grey area . when we consider fault protection, generally it's a fault (short) from a live part to an exposed conductive (earthed ) part. in which case, the RCD is performing the role of fault protection, and Zs is the resistance of the earth fault path. when we come to L-N shorts, the MCB will operate irrespective of the Zs and high Ze due to TT. bear in mind that the ELFI relates to L-E faults, not L-N.
We need to check everything will be safe and disconnect in all conditions, under normal circumstances, the r1+r2 is slightly larger than r1+rn due to smaller cpc, so if the mcb works and disconnects for r1+r2 it will do so for r1+rn.
However, in the TT case, the r1+r2 is so much bigger than r1+rn that we have to provide specialist protection to cater for it - the rcd. But the previous assumption that if the mcb will operate for earth faults, it will also operate for line-neutral faults just isn't the case here, therefore we must confirm it is ok.
So check of loop to ensure mcb operation for line-neutral faults, and check of rcd operation for faults including earth.
Hence why we must check the loop against the mcb - using a valid mcb max loop impedance (which if close, you could legitimately use r1+rn) and the rcd by confirmation that it operates via actually testing it. (The cross check via comparing resistance values is basically because we can't actually test the mcb)
However, in the TT case, the r1+r2 is so much bigger than r1+rn that we have to provide specialist protection to cater for it - the rcd. But the previous assumption that if the mcb will operate for earth faults, it will also operate for line-neutral faults just isn't the case here, therefore we must confirm it is ok.
So check of loop to ensure mcb operation for line-neutral faults, and check of rcd operation for faults including earth.
Hence why we must check the loop against the mcb - using a valid mcb max loop impedance (which if close, you could legitimately use r1+rn) and the rcd by confirmation that it operates via actually testing it. (The cross check via comparing resistance values is basically because we can't actually test the mcb)
Yea I thought that , but been told to enter 200 ohms,relevant for the Max on TT System.
The actual circuit Zs measured 3.79 ohms obviously using the loop tables for TN system well over compliance.
Agree its a bit of a grey area. The Max Loop Impedance (Zs) refers to the Line-Earth Loop so should be a big clue!
Some lines of thought are that it should be the maximum for the RCD (50/0.03) =1667ohms.
Some, the maximum for MCB at 0.2secs disconnection time even though its unlikely to meet that requirement if TT.
Never thought to enter 200hms!
BS7671 suggests 200 ohm for earth electrode resistance and anything above this may be unstable. Why you should enter this on a schedule is odd as it is external and does not include the circuit impedance.
L
loz2754
This is not really a grey area.
BS7671 gives us table 41.5 - Maximum earth fault loop impedance (Zs) for RCDs. This is for TT systems.
For an RCD with a rated residual operating current of 30mA, the maximum earth fault loop impedance (Zs) is given as 1667 ohms. We are referred to note 2 below, which states that a value exceeding 200 ohms may not be stable.
It would therefore not be unreasonable to use this figure of 200 ohms in the Max Zs column on the EIC. It would also not be wrong to state the maximum value given in table 41.5, ie 1667 ohms, because that's what's written in black and white.
If we were to use the Max Zs tables for MCBs to BS60898 and put down the figure of 1.37 ohms or the 1.1 ohms corrected figure, and we enter the measured value of Zs for the circuit as (for example) 27.5 ohms, we would be effectively certifying that this particular circuit does NOT comply with the required disconnection time.
Also, as has already been stated by others, the Max Zs column is applicable only to an earth fault, not a short circuit from Line to Neutral.
@Julie. makes a good case for considering the effective disconnection of the circuit by the MCB in the case of a line to neutral fault in a TT system, in relation to what would in fact be the L-N fault loop impedance. As far as I'm aware, BS7671 does not give any tables for L-N fault loop impedance. However, the charts giving the time/current characteristics for MCBs can be consulted for this purpose.
BS7671 gives us table 41.5 - Maximum earth fault loop impedance (Zs) for RCDs. This is for TT systems.
For an RCD with a rated residual operating current of 30mA, the maximum earth fault loop impedance (Zs) is given as 1667 ohms. We are referred to note 2 below, which states that a value exceeding 200 ohms may not be stable.
It would therefore not be unreasonable to use this figure of 200 ohms in the Max Zs column on the EIC. It would also not be wrong to state the maximum value given in table 41.5, ie 1667 ohms, because that's what's written in black and white.
If we were to use the Max Zs tables for MCBs to BS60898 and put down the figure of 1.37 ohms or the 1.1 ohms corrected figure, and we enter the measured value of Zs for the circuit as (for example) 27.5 ohms, we would be effectively certifying that this particular circuit does NOT comply with the required disconnection time.
Also, as has already been stated by others, the Max Zs column is applicable only to an earth fault, not a short circuit from Line to Neutral.
@Julie. makes a good case for considering the effective disconnection of the circuit by the MCB in the case of a line to neutral fault in a TT system, in relation to what would in fact be the L-N fault loop impedance. As far as I'm aware, BS7671 does not give any tables for L-N fault loop impedance. However, the charts giving the time/current characteristics for MCBs can be consulted for this purpose.
Last edited by a moderator:
The operating curve/required time to operate is the same for the mcb aspect irrespective of if the fault is L-N or L-E/CPC.
In this case then, with a measured value of circ 3.79 ohm, which could be around 4.44 ohm when running hot, we could have a fault current (@95% of voltage - the minimum allowed) of 230 x 0.95 / 4.44 = 49A or so.
For a standard 32A mcb this represents a trip time around 200 seconds - close to 3 1/2 mins!
Even at full voltage, and a cold cable, it's still several mins before it trips.
Does this sound OK?
In this case then, with a measured value of circ 3.79 ohm, which could be around 4.44 ohm when running hot, we could have a fault current (@95% of voltage - the minimum allowed) of 230 x 0.95 / 4.44 = 49A or so.
For a standard 32A mcb this represents a trip time around 200 seconds - close to 3 1/2 mins!
Even at full voltage, and a cold cable, it's still several mins before it trips.
Does this sound OK?
Agree, & this becomes a problem, the 3.9 reading refers to line & earth not L & N,I’ve done loads of calculations for this scenario,& it’s a worry,when circuits on TN System are not complying to the Zs required, & a rcbo is then used.
No is the answer to your question,sorry.
Agree, we are furiously agreeing with each other.
It depends on how the Zs is made up, if the r1 is 0.5 ohm, r2 = 0.83 ohm, and Ze =
2.46 ohm (but Z1 [line-neutral] = 0.1 ohm)
Then although the Zs (earth fault) would be 3.79; recalculating for r1+rn +Z1 = 1.1 ohm - which is just about ok.
Made up figures here of course.
But completely agree with your worries, I think the same, it's unfortunate that the process, which is a shortcut very suitable for TNx systems is applied to TT, it completely misses out valuable checks.
I really don't like substituting a proper design by "just add a rcd - that will clear it" type logic, in my mind the design should try to achieve everything properly and only rely on rcds as a last resort.
Unfortunately this view isn't shared by all, with so many just relying on a single rcd to perform magic!
Last edited:
Yep....it seems more in theses times with the younger generation,just fit a Rcd/rcbo not thinking of the implicactions involved.Off to work have a good day.
If you verify that the PSSC is enough to meet disconnection time you also have, in a sense, verified the volt drop is reasonable. Not necessarily < 5% (which would imply > x20 short to nominal current, enough for a type D MCB) but also it might not need to be that as the volt drop applies at the design current which could be a lot less than the MCB.
Meeting disconnection times comes under chapter 41 - Protection against electric shock, which I don't think would be a concern for L-N faults (or perhaps it might be, on a TN-C-S?). So to disconnect in <= 0.2s for L-N fault on TT is not necessary.
There are of course other considerations - adiabatic and cable rating.
There are of course other considerations - adiabatic and cable rating.
Correct, there isn't a requirement to operate for L-N or L-L faults within any set time but 434.x does say that protection against fault current is required, to me that does mean it has to operate in a timely manner, not necessary 0.2/0.4 sec - or even 5 sec but not several mins.
My concern wouldn't really be the cable, but the consequences of the fault itself - if this is a water/rodent/whatever issue causing the fault, the likelihood is a strong arc persisting for a long time before finally being cleared some time down the line. This could cause thermal damage/fire.
Perhaps this is why the 18th amendment 2 is looking for AFD devices - just like the premature wiring collapse, caused by bad practices - instead of fitting trunking properly Mr 'Bodgeit and Scarper' (Ltd of course) just stick the double sided tape on a dusty flaking paint surface, so by the time they leave site it's already fallen off! - Result, we have to use metal cable clips et al. same here - rather than people thinking about the consequences of high loop impedance, they just assume an RCD will sort it - hence down the line we have to start adding additional devices:
Fuse/MCB - Primary protection
RCD - Additional protection
AFDD - Additional Additional protection
????? (TBA) - Additional Additional Additional protection
Not very convincing, I don't think anyone has answered the OP original question!
If its not so 'grey' it should be pretty obvious!
L
loz2754
Telectrix answered correctly, 1667 ohms.
I agreed, but also said there would be nothing wrong with putting 200 ohms.
Then as usual, the thread took an interesting and most thought provoking turn, thanks mainly to @Julie.
I'm not sure it would ever take anywhere near that long to disconnect in any healthy circuit. The OP's measured Zs of 3.79ohms has to be the earth loop, not L-N loop, or it would be ~370m long at 4mm².
There is a mention somewhere in the OSG saying that the disconnection times should be met for L-N faults, but it doesn't link to a regulation. I'll try to find where it says this. Some of the final circuits in chapter 7 are limited by short circuit, but very few. I can't understand why this is, if it is not required by regs.
Here it is. Chapter 7, P77. Talks about 5s for some reason:
No, I strongly suspect that the actual L-N loop would be much less than Zs in a TT as you say.
But really my point is that we shouldn't be guessing - we should be checking.
On a TNx usually the Zs is higher than the L-N loop purely because the cpc portion is 2.5/1.5× or 4/2.5x the neutral resistance, which isn't a great difference; however in TT they are vastly different, - could be orders of magnitude.
So actually using 1667 or 200 is totally irrelevant, you wouldn't be cross checking this anyway, you would be using the rcd actual tests to prove it's OK.
As for the osg saying it should meet the same disconnection times, I am not sure it does, but I am more familiar with the regs, than the osg as I don’t need it that often, so happy to be wrong.
But that is sort of what I am saying, there is no MUST disconnect, but it should disconnect within a reasonable time frame. (if that's within the same times, great)
Exactly my point! ....... Is it 1667 ohms or 200 ???
OR Something else??
Confusing for the OP!
Cheers,yea it’s the earth loop.
Agree, it should disconnect sooner rather than later. But it is a struggle to find a circuit where SC loop impedance would the limiting factor, at least for the type of work I do. Voltage drop almost always gets there first. Very lightly loaded, very long lighting circuits perhaps.
The OSG paragraph is in my previous post #24, but doesn't link to a regulation, and I can't find anything in the regs that directly back it up.
On TT yes, I believe so.
In my opinion this should replace the earth fault loop test/calculations, as earth fault is covered by testing the rcd.
On TNx, not required, the earth fault loop covers it, if the rcd is additional protection
BUT, if the rcd is applied to the TNx because the earth fault loop cannot meet the disconnection time, then I think both sets of loops/calculations are needed.
Well I actually dug my osg out last night, dusted it off and cleaned the cobwebs away, and yes it's in there!
That note and looking at the table suggests the loop impedance is the most common limiting factor without a rcd, and my experience would be the same, although the only real occasion in my experience is long remote lighting with small loads.
I don't know of a regulation which determines a required trip time, but obviously there's a subsection for fault protection which doesn't seperate a need between line-line, line-neutral, or live-cpc/earth, basically demanding protection for all types of fault current, whist live-cpc/earth faults do have a specific set of disconnection times
Just thinking out loud:
434.5.2
A fault occurring at any point in a circuit shall be interrupted within a time such that the fault current does not cause the permitted limiting temperature of any conductor or cable to be exceeded.
it then goes on to give a rearrangement of the adiabatic to calculate that time. The adiabatic is only accurate for faults lasting up to 5 seconds. Perhaps this is why the OSG wants us to limit the duration of any fault to 5s?
Zs is measured to determine whether or not fault protection is achieved. The requirements for fault protection do not concern themselves in any way, shape or form with short circuit protection.
Section 434 is where you will find the requirements for disconnection under short circuit conditions.
To answer the OP's original question, 1667 ohms is the correct value as the sole means of fault protection is the 30mA RCD. If stable TN values have been achieved then the correct value to input would be 7667 ohms.
Section 434 is where you will find the requirements for disconnection under short circuit conditions.
To answer the OP's original question, 1667 ohms is the correct value as the sole means of fault protection is the 30mA RCD. If stable TN values have been achieved then the correct value to input would be 7667 ohms.
Last edited:
OFFICIAL SPONSORS
These Official Forum Sponsors May Provide Discounts to Regular Forum Members - If you would like to sponsor us then CLICK HERE and post a thread with who you are, and we'll send you some stats etc
Advert
Thread Information
- Prefix
- N/A
- Replies
- 32
Thread Tags
- Tags
- wondering
Advert
Advert
TrueNAS JBOD Storage Server
-
-
Understanding TrueNAS JBOD Servers: A Comprehensive Guide- Started by Dan
- Replies: 8
-
