Calculating maximum Za for Time Delay RCDs

The R_A x I_Δn≤50V requirement is based soley on the rated residual current - so the time delay doesn't come into it. So calculate just as you normally would.


That bit really only makes sure that leakage currents (rather than a hard L-PE fault) can't raise the earthing system to a hazardous voltages (due to likely high earth impedances with TT systems) without anything tripping eventually. There's a separate requirement to ensure disconnection within a given time for normal ADS requirement (which would involve making sure that Zs is low enough to trip the RCD within the required time, based on Uo, taking delays into consideration) - but it's unusal for that to be more onerous than the 50V limit above.


Table 3A (lower part) is for S-type i.e. "selective" RCDs - non-adjustables ones that should discriminate with downstream instantaneous RCDs. Adjustable RCDs are different and you'd have to use numbers based on the setting rather than anything in table 3A.


   - Andy.

Andy,


That’s a great explanation and I think ive got it. 


Say my earth electrode was 20 ohms.


My upfront adjustable RCD is rated at 500ma, with a time delay of 500ms. Although, I may be able to move this back to 1s due to being a TT sub main, but I may be hesitant to do this as the delay is 1 sec and I’d assume there’d be an inherent trip time of the rcd taking it above 1s.


My intermediate RCD is an S type 300ma. My electrode resistance is low enough to achieve the 5x mA figure in the table above and therefore trip in 250ms.


For final circuits I’ve got a 30ma which will trip in 40ms.


I should therefore have a discriminating system. Although, the 300ma rcd is not 3x more than the upfront 500ma rcd. What are your thoughts on this?

Although, the 300ma rcd is not 3x more than the upfront 500ma rcd. What are your thoughts on this?

There's a small risk of lack of discrimination there - since RCDs usually can trip anywhere between >0.5x and ≤1x their rating, so the 500mA unit (if enthusiastic) could trip with residual currents above 250mA where the 300mA unit (if lazy) might not trip until 300mA - so if you had a residual current in the 250-300mA range (say due to moisture getting in somewhere or cumulative protective conductor currents if you have multiple downstream RCDs) there's a possibility of the 500mA unit tripping rather than the 300mA one.


   - Andy.

Thanks for all the help so far Andy.


Last question, what are your thoughts on going above 500ma to avoid your example above. Schneider show an example going to 1a in their guide design efficient earth fault protection with RCDs, however I cannot see anything in BS7671 above 500ma, this is only for a sub main and the 50v would still be adhered too?


thanks again

Last question, what are your thoughts on going above 500ma to avoid your example above. Schneider show an example going to 1a in their guide design efficient earth fault protection with RCDs, however I cannot see anything in BS7671 above 500ma, this is only for a sub main and the 50v would still be adhered too?

As far as I know there's nothing in BS 7671 to stop you using any rating of RCD you like (provided they comply with the specified product standards). Certainly the tables of helpful information shouldn't be read as limiting your choice to just those values. 1000mA is unusual though - you'd have to ensure that Ra was reliably below 50Ω - which might be a challenge in some circumstances (especially given seasonal variability of soil resistance).


The other option is to check manufacturer's data carefully. It's quite common to find that manufacturer's data gives much tighter tolerances than the product standards - so it's possible you might find some brand does offer proper discrimination between their 500mA and 300mA devices for example.


  - Andy.