3 phase problem after spark did 5 year safety inspection

[jimthegent]

L2 in that three phase board loo from the electrical

No, definitely not.


The fault you show is of the top (input) of a single phase consumer unit. A 3phased balance load would not have any bearing on the fault IMHO.
The damaged cable is a 'red herring', I believe as the clamp being insulating material there would nothing to make a circuit and cause a 'short'.
I would think the DP Isolator has failed and there has been a L-N 'short' within the isolator, or maybe a stray strand!

The clamp is mild steel

 

[plugsandsparks]

Only going to know for sure with photos of the back of the clamp and back plate, if there are no marks i would agree and internal failure of the isolator comes into play, then was it caused by a loose connection or an internal failure of the DP Isolator. I saw the split in the tail on L2 supply and blamed the clamp (until proven otherwise) - it possible the heat generated by a loose connection compromised the tail insulation, hence you would need to see the components to work out what went first.
However i think we all agree unless the CB in the TP board is welded shut, then ADS should have saved it from this mess if it was indeed overload. (assuming the SP board is fed from the TP board) - i cannot tell from pic

 

[jimthegent]

L2 in that three phase board looks like its got hot too?

Could that have come from the electrical fire in the box we had it was cleaned up but initially there was a load of soot and detritus all over the box

 

[plugsandsparks]

I cannot see a SP breaker in the TP board to feed the SP board, so do the tails go to the Henleys and then straight to the incoming fuse? - If so i assume (again) this is either a 80A or 100A TP supply?

 

[jimthegent]

Only going to know for sure with photos of the back of the clamp and back plate, if there are no marks i would agree and internal failure of the isolator comes into play, then was it caused by a loose connection or an internal failure of the DP Isolator. I saw the split in the tail on L2 supply and blamed the clamp (until proven otherwise) - it possible the heat generated by a loose connection compromised the tail insulation, hence you would need to see the components to work out what went first.
However i think we all agree unless the CB in the TP board is welded shut, then ADS should have saved it from this mess if it was indeed overload. (assuming the SP board is fed from the TP board) - i cannot tell from pic

Only going to know for sure with photos of the back of the clamp and back plate, if there are no marks i would agree and internal failure of the isolator comes into play, then was it caused by a loose connection or an internal failure of the DP Isolator. I saw the split in the tail on L2 supply and blamed the clamp (until proven otherwise) - it possible the heat generated by a loose connection compromised the tail insulation, hence you would need to see the components to work out what went first.
However i think we all agree unless the CB in the TP board is welded shut, then ADS should have saved it from this mess if it was indeed overload. (assuming the SP board is fed from the TP board) - i cannot tell from pic

We are never going to know that the contractor returned the next day replaced the breaker replaced the burnt out tail and clean the crime scene up they also removed the items which technically still belong to us

 

[plugsandsparks]

OK, but your loads presumably have not changed prior to now, so just measure the load now.
A pic of the SP board breakers may give a clue to whether overloading the board is even possible.

It was rather naughty to not leave the isolator with you, however you still have the board and clamp. ;-)

 

[jimthegent]

OK, but your loads presumably have not changed prior to now, so just measure the load now.
A pic of the SP board breakers may give a clue to whether overloading the board is even possible.

It was rather naughty to not leave the isolator with you, however you still have the board and clamp. ;-)

They probably

 

[MJPD29]

Yes the cover of the board was blown off and there's a lot of black soot all over the switches so clearly a lot of heat was generated

The cover was blown off? Really?

 

[MFS Electrical]

Yes 100% he took them out and put them back in again

Certainly not a standard thing to do on an inspection.

When you do the 2391 (inspection and testing) you are told to disturb as little connections as possible to ensure you don’t create faults.

It’s highly unlikely that the electrician would have removed the tails and reterminated them on an EICR (inspection)

Also, the picture you’ve posted looks like a single phase main switch?

 

[MFS Electrical]

An overloaded board but it's run fine for 5 years and it just happend to blow a few hours after they wiggled the wires plus each phase is perfectly balanced with 9 feeds

So he didn’t remove the tails? But wiggled them?
Sounds more likely that there’s been a problem hiding there for a while either from a loose connection causing the cable to overheat and eventually melt into the clamp or the wiggle has exacerbated a defect already there and with a little load on the cable the cable moves and boom.

Also re the overload; It is certainly possible that an overload has occurred even if it has been running fine for years. An overload of a small magnitude can be enough to heat the cable past its operating temperature and cause it to significantly expand, then upon cooling and contracting loosen the connection. Rinse and repeat for 5 years and you certainly could have a situation like this.

I see it all the time where a shower installed on a 6mm cable ran through insulation in a wall and then under loft insulation is used to power a 9.5kW shower and due to the overloading of the cable (which should have been derated due to the insulation) burns out at the terminations of the switch or the MCB or the shower.
Genuinely, funnily enough it usually takes around 5 or so years of normal use for the problem to come to light by which time the spark that put it in is long gone.

 

[Yozza19]

I don’t think the clamp is the source of the problem it would of been in a dam sight worse condition if it was

I’d say a loose connection that has eventually caused the internals to short

To be honest the whole design of l2 to be loaded up like that is a sh!t show
Another 3 phase db instead and sockets spread across the phases

It’s easy to sit there blaming the spark who could easily of been anyone one here with ought knowing the full circumstances

The biggest thing that stands out to me is he used the tail clamp what puts him/her above 60/70 %of the butchers that get away with it week in week out

 

[marconi]

Good morning. If the cooling system fed by the 3P db did indeed have three phase 3 wire motors, then even though the main L2 fuse ruptured the site‘s L2 conductors would remain energised because it is connected by the motor windings to L1 and L3. The voltage of L2 with respect to N will depend on the L2 load impedances and that of the motors. The current in L1 and L3 will increase as the 3 phase motors run in single phase mode. The motors would not start with only L1 and L3 but they can continue to spin if they were already and a phase was lost.

What happens to the 3-Phase Motor When 1 Out of 3 Phases is Lost? - https://www.electricaltechnology.org/2021/11/what-happens-to-the-3-phase-motor-when-1-out-of-3-phases-is-lost.html

Thus even though L2 fuse ruptured further damage to equipment loads may have happened using current drawn from L1 and L3. What if anything happened to isolate all electricity ie disconnect L1 and L3?

Where the servers fed by UPS and/or voltage regulators?

I concur that the original short circuit was caused by a failure of the termination and squashing of the cable by the clamp. But can see how L2 may have remained energised and significant current might still flow causing further damage.

 

[jimthegent]

They probably

They probably

So he didn’t remove the tails? But wiggled them?
Sounds more likely that there’s been a problem hiding there for a while either from a loose connection causing the cable to overheat and eventually melt into the clamp or the wiggle has exacerbated a defect already there and with a little load on the cable the cable moves and boom.

Also re the overload; It is certainly possible that an overload has occurred even if it has been running fine for years. An overload of a small magnitude can be enough to heat the cable past its operating temperature and cause it to significantly expand, then upon cooling and contracting loosen the connection. Rinse and repeat for 5 years and you certainly could have a situation like this.

I see it all the time where a shower installed on a 6mm cable ran through insulation in a wall and then under loft insulation is used to power a 9.5kW shower and due to the overloading of the cable (which should have been derated due to the insulation) burns out at the terminations of the switch or the MCB or the shower.
Genuinely, funnily enough it usually takes around 5 or so years of normal use for the problem to come to light by which time the spark that put it in is long gone.

Sorry I was being flipent there he did remove the tails and put them back in the day before the outage

 

[plugsandsparks]

MFS makes an interesting point, i cannot recall needing to remove tails from an isolator to do an EICR, plus if they are connected to the Henley rather than a CB off the TP board, they would have been live unless the main fuse was pulled or there is another TP isolator for the whole supply. If you have the ECIR report it will tell us how the board is connected to the supply. I am wondering if he was getting poor readings when testing to the isolator terminals and removed the cables to work out why ?

Anyway, plenty of theories, which do not appear to have a way forward without some evidence to rule in or out various possible causes

 

[jimthegent]

MFS makes an interesting point, i cannot recall needing to remove tails from an isolator to do an EICR, plus if they are connected to the Henley rather than a CB off the TP board, they would have been live unless the main fuse was pulled or there is another TP isolator for the whole supply. If you have the ECIR report it will tell us how the board is connected to the supply. I am wondering if he was getting poor readings when testing to the isolator terminals and removed the cables to work out why ?

Anyway, plenty of theories, which do not appear to have a way forward without some evidence to rule in or out various possible causes

Morning the best evidence I can produce is that we had perfect balanced system un stressed working well and a few hours after someone did work on it the system failed I do have a video of the electrician who came the next day to rectify on the phone to the office saying "it looks like we've nipped the outer casing when we put it back" and her boss saying is the main fuse that blew on the same line as the breaker......they know they've balls'up but that'd why I so great full to you guys as I'm not an electrical engineer and they are trying to lead me by the nose to bankruptcy

 

[suffolkspark]

That single phase board doesnt appear to be fed by the three phase board, so at some point L2 looks to have been loaded up, or loose, with the visible heat damage there, but is possibly un related, theres just so many unknows here for us, which board are you saying blew off the wall? If you could post the report on here with the inspectors details redacted that may help, but im not so sure. You mention data loss, was there no off site backups? Or ups? Or anything to alert you to the supposed high temperatures in that area by the fans not working? I may have detsils mixed this threads getting a bit long now

 

[marconi]

The Sp consumer unit is fed from the Henley block.
Op - Please could you confirm whether the feed to 3P board second from left which looks discoloured is L1 or L2. Is this feed actually discoloured from overheating?

 

[MJPD29]

How do you know your system is perfectly balanced? Why is that board fed in tails? Generally heavily loaded single phase boards would indicate the system is unbalanced in my experience. I can’t really see why the spark had the tails out for testing either, just doesn’t really make sense.

Edit- no answer to my previous question either, did the cover actually blow off? How did that happen with threaded screws on the cover, did they break? Do you have a picture?

 

[jimthegent]

Morning the best evidence I can produce is that we had perfect balanced system un stressed working well and a few hours after someone did work on it the system failed I do have a video of the electrician who came the next day to rectify on the phone to the office saying "it looks like we've nipped the outer casing when we put it back" and her boss saying is the main fuse that blew on the same line as the breaker......they know they've balls'up but that'd why I so great full to you guys as I'm not an electrical engineer and they are trying to lead me by the nose to bankruptcy

How do you know your system is perfectly balanced? Why is that board fed in tails? Generally heavily loaded single phase boards would indicate the system is unbalanced in my experience. I can’t really see why the spark had the tails out for testing either, just doesn’t really make sense.

Edit- no answer to my previous question either, did the cover actually blow off? How did that happen with threaded screws on the cover, did they break? Do you have a picture?

Hi its our job to have precise amps distributed to our servers so we take a great deal of effort to get it bang on the lid was on the floor and covered in black soot when we went in as was the entire board with heat marks all over

 

[jimthegent]

How do you know your system is perfectly balanced? Why is that board fed in tails? Generally heavily loaded single phase boards would indicate the system is unbalanced in my experience. I can’t really see why the spark had the tails out for testing either, just doesn’t really make sense.

Edit- no answer to my previous question either, did the cover actually blow off? How did that happen with threaded screws on the cover, did they break? Do you have a picture?

Also this is a local authority building so the electricians are paid in a way that anyone who's got a local authority contract will understand its in their interest to have everything undone and redone and they also then get the gig to do any remedials and this is done on the nod a couple of photos to the Borough electrician and help yourselves lads look forward to your invoice

 

[jimthegent]

How do you know your system is perfectly balanced? Why is that board fed in tails? Generally heavily loaded single phase boards would indicate the system is unbalanced in my experience. I can’t really see why the spark had the tails out for testing either, just doesn’t really make sense.

Edit- no answer to my previous question either, did the cover actually blow off? How did that happen with threaded screws on the cover, did they break? Do you have a picture?

There's a pic of inside the box there's black all up the rhs and inside it's black and scorched

 

[jimthegent]

How do you know your system is perfectly balanced? Why is that board fed in tails? Generally heavily loaded single phase boards would indicate the system is unbalanced in my experience. I can’t really see why the spark had the tails out for testing either, just doesn’t really make sense.

Edit- no answer to my previous question either, did the cover actually blow off? How did that happen with threaded screws on the cover, did they break? Do you have a picture?

 

[jimthegent]

How do you know your system is perfectly balanced? Why is that board fed in tails? Generally heavily loaded single phase boards would indicate the system is unbalanced in my experience. I can’t really see why the spark had the tails out for testing either, just doesn’t really make sense.

Edit- no answer to my previous question either, did the cover actually blow off? How did that happen with threaded screws on the cover, did they break? Do you have a picture?

There's the best I can get of the burnt tail after it has been taken our

 

[suffolkspark]

View attachment 112423

But thats all hanging off that single phase board, so i dont get why you keep talking about phase balancing? How many amps are each of those servers drawing? If the full or near to the 16A that is mighty loaded, With that board connected straight off the henleys i wonder what the upfront main fuses are. Are they greater than 100A?

 

[jimthegent]

But thats all hanging off that single phase board, so i dont get why you keep talking about phase balancing? How many amps are each of those servers drawing? If the full or near to the 16A that is mighty loaded, With that board connected straight off the henleys i wonder what the upfront main fuses are. Are they greater than 100A?

Hi 5 amps a unit and yes the fuses are 100amp per phase there's a small drawn through the main board to run the fans but it's tiny

 

[jimthegent]

Hi 5 amps a unit and yes the fuses are 100amp per phase there's a small drawn through the main board to run the fans but it's tiny

They servers all run on low power setting

 

[suffolkspark]

They servers all run on low power setting

How many amps each?

 

[jimthegent]

How many amps each?

Hi I can't promise this is dead right but each units is 1500 w

 

[suffolkspark]

Hi I can't promise this is dead right but each units is 1500 w

One on each socket?

 

[jimthegent]

One on each socket?

Yeah and we'd have 5 to 8 running at any one time

 

[jimthegent]

Yeah and we'd have 5 to 8 running at any one time

We are talking 12kw at the top end no more than that a 1kw for other bumfluffery fans main server lights etc

 

[jimthegent]

One on each socket?

I'm not ruining your Christmas here am I mate ? More than happy to leave this till mid week if your dinners going cold

 

[suffolkspark]

I'm not ruining your Christmas here am I mate ? More than happy to leave this till mid week if your dinners going cold

I just dip in and out of here when i get a sec, i really have no further ideas, thats not a massive load if all is well, but hefty if there was a poor main switch or loose wire

 

[jimthegent]

I just dip in and out of here when i get a sec, i really have no further ideas, thats not a massive load if all is well, but hefty if there was a poor main switch or loose wire

True there is a higher draw at start up as we need to get the building down to optimum temp and the fans on the machines work harder until we reach equilibrium but a loose wire is a disaster no matter what the load is isn't it

 

[DefyG]

True there is a higher draw at start up as we need to get the building down to optimum temp and the fans on the machines work harder until we reach equilibrium but a loose wire is a disaster no matter what the load is isn't it

I still maintain its nothing to do with a balanced or unbalanced 3phase load.
I do not think it is an overload (you dont get flash/bangs like that and molten copper from an overload)
IMHO Its a L - E or N 'Dead Short'

 

[suffolkspark]

I still maintain its nothing to do with a balanced or unbalanced 3phase load.
I do not think it is an overload (you dont get flash/bangs like that and molten copper from an overload)
IMHO Its a L - E or N 'Dead Short'

I think we all agree on that, but as it was running when the inspector left, something happened in between, the live side of the main switch has a significantly larger hole in the top compared to the neutral, so id say its sat there loose with 50A running through it getting mighty hot followed by a bit of arcy action first before it truly ---- itself

 

[jimthegent]

I still maintain its nothing to do with a balanced or unbalanced 3phase load.
I do not think it is an overload (you dont get flash/bangs like that and molten copper from an overload)
IMHO Its a L - E or N 'Dead Short'

Yeah and there's a lot of breakers on our system theres breakers in the pdu there are breakers on the main board surely it would just stop working with too much power drawn

 

[DefyG]

This was what is left of a RCD that had been in service for years without being touched (even during an EICR a few months prior).
It was from an installation in a community hall which would have been drawing less than 20A at the time.
The 'Dead Short' caused the cutout fuse (100A) to 'blow'.
We concluded it was an internal failure of the RCD and not caused by recent activity (as there wasn't any) but why? Who knows, one of those things!

 

[HappyHippyDad]

Looks like the Red clamp has squashed the cable and split the insulation.
That's allowed the copper core to touch the earthed metal backplate and or the screw holding the clamp.

If it was the cable lose in the terminal or wrong side of termianl clamp it would have fizzled and overheated for a while before ultimately failing.

As Westwood, it looks like a dead short, especially with the ball of copper

Good spot! It looks exactly that.

 

[freddo]

There have been several reports of Hager RCCBs detonating when the test button is pushed which then blows the cutout fuse, it also happened to me a few years back, really loud.