RCD disconnection times on a TT system

The 30mA figure is certainly slower than is usually achieved. What is the Ze of the installation, because I suspect that it is towards the upper end of the range that one usually sees? It does appear to comply, but I would try another RCD to verify my result.

Can we state that because the 5x test trips at less than 40ms it complies with fault protection, or does the 30mA disconnection time also have to comply?

As with a TN system you need to ensure that disconnection will occur within the required time for a L-PE fault of negligible impedance. So things depend a little on the actual loop impedances involved - but in a typical setup where Ra is under 200 Ohms you'd have an earth fault current of in excess of 1A - so the 5x figure should be adequate with a large margin of safety.


   - Andy.

That figure does seem very high.

Longest time I`ve ever witnessed on an RCD (TN system) was about 90mS (at x1 on a 30mA) and was alarmed at that. For practical reasons the circuit could not be powered down for more than a few seconds but a very quick swap of two RCD circuits revealled the long time stayed with the circuit not the RCD and there were many fridges and freezers on that circuit. I asked my night school 2380 tutor that evening if he agreed with me that the fridges/freezers total was causing this extra time and he that quite likely. A few days later I was able to disconnect all appliances from these circuits and the readings were as I expected . At that time we were taught that circuit itself should be left connected and all appliances disconnect in order that the circuit wiring comes into that test (if a ring then at near midpoint too) . The bloke who was testing was on the 2391 course at the time and his tutor agreed likely too.

I do not do many TT installs.

I`d suggest trying another RCD if possible and also a different test meter.

Although BS (EN) allows 300ms and the old BS allowed 200ms I never saw any RCD that was not far faster than both allowed on a non delayed RCD.

I think your reading obtained would cause me concern.

Having said that I have tested only a limited range of makes and it`s possible that some other makes do tend to have slower times

You say it’s a Type A RCBO, so my first thought is Hager.


My second thought is, are you using a tester with a Type A test?


Andy Betteridge

How many times was it subject to test? Many times I have got +200ms on x1 or even no trip on first assault. Hit the test button a few times and then different results. If you have already done this then given your results, there is nothing to be concerned about.

Did you test the device properly, that is with its outgoing wiring disconnected? As the tester doesn’t really know when the red opens, it has to look for the voltage to fall away. If you have something that retains voltage, such as something with capacitors in, or long lengths of cable, this will trick the tester into believing the rcd has opened late. 


Regards,


Alan.

Thanks for all the replies! I had to go out on site Friday so didn't have a chance to read through until this morning.


To answer some of the questions, it wasn't ourselves that tested it so have to reply on the customers response:

The switchgear is all Eaton type A RCBOs

The RCBO only supplies an outdoor 16A socket, so no loads were connected at the time of test.

There are numerous other RCBOs of the same type, all giving timings of 160 to 199ms.

All Zs readings are "well below maximum permitted values"

They were tested with multiple test instruments, getting the same readings, so were cycled many times.


 

disconnection will occur within the required time for a L-PE fault of negligible impedance.

I had missed those last two words, so as the 5x tests are all under 40ms I don't think we have a problem.


Many thanks again