IET logo
 
IET
Decrease font size
Increase font size
Topic Title: Fluctuating Zs Query
Topic Summary: EICR conundrum
Created On: 02 September 2017 11:59 PM
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.
 02 September 2017 11:59 PM
User is offline View Users Profile Print this message



inHen

Posts: 3
Joined: 11 December 2016

Whilst conducting an EICR on a property in a tower block, supplied by 25mm split con, 80A main fuse, 61008 100mA RCD, ringmain on a 32A C type 60898 wired in 2.5/1.5 readings fluctuating from 0.32ohms to 1.44ohms with a 1651b! Test repeated at several sockets resulting in similar results.

A colleague repeated the test with a 1662 multifunction tester and recorded slightly lower results but in a similar range.

The test was repeated with all other circuits and bonding conductors isolated and results were very similar.

Tests on the other ringmain in the property produced similar results as did a test on a ringmain in a completely separate property on a different floor.

Obviously I recorded the highest readings a full 1.1ohms above calculated with a C2 outcome and FI caveat.

Any insight would be welcome.
 03 September 2017 01:16 AM
User is offline View Users Profile Print this message



Alcomax

Posts: 212
Joined: 12 November 2009

A bit more info required such as :
1. Supply Meters at beginning or end of split con?
2: Assume "fluctuating" Zs means different readings on repeated tests. If so, how stable were results before and after RCD and stability on a radial such as lighting?

These were fluke multifunction testers? Not sure, but numbers ring a bell. In any event with no trip testers generally, you will get "odd" test results due to local factors such as the network distribution, distance of TX and general noise due to complexity of distribution . There can also be an uplift of results upstream of RCD.

If you have access or can borrow a more old skool high current loop tester, try this before the first RCD and compare with your meter readings.
 03 September 2017 08:05 AM
User is offline View Users Profile Print this message



leckie

Posts: 4263
Joined: 21 November 2008

Well as said a bit more info required.

So what about Ze, Zsdb, are those results stable?

Do you mean you are getting test results variances around different points of the ring, or test results are jumping around at individual test points?

Sketch out the property and socket outlet test position, record results as you take them.

End to end continuity tests results, and figure of eight results - what do they tell you. If you look at the results for each point do they make sense as a ring or are there spurs off the ring? These results should be more reliable in terms of not being affected by noise, low current Zs testing or the effects of the RCD.

So you should be able to access what the measured Zs should approximately be at each point from the above information. If it varies from what you are reading for Zs compared to what you expect from Zsdb + R1 + R2 at each point, is it consistent? If it is it's probably the effect f the RCD or the low current test method of the meter. F it's all over the place then it's more likely to be noise. My meter ((Megger) shows a symbol indicating noise during the test if present. When that shows all bets are off.

For testing purposes only, and if safe to do so, you could defeat the RCD and test again at high current - this should give you a more reliable result if you are concerned.

Giving a C2 for a high Zs might not be the correct thing if the reason for the results it the test method and/or noise.
 03 September 2017 10:39 AM
User is offline View Users Profile Print this message



John Peckham

Posts: 8773
Joined: 23 April 2005

in Hen

As a pedant and a lecturer in inspection and testing I have a few problems with your post.

1. It is a final ring circuit not a ring main.
2. Installations should not be energised with protective conductors disconnected. That is a fail on a 2394/2395 practical. See GN3.
3. You can't have a C2 with an FI. One or the other. Are you using old PIR forms?

Given you have tried 2 different meters is not a meter error given all low current ranges on test meters will give a wide range of test results. As Leckie said noise on the installation will just adds to the variability of readings. The loop tester holy grail is a no trip test facility that is consistent and reliable. My old Robin 4120DL with DLOK is very good is the best on RCD protected circuits but trips RCBOs. I have tried many different makes and models of loop tester and the Kewtech KT65DL "Triplok" is the best on the market even beating my much loved Megger LTWs.

My money would be on cheap Chinese sockets or old worn sockets. Releasing the socket from the back box and testing with probes on the terminals will probably give a steady more accurate reading. You get the same variability on R1 + R2 tests using a socket break out lead.

-------------------------
John Peckham

http://www.astutetechnicalservices.co.uk/
 03 September 2017 11:20 AM
User is offline View Users Profile Print this message



leckie

Posts: 4263
Joined: 21 November 2008

I had exactly what JP has just mentioned a couple of months back. I did a few Zs tests at sockets at survey stage around an office before I actually started the EICR just to get a feel for the job that I was doing over a weekend. A couple of the sockets gave very high test results which was alarming. I started the EICR the next day and went straight to this circuit.

The continuity tests were consistent on the end to end tests of the ring final circuit. I reconnected and went back to the sockets that I had got the high Zs test result and tested again, still high. I dropped the sockets off and tested at the terminal; a much lower test result in line with what I would have expected. I replaced the socket and test from the socket outlet and the result was fine.

I opened up the socket on my bench later, along with a few others with the same problem. The earth pin entry point was a bit of sprung metal that had lost its spring! So the contact pressure was very poor on non existent. the only thing is said regarding manufacturer was Classic. Never head of them, they had been installed for about nine years.
 03 September 2017 12:38 PM
User is online View Users Profile Print this message



AJJewsbury

Posts: 15714
Joined: 13 August 2003

When I've had fluctuating loop readings it's been down to a loose/tarnished connection somewhere - most recently the DNO's clamp on a lead sheath of the supply cable (you could actually hear the crackle during the high current test!).
- Andy.
 03 September 2017 01:13 PM
User is offline View Users Profile Print this message



leckie

Posts: 4263
Joined: 21 November 2008

Yes Andy, that's why I asked about the Ze and ZsDB. If they were stable then it would point to the sockets as the fluctuating readings were on more than one circuit. But if the ZsDB is hopping around then the problem is likely to be upstream.

I had the same as you, a crackling earth connection on a TNS connection to a supply cable. The occupants had been abstracting electricity and overload the supply massively. The pitch had melted out of the old service head and had run down the supply cable! I think they were growing some herbal remedy products.
 03 September 2017 02:59 PM
User is offline View Users Profile Print this message



weirdbeard

Posts: 3110
Joined: 26 September 2011

Originally posted by: inHen


Any insight would be welcome.


In addition to all the great posts above, your test equipment will have a duty rating, it will not be designed to perform repeated EFLI tests consecutively at the same point, there is normally a delay of at least the time it takes to unplug and move to the next point, it's the gubbins inside the tester that get warm and need a little cooling time between tests to give normal results.

-------------------------
:beer)
 03 September 2017 04:50 PM
User is offline View Users Profile Print this message



mapj1

Posts: 9258
Joined: 22 July 2004

you may be discovering the reason for the 100mA RCD
Joking aside, you say split con - so TN-S at that point, my thought/fear is it may be derived from TNC-S further back.
Please check that the PSSC stable, i.e. is the L_N loop OK. ?

I'd be worried by a real ohm in the L-N loop of an 80A supply, and I'd not be very pleased with it on a 32A final cct either, as it represents 30V drop on full load. If it only looks like an ohm on the E part of the loop then while it is interesting, it is not a show stopper, as the RCD will do ADS for you, even if the larger MCB breakers would be a bit slow off the mark

-------------------------
regards Mike
 13 September 2017 04:22 PM
User is offline View Users Profile Print this message



inHen

Posts: 3
Joined: 11 December 2016

Originally posted by: Alcomax

A bit more info required such as :

1. Supply Meters at beginning or end of split con?

2: Assume "fluctuating" Zs means different readings on repeated tests. If so, how stable were results before and after RCD and stability on a radial such as lighting circuit


1 supply meters at the end of split con, digital key type meter.

2 readings stable before rcd at 0.05ohms. Radial results stable.

As a pedant and a lecturer in inspection and testing I have a few problems with your post.

1. It is a final ring circuit not a ring main.
2. Installations should not be energised with protective conductors disconnected. That is a fail on a 2394/2395 practical. See GN3.
3. You can't have a C2 with an FI. One or the other. Are you using old PIR forms?


1 apologies, I use the colloquial when dealing with customers so often it becomes habit.
2 protective conductors were temporarily disconnected to see if they were effecting results, obviously it was a long shot and only done once all other final circuits had been isolated and results remained the same.
3 the C2 and FI was more of a flag to the main contractor that a revisit would be required and was not submitted on the final certificate, which wouldn't have been possible on the esycert software.

The L-N loop was fluctuating to a lesser degree but still not consistent enough to be discounted, in my opinion. Readings were taken at terminals behind sockets. All sockets were unplugged and only one redundant spur was present.

@ weirdbeard I think you may have a very valid point regarding the duty rating of the test equipment but the higher readings were what prompted the retests resulting in lower and then fluctuating results. If heat was a factor in the test kit I would expect raising results on the second test but it was the fluctuating results from high to low etc that baffled us.
 13 September 2017 04:54 PM
User is offline View Users Profile Print this message



KFH

Posts: 516
Joined: 06 November 2010

When I get unexpected high R1R2 or Zs on a socket using a plug I operate the switch a few times and unplug and plug the plug in a few times. Usually reduces the values to an acceptable level. If not then then test at the back of the accessory as suggested.
Note to the customer that the accessory may need replacing.
 13 September 2017 09:05 PM
User is offline View Users Profile Print this message



Alcomax

Posts: 212
Joined: 12 November 2009

I would plug for low current "no trip" loop tester "issue". Even though things appear more consistent on a radial and all circuits are on the same 100mA RCD [ I presume from info given ], it is still "uplift" of Zs results down stream of RCD and the nature of the ring final circuits [ i.e. going around in a circle all over the place ] is giving the particularly weird readings at the sockets. You say this is happening in the same building in other flats, so more evidence that the "ring" configurations are particularly affected. But be careful of condemning on excessive loop reading grounds. You will have to check the end to end ring resistances, calculate an R1 + R2 [ !!!!!] add to your Zs at the DB [ ZdB ] and see if those are excessive. I bet they are not.

If the RCD is an AC type [ very likely ] you would save lots of heartache by getting hold of an old skool high current loop tester with D lok or similar. Robin 4120DL as John suggests and also kewteck KT42 [ unfortunately no longer in production ].
Statistics

New here?


See Also:



FuseTalk Standard Edition v3.2 - © 1999-2017 FuseTalk Inc. All rights reserved.

..