symptoms of lost neutral?

[Gavin A]

I'd appreciate some second opinions here, as I've been sat here assessing the various symptoms of a situation, and coming to the conclusion that it seems to add up to a either a lost neutral, or high resistance neutral situation, and trying to work out if I need to be giving the building owner a ring over the weekend...

Situation is a factory unit, 3 phase, TN-C-S, 100amp per phase supply, at the end of a fairly long supply line at the extremes of it's capacity, and I think with several other units being fed from the same line prior to the factory, the DNO installations is probably 50+ years old.

Symptoms are

1 - extreme imbalance of the phase voltages, with the phases eg 250, 238, 220V respectively during the day when the 3 phase printing machines were operating relatively well balanced load, then just as I was being ushered out of the door with all the machines off and only a small single phase load for office lights, fridge, etc, I went to check why our inverter wasn't working and the voltage readings were showing something like 280, 250 and 180V respectively.

2 - Getting a few minor shocks between neutral and the earthed metal trunking on a little sub board with the 3 pole breaker feeding that board switched off.

We'd fitted a 30kWp solar PV system a month or so ago, but this was switched off when getting the neutral - earth shocks, and hadn't fired the PV generation up when I was monitoring the later most extreme phase voltage differences. Their regular electrician had also been in at some stage to apparently rebalance the phases.

Unfortunately I only went in there to set up some monitoring kit this afternoon, and was being rushed to get out of the door at the end by their guy wanting to lock up when I spotted the last voltage readings, so didn't have time to actually check or do anything other than turn the inverter circuit off that I'd been working on in case I'd done something on that circuit.


ps yes I'm only short course trained, but have been working alongside a proper spark for the last 4 1/2 years who did all the wiring and testing, I was just setting up the monitoring kit for the solar in a little single phase sub board we'd already installed, as I'm better at that stuff. Planning to go back first thing Monday morning with our spark to fully test it and work out what's going on, but I'm trying to work out if we actually should be trying to get hold of the owner to get in on Sunday before they actually fire up all the machinery on Monday morning.

Any thoughts appreciated

 

[Gavin A]

To pre-empt the questions, I didn't have chance to do any testing, I only just had change to knock the circuit off after seeing the readings before having to vacate the building. We'll either be back on Monday morning to test it properly, or if it's more urgent will see if we can get in over the weekend. It really wasn't obvious the extent of the problem was until all the 3 phase machinery had been turned off just before they were wanting to kick out. Turning the machines off seemingly transformed the situation from being a bit odd to seriously bad.

 

[Darkwood]

Loss of N on a 3ph installation can have some very strange and head scratching effects as you get a floating star point meaning in certain cases the single phase loads can carry on seemingly unaffected as they use motor windings, lighting control gear etc etc as a false N and you can find motors become noisy and problematic, banks of lighting work up to a point and the rest flicker or refuse to work and yes a very dangerous set-up .... many people relay a single PH neutral fault logic to 3ph and don't realise the situe' isn't always the same (depending on where the N is lost)

 

[Marvo]

The symptoms you describe do sound like a high impedance neutral and higher loads don't necessarily mean larger volt differences between phases, it's more about the balance of the phase currents.

Surely though whatever the issue is you're just going to throw it back to the DNO. Just the phase voltages you've stated would be sufficient reason to get the supply authority involved.

 

[topquark]

See if you can test at the service head, with the installation disconnected (not just phases isolated). If you see the same there contact the DNO for your area. Otherwise you could be chasing your tail fault finding if its a supply issue. Sounds a bit like a floating neutral.

 

[Gavin A]

The symptoms you describe do sound like a high impedance neutral and higher loads don't necessarily mean larger volt differences between phases, it's more about the balance of the phase currents.

Surely though whatever the issue is you're just going to throw it back to the DNO. Just the phase voltages you've stated would be sufficient reason to get the supply authority involved.

cheers - it's actually the lower loads that it was worst at, basically because the big loads are all fully balanced, whereas once the machinery is switched off it was just the single phase supply to the office, canteen, lights etc that was left with power through it, so it was far more imbalanced at that point.

That was what made me realise as I was driving away what I'd probably just witnessed.

It probably will be a DNO issue, but I want to repeat the voltage checks at the incomer, and also double check that it's not something stupid like my spark, or their spark failing to tighten the main neutral back up after doing his testing or something silly.... I never like to give the DNO engineers the opportunity to take the **** / like to rule everything out first.

 

[Gavin A]

I was already going to email the DNO about the voltage imbalance, but it's a bit more urgent if it's actually a floating neutral situation causing it.

Thanks for the confirmation all 3 of you, I'll get in touch with the building owner then.

 

[Engineer54]

This sounds more like a high impedance neutral fault, rather than a floating neutral situation. That fault could be on or around the main cutout, or on the DNO supply side....

Is this a DNO overhead supply to the factory unit or underground cable supply??

 

[Gavin A]

This sounds more like a high impedance neutral fault, rather than a floating neutral situation. That fault could be on or around the main cutout, or on the DNO supply side....

Is this a DNO overhead supply to the factory unit or underground cable supply??

That was what I was thinking as well, as the voltages while pretty extreme, aren't

UNderground cable.

Thing is, the DNO had already said they were concerned about the capacity and volt drop on the cable, but gave the go ahead for 16amps per phase solar connection - they wanted to upgrade the cable for anything more, basically stating that it was already really over capacity so they had to have the transformer at the highest tapping to maintain voltage at peak demand levels.

First thing I want to double check is that the neutral connections have all been remade properly in the distro boards after testing, but to do that I have to remove the lid to the board, which in turn means powering down the entire factory just to get the lid off.

 

[topquark]

Given the neutral/earth shocks quite likely Eng. I'd still try to get a quick internal versus external root cause by disc and checking at the head before doing too much diag elsewhere. If the neut/earth impedance is high, thats prob most of the prob, if not all. I would also check the TNC-S assertion given the age, whilst you're disconnected. It's obv been converted to PME at some point if it is TNCS. Worth checking the earthing/bonding arrangements for the other boards after that.

 

[Gavin A]

hmm, well it turned out to be a faulty component with continuity between L1 and L3, and only really affecting our circuit (luckily).

Really bizarre fault, I can't really get my head around how it was causing the symptoms, but as soon as we removed it the voltages sorted themselves out.

Also found 0.33MOhms on the insulation resistance test through our generation meter between the phases and neutral, so will have to replace that as well.

And to cap it all off, as I was packing up my sodding bits box came flying open and spewed the contents across the floor of the office. Today was not a good day.

 

[Marvo]

hmm, well it turned out to be a faulty component with continuity between L1 and L3, and only really affecting our circuit (luckily).

Can you elaborate? What was the component and if there was continuity between 2 phases to the point where the phase voltages were being radically affected why didn't the OCPD disconnect? Also how would this type of fault cause an elevated neutral voltage.

I had far less questions about this before you made your last post and confused the hell out of me......

 

[Darkwood]

hmm, well it turned out to be a faulty component with continuity between L1 and L3, and only really affecting our circuit (luckily).

I also would like to know what component you speak of and what test you used to establish this?
Continuity is usually associated with low resistance and this would lead to a near if not dead short circuit.

Really bizarre fault, I can't really get my head around how it was causing the symptoms, but as soon as we removed it the voltages sorted themselves out.

Also found 0.33MOhms on the insulation resistance test through our generation meter between the phases and neutral, so will have to replace that as well.

You insulation tested the meter ???... Are you not surprised this is the reading you got and should you really be insulation testing the device!

And to cap it all off, as I was packing up my sodding bits box came flying open and spewed the contents across the floor of the office. Today was not a good day.

It seems you have not really got to the bottom of this and from your wording would seem you are putting IR tests through connected loads and wondering why you get low readings... I feel you haven't yet got to the bottom of this.

 

[Gavin A]

Can you elaborate? What was the component and if there was continuity between 2 phases to the point where the phase voltages were being radically affected why didn't the OCPD disconnect? Also how would this type of fault cause an elevated neutral voltage.

I had far less questions about this before you made your last post and confused the hell out of me......

I'm pretty unclear about this myself, there is the potential for there to have been 2 faults occurring at the same time as our 3 phase generation meter also had very low resistance through the meter between all phases and neutral - 0.33 Mohms, something that appears to have happened since the original installation a month ago as the IR tested out fine at the time (according to our test sheets, still trying to confirm that this was done right at the time).

In terms of the faulty equipment though, we were able to isolate the different components and monitor the voltage as we switched each item on in turn. As soon as this kit was switched on the voltage swang wildly to 300+V on L3, and 80V on L1 and maybe 250V on L2, something like that, we didn't stop to write the voltages down when we got those readings, though we did repeat it with a 2nd tester just to be sure.

Pulled the unit out and the voltages settled down to within 6V of each other across the 3 phases at around 234V.

Tested continuity between the phase terminals on the unit and had continuity between L1 and L3 phase terminals through the unit without anything else wired into it.

The unit was a din rail mounted £500 piece of monitoring kit designed to monitor generation, consumption, export etc via 6 x CT clamps. It does have a relay function that we had wired in to something, and an alarm input from the inverter, so possibly it was something to do with them, but I don't see how.

It's baffled me as well, as the unit was protected via a 6amp 3 pole MCB, so if there were an actual short circuit running through the unit then I'd have expected the MCB to trip.

All I know is that as soon as we pulled that unit out of the system all the problems disappeared.

We have left it locked off though until we can replace the meter, as at the point it was working we'd bypassed the neutral on the meter due to the low IR reading, and i didn't try it with that neutral back through the meter, so it's possible that this was the original cause of the problem, and the other unit had been damaged by the voltage imbalance... will test that out when we return to replace the meter.

It was all very odd - another test we'd done previously was all phases to neutral, and all phases to earth. To neutral we were getting seriously wrong voltages, but testing to earth and we were getting approx 234V per phase pretty much in balance.

Glad it's not just me being a bit confuddled on this one, my suspicion is that there was / is more than one problem going on here, possibly a bit of a cascade failure scenario that we caught in the middle of it going much more badly wrong. Will be back in a few days with replacement components, see if we can properly work out what was going on.

 

[Marvo]

I'm pretty unclear about this myself, there is the potential for there to have been 2 faults occurring at the same time as our 3 phase generation meter also had very low resistance through the meter between all phases and neutral - 0.33 ohms, something that appears to have happened since the original installation a month ago as the IR tested out fine at the time

Apologies in advance, I haven't got time to go thru your entire post now but your opening remark really makes no sense to me. If the meter is low resistance between phases (0.33 ohms) there would have been a fault current in excess of 1kA which would certainly have been sufficient to cause circuit interruption by any OCPD.

How did you get this low IR result? Can you give a blow by blow explanation of how you tested and can you say what instrumentation you were using? There's definitely something hinkey going on.

 

[Gavin A]

You insulation tested the meter ???... Are you not surprised this is the reading you got and should you really be insulation testing the device!

just our generation meter, and only at 250V.

no loads were connected, it was isolated at both ends of the cable when we tested through the meter.

Well aware we've not completely got to the bottom of this, but I also know that when we remove the meter and other faulty unit from the circuit that all symptoms of the problem ceased completely.

We were also on time constraints and stuck in a tiny space in the corner of a busy office, so figured it was best just to remove and replace the components that seemed to be causing the fault rather than investigating on site any further at that stage.

I'll see if we can get time to test the faulty unit any further, just a bit concerned about voiding the warranty.

 

[Gavin A]

Apologies in advance, I haven't got time to go thru your entire post now but your opening remark really makes no sense to me. If the meter is low resistance between phases (0.33 ohms) there would have been a fault current in excess of 1kA which would certainly have been sufficient to cause circuit interruption by any OCPD.

How did yu get this low IR result? Can you give a blow by blow explanation of how you tested and can you say what instrumentation you were using?

isolated the circuit at the supply side via a 3 pole isolator, and removed the neutral from the neutral bar. Isolated the citcuit at the inverter end via an all pole isolator.

Took IR readings at 250V between L1>Neutral, then L2>Neutral, then L3 to Neutral actually on the meter terminals. IR phase to phase was >500 (or >1000, I forget which it is on that meter, no continuity anyway).

We did this because we'd originally tested the cable, and came up with the 0.33 MOhms reading from phase to neutral on the cable, so split the cable at the generation meter and tested each side of the cable with the result that both sides of the cable tested out fine, so we then tested the meter itself with both neutral cables actually disconnected from the meter and got the 0.33Mohms reading.

I have to admit, I'm not 100% clear that this isn't just some function of the generation meter as I don't usually do the testing, but it's never cropped up before, and we have a couple of spare meters in the warehouse I was intending to repeat the test on just the be sure.

I think we were pretty careful to remove any potential for false readings - I did wonder about the surge protection devices in the inverter, but not when we repeated the test on the meter with both neutral cables removed.

eta - the main meter used was a Fluke MFT, unsure of the model. Double checked some of the readings with a metrel MFT, and did the final continuity tests and double checked the voltages with a multimeter / clamp meter that we mainly use for the DC circuit testing, I forget the model but it was £100 or so, so not a cheap Argos job.

 

[Marvo]

Then the reading you got has to be something internal with the circuitry in the meter itself. I think we can safely assume if there was over 1kA flowing for any length of time and the meter was the only load then you'd have noticed the symptoms of this ie the explosion and the ensuing fire not to mention the disruption in electrical supply to the surrounding area .

 

[Darkwood]

His original reply stated 0.33ohms he's amended it to 0.33Mohms - the electronics in the meter would clearly account for this and the last thing Id really want to do is Megger 500v through a £500 set-up. the meter shou;d have been isolated before testing, the meter is a load you can damage it, it will have electronic circuitry!

 

[Gavin A]

Then the reading you got has to be something internal with the circuitry in the meter itself. I think we can safely assume if there was over 1kA flowing for any length of time and the meter was the only load then you'd have noticed the symptoms of this ie the explosion and the ensuing fire not to mention the disruption in electrical supply to the surrounding area .

ah right, ok I made a typo in that post, but I had referred to it as 0.33 Mohms consistently throughout this thread before and since that post.

The figure for current flow from any phase to neutral with resistance of 0.33 Mohms would be 230 / 330000 ohms = 0.00007 A, or 0.7mA, or thereabouts.

so thanks for the ****taking, but tbh I'm not really in the mood.