Garo PME unit for metal shed supply

[chris day]

Metal shed In Garden with no extraneous conductive parts
House electrics PME supply
Garo make a 40a PME fault detection unit for EVs and I was thinking of exporting the PME via the garo init at the supply end and swapping out the factory 30ma to a 100ma time delay
Then use 30ma RCBOs in the final circuits, reason being is I don’t like using TT rods and relying on RCDs as the means of ADS as they always become locked up over time.
The garo unit seems to be a sound way to protect the shed supply in the event of a PME fault and I get to use a superior earthing system

What are your thoughts

 

[Risteard]

Disconnecting the cpc is generally prohibited. This is not a Section 722 installation.

 

[pc1966]

As a metal shed it might well constitute an extraneous conductive part in itself, so would require bonding anyway and if you retain the TN-C-S supply that implies at least 10mm CPC size. Yes, it is unlikely to be that conductive, and potentially an open-PEN unit could limit fault duration, but regs still demand that 10mm size.

So unless you are running 10mm 3-core SWA or equivalent out there you would be better going TT to save that extra expense. Considering the cost of the PME fault detection unit you might as well go for a local CU with 100mA delay and 30mA RCBOs as then no single point of failure in the RCD electronics, and rod is still only 200 ohms to be met.

Sub-main out can probably be on 2-core SWA depending on the cable end Zs and protection at supply. Usually fused-switch gives better selectivity with final RCBOs than an MCB, and higher Zs to be OK on 5s disconnection time for TN source, but unless shed has unusual use I don't think hard fault selectivity is a big issue.

If it is, then some emergency LED lights would avoid someone's workshop going dark with power tools still running, etc.

 

[timhoward]

Then use 30ma RCBOs in the final circuits, reason being is I don’t like using TT rods and relying on RCDs as the means of ADS as they always become locked up over time.

I'm more comfortable where there are two of them, and in fact the other day an S-type 100ma tripped when I accidentally touched N and E, revealing the downstream 30ma was indeed locked up.

 

[chris day]

As a metal shed it might well constitute an extraneous conductive part in itself, so would require bonding anyway and if you retain the TN-C-S supply that implies at least 10mm CPC size. Yes, it is unlikely to be that conductive, and potentially an open-PEN unit could limit fault duration, but regs still demand that 10mm size.

So unless you are running 10mm 3-core SWA or equivalent out there you would be better going TT to save that extra expense. Considering the cost of the PME fault detection unit you might as well go for a local CU with 100mA delay and 30mA RCBOs as then no single point of failure in the RCD electronics, and rod is still only 200 ohms to be met.

Sub-main out can probably be on 2-core SWA depending on the cable end Zs and protection at supply. Usually fused-switch gives better selectivity with final RCBOs than an MCB, and higher Zs to be OK on 5s disconnection time for TN source, but unless shed has unusual use I don't think hard fault selectivity is a big issue.

If it is, then some emergency LED lights would avoid someone's workshop going dark with power tools still running, etc.

This is actually exactly as I am now going to do it as it’s fully compliant and cheeper, thank you for the quality response