RCD/RCBOs DC Fault

[Mike893]

Evening all, after a bit of advise or reference to look towards.

We protect our installations with RCD/RCBOs to protect against earth faults and stop people getting zapped. However with smart homes on the rise, and most houses now having LED lighting, plus USB sockets. The RCD/RCBOs we generally fit will only protect against AC sine wave. Why are we not installing ones that protect against DC? And why if it is an issue hasn't it been pushed in the 18th edition?

For the benefit of the ones who want to massage their ego and pick faults in my question, lets assume I'm a week 1 day 1 apprentice and I got my ECS card out of the coco pops box

Cheers

Mike

 

[dodger421]

Are you sure that’s not the tripping curve for the overload side of the RCBO? I’ve only really come across type A or AC B/C curve RCBO’s.

The terminology just isn’t confusing enough these days ?

 

[pc1966]

The IET did a whole article on it which answers your questions better than I ever could;

That article has a mistake in it - it implies that all type-S (i.e. selective) RCD are AC types, which is wrong.

The 'S' part simply means there is enough of a delay to be selective with normal 30mA "instant" RCDs (which really means <= 300ms at 30mA or <= 40ms at 150mA) and that, along with having a factor of (typically) x3 in trip current, allows isolation of downstream faults that are slowly rising (i.e. current level allows selectivity) or sudden high-current faults (where the time delay allows the downstream RCD to open before the upstream one decides to act.

You can (and should) get type A RCD with selective time-delay characteristics for board incomers, etc.

 

[Dartlec]

I believe the DPC of AMD2 says that Type A will be the standard for general circuits going forward.

RCBOs with type B for RCD are quite rare, most are type B for the magnetic over-current trip.

Yes, that is a whole tin of WTF over the choice of identification!

I expect that soon you will only see type A RCD/RCBO being sold as the UK moves towards the same sort of guidance as the EU where type AC is pretty much banned due to the high likelihood of a DC component in fault currents.

However, it is not as simple as Type AC not tripping on a pulsed current DC/AC waveform, sometime they will trip but not always as specified (maybe higher current or slower, etc). The point is type AC is not warranted to work under those conditions, and can be blinded by high DC components saturating the sense transformer, so you should not be using them where equipment is rectifying the mains for its application, and these days it is a lot of stuff.

However, we also have the rather stupid situation of electrical wiring having to fit RCD capabilities suitable for general purpose sockets and no guarantee of what sort of a load can be attached to them!

I believe the DPC of AMD2 now says Type A is to be used for general circuits, and even Screwfix have some boards with type A now, so it looks like the issue will be resolved with time.

It will leave the whole issue of what code to give type AC rcds in EICRs - and there will be a lot due in 5 years with the rental requirements.

Plus who's to say that as manufacturers develop their products (aka cut corners or find ways to make things cheaper), that Type B won't be needed in 5-10 years, especially if it's a retrofit RCD covering several circuits.

 

[nicebutdim]

That article has a mistake in it - it implies that all type-S (i.e. selective) RCD are AC types, which is wrong.

The 'S' part simply means there is enough of a delay to be selective with normal 30mA "instant" RCDs (which really means <= 300ms at 30mA or <= 40ms at 150mA) and that, along with having a factor of (typically) x3 in trip current, allows isolation of downstream faults that are slowly rising (i.e. current level allows selectivity) or sudden high-current faults (where the time delay allows the downstream RCD to open before the upstream one decides to act.

You can (and should) get type A RCD with selective time-delay characteristics for board incomers, etc.

I've commented before about the odd situation where protective devices have overlapping designations for different characteristics - it's absurd.

I note that Lewden and Doepke use the respective designations SA & AS for their time delayed 'Type A' RCCBs, but also recently discovered that Wylex's range of 'WRMT' Type S RCDs are now available as both Type A and AC - seemingly under the same part number.