RCBOs

[rumplestiltskin]

Just started looking at these as July is approaching fast.
They seem to be sold commonly as type A RCDs with type C MCBs.
I gather the curve for the MCB is not so important for personal protection as the RCD will do that job quicker.
As the C curve will give less nuisance tripping is this then going to become the norm for domestic installations instead of the more usual B? or have I got the wrong idea?

 

[JohnD]

My guess is that it is because RCBOs have in the past been more often used in Industrial and Commercial installations.

Of course, if you buy MEM, their unique Pod design means you can assemble an RCBO for any rating combination you desire

(I have no connection with the company other than as satisfied user)

Although a C-curve MCB should be more resistant to nuisance tripping (domestically, this is generally when a filament lamp burns out) I have not found it helps. And you have to test the EFLI before going to a C.

p.s. RCBOs are great!

 

[Spark123]

I gather the curve for the MCB is not so important for personal protection as the RCD will do that job quicker.

In a normal TN installation the RCD part of the device is only providing supplementary protection, the circuit still needs to be designed for the MCB part to cause automatic disconenction.
The RCDs in these are normally electronic, in under fault condition there may not be enough voltage present to drive the RCD circuitry stopping the RCD side of it from working correctly.

 

[Taylortwocities]

Hmm, been thinks about that.

What's the problem with the EFLI measurement when using an RCBO. It will trip if the earth fault current exceeds 30mA. Way before the the rated trip current.
Maybe we should be checking that R1+Rn is low enough to achieve 0.4sec trip and ignore R1+R2?

 

[Spark123]

Doesn't matter as much about R1+Rn, you need to ensure it disconnects quickly enough under phase - earth fault condition. The RCD may be inoperable if there is a dead short p-e, i.e. there is not enough voltage there to drive its electronics.
Best way to do the measurement is with a no trip loop tester such as the Megger MFT 1552
Failing that, Zs = Ze + (R1+R2)

 

[JohnD]

I meant, if you change a "B" MCB for a "C"

 

[Spark123]

The same rules apply as they do for changing a type B for a type C MCB, the efli needs to be checked to ensure it disconnects the MCB side of the device.

 

[securespark]

The EFLI values for C types are half that of the B's, so may circuits will not achieve a low enough EFLI figure to allow a swap to a C type.

 

[rumplestiltskin]

Hmmm....food for thought.

With an installation with a seperate RCD and MCB it would make sense to ensure the EFLI was low enough to suit the C MCB in case the RCD was ever removed or bypassed (note - the EFLI should always be checked at this point). But an RCBO is not an RCD and an MCB, it is an individual device which performs both functions, and so the RCD function will always beat the C curve function of the device on an earth fault.

There is a valid argument (in my opinion) that if the EFLI is too high then the RCD won't trip, and so this deinitely matters, and that very high value needs to be dealt with. However I'm not seeing a change from a B to a C making any effective difference to the trip time as the RCD function (if it is going to work) will always beat the MCB function on an RCBO.

 

[JohnD]

RCDs sometime fail to trip

Much more often than MCBs

And in e.g. the MEM RCBO, it is definitely made in two parts - the RCD Pod clips onto an ordinary MCB off the shelf (though of course you can buy them ready-assembled)

 

[Spark123]

There is a time when the RCD will not function as the supply has collapsed under fault condition - it is electronic and needs a supply to function.
The Max efli for a 30mA RCD is normally 1666 ohms, however as the scenario given above exists the MCB needs to perform the function of disconnecting the supply in the event of an earth fault when the supply collapses which is why the max efli figures given in BS7671 are for the MCB side of the device.
Type B to type C is talking about the magnetic trip of the MCB side of the device, often used where there are heavier start up currents i.e. transformers, lots of fluorescent fittings etc.
The max efli of a type C as Securespark says is half that of a type B.
I am wondering if you are confusing type A RCD, type A indicates the device will trip on AC or pulsating DC leakage.

 

[rumplestiltskin]

No I'm not considering the type of RCD.

Yes I know the difference between B and C curves.

I would agree RCDs (and at this point in time the RCD function of an RCBO) fail more often than MCBs so there is a valid case for the underlying EFLI to be correct for the type of MCB or RCBO fitted, although this leads me to wonder why major suppliers are supplying C curve RCBOs with no option to specify the type you want. For example order this www.tlc-direct.co.uk/Products/WYNSBS32slash1.html and you will get an RCBO with a type A RCD function and a C curve MCB function.

I'm not really too sure what is meant by the supply collapsing under fault conditions. Would the MCB still trip (or have tripped) in these conditions?
Is this a L to N short circuit? I'm assuming a L to E short would have tripped the RCD anyway.

Not that it really forms much of this discussion, but as far as I am aware the RCD pod which clips onto the MCB is not removeable (so it says but have never tried it) and so it becomes one item when fitted.

 

[JohnD]

not removeable

unless you have the special skills and the special tools

you're right, they snap together and are not intended to be separated once they have been assembled. you would normally damage one or both parts if you tried to separate them.

 

[Spark123]

I'm not really too sure what is meant by the supply collapsing under fault conditions. Would the MCB still trip (or have tripped) in these conditions?
Is this a L to N short circuit? I'm assuming a L to E short would have tripped the RCD anyway.

I mean the electronic circuit may require somewhere in the region of 50v for it to remain operational. If there is a dead short L-E then the voltage of the CPC can rise, and the voltage of the phase conductor can fall owing to the impedances of the conductors for the duration of the fault, this may cause the 50v to be no longer be available to the circuitry.
The MCB part is not affected as it does not require a supply voltage to operate, as you should have ensured the EFLI is low enough to cause enough current will flow around the circuit the MCB side of the RCBO should disconnect the supply.
I have come across this type of electronic RCD before which has failed, possibly owing to a surge.

 

[rumplestiltskin]

There should still be enough L to N voltage which operates the RCD function, so I'm not sure that scenario would necessarily cause the RCD function to fail to operate. There is a possibility with TN-C-S that this situation could occur, but I've found RCDs always operate on a short to earth.
Also the situation described would not be affected by replacing a B curve with a C curve.