Why is split RCD not appropriate for TT?

[swaRRR]

OSG Pg.40 says:

'[Split CU with 2x RCCBs] is not suitable for an installation forming part of a TT system as there is insufficient fault protection of the single-insulated conductors that connect the load side of the double-pole main switch to the supply-side of the RCCB's.'

Why is there insufficient fault protection here?

I've seen loads of TT systems with RCD protection in them. What would you use instead? I know you can use RCBO's but they're not standard.

 

[cireland]

OSG Pg.40 says:

'[Split CU with 2x RCCBs] is not suitable for an installation forming part of a TT system as there is insufficient fault protection of the single-insulated conductors that connect the load side of the double-pole main switch to the supply-side of the RCCB's.'

Why is there insufficient fault protection here?

I've seen loads of TT systems with RCD protection in them. What would you use instead? I know you can use RCBO's but they're not standard

I think it's because of the split cu. The cable connecting from the bottom of the main switch to the top of the RCCB's aren't RCD protected.
If you have an RCD main switch or Isolator the it's ok.
That's my interpretation of it anyway, I may well be wrong.

 

[mainline]

OSG Pg.40 says:

'[Split CU with 2x RCCBs] is not suitable for an installation forming part of a TT system as there is insufficient fault protection of the single-insulated conductors that connect the load side of the double-pole main switch to the supply-side of the RCCB's.'

Why is there insufficient fault protection here?

I've seen loads of TT systems with RCD protection in them. What would you use instead? I know you can use RCBO's but they're not standard.

Typically, with a split-board consumer unit having one or more 30 mA RCDs for the final circuits, the front-end RCD would be 100 mA, type S (time-delayed) to permit selectivity between RCDs.


However, the cables from the origin of the supply to the input terminals of the RCD were a concern. This was because RCDs don’t protect the cables which supply their input terminals. This means that a line to earth fault on the supply side of the front-end RCD could go undetected.

Where a metal consumer unit is installed, there’s always the risk, however small, that the metal enclosure – and other exposed-conductive-parts – will make contact with a live conductor, causing the metalwork to rise to a potential approaching that of the line voltage of the supply.


The relatively high external earth loop impedance in TT systems might mean that the service cut-out fuse would not operate within the maximum permitted time of five seconds, permitting a dangerously high potential to remain on all the earthed metalwork of the installation.

Consequently, you needed to be sure that any exposed metalwork couldn’t be made live by a fault ‘before’ the RCD, caused either by an insulation fault or by one of the conductors becoming loose.

 

[loz2754]

I know you can use RCBO's but they're not standard.

They are standard for me, and plenty of other sparks. I will not install a split load board anymore, as the RCBO boards are so much cheaper to buy now, and are infinitely superior.

 

[swaRRR]

Typically, with a split-board consumer unit having one or more 30 mA RCDs for the final circuits, the front-end RCD would be 100 mA, type S (time-delayed) to permit selectivity between RCDs.


However, the cables from the origin of the supply to the input terminals of the RCD were a concern. This was because RCDs don’t protect the cables which supply their input terminals. This means that a line to earth fault on the supply side of the front-end RCD could go undetected.

Where a metal consumer unit is installed, there’s always the risk, however small, that the metal enclosure – and other exposed-conductive-parts – will make contact with a live conductor, causing the metalwork to rise to a potential approaching that of the line voltage of the supply.


The relatively high external earth loop impedance in TT systems might mean that the service cut-out fuse would not operate within the maximum permitted time of five seconds, permitting a dangerously high potential to remain on all the earthed metalwork of the installation.

Consequently, you needed to be sure that any exposed metalwork couldn’t be made live by a fault ‘before’ the RCD, caused either by an insulation fault or by one of the conductors becoming loose.

Ok but how is this different with RCBO's?

And what's the difference between a normal 100a main switch in a CU and an S Type RCD?

It's amazing how many of these i've worked on without knowing anything about them.

 

[mainline]

They are standard for me, and plenty of other sparks. I will not install a split load board anymore, as the RCBO boards are so much cheaper to buy now, and are infinitely superior.

The only sparks to be fitting them now are the John Wayne ones.

The likes of screw fix and tool station still punting them out cheap as chips type AC rcds no spd they should be ashamed.

I like the Fusebox range.

 

[cireland]

Ok but how is this different with RCBO's?

With RCBO's the supply is via the busbar. It's in a fixed position so can't come into contact with the casing.

 

[mainline]

Ok but how is this different with RCBO's?

As each RCBO is supplied by a busbar, the risk of the busbar becoming loose and making contact with the metal enclosure is minimal.

 

[swaRRR]

With RCBO's the supply is via the busbar. It's in a fixed position so can't come into contact with the casing.

I dunno about that, some of the installs i've seen...

 

[swaRRR]

The only sparks to be fitting them now are the John Wayne ones.

The likes of screw fix and tool station still punting them out cheap as chips type AC rcds no spd they should be ashamed.

I like the Fusebox range.

I worked on a 1700-home council refit earlier this year where every single board was split RCD.

I also worked on several premium new builds (you know the ones where they use nice insulation instead of the stuff that itches and it comes with triple-glazing?) that used them too.

Scandalous really.

 

[cireland]

I dunno about that, some of the installs i've seen...

Alas, me too.

 

[swaRRR]

Alas, me too.

I worked on one where the busbar had been bent outwards towards the lid, because they couldn't be arsed chopping it down and the main incomers were in the way, the way they'd been brought in. Luckily the CU was still plastic.

 

[westward10]

Does the OSG cite a Regulation.

 

[Baddegg]

They are standard for me, and plenty of other sparks. I will not install a split load board anymore, as the RCBO boards are so much cheaper to buy now, and are infinitely superior.

Agree havnt fitted a split load for at least 4 years I reckon

 

[swaRRR]

Does the OSG cite a Regulation.

No, but this seems to be a regular occurrence where the OSG says something and it turns out it's not actually a reg. Is the OSG fit for purpose?

 

[westward10]

Regulations now require residential to be rcbos.

 

[westward10]

No, but this seems to be a regular occurrence where the OSG says something and it turns out it's not actually a reg. Is the OSG fit for purpose?

Not looked at an OSG for over 20 years your book of choice is BS7671.

 

[cireland]

Does the OSG cite a Regulation.

Wouldn't it be covered by the same regulations stating TT systems have to be RCD protected. This wiring is still on the consumer side of the main switch albeit within the consumer unit itself.

 

[swaRRR]

Not looked at an OSG for over 20 years your book of choice is BS7671.

I know that but OSG is supposed to be a bite-sized, easy reference for whats in the big brown book. Seems like the OSG talks a lot of gibberish.

Here's another one.

OSG page 61.

'Other than at the origin of the installation, every circuit...that may have to be isolated without interrupting the supply to other circuits should be provided with its own isolating device. The device must switch all live conductors in a TT system and all line conductors in a TN system.'

This basically says all breakers have to be double pole in a TT, no?

Yet you go to 132.15.201 which the OSG references and it says absolutely nothing about this.

 

[mainline]

No, but this seems to be a regular occurrence where the OSG says something and it turns out it's not actually a reg. Is the OSG fit for purpose?

Think it stems from
Regulation 531.3.5.3.2.201