not quite the same question but at say 5 times the nominal trip current,  100ms is typical , and less than 50ms is a fail and 150msec is also. From here  

If cascading earth fault relays I tend to use a rule of 3 for current and time as a starting point for modifications.

so instant 30msec fed from sub-mains supplied by perhaps 100A circuits with protection set to 100mA 0.1second, fed by generator of a few hundred amps, earth protection set to 300mA and 0.3 seconds. There are always exceptions, but it is a place to start from .

mike

thanks @mapj   -   i wonder how that table relates to this: 

when I asked BG for the actual built in “time delay” setting, with regard to  their 100mA type-s,  they supplied 34 - 36 ms.  

There’s a table towards of the back of the regs book , which I don’t have with me, that gives the range of the standard time delays.

Sparkingchip: 
 

There’s a table towards of the back of the regs book , which I don’t have with me, that gives the range of the standard time delays.

Page 363. S type 130 to 500mS. (100 to 500mA)

Z.

@zoom  that's the trip time range I believe (of which I was aware) ?   The actual applied time delay is what I was referring to   (see my comment about the figure provided by BG for their device).

I will have another check in Regs.  but I think it might be one of those manufacturer things along side ensuring it all trips out within that specified time range.

The EIC seems to ask for the rated time delay (like one might set on an adjustable unit) … but that's not the same as tripping time range in my mind ( or is it ? ).

re: the table posted by @mapj -

is the ‘non-actuating’ value effectivel ythe  max. allowed ‘delay’, so that for example, 130ms is the max delay that can be applied at Idn?

edit: to answer my own question, “no it isn't”  as it appears to be the trip time range  that is in BS7671.  

I wonder if that box was intended for the likes of big boy's RCDs - earth leakage relays or add-on modules for MCCBs - that can have a definite (and likely constant) delay time. 

S-types are just meant to be selective with instantaneous RCDs so often have a delay that varies with the residual current - so there is no one number you can sensibly write down.

I'd almost be tempted to write “S-type” and cross out the "ms".

   - Andy.

@AJJ 

I think that is favourite …   where a time delay can be selected (adjusted).

Interesting that BG gave a figure when I asked though (see previously) and it wasnt current dependent (or at least they did not mention)… because perhaps of it's main purpose for simple selectivity.

100 Amp 100mA Time Delayed RCD 2 Pole - Wylex

5013601017570

WRMT100/2

RCD terminal capacity,
35 mm

Wylex
100A 100mA Time Delayed RCD
2 Pole

100mA time delay

Suitable for use when total RCD protection is required to supplement
the system earthing and where local 30mA RCDs are used to give
additional protection against direct contact.

The time delay RCD will discriminate with the 30mA RCD.

• All Wylex RCDs incorporate a filtering device to provide protection against transient surges in the supply to the unit, thus reducing the occurrence of unwanted tripping.
• Wylex RCDs must never be used as the sole method of direct contact protection, but are invaluable in providing supplementary protection in high risk environments where damage may occur.
• The operating dolly may be locked in either the ON or OFF position without affecting the ability of the trip mechanism to operate, i.e. the RCD is 'trip-free'. It is not possible to hold the contacts closed when a fault condition exists.
• When in the OFF position a contact gap of 4 mm is present, enabling Wylex RCDs to be used as isolating switches where appropriate.

Well, if the S type is slower in operation than 30mA instant types downstream, who cares about the actual trip time. As long as it is slower than say 40mS or longer?

Z.

was it worth posting  all that @zoom , to then type “who cares about the actual trip time” .  :-)

My question that I was caring about, was in relation to the space on the EIC for "rated  time delay" and it seems reasonable to conclude that it is not really intended for s-types -  so in that respect it's sorted.