So I think the likelihood of a single CPC melting during a fault is low.

Actually meting the copper is highly unlikely - it would have to reach well over 1000ºC - but many problems can occur at much lower temperatures. A hot c.p.c. is likely to start melting insulation in its vicinity, even acting like a cheese wire cutting through the cable construction where there's any stress (e.g. at bends) or causing thermal damage to to internals of accessories - and not just at the point of the fault but all through the circuit upstream of the fault - with consequential dangers ranging from shock to fire.

So a 'complete ring' cpc might be necessary to meet the thermal constraints.

I'm not following the thinking here. Ring design normally only considers a single conductor as far as faults are concerned - as there's no guarantee that the fault current will divide significantly between the two legs - either because the fault is close to the origin of the circuit on one leg so the vast majority of the current passes along that leg - or because it's on a spur and there is only one conductor anyway. To my mind the comparison of measured Zs to acceptable values takes into account ADS, thermal withstand and differences in conductor temperature - e.g. in simple terms it satisfies the OSG Zs tables rather than just BS 7671 chapter 41 ones.

In a way, the increased Zs due to a broken c..p.c. in a ring is no more onerous than having a longer complete ring that measured the same at the end of a spur.

So say we'd identified the problem as G/Y sleeving having slipped into a terminal before it was tightened (and were under orders not to correct faults found) - and Zs at all points was satisfactory and no requirement to handle high protective conductor currents - certainly doesn't meet all BS 7671 requirements - but where is the danger (potential or otherwise) - that could occur from this situation that could justify a C2?

   - Andy.

So I think the likelihood of a single CPC melting during a fault is low.

Actually meting the copper is highly unlikely - it would have to reach well over 1000ºC - but many problems can occur at much lower temperatures. A hot c.p.c. is likely to start melting insulation in its vicinity, even acting like a cheese wire cutting through the cable construction where there's any stress (e.g. at bends) or causing thermal damage to to internals of accessories - and not just at the point of the fault but all through the circuit upstream of the fault - with consequential dangers ranging from shock to fire.

So a 'complete ring' cpc might be necessary to meet the thermal constraints.

I'm not following the thinking here. Ring design normally only considers a single conductor as far as faults are concerned - as there's no guarantee that the fault current will divide significantly between the two legs - either because the fault is close to the origin of the circuit on one leg so the vast majority of the current passes along that leg - or because it's on a spur and there is only one conductor anyway. To my mind the comparison of measured Zs to acceptable values takes into account ADS, thermal withstand and differences in conductor temperature - e.g. in simple terms it satisfies the OSG Zs tables rather than just BS 7671 chapter 41 ones.

In a way, the increased Zs due to a broken c..p.c. in a ring is no more onerous than having a longer complete ring that measured the same at the end of a spur.

So say we'd identified the problem as G/Y sleeving having slipped into a terminal before it was tightened (and were under orders not to correct faults found) - and Zs at all points was satisfactory and no requirement to handle high protective conductor currents - certainly doesn't meet all BS 7671 requirements - but where is the danger (potential or otherwise) - that could occur from this situation that could justify a C2?

   - Andy.

AJJewsbury said:

I'm not following the thinking here. Ring design normally only considers a single conductor as far as faults are concerned

Not quite true, because we use the ring-value of R2 to calculate anticipated Zs for circuit length, not the single-conductor value, i.e. we use [(R2/2)//(R2/2)] and not [R2/2]  ... although as I said in the earlier post, for 1.5 sq mm, in general Za > Zs (terms as defined in EIDG) and therefore shouldn't be an issue.

But it it's 1 sq mm, it won't meet the criteria  for breakers, regardless of whether Zs is met for ADS.