IET
Decrease font size
Increase font size
Topic Title: Time delay rcd
Topic Summary:
Created On: 26 March 2019 10:19 AM
Status: Read Only
Linear : Threading : Single : Branch
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 26 March 2019 10:19 AM
User is offline View Users Profile Print this message



Baldyhugh

Posts: 413
Joined: 01 June 2007

Hi all

I have a customer with a TP&N supply to a farm and have an rcd query that I would like some advice on. At the incomer is a TP&N distribution board with a 100mA rcd main switch. There is a single phase distribution circuit from this board to an outbuilding containing two consumer units with 100mA main switch in each. As you can guess, an earth fault in the outbuilding takes out both the rcd in the outbuilding and at the incomer position. To rectify this would it be acceptable to replace the 100mA TP&N rcd at the incoming position with a time delay version? Would this still provide adequate fault protection to the other circuits on the TP dist board?

Thanks
 26 March 2019 11:52 AM
User is offline View Users Profile Print this message



mapj1

Posts: 12039
Joined: 22 July 2004

Well, given the late hour, be prepared to move this to the new forum, the clock is ticking, but meanwhile, the answer is "probably, but it depends"

What else is protected by the RCD you propose to make time delay ? (type S I presume ?) So long as it is SWA submains to other boards with instant tripping RCDs or RCBOs at that end, or fixed loads hard wired that don't need fast RCD protection, then you are fine.

If it is twin and earth or soft flex feeding 13A or 16A sockets, or really anything portable, then more thought may be needed, and it may be you need to split off differently so there are two boards, one with fast RCD protection for the local loads and a new one that is just for the outbuildings, or consider adding an odd RCBO to the board if it is the sort where that is available, and there are only a few loads requiring fast protection.
In any case 100mA instant is too high for a number of situations like sockets for plug in hand held things, and some final circuits should really be 30mA.
Actually I find 100mA instant RCDs a pain, and there are not many situations they are needed - 30mA instant for shock, and 100mA S (~ 0.1 sec) type for distribution, and if need be something slower still at 300mA behind that, but often splitting at source over more than one 100mA S type is better.

-------------------------
regards Mike
 26 March 2019 12:54 PM
User is offline View Users Profile Print this message



Baldyhugh

Posts: 413
Joined: 01 June 2007

Hi Mike

There are three or four, I cant remember offhand final circuits on the dist board that I plan to swap the rcd in. I think a couple are motors but from memory, they all are fixed loads, no socket outlets or anything that will require 30mA rcd protection
 26 March 2019 05:42 PM
User is offline View Users Profile Print this message



Weirdbeard2

Posts: 753
Joined: 29 November 2017

Hi Hugh, have you been specifically tasked with improving the reliability of the supply due to problems caused by loss of supply caused by the existing arrangement, and what's the customers ball park budget?
 26 March 2019 07:19 PM
User is offline View Users Profile Print this message



Alcomax

Posts: 450
Joined: 12 November 2009

mmmmm...reading the OP.

Changing the front end 100mA to a 100mA[S] is not going to change much, if anything, if there are 100mA RCDs on the downstream distribution boards. They will still trip. Selectivity/ discrimination is 3:1. So 100m[s] : 30mA , for example. Also 300mA[S] : 100mA.

As WB2 points out, is there an actual need to change anything or instruction to sort out some unwanted outcome of the existing arrangement?
Farms are hoary old installs at the best of times, trying to get some "selectivity" may end up opening a can of worms. For example, latent defects may not be tolerated by 30mA RCDs.

Ideally you may want a 300 [s] upfront......but that will depend on a nice lowish Ze.
 26 March 2019 08:12 PM
User is offline View Users Profile Print this message



Baldyhugh

Posts: 413
Joined: 01 June 2007

I've been asked to sort it out by the customer as tripping the front end rcd is a headache for him as the other circuits on the TP board going down during milking is a bit of a nightmare. I was dubious if a time delayed rcd would hold on to let a non time delayed rcd of the same rating downstream trip
 26 March 2019 08:19 PM
User is offline View Users Profile Print this message



chrispearson

Posts: 1095
Joined: 15 February 2018

Why not type S RCD as the incomer at the main DB and RCBO outgoers with ordinary main switch at the final DBs?
 26 March 2019 08:52 PM
User is offline View Users Profile Print this message



mapj1

Posts: 12039
Joined: 22 July 2004

Time delay alone will still help, if the fast one disconnects the faulty branch.
The bigger question may be why does it trip at all, and how close is it to the edge even when it appears to be running OK. The belt and braces is to split the pre-RCD incomer to another box and RCD, so outbuilding and other circuits are separated.
I have seen a system with 3 S-type RCDs and nothing else in a box that in another place would normally have been an isolator, and then close by, 3 boxes of breakers, one fed from each RCD.

-------------------------
regards Mike
 28 March 2019 10:01 AM
User is offline View Users Profile Print this message



AJJewsbury

Posts: 17795
Joined: 13 August 2003

Time delay alone will still help, if the fast one disconnects the faulty branch.

It's a bit of a gamble still though - any 100mA RCD should disconnect somewhere between 50mA and 100mA but where exactly depends on the particular unit. So say you had 70mA leakage - you could still have the up-front S-type trip (after a delay) but the downstream 100mA still hold.

Plus of course where have several downstream RCDs there's the issue of cumulative residual currents - not only by simple addition (30mA from each of three submains would likely take out a front-end 100mA S-type even though none of the downstream ones would trip), but also the additional leakage currents reducing the delay on the S-type (at 2x an S-type could be in a similar tripping time region to an instantaneous at 1x, at 5x it could in theory at least be faster)

- Andy.
 28 March 2019 10:06 AM
User is offline View Users Profile Print this message



mapj1

Posts: 12039
Joined: 22 July 2004

agree - the real fix is to partition fully or have a 10:3:1 tiered approach in time and current.

-------------------------
regards Mike
 28 March 2019 05:23 PM
User is offline View Users Profile Print this message



Weirdbeard2

Posts: 753
Joined: 29 November 2017

Hi Hugh, thanks for the reply, it might be worth going over to the new forum where you can post pics directly to a topic of the current arrangement if you have them?

as per the great posts above, you ideally need to know what is causing the existing trip problem and how much inconvenience in £ does each trip equate to. IE, does it trip once every 6 months for no reason and the inconvenience is having to walk 500 yards to see which trip has gone or is it more regular and causes high production losses?
Statistics

New here?

  • To participate in discussions, please log in and introduce yourself.

See Also:



FuseTalk Standard Edition v3.2 - © 1999-2019 FuseTalk Inc. All rights reserved.

 
..