high readings on RCD test

[vbyrne]

just a quick one lads, I was sent to a job to change a trip in a pirilli box that was covering the cable going into the house. Then I changed the trip in the house that was covering the house its self. My foreman give me 100Ma time delay for outside and a 30Ma for inside, I thought that we had to use a 30Ma time delay for all domestic installations, I said this to him but he was having none of it, I was not putting my name to anything for this job s I just did what I was told. Also the main earth was a 6mm cable serving a TT system and I was getting a reading of 805ohms for Ze. I also did a tripping test on my new trips but the readings were mental I made a note of all this and reported back to the office with my findings.I made recomendations to change the earth up to 16mm and do a full test on the installation. Would the situition with the earth affect the RCD times reading and is there anyadvice about testin a trip with time delay, still learning about testingmg_smile:

 

[malcolmsanford]

I don't think I have ever seen a 30mA time delayed or S type RCD, but not to say that there isn't one. But your boss as done the normal practice of an upfront 100mA type S and a 30mA for additional protection and discrimination.

As for you Ze(Ra) well you can put a 240mm cable onto it and not get a better result, you need to address the rod not the cable size.

 

[spark 68]

As Malcom has said this is normal practice, the upstream RCD has to be => 3X the downstream rcd to achieve discrimination, and as 100mA is > 3X 30mA this achieves that.

Also as Malcom says TT can be fine on 10mm^ or even 6mm^ cable, but if the rod is crap putting thicker cable will not solve your problem with the Ze (Ra) readings.

Ps. I haven't seen an S type 30mA RCD either

 

[telectrix]

it's all been said. sectioned rods driven deep and/or relocated.

 

[JUD]

As Malcom has said this is normal practice, the upstream RCD has to be => 3X the downstream rcd to achieve discrimination, and as 100mA is > 3X 30mA this achieves that.

It's my understanding that it's the delay in the S type RCD that provides the discrimination, not it's rating. We check 30mA RCDs at 5 x IΔn and they should trip in ≤40ms, however this current would also take out a 100mA non-delayed RCD, hence the need for the delay.

 

[spark 68]

True Jud,

It's my understanding that it's the delay in the S type RCD that provides the discrimination, not it's rating. We check 30mA RCDs at 5 x IΔn and they should trip in ≤40ms, however this current would also take out a 100mA non-delayed RCD, hence the need for the delay

As 5 x I delta n @ 30 mA = 150mA, but this is only for 30mA RCD's or less (which is probably why I haven't seen an S-Type of this value) which protect the final circuit ( 415.1.1 RCD's not exceeding 30mA used for additional protection), the only RCD's as far as Iam aware to be tested @ X5 and have to trip in less than 40mS, but my statement is also true to an extent in the fact that I haven't seen any of the higher trip rating ones that are not S-type, it was just a rule of thumb and I was keeping it simple

 

[Deleted member 26818]

Crikey, you think it's just a quick one?
There are two main reasons for providing RCD protection.
One is to provide additional protection for circuits of special locations, socket-outlets and cables concealed in walls.
This applies irrespective of whether the installation is TT, TN-S or TN-C-S.
BS7671 requires that such a device be rated at 30mA and operate within 40ms when the operating current is 150mA or greater.
Another reason for using RCDs, is where the disconnection times as specified in BS7671 cannot be achieved by use of an Overcurrent Protective Device, due to high Zs.
Typically in TT installations where distribution circuits require 1s and final circuits require 0.2s (200ms).

RCDs are very sensitive, and many appliances have earth leakage in normal operation.
To prevent the cumulative earth leakage of a number of appliances from causing nuisance tripping, we are required to divide installations into a number of circuits.
However this still leaves distribution circuits requiring the use of RCDs to achieve disconnection times being subjected to the earth leakage of a number of appliances.
To prevent nuisance tripping on these circuits, a higher rating is used, typically 100mA.
However, by doing such, we would have RCDs in series, which will cause discrimination problems in the event of an earth fault. Something that BS767 also requires us to address by dividing installations into separate circuits.
Generally earth faults are of such a high amperage, that they would cause any RCD to operate whatever the rating.
To provide discrimination, for earth faults, time delayed RCDs are used.
As such this would entail a 100mA time delayed RCD protecting distribution circuits, with 30mA RCDs protecting final circuits and socket-outlets.
If the distribution circuit also required additional protection due to being concealled in a wall, another method of additional protection would have to be provided.
To test time delayed RCDs using most MFTs, you have to set the device for RCD test, then toggle through the options for type, rating and time for the particular RCD being tested.

 

[vbyrne]

Thanks lads that is a lot of info ,thanks for reading my thread and taking time out to answer my questions

 

[Engineer54]

As Malcom has said this is normal practice, the upstream RCD has to be => 3X the downstream rcd to achieve discrimination, and as 100mA is > 3X 30mA this achieves that.

Rarely if ever, will a 100mA RCD give upstream discrimination to a 30mA RCD device, no matter what it says in the Reg's or anywhere else!! The only sure way of achieving discrimination is via a ''time delay'' feature on S type RCD devices. I've seen 500mA RCDs tripping with a 30mA downstream RCD...

 

[SPARTYKUS]

I don't think I have ever seen a 30mA time delayed or S type RCD, but not to say that there isn't one. But your boss as done the normal practice of an upfront 100mA type S and a 30mA for additional protection and discrimination.

As for you Ze(Ra) well you can put a 240mm cable onto it and not get a better result, you need to address the rod not the cable size.

http://www.protekuk.co.uk/rcd-data-sheet.pdf - page 2 at bottom

 

[spark 68]

Hi Eng.

You are of course correct, what I omitted from my earlier post, was : " to achieve discrimination between RCD's the up stream device should be => 3X the next downstream device of the same type".
Which as you say the upstream device/s should be S-types for distributution/fire protection TT etc...etc to give discrimination for the ones downstream giving additional protection, and also the ones rated at not exceeding 30mA for both additional protection and for supplying socket outlets to BS 7288, and for all of the other reasons incl. special locations, high Zs and cables installed in walls etc.... etc.

This was just poor wording on my part :hanged:

 

[Deleted member 26818]

I guess that if the Zs can't be reduced below 500Ω then a 30mA time delayed RCD will be an option.
However it will still be subject to nuisance tripping if the cumulative earth leakage is too high.
Might be better to just make the distribution circuit double insulated.

 

[vbyrne]

what do you mean 'make the distribution circuit double insulated,'

 

[Deleted member 26818]

what do you mean 'make the distribution circuit double insulated,'

Use conductors that have double or reinforced insulation, and equipment that is double insulated, as per Section 412 in BS7671.

 

[spark 68]

Cheaper to upgrade the rod/s though Spin