risks.

 

[James]

In that case how do yo feel about a "S" RCD upfront in a plastic unit before the main CU.

Depending upon all the other circumstances then I may well be ok with it.

 

[nicebutdim]

In that case how do yo feel about a "S" RCD upfront in a plastic unit before the main CU.

How would you feel about a metal enclosure when PSCC/PFC is likely insufficient to take out a main fuse?

 

[DefyG]

If incorporated inside a fire retardant meter box, plastic enclosure would be compliant to Regulation 421.1.20 .
Not sure all meter boxes are fire retardant though!
Wording like 'similar switchgear' only helps with various interpretations!

Mitras Fire Retardant Recessed Electricity Meter Box

The standard "Fire Retardant" recessed electric meter box is designed to house any domestic electric meter. This electric meter box is manufactured from DMC material and complies with BS476 part 7 class 2. This product is available nationally through Builders' Merchants. Click the 'Find Your...

www.meterboxes.co.uk

 

[James]

If incorporated inside a fire retardant meter box, plastic enclosure would be compliant to Regulation 421.1.20 .
Not sure all meter boxes are fire retardant though!
Wording like 'similar switchgear' only helps with various interpretations!

Mitras Fire Retardant Recessed Electricity Meter Box

The standard "Fire Retardant" recessed electric meter box is designed to house any domestic electric meter. This electric meter box is manufactured from DMC material and complies with BS476 part 7 class 2. This product is available nationally through Builders' Merchants. Click the 'Find Your...

www.meterboxes.co.uk

That is not non combustible

 

[brianmoooore]

That is not non combustible

Indeed. IIRC, when this ridiculous regulation first made the light of day, there were several questions and proposals regarding the use of non combustible plastics, but it turned out that the regulation was framed such that only metal would meet the requirements, without resorting to exotic and expensive materials.

 

[Yellowhammer]

I would/have put a plastic enclosure up front with an S RCD in it on a TT install. Tails torqued correctly and ideally checked again the following day, suitably glanded in and out with a tails gland, with the earth hole blanked. I think this is a preferable/less risky scenario than having the metal enclosure with only the main 100A fuse protecting against something shorting to a metal enclosure with rather likely a PEFC of under 10A..

 

[oscar21]

How would you feel about a metal enclosure when PSCC/PFC is likely insufficient to take out a main fuse?

Is there not some sort of guidance or similar issued by NAPIT or the NIC that states tails that are securely clamped and glanded in a metal enclosure aren't classed as a risk in a metal TT enclosure.

 

[oscar21]

Found it, I think.

RCDs, metal consumer units and TT installations | NAPIT

NAPIT's Bill Allan answers another common contractor query. Contractors sometimes ask whether it's necessary to install an RCD in a TT installation before a metal consumer unit (i.e. a 'front end' RCD

professional-electrician.com

you can avoid 10 minutes reading it if you want and skip to the last line, which is what NAPIT could have done.

Do you need to install an RCD on a TT installation before a metal consumer unit if all the outgoing sub-main and final circuits are RCD/RCBO protected? The answer is no

 

[SparkySy]

Found it, I think.

RCDs, metal consumer units and TT installations | NAPIT

NAPIT's Bill Allan answers another common contractor query. Contractors sometimes ask whether it's necessary to install an RCD in a TT installation before a metal consumer unit (i.e. a 'front end' RCD

professional-electrician.com

you can avoid 10 minutes reading it if you want and skip to the last line, which is what NAPIT could have done.

Do you need to install an RCD on a TT installation before a metal consumer unit if all the outgoing sub-main and final circuits are RCD/RCBO protected? The answer is no

It is indeed no, but I still go for one just in case the 30ma protection fails, at least there is still some protection as a last resort back up!

 

[grumpyjohn01]

It is indeed no, but I still go for one just in case the 30ma protection fails, at least there is still some protection as a last resort back up!

I totally agree, on a new TT install or a replacement CU I pretty much always install a Wylex 'S' type RCD upfront in a Wylex Isolator enclosure (plastic) which fits perfectly and has restraint for the tails built in, has the bonus of fault protection for submains or other circuits you might want to have on MCBs or fuses downstream.

 

[Simon47]

Indeed. IIRC, when this ridiculous regulation first made the light of day, there were several questions and proposals regarding the use of non combustible plastics, but it turned out that the regulation was framed such that only metal would meet the requirements, without resorting to exotic and expensive materials.

Being pedantic, only ferrous metal mets the requirements.
There are probably many materials that would meet the requirement if they had turned on some brain cells and referenced a standard for what "combustible" means. Pity someone didn't tell the fire brigade something like "tell you what, leave us to do the electrical engineering, and we'll leave you to the fire fighting"

 

[Yellowhammer]

I've read that article before, it does make some sense, though I'm not sure, personally that fitting the upstream S has downsides to be avoiding.
Often on a TT it is the provided a means of isolation before the board which is always handy.

It's also handy that it helps you meet the disconnection times for a submain on the install, so you can fit say a garage supply from an MCB in the main board (assuming it's not buried <50mm in a wall, or any other means of installation requiring 30ma RCD protection). You can then have a couple of upstream RCBO circuits in the garage and still maintain selectivity. I'm a fan, personally of local RCBOs in the garage as I want to lights to stay on ideally if the sockets were to go while I have a welder or grinder or similar in my hands. (Of course the downside being any fault on the submain cable itself will take out the entire house, given the MCB is as likely as good as useless).

 

[mainline]

I've read that article before, it does make some sense, though I'm not sure, personally that fitting the upstream S has downsides to be avoiding.
Often on a TT it is the provided a means of isolation before the board which is always handy.

It's also handy that it helps you meet the disconnection times for a submain on the install, so you can fit say a garage supply from an MCB in the main board (assuming it's not buried <50mm in a wall, or any other means of installation requiring 30ma RCD protection).
You can then have a couple of upstream RCBO circuits in the garage and still maintain selectivity.

I'm a fan, personally of local RCBOs in the garage as I want to lights to stay on ideally if the sockets were to go while I have a welder or grinder or similar in my hands. (Of course the downside being any fault on the submain cable itself will take out the entire house, given the MCB is as likely as good as useless).

An upfront rcd whether it be delayed or not will only give partial selectivity.

 

[Simon47]

An upfront rcd whether it be delayed or not will only give partial selectivity.

Absent faults or incorrect design, it should give full selectivity.
I.e., regardless of the fault current the RCDs are exposed to, the downstream unit should have opened the circuit and removed the fault long before the delayed unit will trigger.

 

[Risteard]

Regulation 421.1.20 of BS 7671:2018 states, for consumer units and similar switchgear assemblies in domestic premises, to be manufactured from non-combustible materials or be enclosed in a cabinet constructed from non-combustible materials.

That presumes that an isolator is a "similar switchgear assembly" to a "consumer unit" (single phase type tested DBO) which is doubtful.

 

[mainline]

Absent faults or incorrect design, it should give full selectivity.
I.e., regardless of the fault current the RCDs are exposed to, the downstream unit should have opened the circuit and removed the fault long before the delayed unit will trigger.

A 30 mA device would require a minimum of a 100 mA Type S RCD upstream to provide selectivity.

An rcbo board could have accumulated leakage and not provide full selectivity.

 

[SparkySy]

A 30 mA device would require a minimum of a 100 mA Type S RCD upstream to provide selectivity.

An rcbo board could have accumulated leakage and not provide full selectivity.

I have installed this arrangement many times over the years and never had an issue with accumulated leakage on a domestic installation to that level.
I have however come across this issue on dairy installations where the vacuum pumps have been upgraded and changed to vfd driven. The vfd's were causing the milking shed to trip and shut the whole lot down!