First question - Yes, 5 s is the disconnection time, use Table 41.6 where OCPDs are used for ADS, or effectively Table 41.5 where RCDs provide fault protection (see 411.8.3 and page 40 of the 2022 edition of Guidance Note 5 Protection against electric shock). Need to make sure the RCD (and particularly its test button) are rated for the RLVS system.

Second question - Not sure why we are risk assessing? It's a requirement of BS 7671 to be able to achieve 5 s disconnection time, AND meet the requirements for protection against overcurrent for all possible faults of negligible impedance, e.g. for single-phase that would be L1-L2, L-PE and L2-PE.

Thanks Gary, 

The challenge has came to evaluate the risk since the owners simply don't want to isolate (Which I disagree with), however they're coming back with class II appliances etc. Which regardless won't protect against a L1-L2 as a stop gap until double pole RCBOs can be sourced, the phrase used by the decision makers was "Low likelihood of fault in a short period of time"

It sounds to me like the 'decision makers' have already risk assessed the job, so why undertake a risk that is theirs to own?

Personally I'd run a mile.

If by RLV you are referring to the split phase 55-0-55 supplies found on building sites etc then when they were devised the safety case was intended not to need such reliable ADS - the maximum voltage exposed to ground is 55v (though maybe higher in some 3  phase derived systems if the two lives are not from the same phase) In any significant fault the exposed voltage is pulled below 50V then in principle from a safety of life point of view ,then there is no need to disconnect ever - there is a reason that RCD currents and rod resistances are coordinated around a 50V rise ;-)  The ' turn off in half a heartbeat or fibrillation starts' consideration that drives the safe-ish ADS times in a normal 230V TN system does not really apply. So we are left with damage or injury from things getting hot. Hence the 5 seconds, not 0.4 or 0.8 or something. It was also agreed that there would not be a distributed neutral, and all loads would be wired live to live.

Since all that was decided attitudes have changed, and RCDs are now de-riguer,  and appear on these 55-0-55 systems even though the actual fatal shock risk is much lower. An RCD is likely to be much faster than you need anyway.

Mike.

If formula for (K^2 * S^2)/(I^2) is satisfied

If disconnection isn't happening within 5s then the adiabatic starts to loose accuracy. Generally things are OK when In ≤ Iz for both overloads and faults (providing of course the breaking capacity of the protective device is suitable for the PFC).

  - Andy.

AJJewsbury said:

If disconnection isn't happening within 5s then the adiabatic starts to loose accuracy.

Agreed ... but it over-estimates erring on the side of safety.

mapj1 said:

Since all that was decided attitudes have changed, and RCDs are now de-riguer,  and appear on these 55-0-55 systems even though the actual fatal shock risk is much lower

There has never been a fatal electric shock on such systems since their introduction (Paul Cook Commentary on the Wiring Regulations 17th Edition)