My new Megger MFT 1730

[eveares]

In regards to EddieM, another reason I wanted my own MFT is so I can ensure that my wiring is electrically sound and everything is as it should be.

I am rather grateful for having it now, as without it I would not have discovered that I have a higher than permitted Zs for my kitchen radial socket circuit. My kitchens socket circuit is a radial circuit protected by a 32A type B MCB that is wired in about 20M of 4mm² T&E and has a system loop impedance (Zs) of 1.23Ω even though the PSCC was 187A. On a another circuit what is a 20A Radial wired in 2.5mm² T&E, I got a system loop impedance (Zs) of 1.16Ω on a socket what was about the same distance from the DB as the socket in the kitchen with the high Zs value.

I suspect the high loop impedance reading in the kitchen is down to where the first item/s on the circuit are 5 FCU's for the fixed kitchen appliances (4mm² T&E loops in and out of each FCU). I remember my electrician having a hard time installing the FCU's while at a awkward angle, plus we had settlement in that area of the house where the floor dropped by about 40mm few years ago. All is nice and tight in the DB/CU.

Reference for permitted Zs values: https://www.niceic.com/Niceic.com/media/Pocket-Guides/Pocket-Guide-18.pdf

Also, is this outdated as it has different Zs values than the above link: http://www.electricalsafetyfirst.org.uk/mediafile/100106752/master_EARTHFAULTLOOPtable.pdf

Can anyone scan and attach the page of permitted Zs values from Amendment 3 of BS7671:2008.

Regards: Elliott.

 

[EFLImpudence]

I agree with the second link. NIC's 1.08 seems a bit low.

I will scan them tomorrow for you, if you want but the second link looks ok (at a quick glance).

You can work them out yourself if you want.

230V ÷ 5 x the B type MCB rating (10 x C type rating or 20 x D type rating)

so

230 ÷ (5 x 32) = 230 ÷ 160 = 1.4375 call it 1.44Ω

Then divide by 1.2 to account for the higher impedance when the conductor is 70°

1.44 ÷ 1.2 = 1.2Ω (that's a coincidence)

Then 95% of that to allow for the (nearly) lowest permited voltage

1.2 x 95% = 1.14Ω


That's the worst case scenario - your voltage won't be that low and the conductors are unlikely to be that hot.

 

[JohnW2]

In regards to EddieM, another reason I wanted my own MFT is so I can ensure that my wiring is electrically sound and everything is as it should be.

Indeed, and I've already made that point to him. As I told him, because I have the equipment and knowledge to do it, I am able to check my electrical installation far more frequently than I could reasonably afford to have an EICR undertaken by a third party.

That could be particularly relevant in terms of RCDs. IF one believes that they are worth having, and given that they are said to have a significant in-service failure rate, to have them properly tested once every 5 or 10 years (if that!) is really not very satisfactory. After all, an MOT on a vehicle every 5 or 10 years (or never!) would not be regarded as acceptable in terms of safety considerations.

Kind Regards, John

 

[JohnW2]

I agree with the second link. NIC's 1.08 seems a bit low.

It does, but only slightly in these Amd3 days. However, those NICIEC figures (dated Jan 2015) are presumably meant to be pre-Amd3 ones, hence should be appreciably higher than present-day ones.

You can work them out yourself if you want.

One can, but I think that, in terms of the method described in the regs, there is a small error in the calculation you present. I think that the temperature correction factor one divides by should probably be 1.25 (expressed as "multiply by 0.8", as in Appendix 14 of regs), not the 1.2 you used. Hence I think your calculation should read:

230V ÷ 5 x the B type MCB rating (10 x C type rating or 20 x D type rating)
so
230 ÷ (5 x 32) = 230 ÷ 160 = 1.4375 call it 1.44Ω
Then divide by 1.2 1.25 to account for the higher impedance when the conductor is 70°
1.44 ÷ 1.25 = 1.2Ω (that's a coincidence) 1.152Ω
Then 95% of that to allow for the (nearly) lowest permited voltage
1.2 1.152 x 95% = 1.14Ω 1.0944Ω

If you had not rounded 1.4375Ω to 1.44Ω, then the answer would have been 1.0925Ω

My corrected calculation gives an answer very close to the figure based on Table 41.3 of the regs. That gives 1.37Ω (at 70°C) for a B32, which when multiplied by the temperature correction factor of 0.8 (or divided by 1.25, per App 14) gives 1.096Ω.

In that regard, I have to say that I've always felt that Tables 41.2 - 41.4 of the regs are potentially misleading (at least for people who do not read them, and their footnotes, very carefully), since I suspect that at least some people (nearly all of whom will measure Zs 'cold') will look at them, without reading the footnotes or Appendix 14, and will assume that those Tables present the maximum permissible Zs values in terms of what they measure (usually at around ambient temp').

Kind Regards, John

 

[scousespark]

I was not querying your point about testing, I doubt any electrician (or knowledgeable DIYER) would argue against stating that the importance of testing.
I was puzzled by you asking why Eddie would be interested in this forum. Surely people like Eddie would benefit most from advice here as he has more to learn.
As to the spec, it was just an observation that the 1730 is a high spec bit of kit for somebody not using it on a daily basis.

 

[EddieM]

I was not querying your point about testing, I doubt any electrician (or knowledgeable DIYER) would argue against stating that the importance of testing.
I was puzzled by you asking why Eddie would be interested in this forum. Surely people like Eddie would benefit most from advice here as he has more to learn.
As to the spec, it was just an observation that the 1730 is a high spec bit of kit for somebody not using it on a daily basis.

Quite so, the information in this forum is invaluable, even if that shows that one doesn't know much about the subject, the mere fact you know you don't know much is invaluable. Really I am not "having a go" at anyone and as John said I too hope Elliot enjoys his purchase and gets many years of service from his MFT.

 

[ban-all-sheds]

Can anyone scan and attach the page of permitted Zs values from Amendment 3 of BS7671:2008.

You've splashed out on an MFT because you want to do things properly.

So here's another investment to make (only £65 this time)

 

[JohnW2]

I was not querying your point about testing, I doubt any electrician (or knowledgeable DIYER) would argue against stating that the importance of testing.

Indeed. As I said I would not have expected you to disagree with the initial answer I gave.

I was puzzled by you asking why Eddie would be interested in this forum. Surely people like Eddie would benefit most from advice here as he has more to learn.

Fair enough. I suppose it depends upon how much time he has. I'm a great believer in learning (anything), but I don't think I'd personally have the time to learn (to any significant extent) about things that I was only going to do 'once in a blue moon'.

As to the spec, it was just an observation that the 1730 is a high spec bit of kit for somebody not using it on a daily basis.

That may be true but, as I said, I thought Eddie's point was a general one about DIYers wanting/having any MFT. He probably did not know anything about the spec of the particular one in question.

Kind Regards, John

 

[EFLImpudence]

One can, but I think that, in terms of the method described in the regs, there is a small error in the calculation you present. I think that the temperature correction factor one divides by should probably be 1.25 (expressed as "multiply by 0.8", as in Appendix 14 of regs), not the 1.2 you used. Hence I think your calculation should read:

I think the 1.2 is correct and the 0.8 used is merely an approximation - or plain mistake.
All the calculations in the NICEIC table appear to be done with approximations and rounding down at every stage.

The temperature coefficient of copper is 0.393% per degree.
Assuming 20°C when measured then a 50° rise to 70° is 50 x 0.393 = 19.65%.

Therefore the resistance measured should be corrected by multiplying by 1.1965.
The reciprocal 1.1965 is 0.836 (using 1.2 it is 0.833).

My corrected calculation gives an answer very close to the figure based on Table 41.3 of the regs. That gives 1.37Ω (at 70°C) for a B32, which when multiplied by the temperature correction factor of 0.8 (or divided by 1.25, per App 14) gives 1.096Ω.

That is mathematically true but I think 0.8 is not the correct figure.

1.37 x 0.836 = 1.145
1.37 ÷ 1.1965 = 1.145

 

[flameport]

Tested my External loop impedance (TN-S Arrangement) with 2-Wire high current and got a nice reading of 0.16Ω for Ze.

20M of 4mm² T&E and has a system loop impedance (Zs) of 1.23Ω

1.23 is implausible given the first result.
0.4 would be a more reasonable figure.

If you are using the plug adaptor in a socket, get another decent plug (such as an MK one) and shove it in/out of the socket 10 times, then test again.
Or test on the terminals on the back of the socket.

A dead test of the circuit resistance would also be revealing.

If you have solar panels, disconnect them before testing.

 

[JohnW2]

I think the 1.2 is correct and the 0.8 used is merely an approximation - or plain mistake.

Hmmm. Interesting. Although Appendix 14 of the regs is theoretically only 'informative', the footnote to Table 41.3 clearly says that if Zs is not measured at the maximum permitted operating temperature of the cable, then the figures in the Table must be adjusted as per Appendix 14 - so I'm not clear as to whether that becomes a 'requirement' (of the regs) or not. If it does, then using 0.833 (aka "1.2") could make the Zs of a circuit certainly appear 'OK' when it was, in fact, non-complaint with regs (plus App14).

The temperature coefficient of copper is 0.393% per degree. Assuming 20°C when measured then a 50° rise to 70° is 50 x 0.393 = 19.65%. Therefore the resistance measured should be corrected by multiplying by 1.1965. The reciprocal 1.1965 is 0.836 (using 1.2 it is 0.833).

I can't disagree with any of that but, as above, I'm not clear as to whether the regs are 'requiring' the use of 0.8 - which leads to a more demanding max Zs.

I still think its a little odd that the regs 'major' on Tables which give the maximum Zs when measured at a conductor temperature which no-one would measure it at!

Kind Regards, John

 

[EFLImpudence]

The alternative, where 1.25 would be correct, is that measurements were and are taken at 7°C.

Appendix 14 does say that if the measurement exceeds that arrived at by using the formula then "a more precise assessment may be made".

I will be charitable and say that 0.8 is just an approximate figure for first use, otherwise I don't understand why it does not simply use "divide by 1.2".

I could go as far as to say the 0.8 has been arrived at out of misunderstanding of the maths, in that the reverse of increasing a figure by 20% (which must have been done to reach the 70° values from 20°) is to reduce the product by 20% (i.e. 80% of it).

Just as people think VAT is 20% of the final price.

 

[eveares]

Just bought a copy of BS7671 from TLC. 1.37Ω is the maximum earth fault loop impedance value for a Type B 32A MCB in the table on page 58, what all ready has the 0.95 voltage factor applied.

So 1.37Ω x 0.8 = 1.09 permitted Zs

Also, what pages does the 0.8 temperature correction factor show up in BS7671 other than on page 452 under Appendix 14?

Flameport, poor contact with the kitchen socket was on my mind for causing the high readings. It is a kitchen with steam, grease and oil from cooking after all. Will try a a variety of socket adapters and plugs a number of times and hope for a better reading before I start taking everything out from under the boiler cupboard where the 5 FCU's are.

 

[JohnW2]

Appendix 14 does say that if the measurement exceeds that arrived at by using the formula then "a more precise assessment may be made".

It does, so that does enable one to do it 'properly'.

I wonder how precise people get about this. I would imagine that, for example tests undertaken on a new installation in a new build are usually undertaken in an unheated building - which, at some times of year, could presumably be considerably cooler than 20°C.

I will be charitable and say that 0.8 is just an approximate figure for first use, otherwise I don't understand why it does not simply use "divide by 1.2".

Yes, that's possible.

I could go as far as to say the 0.8 has been arrived at out of misunderstanding of the maths, in that the reverse of increasing a figure by 20% (which must have been done to reach the 70° values from 20°) is to reduce the product by 20% (i.e. 80% of it). Just as people think VAT is 20% of the final price.

... and, yes, that possibility did occur to me, given that it is so common a mistake. Indeed, I confess that, despite my 'mathematical mind' and education, I had to pause for a few seconds just to satisfy myself that "multiply by 0.8" and "divide by 1.2" were not the same thing!

However, I think I may have discovered (roughly!) what is going on. I've just looked at my OSG (albeit I only have a red one). It carefully explains why the maximum Zs figures it gives are different from those in Tables 41.2 - 41. of the regs, and then gone on to say that the figures in the OSG relate to measurements undertaken at 10°C (maybe they are thinking of the 'new builds' above!), which they say uses a divisor of 1.24 (corresponding to a multiplier of 0.806). That is very close to the Appendix 14 figure of 0.8, so I suspect (I've just noticed that they don't say) that the 'deemed to satisfy' 0.8 figure in App 14 relates to measurements at 10°C, rather than the 20°C figure (more reasonable for inhabited premises!) you used for your calculations.

Kind Regards, John

 

[JohnW2]

Just bought a copy of BS7671 from TLC.

I hope that you didn't pay too much for it, given that it's only got about 9 months' life left.

1.37Ω is the maximum earth fault loop impedance value for a Type B 32A MCB in the table on page 58, what all ready has the 0.95 voltage factor applied. So 1.37Ω x 0.8 = 1.09 permitted Zs

Indeed - that's roughly what I said (using the App 14 0.8 temperature correction factor).

Also, what pages does the 0.8 temperature correction factor show up in BS7671 other than on page 452 under Appendix 14?

That's the point - it doesn't appear anywhere other than in Appendix 14. However, see what I've just written to EFLI - it appears that the 0.8 figure relates to measurements taken with an ambient temperature of (or about!) 10°C.

Kind Regards, John