RCBO 5In trip times.

[ste mac]

Just installing a new consumer unit for a customer. Split load board with 1 half RCD & the other half RCBO's. RCD & all but 1 RCBO has had good trip times, all sub 40ms at 1In & 5In, but the 6A RCBO for the kitchen lighting circuit is tripping at 40.8ms & 40.7ms at 5In. If I disconnect the outgoing live & retest the the trip times are 18.5 & 17.2. Anyone shed any light as to why this is happening & is it acceptable to test the RCBO without the outgoing ways connected?

 

[Guest55]

........ is it acceptable to test the RCBO without the outgoing ways connected?

Yes , and its the prefered method anyway.
Theres a good chance electronic transformers / ballasts are buggering up the readings at the first attempt at testing so youre good to go with recording the 18 / 17 ms trip times

 

[somersetsparks]

^^^^agreed, RCBO's should always be tested at their outgoing terminals with loads disconnected

 

[Jimmy Boy]

Biff/Somerset..Why ? Where is this written I can't say I have seen this anywhere although granted I don't own every GN book published, this is a serious question not a dig.
J

 

[Guest55]

Biff/Somerset..Why ? Where is this written I can't say I have seen this anywhere although granted I don't own every GN book published, this is a serious question not a dig.
J

No probs with asking jimmy lol.
The rcd manufacturers recommend testing of the devices in this way for a start.
Plus the test is just to acertain the trip time of the rcd , so theres no need for any other part of circuit to be connected.
And finally , electronic gizmo's can interfere with the readings , as shown by ste macs query.

 

[Guest55]

ps , i dont own any gn book lol.

 

[Crosswire]

Sounds like a bad earth on the kitchen lighting circuit to me.

Am not sure why you are disconnecting the lighting circuit to test the RCBO. I thought you had to measure disconnection times from the furthest point ( if not every point) on the circuit, so as to test the cabling of the entire circuit, and not simply the RCBO on it's own.

Or am I missing something?

 

[Guest55]

Sounds like a bad earth on the kitchen lighting circuit to me.

Am not sure why you are disconnecting the lighting circuit to test the RCBO. I thought you had to measure disconnection times from the furthest point ( if not every point) on the circuit, so as to test the cabling of the entire circuit, and not simply the RCBO on it's own.

Or am I missing something?

Er no.
Why test through loads of cable / switches / fittings which can lead to inaccuracies when you can just do it at the CU ?
Have a read of my post #5 for more info.

 

[Crosswire]

Surely if you test at the CU you will get a perfect reading every time?

If you test at the end of the circuit you will ensure a) the earth is continuous and b)the EFLI is low enough to operate the device in the required time.

Am not sure why you would want to ensure that the RCBO is fine, whilst the circuit and accessories that the device is meant to protect may harbour a fault that causes the (perfectly fine) RCBO not to operate in the required time?

I am not trying to be confrontational, I am simply confused as this is the first time I have encountered this method of recording trip times. I always record from the end of line.......

 

[Jimmy Boy]

Well that was my thoughts as well .
J

 

[Guest55]

Well you asked , i explained , do as you wish , makes no odds to me.

 

[Guest55]

Surely if you test at the CU you will get a perfect reading every time? Only if the rcd is functioning correctly , if its not then it wont pass the test no matter where you do the test.

If you test at the end of the circuit you will ensure a) the earth is continuous and the EFLI is low enough to operate the device in the required time. Eh ? this is a rcd test , what EFLI got to do with it ? that test is done seperately ?

Am not sure why you would want to ensure that the RCBO is fine, whilst the circuit and accessories that the device is meant to protect may harbour a fault that causes the (perfectly fine) RCBO not to operate in the required time? Then you have a problem with the wiring or accessories , not with the rcd , and you would need to investigate this seperately - your confusing what the test is for to be honest and the rcd is there to protect the user / occupier not the accessories.

I am not trying to be confrontational, I am simply confused as this is the first time I have encountered this method of recording trip times. I always record from the end of line.......

.......................

 

[Crosswire]

Don't know what to say to that........No offence was intended, purely questioning a technical point.

 

[Guest55]

No was offence taken , just trying to answer all your questions.

 

[Crosswire]

P.S. If the device is operating correctly with no circuit connected, but fails to operate in the required time with the circuit connected, the device has become pointless...Why have disconnection device that doesn't disconnect (in the real world) ?

I would check the earthing of this circuit again.

 

[Guest55]

My turn to ask a question ,
Why would earthing issues affect the trip time of a rcd ?

 

[Crosswire]

As a path for current leakage, which would imbalance the coil. A high impedance on this path (or no path) would reduce ( or negate) the ability of the device to operate in the required time.

No?

 

[just add water sparks]

Surely if you test at the CU you will get a perfect reading every time?

If you test at the end of the circuit you will ensure a) the earth is continuous and b)the EFLI is low enough to operate the device in the required time.

Am not sure why you would want to ensure that the RCBO is fine, whilst the circuit and accessories that the device is meant to protect may harbour a fault that causes the (perfectly fine) RCBO not to operate in the required time?

I am not trying to be confrontational, I am simply confused as this is the first time I have encountered this method of recording trip times. I always record from the end of line.......

I agree most of these regs are ambiguous,its good to question a lot of things as it shows we are thinking and not following like sheep but testing at the farthest point is not necessary.Quote from wiring matters Spring 2010 "Whether the RCD in question protects a number of circuits,eg on a split load board,or an RCBO is used to protect a single circuit,or an RCD is installed nearer the item of equipment being it is protecting,the tests can be carried out immediately on the load side of the device.Often it is suggested that an RCD protected circuit should be tested at the extremity,ie. at the same point at which the Zs measurement is taken.There is LITTLE TO BE GAINED in testing the RCD at the extremity of the circuit it is protecting.Firstly , the test is required to prove the RCD operates correctly.As the circuit in question will meet all other requirements ie earth loop impedance and volt drop,add little to the operating time of the device" Hope this helps.

 

[Guest55]

^^^which is what i've been trying to explain for the last half hour lol.

 

[Guest55]

No?

You dont sound sure ?

 

[Crosswire]

^^^which is what i've been trying to explain for the last half hour lol.

Maybe, but not very clearly.

But thanks for that, it will definitely save me a lot of time in the future.

The only issue that remains is that of the OP. Why the difference in disconnection times with the the circuit connected or disconnected?

Any ideas?

As regards to ballasts etc "corrupting" the readings, if they are part of the circuit, they should conform to disconnection times.

Or am I wrong again?

 

[Guest55]

FFS
I answered that in my first post.
Why do i bother.

 

[tony mc]

I have to agree with biff on the above.

RCD test should be done with loads disconnected and a EFLI test prior to RCD testing.

As mentioned the times can be alot different if equipment is left connected.

GN 3 Rcd testing also makes mention of this.

 

[Crosswire]

FFS
I answered that in my first post.
Why do i bother.
View attachment 13794

Well not really. You said components of the circuit would "bugger up" the readings. I was under the impression the entire circuit should be tested, and conform.

Perhaps I am wrong and am willing to admit that. However, the NICEIC forms I fill in have a seperate box for the RCD disconnection times for every circuit. So when I fill these forms in, I measure the RCD connection times for each circuit, and record them appropriately.

Perhaps I should correct for ballasts and switches corrupting my results in the future. Do you have the correction factor for this as it will be very helpful?

 

[Crosswire]

I have to agree with biff on the above.

RCD test should be done with loads disconnected and a EFLI test prior to RCD testing.

As mentioned the times can be alot different if equipment is left connected.

GN 3 Rcd testing also makes mention of this.

Ok fair enough. I have pulled the GN3 off the shelf and it does say the load should be disconnected for the test. I was wrong on this point but my instinct still tells me the whole circuit should be tested. In the future i will test at the CU on a standard split load board and simply duplicate my results for the number of circuits that the RCD is protecting.

Sounds like a cowboy job to me, but if that is what is required then that is what I will do. Thanks to everyone who set me straight, as I was doing a lot of (apparantly) needless work.

 

[Strima]

We're only required to test the device, not the circuit for disconnection times of RCD/RCBO. I can see the reasoning behind testing the whole circuit but that is why we do R1+R2 tests to ensure we have continuity that will meet with disconnection times.

If you start taking into account ballasts & transformers where will it end?

 

[Strima]

It's just clicked, the test meter sends a 'pulse' to initiate the trip. Ballast/transformers/loads can absorb some of this 'pulse' and give spurious readings.

I think...

 

[Crosswire]

It's just clicked, the test meter sends a 'pulse' to initiate the trip. Ballast/transformers/loads can absorb some of this 'pulse' and give spurious readings.

I think...

Sounds reasonable. And I know I was (am still) wrong, But if they absorb leakage in a test situation, might they also absorb leakage in a genuine fault?
I am not trying to flog a dead horse but just think testing an RCD in 'laboratory conditions' defeats the point of the testing.
Anyhow, I concede defeat on this issue as the powers that be have ruled otherwise.
Saves me a lot of time

 

[Strima]

Consider putting a nail through a cable, the fault will be greater than 30mA.

 

[Johnny-G]

However, the NICEIC forms I fill in have a seperate box for the RCD disconnection times for every circuit. So when I fill these forms in, I measure the RCD connection times for each circuit, and record them appropriately

Whilst each line has provision for RCD times you only have to record the times on the line the RCD itself is on, unless you're saying that you don't actually have RCD's on their own line?

 

[PEG]

With Biff on this one,RCD test is just that,a test on a specific device.IR and R1 / R2 tests confirm satisfactory operating criterea of device.If you had to test it with loads connected,which would you choose to be connected/on/off or any combination?

 

[Crosswire]

I do have them on their own line, but have always recorded the disconnection times on the circuit line, as this is where I measured it , Do you record it on the RCD line and put N/A on each circuit line?

 

[Guest55]

Ok fair enough. I have pulled the GN3 off the shelf and it does say the load should be disconnected for the test.

So you had the info all along and didnt bother to follow it ????

Then you claim my explainations were unclear in a completely ungrateful manner after i spent over an hour on the thread ???

Now youre still reluctant to do the testing as advised even though its alot quicker and easier just because you dont understand the science behind it ???

You're just all over the place to be honest.

 

[Johnny-G]

I do have them on their own line, but have always recorded the disconnection times on the circuit line, as this is where I measured it , Do you record it on the RCD line and put N/A on each circuit line?

I just leave it blank, but that's because I only test the RCD, not each circuit

 

[Crosswire]

With Biff on this one,RCD test is just that,a test on a specific device.IR and R1 / R2 tests confirm satisfactory operating criterea of device.If you had to test it with loads connected,which would you choose to be connected/on/off or any combination?

Usually on as this mirrors a real life situation, where a fault may potentially develop.

 

[Crosswire]

So you had the info all along and didnt bother to follow it ????

Then you claim my explainations were unclear in a completely ungrateful manner after i spent over an hour on the thread ???

Now youre still reluctant to do the testing as advised even though its alot quicker and easier just because you dont understand the science behind it ???

You're just all over the place to be honest.

Get over yourself. This forum is a place for discussion of technical issues. Please do not take every question raised as a personal attack. I simply debated a matter of recording disconnection times in the hope it would stimulate an intelligent debate.

If you find this offensive, simply do not answer. I do not owe you any kind of gratitude for joining the discussion. I queried a method of recording disconnection times which was raised by the original poster, people posted various answers, I re-checked my literature on the subject, and admitted I was wrong.

What do you want ? A bunch of flowers?

With regards to your comment, there is no 'science' behind sticking probes onto the RCD at the CU. A child of three could do it.The most complicated thing about electrical installations are the regulations that govern it.

 

[Guitarist]

biff and the boys are totally right here...
1) Test your R1+R2 to ensure the circuit complies with EFLI
2) Test the RCD at the CU with loads disconnected
3) Assuming they conform, put the readings on the test sheet. Personally I put the same reading down for each line related to said RCD, tho many do the " sign. Don't put N/A, as it IS applicable.

All sorted now, sleep well

 

[Guest55]

Get over yourself. This forum is a place for discussion of technical issues. Please do not take every question raised as a personal attack. I simply debated a matter of recording disconnection times in the hope it would stimulate an intelligent debate.

If you find this offensive, simply do not answer. I do not owe you any kind of gratitude for joining the discussion. I queried a method of recording disconnection times which was raised by the original poster, people posted various answers, I re-checked my literature on the subject, and admitted I was wrong.

What do you want ? A bunch of flowers?

With regards to your comment, there is no 'science' behind sticking probes onto the RCD at the CU. A child of three could do it.The most complicated about electrical installations are the regulations that govern it.

Maybe you should give testing a miss till you know what youre doing.
And i prefer chocolates.

 

[kingeri]

Been an outside observer on this one til now.....but the Crosswire has certainly lived up to his name in this thread! The RCD testing we do is just that....a test to confirm that the RCD works! The circuit itself has already been tested by the time you do the RCD test. The testing done on the circuit to verify EFLI ensures that the fuse or MCB will operate (or circuit breaker element of the RCBO) within the required time. The RCD is there as additional protection, and as long as you check that it's working, then everything's fine!

 

[Crosswire]

Maybe you should give testing a miss till you know what youre doing.
And i prefer chocolates.

Choccies it is then