Live Or Dead Testing - A Question For Members

[ProDave]

Yes strictly a L-N fault is not directly related to Zs, but if Zs is too high, there's a very high chance the L-N loop impedance will also be too high to trip the fuse or circuit breaker in time in the event of a L-N fault. So further investigation and testing is needed.

 

[says-les]

No, ProDave. There is no need to get involved in our thinking about time to clear a L-N fault. We as inspecting sparks have no control over such timing, and the main thing to clear up here with every one is that Zs and time to clear is all to do with Shock Protection.

 

[apprentice87]

Hi All, Now here is MY [newbie] understanding of it all. There are TWO elements to circuit protection that we need to worry about. One is disconnection times for shock protection. There is nothing wrong with using an RCD to do this, if Zs is too high for a MCB or fuse to clear the fault in the required time, in fact, there is a bit in the regs that specifically states that this is ok, so Les is PERFECTLY correct in what he says. The second is overcurrent protection. Now, this is where it gets interesting....

ProDave is also perfectly correct too!!!! So far as i know, there is ABSOLUTELY NOTHING in the regs that specify ANY disconnection time requirements WHATSOEVER for a L/N fault, ONLY L/E ones. BUT and this is a GREAT HUGE BIG BUT, as ProDave says, we still need to be mindful of L/N loop impedance so our mcb or fuse can clear the fault in the required time, YES, but in what time?? Well, part of the job of designing a circuit is to check the cable size from the point of view of thermal effects regarding cable heating in the event of a fault.

Now, in the regs, it gives all kinds of tables, [values of "K"] to help us do this, based on the type of cable, the problem is, as it says in the small print, all this is based on a MAXIMUM DISCONNECTION TIME OF FIVE SECONDS, so if our overcurrent protection will not clear the fault in five seconds, then we will have a problem. We will have to work out our own "custom built" value of "K" instead, and i believe this is given in BS 7454 Obviously, this will not be fun. This is why i always check Zs and L/N loop impedance too, because, as pointed out by others, it is perfectly possible for earth loop impedance to be HIGHER than line/neutral loop impedance.. Hope i have not bored you all to death!!! john...

 

[Sidewinder]

Yes there are calculations in 7454, just £82 for 14 pages including covers etc. if you want it!  :slap

 

[apprentice87]

Hi Paul, Cannot see a sudden rush to buy them then!!! john...

 

[says-les]

Thanks apprentice87. K was introduced in the 16th edition or earlier for the ADIABATIC equation. Again, this has nout to do with L-N loop but only L-E loop where the equation has to be checked at the design stage to see that any reduced CPC such is in a PVC Sheathed Cable does not suffer overheat damage. The Ipscc is the only real guarantee that fuses or breakers will clear a short circuit fault.

I think we need to close this post now unless anybody has any more to add. Thanks to all who contributed. I appreciate it.

 

[Technician]

Just answering Says Les;  The difference between those prospective

fault current reads concerned me to a degree that made me raise the

issue with the guys upstairs.  We know that the larger of them is the more

important when considering the selection of the protective device that

protects that circuit but, should the other be so low that disconnection time

might be extended as a consequence, it does warrant further investigation.

As Doc says, this cannot be done until the circumstances he has specified

would permit it.

I have enjoyed reading these posts.

 

[Steve3948]

Yes it has been a good post and thread, enjoyed reading.

 

[says-les]

To Technician and Steve3948,

Thanks for expressing some enjoyment/interest in this original post which blossomed into the whole

subject of testing Zs and why we do test Zs. What has been an eye-opener has been the difference in

opinions/interpretations and even conceptual understanding.

I'm not sure what Technician means by, "We know that the larger of them is the more

important when considering the selection of the protective device that

protects that circuit but, should the other be so low that disconnection time

might be extended as a consequence.....".

Cheers

 

[professional]

 
  415.1           Additional protection: Residual current devices (RCDs)
415.1.1           The  use of RCDs  with  a rated  residual   operating   current  

 not  exceeding   30 rnA and  an operating time  not  exceeding   40  ms  at a residual   current   of   

  is recognized    in a.c.  systems   as additional   protection   in the event  of failure  of the provision   for basic  protection   and/or   the provision   for fault protection   or carelessness   by users.

415.1.2          The  use  of RCDs   is not  recognized    as  a sole  means   of  protection    and  does  not  obviate   the  need  to apply  one  of the  protective   measures   specified   in Sections   411  to 414.

I would suggest that the above regulations indicate that protection by the use of an RCD does not obviate the use of design criteria for an existing circuit.
 
The fact that YOU have failed to apply design criteria correctly does not mean that "bunging" an RCD onto the circuit covers your a**e because you have not met the requirement for ADS/ADOS or whatever the term is this week!

 

[says-les]

I have to disagree with you, Professional. Remember this is not a new job that I or anyone has just installed, but rather it is a 20 year old building that I have taken an interest in as member of the organisation that belongs to the building. Now, in order to add additional protection I added pods to the C-Type breakers. Now a professional engineer agrees with me that as soon as that is done, i.e. an RCD fitted then any/all such circuits need not rely on the CB to clear an earth fault. Now in any job if the correct rating of protective device matches the current rating of the protected cable in that circuit, then no one need be concerned if that overcurrent protective device will protect the circuit from short circuit faults. You are, or one spark replied that he is concerned about testing the L-N loop!!!! Nowhere does any Reg say that this L-N loop has to be tested. It will protect against short circuit faults if it has been rated to suit the size of cable, taking rating factors into account.

The regs do require additional protection now for certain circuits as you know such as where cables are less than 50mm from the surface of plastered walls, and in every socket-outlet circuit, and any type of circuit in bathrooms and other special locations. In these cases this is classed as additional protection, and only the RCD test need be applied but the Schedule of Test Results must be filled in for Zs and R1+R2 and Ze. If Zs is higher than normal for any reason and not your design of installation then fitting an RCD is additional protection BY CHOICE - nout wrong with that, and you will have read that a 30mA RCD is effective with a Zs of up to 1667 Ohms.

Cheers

 

[Sidewinder]

les,

You have also had professional engineers disagree with you!

I would never advocate total reliance on the RCD, a circuit should be designed to meet ADS without requiring additional protection as I have suggested.

It is taken as "OK" as I have already said, but it is not right.

I don't quite follow why you did not replace the C's for B's?

Whilst it could be argued that the RCD's are required for additional protection then that is the role they should fulfil.

Also one does have to test the L-N loop, or otherwise establish what the PSCC is to ensure that the protective device for the circuit is adequately rated.

 

[steptoe]

headbang

why oh why do people insist on just bunging an RCD in and it will be right?

either do the job right or dont do it at all.

 

[Steve3948]

Which people ?

 

[says-les]

Thanks to Sidewinder for his last response. Taking, for the moment the L-N loop, I agree that Ipscc must be recorded, but that is only to verify that the short circuit rating of any protective devices exceeds this value in kA. I get the impression from one or two of the guys who have got into this thread that L-N loop involves time similar to the 0.4 second earth fault clearance time. I sincerely hope that any protective device will clear a short circuit but not in the same sense as an earth fault. In other words the breaking capacity is different from ensuring adequate shock protection. We as sparks have no control over whether a fuse or CB will clear a short circuit quickly but we have to check that its breaking capacity is greater than the Ipscc otherwise it will explode trying to clear the short circuit. Testing for earth fault current is all about protecting persons or animals from shock by ensuring a fast enough clearing time can be achieved, to prevent serious shock risk to persons touching earthed metalwork at the time.

Another thing that is stated in the Regs is that where Zs for a non-RCD protected circuit exceeds the max safe Zs value, local equipotential bonding should be installed to lower the Zs sufficiently to comply.

The issue of the RCD only being additional protection and still requiring a Zs less than the max value in the Table 41.3, modified by the 0.8 factor, is a point that not all sparks agree on so we need a professional engineer to contribute to this interpretation.

Now I have asked one and here is his quoted reply,"Zs is 'irrelevant' where rcd is provided..(As long as it doesn't exceed table 41.5 value; 1667 Ohms for 30mA rcd). Therefore table 41.3 values r irrelevant. There's a reg that tells u this in part 5. It used to b on page 113 (or 112) in red book. Not sure of pn in green book"

Is there anyone out there who could clear this up. I notice that the IET who wrote the 17th Edition state somewhere that they will no longer mediate in any dispute re interpretation of an aspect of a Reg! So where do we turn?

I asked the moderator of this forum but they do not hold themselves as an authority in any such need for guidance.

Maybe you reckon you don't need to clear up how this should be interpreted but I would like to.

Thanks again.

 

[steptoe]

Zs is over that permitted in BS7671 so just stick an RCD on and magically we are allowed 1667 ohms !

 

[Steve3948]

I love Magic

 

[says-les]

But that seems to be the case. If a 30 mA RCD experiences 5x this (150mA) leaking to earth or more it will clear in less than 40mS and we have to check this. Meantime the Cb associated with it, in the case of a converted MCB with a pod added to make it an RCBO' then there is not only no time for the tripping current of it to be reached, but the earth fault will be cleared by the RCD anyhow. So what is your reasoning that leads you to make sure Zs is as per Table 41.3???

 

[apprentice87]

Hi Again,

The L/N loop DOES very definitely, for all practical purposes, have a maximum fault clearance time.. Unless you really do want to buy the £85 book and calculate you own value of "K"

The regs clearly state "A fault occuring in a circuit at any point shall be interrupted within a time such that the fault current does not cause the permitted limiting temperature of any condcutor to be exceeded" no mention of "earth fault current" just "fault current" which is definined as " A current resulting from a fault" whereas "earth fault current" is defined as "An overcurrent resulting from a fault of negligible impedance between a line conductor and an exposed conductive part or a protective conductor"

The formula for checking this, is DIFFERENT from the usual adiabatic one for checking CPC csa and needs to be checked separately... john...

 

[says-les]

To Steptoe, why is an RCD now to be installed, protecting sockets unless in the care of competent people? Although it is classed as additional protection, we must assume the reason is to give closer protection against shock by rapid tripping (40mS instead of 400mS as for a CB or Fuse). You cannot say that it is additional in case the CB or Fuse fails to clear an earth fault, and so you cannot say that the RCD is less reliable than the CB or fuse any more than saying that a CB is more reliable than an RCD.<br />So unless someone gives firm ruling AND logic which no one has as yet, then where are we?<br />Even forgetting the C-Type breakers. I amateurs wired this place and used b-Type breakers and Zs was high then we have the choice of adding local equipotential bonding or adding RCD in addition. We are no by doing the latter RELYING on the RCD solely as suggested by someone's logic, but rather ensuring that shock does not occur. Bad design, I know perfectly well, but who is going to tell the owners that the whole installation is badly designed? and it is!<br />Please don't think I am being argumentative. I am just wanting better logic as to why I am wrong in my logic. Put it this way, if EVERY CB was replaced by an RCBO Zs would not need to comply with Table 41.3!<br />Is there anyone out there who knows how to ask Tony Cable of the NICEIC?<br />He once gave me his card but I cannot find it.<br />Anyone not know who I am referring to?