TN system - Max Zs Vs tripping time in the Presence of RCD protection

No you can't completely ignore Zs, but using an RCD it can be allowed to be much higher than if relying on overcurrent devices. E.g. to guarantee a 30mA device will trip in 0.4s you need at least 30mA to flow and hence max Zs needs to be below 230V/30mA = 7666Ω. (For a TT system with a 0.2s requirement, you'd be looking at needing 60mA and hence about 3.8kΩ). There's also a requirement to keep the touch voltage below 50V where the fault/leakage current is too low to guarantee tripping the RCD - so Ra < 1667Ω (which can be taken to mean Zs < 1667Ω if you don't have any more particular data).

So yes, small excess values of Zs might be dealt with by adding a suitable RCD.

Do check if Zs is reasonable for the circumstances - on a short circuit a high Zs may well be caused by corrosion or a loose connection - which means it can't be guaranteed, even to stay within the generous RCD limits. Also check that voltage drop limits aren't exceeded.

    - Andy. 

RCD trip time is 20ms 

Just to add that test results can't really be used for calculations - just because a device tripped in 20ms a week last Wednesday in warm dry weather doesn't mean it'll necessarily do the same under different conditions (cold, damp, the device 5 years older and not been exercised for ages) - all the test really tells you is that it's likely to be working within specification - so you still need to use the general data for calculations - i.e. 300ms at 30mA, 150ms at 60mA and 40ms at 150 or 250mA.

  - Andy.

For reference the worst case Zs for RCDs can be found in BS 7671 in Table 41.5.

Thank you very much

Thank you so much

Seems like a bad design to rely on (RCD) Residual Current Device unless there is no other option,I have run parrellel cpc in the past, does it need to be C curve.

Also you can't use 100% values of 1.09,you need 80% values at 0.87 unless your conductors are at 70C

Newfutile said:

Seems like a bad design to rely on (RCD) Residual Current Device unless there is no other option,I have run parrellel cpc in the past, does it need to be C curve.

Why? Since 2008 (introduction of 17th Ed), there has been no problem using RCD for automatic disconnection of supply, and Table 7.1(ii) of the OSG - which before 2018 used to be 7.1(i) - includes a "with RCD" max circuit length, as well as a "without RCD".

Provided fault current protection is OK (adiabatic), and volt drop, there really is no problem as long as the RCD is correctly selected and is of a type that provides fault current protection (some voltage-dependent RCDs, with marking E2 or E3, whilst they do provide additional protection, might not provide protection for ADS when the voltage collapses during a fault of negligible impedance, this being one reason why the FE lead is present on the ones that do ... also, as well as their position in the circuit, a reason why in general SRCDs and FCURCDs are only suitable for additional protection, not ADS).

Or, instead of RCDs, perhaps employ supplementary protective bonding.

There also used to be an acceptable method of limiting the volt drop along the cpc to 50v as referenced to the connection of the main protective bonding thereby allowing extension of disconnection times to 5s. Perhaps that fell out of favour as MCBs (0.1s disconnection) replaced fuses in many circumstances.  

lyledunn said:

method of limiting the volt drop along the cpc to 50v as referenced to the connection of the main protective bonding

This might sound very pedantic, and I mean no disrespect to lyledunn , but I feel it's important to correct a couple of points to avoid common misconceptions about supplementary protective bonding.

Using this method (Regulations 415.2 and 419.3), the voltage is not actually limited to 50 V for every possible fault.

It is limited to 50 V for currents up to and including the current which causes operation of the protective device (or, where multiple circuits share the same protective bonding, the largest of the protective devices) within 5 seconds.

Example: I have a B32 circuit-breaker to BS EN 60898. The current guaranteed to cause disconnection time within 5 s is 160 A, therefore supplementary protective bonding at the limit specified in Regulation 415.2 is guaranteed to limit to 50 V for fault currents up to 160 A. For fault currents exceeding 160 A, the touch current will exceed 50 V, but the circuit will be disconnected within 5 s.

Certainly, where supplementary protective equipotential bonding is applied in cases where a disconnection time < 5 s cannot be met (e.g. 0.4 s), but a higher disconnection time of 1 s or 5 s could be met, the touch-voltage is not guaranteed to be limited to 50 V.

As Lyle points out, supplementary protective bonding is not really useful for B curve and C curve circuit-breakers, but is useful for fuses and D curve circuit-breakers.

lyledunn said:

thereby allowing extension of disconnection times to 5s

This is also not the case. There is no maximum disconnection time to be applied for protection against electric shock  where supplementary protective bonding is applied according to all of the requirements of BS 7671 (at least in "dry condition"). This is because, for currents below that causing 5 s disconnection time (i.e. leading to longer disconnection times), the touch-voltage does not exceed 50 V by virtue of the formula in Regulation 415.2.

Disconnection within a certain time for protection against overcurrent (in this case fault current) may, however, still be necessary dependent on the characteristics of the overcurrent protective device for the circuit, and the cross-sectional areas and materials of the line conductor and cpc.

I know BS7671 allows this, perhaps bad design was a poor choice of words.i would still argue that meeting disconnection times with the mcb is a better design,and have rcd for additional protection.

I always assumed that BS7671 permitted rcd for ADS so that TT systems can be used.

Most remedial repairs I do are for failed rcd, probably because no one has ever pressed the test button.