More 'RCD Type' uncertainty/confusion

[ericmark]

From reasonable early days I knew there were different types of RCD, however I thought the type A was designed to work with

a simulated sign wave, and was under the impression used in narrow boats and the like, and also the whole idea of the active RCD was so if there was a volt drop it would fail safe, a RCD used with a 55-0-55 volt supply may fail to work as the voltage is too low, and finding a RCD rated at 55 volt to earth was not easy.

However it seems I was wrong, and with solar panels and electric cars we have a large DC device which under fault conditions could impose DC onto the mains supply, pulsed DC has been with us for years, there are many devices that used half wave rectifiers, what is new is some thing which under fault conditions can put a large smooth DC current onto the mains, I have seen this once, with a ward leonard control on a crane, here the regenerative braking when dropping the load caused a problem with the cycles shown on the frequency meter hitting over a 100 Hz, but from the sound of the engine the Cat diesel was clearly not doing over 6000 RPM it was more like 3000 RPM so clearly not really 100 Hz, however the fault caused the brake not to engage and the crane dropped its load, which resulted in all the tower cranes on building of T5 being stood down until the fault was found.

This was at the same time as my accident so I did not find the outcome.

However it is clear like electric cars and solar panels motors can also generate DC which can end up on the supply, however ward leonard is not used that much and not in domestic, it may be the odd lift could cause problems, but in the main looking at solar panels and car batteries. I can't thing of anything else with stored energy, but as said the John Ward demo raises more questions than it answers, I had thought type B worked with DC but he put a 50 mA and 250 mA current through the RCD without it tripping, I suspect maybe due to no voltage to work electronics, but clearly from the demo they both failed to work at 30 mA. Yes one worked with 30 mA AC, but not with 250 mA DC.

But it seems likely unless there is some stored energy, be it a crane with a suspended load, car battery, or solar panel, the chance of have smooth DC in a domestic installation is very low. But in an industrial environment with motor drives then it is very different. OK we do have some in domestic, washing machine and refrigeration both use inverter drives, don't think too much of a problem with refrigeration but a washing machine can take some time to stop, so it could delay a RCD tripping by 3 minutes or so.

So if we have a earth leakage at the same time as doing the washing then it could cause the RCD not to trip. But with anything with regenerative braking one has to ask what happens if a RCD trips? Would fitting a RCD be dangerous?

 

[ericmark]

From reasonable early days I knew there were different types of RCD, however I thought the type A was designed to work with View attachment 229579a simulated sign wave, and was under the impression used in narrow boats and the like, and also the whole idea of the active RCD was so if there was a volt drop it would fail safe, a RCD used with a 55-0-55 volt supply may fail to work as the voltage is too low, and finding a RCD rated at 55 volt to earth was not easy.

However it seems I was wrong, and with solar panels and electric cars we have a large DC device which under fault conditions could impose DC onto the mains supply, pulsed DC has been with us for years, there are many devices that used half wave rectifiers, what is new is some thing which under fault conditions can put a large smooth DC current onto the mains, I have seen this once, with a ward leonard control on a crane, here the regenerative braking when dropping the load caused a problem with the cycles shown on the frequency meter hitting over a 100 Hz, but from the sound of the engine the Cat diesel was clearly not doing over 6000 RPM it was more like 3000 RPM so clearly not really 100 Hz, however the fault caused the brake not to engage and the crane dropped its load, which resulted in all the tower cranes on building of T5 being stood down until the fault was found.

This was at the same time as my accident so I did not find the outcome.

However it is clear like electric cars and solar panels motors can also generate DC which can end up on the supply, however ward leonard is not used that much and not in domestic, it may be the odd lift could cause problems, but in the main looking at solar panels and car batteries. I can't thing of anything else with stored energy, but as said the John Ward demo raises more questions than it answers, I had thought type B worked with DC but he put a 50 mA and 250 mA current through the RCD without it tripping, I suspect maybe due to no voltage to work electronics, but clearly from the demo they both failed to work at 30 mA. Yes one worked with 30 mA AC, but not with 250 mA DC.

But it seems likely unless there is some stored energy, be it a crane with a suspended load, car battery, or solar panel, the chance of have smooth DC in a domestic installation is very low. But in an industrial environment with motor drives then it is very different. OK we do have some in domestic, washing machine and refrigeration both use inverter drives, don't think too much of a problem with refrigeration but a washing machine can take some time to stop, so it could delay a RCD tripping by 3 minutes or so.

So if we have a earth leakage at the same time as doing the washing then it could cause the RCD not to trip. But with anything with regenerative braking one has to ask what happens if a RCD trips? Would fitting a RCD be dangerous?

Isn't that the rub though?

I found a video that describes it perfectly!

Does that video really explain what is going on? it is same rig as John Ward used it shows that 250 mA failed to trip a type B.

 

[JohnW2]

Isn't that the rub though? I found a video that describes it perfectly!

That video, and particularly the demonstration, is almost identical to flameport's (John Ward's) one which we have already discussed, and I have exactly the same reservations about it. In particular...

Firstly, the constant ('smooth') DC current is totally 'unbalanced', since it only goes through the RCD in one direction. If, in a real-world situation, it were flowing through the L& N supply conductors (hence RCD) in opposite directions (in addition to the AC), the electromagnetic effect of the DC current in the two conductors would more-or-less cancel ... and, as I've said, I'm not really sure what real-world situations would/could result in an 'unbalanced' DC component analogous to what was done in that 'demo'.

Secondly, as I've said, explanations of the functionality of Types A, F & B RCDs such as that in BS7671 say nothing about the effects of superimposed DC current on the ability of the device to trip in response to a residual (~sinusoidal) AC current (AC L-N imbalance). However, in terms of 'residual pulsating DC current' it says that (30 mA) Types A, F and B will trip (with 30 mA 'residual pulsating DC current') in the presence of superimposed DC of "up to" 6 mA, 10 mA and 12 mA respectively - the implication presumably being that they may not trip (with 30 mA 'residual pulsating DC current') if the superimposed DC current is greater than those (very small) figures.

That seems to mean that the video was showing that a the ability of a type AC RCD to trip in response to a 30 mA AC residual current was impaired by a superimposed (and 'unbalanced') DC current 20 times more than the current which might stop a Type F from working properly and 40 times more than the current which might stop a Type A one working properly - which results in my response of "So What?" ... or am I misunderstanding or missing something?

Kind Regards, John

 

[Swwils]

Now we know why Type B cost so much! DC current detection and earth leakage current at frequencies higher than 50/60Hz.

RDC-DD tripping at 6ma not going to offer the greatest continuity of service in comparison.

 

[JohnW2]

Now we know why Type B cost so much! DC current detection and earth leakage current at frequencies higher than 50/60Hz.

As I've said before, it would not surprise me if there has to effectively be two (or more) 'linked devices' within a Type B, since I'm not sure that it would be possible to create one mechanism that could detect all the things it is meant to be able to detect!

Kind Regards, John