High RCD Trip Time Despite Good EFLZ

[SidTheSparky]

Replaced an old Wylex fuseboard (BS3036 fuses) the other day.

TNCS supply with Ze 0.14ohm.

I was expecting fast rcd trip times on x1 and x5, but x1 are surprisingly high. Tried 0 and 180.

Examples of some of the readings:

Sockets (ring) Zs 0.28ohm, RCD 43ms@x1 and 22ms@x5.

Sockets (radial) Zs 0.32ohm, RCD 53ms@x1 and 22ms@x5.

Upstairs Lights (radial) Zs 0.69ohm, RCD 39ms@x1 and 8.5ms@x5.

Used a Crabtree Starbreaker dual RCD board with 30mA RCDs.

Tried swapping circuits between the RCDs but got similar readings.

Tested with Fluke 1653 MFT.

BTW: the upstairs and downstairs lights seem to be sharing a neutral, so they've had to go on the same RCD. Insulation tests all gave 500Megohm.

Very quirky results.

Anyone ever had a similar problem?

Not really sure what to do. I've told customer that his RCD disconnection times are outside of BS7671 and I'll be stating that on the EIC. Trouble is I haven't sussed out exactly why.

 

[steptoe]

How are they outside? They are well inside and perfectly fine.

BTW, you need to put both lights on one MCB, not just one RCD.

 

[Theorysparky]

sometimes on first RCD testing you get higher than expected results

i normal recycle the rcd a few times,push the test button just to clean the contacts

get better readings

 

[SidTheSparky]

How are they outside? They are well inside and perfectly fine. BTW, you need to put both lights on one MCB, not just one RCD.

Both lights on one MCB. Good point. I'll correct that. Thanks.

As for trip times - they ought to be <40ms @x1. Only the lights seem to be managing this. What puzzles me more is that they have a higher impedance than the socket circuits so they should take longer to trip. I've always taken it for granted that trip times would be higher for higher impedances. My recorded readings are somehow quirky.

 

[Sidewinder]

Sid,

Where are you taking these readings?

 

[SidTheSparky]

Both lights on one MCB. Good point. I'll correct that. Thanks.As for trip times - they ought to be <40ms @x1. Only the lights seem to be managing this. What puzzles me more is that they have a higher impedance than the socket circuits so they should take longer to trip. I've always taken it for granted that trip times would be higher for higher impedances. My recorded readings are somehow quirky.

Thanks Theorysparky. I too have noticed that first time trips are often all over the place, so I tend to do give new RCDs a few activations before recording results.

---------- Post Auto-Merged at 17:20 ---------- Previous post was made at 17:14 ----------

Sid,Where are you taking these readings?

At the ceiling rose: on the switch wire.

At the socket: with plug-in lead, also double checked with probe-leads on back of socket.

I believe you can test an RCD at the RCD itself as a functional test (I might be mistaken), but i think that is not ideal.

 

[Theorysparky]

Both lights on one MCB. Good point. I'll correct that. Thanks.As for trip times - they ought to be <40ms @x1. Only the lights seem to be managing this. What puzzles me more is that they have a higher impedance than the socket circuits so they should take longer to trip. I've always taken it for granted that trip times would be higher for higher impedances. My recorded readings are somehow quirky.

thats for x5

 

[SidTheSparky]

thats for x5............

Oh!

Seems all is well and good then.

 

[Sidewinder]

You should not take rcd trip times at the extremity of the circuit as there are other factors that can interfere with the reading, anyone...

 

[kme]

All is perfectly fine & dandy.

It IS better to take the readings at the RCD, IMO - there are no spurious capacitative effects that can affect the readings; and you can easily remove all loads from the cct for testing.

If you`re going to do the job, and have the correct gear, etc; how hard can it be to actually open the on site guide (or BRB) to check the trip times, if you aren`t sure?

What in the heck is an "EFLZ" anyway? I`ve not heard of that before.

BTW, you can have a TT install with >1K ohm EFLI; and your RCD should still trip within the recommended parameters.

What did you think the 5x trip times were? Just out of interest.

KME

(ruddy pushy-shovey by the snakehipped one....again!!!!! ] )

 

[gordy71]

You should not take rcd trip times at the extremity of the circuit as there are other factors that can interfere with the reading, anyone...

I always test from board cos as long as zs are ok then rcd is just functional test (have to be within times of course)

 

[batty]

I personally test were it easiest as I mainly use RCBO's and some RCBO's like MEM cannot be tested at the RCBO so need testing at the end or in the circuit.

 

[gordy71]

I personally test were it easiest as I mainly use RCBO's and some RCBO's like MEM cannot be tested at the RCBO so need testing at the end or in the circuit.

Just curious as to why you use rcbos mainly? I know rcbo are better to segregate the fault current but how do you sell this to a customer when they are 10x more expensive?

 

[Revved Up Sparky]

Off the top of my head, if I remember right, x1 needs to be < 200ms if a BS4293 or < 300ms if a 61008/61009

 

[batty]

Just curious as to why you use rcbos mainly? I know rcbo are better to segregate the fault current but how do you sell this to a customer when they are 10x more expensive?

Explain if they have a fault they will loose half of there circuits until fault is found this usually sells it.

 

[Riggy]

Explain if they have a fault they will loose half of there circuits until fault is found this usually sells it.

You don't bother explaining how (in)frequently a genuine fault will occur on a reasonable installation I take it?

 

[Andy™]

Off the top of my head, if I remember right, x1 needs to be < 200ms if a BS4293 or < 300ms if a 61008/61009

correct

 

[batty]

You don't bother explaining how (in)frequently a genuine fault will occur on a reasonable installation I take it?

Why would I want too I think RCBO's are a better way forward so why would I want to push a two RCD board. People have the choice they probably will only have cu changed once so RCBO boards may be a bit more expensive but to me it is the better way. Take a rented property you do a cu change put a twin RCD board in if there is a fault it will then be an emergency for landlord to fix it if you put an RCBO board in if there is a fault it probably will not be an emergency to fix it so tenant will probably be happier and landlord probably will be happier to.

 

[SPECIAL LOCATION]

Very quirky results.

Anyone ever had a similar problem?

Not really sure what to do. I've told customer that his RCD disconnection times are outside of BS7671 and I'll be stating that on the EIC. Trouble is I haven't sussed out exactly why.

Best course of action...

1/ Get approx

 

[steptoe]

I personally test were it easiest as I mainly use RCBO's and some RCBO's like MEM cannot be tested at the RCBO so need testing at the end or in the circuit.

how come?

I have never had issues testing any at source.

what issue are you having?

---------- Post Auto-Merged at 23:41 ---------- Previous post was made at 23:37 ----------

Why would I want too I think RCBO's are a better way forward so why would I want to push a two RCD board. People have the choice they probably will only have cu changed once so RCBO boards may be a bit more expensive but to me it is the better way. Take a rented property you do a cu change put a twin RCD board in if there is a fault it will then be an emergency for landlord to fix it if you put an RCBO board in if there is a fault it probably will not be an emergency to fix it so tenant will probably be happier and landlord probably will be happier to.

that will be because they are mostly single pole and the rest of us are stuck in the dark ages with DP isolation in a fault scenario then.