RCBO curve

[Greenarmy]

I’m probably going to look stupid here, but it will be worth it if I get the answer , as I can’t fathom it out.
I’ve never really thought of it before, but…..
A circuit breaker comes in curves B, C, and D. Now we all know it’s mostly to do with the in rush of current ( layman terms) but the ELI value has to be lower to achieve tripping times.
However, with a RCBO , your tripping time will be around let’s say 18 milli seconds. So why are all RCBO not at least C type? What’s the point of a B type if you have already achieved the task of the device tripping in the desired time under fault conditions?
No doubt someone will tell me what I’m missing . I look forward to it .

 

[nicebutdim]

RCBOs incorporate two different types of protection which deal with distinctly different types of fault.

B, C & D curves deal with overload/overcurrent situations, with trip time varying. A high level of fault current (perhaps short circuit) will instantaneously trip the magnetic component, whereas a more moderate overload current might take some time to trip the thermal component which takes the form of a bi-metallic bending as heat increases. This part of an RCBO functions in the same way as an MCB.

The RCD component deals with earth leakage currents and should trip somewhere in the region of the figure you quote, when current leakage to earth approaches a set level (generally 30mA).

 

[westward10]

The old Federal Electric Stab-Lok RCBOs manufactured to BS4293 had no type characteristics. They had an overcurrent and residual current rating and that was it.

 

[Greenarmy]

Thanks for the quick response, but I’m still not convinced.
I understand how b, c, and d curves work. And the advantage of a d curve for example, ignoring an initial in rush of current , becomes a disadvantage when you have to achieve a lower ZS to achieve tripping times in a fault condition.
However, the rcbo will see the fault condition and disconnect in milli seconds would it not? Even in a live to neutral short circuit ?
So for example if you have a d curve rcbo , and your ZS is .1 ohms too high, changing to a c curve rcbo of the same amp rating, is not going to make it trip any quicker under fault conditions , as it should already trip in around 18 milli seconds.
Or am I being thick ?

 

[westward10]

If your Zs is too high it is perfectly acceptable to use the RCD side for fault protection.

 

[pc1966]

One of the major advantages of RCD protection is it is totally unrelated to the load's operating current - you can define the MCB arrangement to suite the load profile, and the RCD side to deal with faults to earth which could be a far higher impedance (e.g. in TT earth case).

Having said that, in reality it is not so simple as most RCD take 10ms or more to disconnect and often cannot cope with fault currents above 1-1.5kA, so for high energy faults you need the MCB side to offer faster disconnection & energy limiting, and to take the worst of the arcing on opening the faulted circuit so the RCD survives.

Also there is a reliability/safety advantage to also being able to disconnect on a hard fault using the MCB side just in case the RCD fails to operate and there is no upstream RCD, etc. But the regs do not demand it, just that you meet ADS times by some approved means.

 

[Pretty Mouth]

It's a good question. As you say, a L-E fault will disconnect either because of the magnetic trip, or the RCD, so disconnection times will be met regardless of B, C, or D curve.

L-N faults, however, will not operate the RCD, so rely on the magnetic trip. Some faults that would disconnect a B curve instantly may take several seconds for a C curve. I guess this could potentially be a problem from a thermal damage point of view.

 

[nicebutdim]

Thanks for the quick response, but I’m still not convinced.
I understand how b, c, and d curves work. And the advantage of a d curve for example, ignoring an initial in rush of current , becomes a disadvantage when you have to achieve a lower ZS to achieve tripping times in a fault condition.
However, the rcbo will see the fault condition and disconnect in milli seconds would it not? Even in a live to neutral short circuit ?
So for example if you have a d curve rcbo , and your ZS is .1 ohms too high, changing to a c curve rcbo of the same amp rating, is not going to make it trip any quicker under fault conditions , as it should already trip in around 18 milli seconds.
Or am I being thick ?

A short circuit would trip instantaneously, provided a sufficiently high fault current flows and I previously explained this.

Take a look at thermal trip times for RCBOs and tell me why you think a 16A RCBO would trip within a few milliseconds when handling an overload current of 26A.

 

[pc1966]

I guess this could potentially be a problem from a thermal damage point of view.

I think if the MCB's rating is at or below the cable's CCC then the trip time is not an issue, as they have similar/coordinated thermal overload characteristics. Meeting ADS time is a different matter of course!

 

[ipf]

If your Zs is too high it is perfectly acceptable to use the RCD side for fault protection.

As per TT system.

 

[westward10]

As per TT system.

Any system really.

 

[ipf]

Any system really.

True. It's down to the purpose of the RCD, I suppose.

 

[Julie.]

I’m probably going to look stupid here, but it will be worth it if I get the answer , as I can’t fathom it out.
I’ve never really thought of it before, but…..
A circuit breaker comes in curves B, C, and D. Now we all know it’s mostly to do with the in rush of current ( layman terms) but the ELI value has to be lower to achieve tripping times.
However, with a RCBO , your tripping time will be around let’s say 18 milli seconds. So why are all RCBO not at least C type? What’s the point of a B type if you have already achieved the task of the device tripping in the desired time under fault conditions?
No doubt someone will tell me what I’m missing . I look forward to it .

You need to think about each of the fault types individually, as the response is different for each case.

A L->L and L->N fault does not have to meet the Zs disconnection times, it has to not trip under inrush; but does for overload, and fault conditions. Ideally it should coordinate with both upstream and downstream devices - a higher instantaneous will be better for downstream coordination, lower would be better for upstream. Having a selection of "curves"* gives flexibility.

A N->E fault doesn't need to meet disconnection times, nor overload conditions, only fault (only realistically detectable by the RCD aspect). It would be good to achieve coordination with upstream and downstream devices, however this can only be achieved if the RCD aspect has a delay feature - which is contrary to the normal requirement.

A L->E fault needs to disconnect within the Zs time, but does not have to deal with inrush, or overload, again it would be good to achieve coordination with upstream and downstream devices, this can be achieved for the MCB characteristics, but for the RCD aspect this can only be achieved if it has a delay feature - which is contrary to the normal requirement.

In this case you can use the RCD aspect to meet the disconnection times if the MCB aspect doesn't meet it.


* the curve is actually identical in all the MCB characteristics, it is only the instantaneous "cut-off" point that changes.

Just having a single MCB characteristic would reduce the ability to coordinate with other devices.

 

[Greenarmy]

A short circuit would trip instantaneously, provided a sufficiently high fault current flows and I previously explained this.

Take a look at thermal trip times for RCBOs and tell me why you think a 16A RCBO would trip within a few milliseconds when handling an overload current of 26A.

Thanks to all for the input, much appreciated all the views.

I’m not really talking about overload current.
Let me explain slightly deeper.
The situation I was asked about was a few circuits had failed an EICR , because the D 32amp RCBO,s had a zs of around .5 when the maximum should have been around .4 ( I think)
The company doing the testing had suggested changing the RCBOs to type C , but both type of RCBOs would ( in my understanding) trip in milli seconds under fault conditions.
The cables supplying these outlets are 6mm SWA , so cable loading is not the issue.

Does anybody think changing to type C would improve the installation?
I’m not involved in the job, someone asked my opinion and it got me thinking, and I don’t think it would change matters ( except maybe trip on the inrush of current ) Apparently the RCBOs are £100 a go and he would need around 10 of them. ( I don’t know the make). I just don’t want to give him the wrong advice, and also want to see if I’m missing something.
Cheers again folks, hope I’ve explained it in more depth.

 

[westward10]

This is a common area to be incorrectly noted on an EICR, there is no issue whatsoever. It should be itemised fault protection is afforded by the RCD. Unfortunately our industry is plagued by non competent people carrying out these works.