Can anyone explain what a time delay RCD is needed for?

[rewireIT]

I've scoured the internet for what these do and am coming up short on finding information.

What are these for? (ie what specific problem are they trying to solve) and why do they need to be time delayed?

Thanks in advance

 

[nicebutdim]

Time delayed element provides selectivity between this device and downstream RCD protection - without that delay, it's quite likely the upstream protection would trip and cut supply the entire premises.

If ADS could not be met, then this may have been the only means of eliminating (reducing) the risk of fatal shock. While there may be fewer circumstances in which they are required in modern installations, there's a solid argument for using them as additional protection on TT installations, where ADS can not be met and protection relies entirely on a device (or devices) that may fail.

 

[DPG]

I can't believe there's nothing on the Internet to answer your question.

But consider why you might not want an up front device to trip, but only the device that is feeding the circuit which has developed a fault.

 

[pc1966]

The most common reason for time-delay RCD is to achieve selectivity when you have a cascade of RCDs in a large system that needs them for some reason. Most obvious case is TT earth where you are unlikely to meet disconnection on over-current alone, but it could be for fire mitigation in agricultural situation where rats chewing cables in combustible areas is a risk.

So if you have a main DB that has an RCD incomer for that sort of reason and then your final circuits need 30mA non-delay RCD protection for shock protection then your up-stream RCD has to have two aspects:

• The trip current has to be high enough that the worst case tolerance of up/down (and any leakage from other multiple circuits) do not overlap. Typically this is a 3:1 ratio or so, hence the common options for 30mA / 100mA / 300mA, etc. This takes care of a slowly rising fault current as the down-stream is going to trip first.
• In addition you need to allow for a high fault current that exceeds both RCD's thresholds. In this case you need the down-stream RCD to open before the up-stream one decides to trip. That is what the delay is for, and typically it is around 200ms added for each stage (so 30mA+none, 100mA+200ms, 300mA+400ms, etc.)

EDIT: Note the "selective" (type S) RCD are all much the same +200ms sort of delay, you get ones called "delay" (type R, I think?) which are a level above, but usually when you get beyond instant/selective you are in to the realms of expensive adjustable RCD where you can dial in trip threshold and trip current.