MAJOR electrical issues in 5 year old house - any ideas?

[jack]

6 minutes ago, ProDave said:

The blame game cannot start until the fault is identified.  If it is a DNO fault, I am absolutely sure the DNO will pay for items damaged in the home. Your insurer may be the vehicle to facilitate that as I am sure they will just pass on the claim. The danger though in going through your insurer is they will deduct the policy excess from the claim, wheras going straight to the DNO (IF it turns out to be their fault) you would likely get the full claim.

 

Exactly.

 

In fact, my understanding is that they only contacted their insurer to let them know what's going on, and not to lodge a claim at this stage. They just heard that the insurer was sending someone anyway. I don't know more than that.

 

More generally, I'm encouraging them to focus on identifying the problem before doing anything else. Until that happens, it's pointless (albeit human nature) trying to plan the next move.

[ProDave]

3 minutes ago, jack said:

 

Exactly.

 

In fact, my understanding is that they only contacted their insurer to let them know what's going on, and not to lodge a claim at this stage. They just heard that the insurer was sending someone anyway. I don't know more than that.

Careful, a claim is probably in progress.

 

I had a VERY minor car accident. Driving on a single track road, oncoming car, passing place on his side of the road, so I stopped just short of the passing place to allow him to pull in, but he didn't pull in he carried on trying to force 2 cars to pass in a single track road (remember I was starionary) and clipped wing mirrors. Mine just cracked the plastic casing, his was left hanging by the wires.  He was adamant it was my fault (how?) and he was going to claim on my insurance, and proceeded to take lots of photographs.  So I informed my insurance company what had happened and made it clear I was not making a claim (I glued the cracked plastic back together, it was an old car)

 

nearly 2 years later I had a phone call to say they never received a claim from the other driver and were closing the claim down. They advised I phoned customer services as I might be due a refund. Sure enough, the 2 renewals in the intervening period had been inflated, and now the claim was closed I got £200 refunded.

[jack]

Thanks Dave, I'll mention it to my parents in law.

[Jeremy Harris]

A couple of times I've chosen to notify an insurer of something, when I was pretty sure there wouldn't be a claim, and both times I did so in writing and made it clear (with a bold heading) that I was not making or initiating a claim, but was merely informing them so that they were aware of the incidents.  I took the view that it was better to be upfront with the insurer as quickly as possible, so as not to jeopardise a possible claim if things didn't go as expected, just in case they used any delay in informing them as a reason not to honour any possible claim that might arise.

 

On both occasions I made sure I closed things off with the insurer afterwards by writing to them saying that the matter had been resolved privately, with no claim, and requesting that they confirm that my claim history hadn't been adversely affected.

[DavidFrancis]

I used to be an insurance broker many years ago and I wouldn't be suprised if the insurers are relucant to cover any losses. I've just looked at an Aviva policy, purely as an example, and without "accidental damage" cover I'd say their policy wouldn't cover these kinds of electrical damage losses and even their accidental damage extension has an exclusion of "electrical or mechanical breakdown". If the original cause of the problems was a fire, an explosion, a falling tree etc then the cover should operate, but for a faulty connection somewhere, there maybe a problem claiming.

 

Not all policies are the same, but in this area I think they are mainly pretty similar.

 

But I did leave the business 30 years ago, so I could be talking out of my posterior.

[jack]

Thanks for that David. I've told them to make doubly clear that they aren't claiming, and I think they're on it (father in law runs a business that includes delivery vans in London, so he's had a lot of experience with insurers over the years!) I don't think they're too concerned if it isn't covered - the main thing at the moment is identifying the source of the problem so they aren't potentially throwing good money after bad.

[AliMcLeod]

Similar to others, i'd be wary of future insurance renewals. We moved into a house 10 years or so ago, and a few weeks later we heard a *very* loud gushing water noise when in our en-suite bathroom, but it was very hard to pinpoint where it was coming from. Our home insurance policy document was in a cupboard somewhere so I phoned our insurance company to query what was covered and they confirmed we were covered for water damage but not for any costs related to finding out where the leak was coming from (Trace and Access cover).

 

In the end, we were lucky and a single floor tile lifted found the culprit - a very bad join in a mains pressure MDPE pipe had separated. We got it fixed it and the tile replaced with no permanent damage

 

Our the next home renewal quote was expensive - when i phoned up to ask why, they referred to the call i had made and said that "incident" was contributing to the increased quote. I went elsewhere.

 

 

Edited October 9, 2018 by AliMcLeod

[jack]

UPDATE

 

I've just returned from the house. My electrician is there with an electrician from the company that installed the wiring and surge protector.

 

It's presently a complete mystery. All measurements are coming back clean and well within range. Loop impedances are all rock steady at around 0.19-0.20 ohms. All connections appear tight.

 

Currently there's some speculation about whether the ASHP could have been failing and somehow dropping a large spike back when turning the compressor motor off. Possible but seems unlikely.

 

Another (probably even less likely) possibility is a joint in the front garden. 5 years ago, the landscaper broke the incoming main, and a joint was made and buried. It seems possible that there's an intermittent fault between neutral and earth at that point, although consensus seems to be that there should be some evidence of that even if it isn't in full fault mode right now.

 

All of this is genuinely just speculation right now, and there's a lot of head scratching going on.

The ASHP guy was running an hour late, and I had to leave just as he arrived. I've asked the electrician to let me know what else they find in due course.

 

Oh, and there's definitely no RCD on the incoming three phase. One of those will be installed as part of all this, although naturally that won't address whatever the actual problem is.

[Onoff]

That joint in the garden...is it intermittently "under water" as in periods of heavy rain?

[Jeremy Harris]

Sounds like a hell of a puzzle. 

 

If the ASHP was the cause, then I doubt it can have been a start up current surge or shut down voltage spike, as, from the description of the damage, it sounds like there was an over-voltage event that lasted for some time, perhaps a few tens of minutes or more, from the description of 3 hours of flickering etc.  If the neutral is OK (and the loop impedance tests imply that it's fine now) then even if there was a phase imbalance problem, perhaps originating from the ASHP, then the phase to neutral voltages would all have stayed within limits; all that would have happened is that the neutral current would have increased, which isn't going to cause a problem if the incoming PEN is OK.

 

The joint in the garden sounds suspicious, especially given the cause.  Makes me wonder if there may be other damage to the cable somewhere, but it seems a bit improbable for this to have somehow fixed itself.  I'd agree with your chap's view, that there should be some evidence of a fault in the incoming cable, most likely a high Ze, and yet the measured impedances seem spot on to me.

 

Fitting an RCD in the 3 phase supply will be a very good move, but it must be worrying to not know the cause.  Let's hope that they find something later that suggests a probable cause, if they don't, then I'd be suspicious of the incoming cable that was damaged and repaired.  I'm assuming the cable is the DNO's incoming supply, is that right?

[Ed Davies]

42 minutes ago, jack said:

It seems possible that there's an intermittent fault between neutral and earth at that point,

Wouldn't neutral and earth be the same thing there, combined as PEN in the outer conductor? But, yes, the possibility of an intermittent break there does need to be considered.

[Ed Davies]

29 minutes ago, Onoff said:

That joint in the garden...is it intermittently "under water" as in periods of heavy rain?

Or the opposite, intermittently dry? Dunno about SE England but September had the ground unusually dry here in the NE of the island.

[jack]

 

8 minutes ago, Ed Davies said:

Or the opposite, intermittently dry? Dunno about SE England but September had the ground unusually dry here in the NE of the island.

 

One of electricians floated that as a possibility - long dry weather followed by rain causing some sort of intermittent tension on the joint. Conclusion was that it's possible but unlikely to now have become completely undetectable.

 

11 minutes ago, Ed Davies said:

Wouldn't neutral and earth be the same thing there, combined as PEN in the outer conductor? But, yes, the possibility of an intermittent break there does need to be considered.

 

I don't really know much about power electrics, so I can't really follow the details of what the electricians were discussing, but could it be that a break in the PEN results in the additional bonded earths at the house becoming the return path for the respective live phases? Not sure...

 

40 minutes ago, Onoff said:

That joint in the garden...is it intermittently "under water" as in periods of heavy rain?

 

No-one's quite sure exactly where it is! But we're on fairly sandy soil here, so water tends not to hang around in the soil for long. I doubt it would remain immersed for long periods, but not sure.

 

36 minutes ago, JSHarris said:

I'm assuming the cable is the DNO's incoming supply, is that right?

 

Yes, that's right.

 

It is a puzzlement, no doubt. All possibilities so far identified seem to be very unlikely, but something has to have caused all of these issues in two separate events separated by weeks. 

 

I'll let you know if we learn anything more once the ASHP guy's done his bit.

 

 

[Jeremy Harris]

4 minutes ago, Ed Davies said:

Wouldn't neutral and earth be the same thing there, combined as PEN in the outer conductor? But, yes, the possibility of an intermittent break there does need to be considered.

 

 

Normally, yes, given the cable is relatively new it would almost certainly be wavecon three core concentric, with the outer copper being the combined PE and N. 

[Jeremy Harris]

5 minutes ago, jack said:

I don't really know much about power electrics, so I can't really follow the details of what the electricians were discussing, but could it be that a break in the PEN results in the additional bonded earths at the house becoming the return path for the respective live phases? Not sure...

 

 

A common arrangement for an underground pot joint would be to add an additional intermediate earth at the time that the joint is made.  This is usually a length of copper strap that's connected to the outer PEN conductor inside the pot joint and then trailed along the trench to provide an additional intermediate earth. 

 

I doubt that the earth conductivity would be good enough, if the PEN wasn't connected properly through the joint,  to explain the low loop impedances measured, though.  I'd not rule it out, but without digging up the incoming cable to physically inspect it and maybe flex it to see if there is a problem, either at the joint or somewhere else where it may have been damaged, it may well be really hard to be certain that this isn't the cause.

 

I wonder if anyone else fed from the same local transformer has experienced any supply problem symptoms?  Might be worth asking around some of the neighbours, as they may not have experienced anything as serious, but may have noticed something like the lights flickering, and that may possibly give some clues for the DNO to follow up.

[jack]

5 minutes ago, JSHarris said:

I wonder if anyone else fed from the same local transformer has experienced any supply problem symptoms?  Might be worth asking around some of the neighbours, as they may not have experienced anything as serious, but may have noticed something like the lights flickering, and that may possibly give some clues for the DNO to follow up.

 

I believe they've asked around, but not sure how far afield they've gone. The next door neighbour also has three phase off the same cable on the street and they've confirmed no problems during the periods when the damage happened.

[ProDave]

An issue with concentric supply cables, is the outer sheath can get punctured, allowing water in and corroding the outer PEN conductor. This will go completely undetected until the PEN fails.  I have seen this on a job where the DNO cut into the cable to make a joint, and water came out.

 

If the landscaper damaged a cable that has to be repaired, what is to say he didn't bruise it in another place that is now allowing water in unseen?

[Mikey_1980]

We had a major issue with the electrics just after we occupied our house. The electrics would trip at random times of the day for no reason. All tests in the house came back as normal and all external tests by SSE came back as normal, this went in for 3-4 weeks, they re cabled our pole, checked out the electrics in all the local barns and farms to try and resolve. . It turned out it was the neighbours sewerage pink was going to earth or something whenever it switched on and trippping our electrics when it happened which was why it was happening at random times. Just thought I would mention as it could be something completely random and un thought of causing this. 

[Jeremy Harris]

25 minutes ago, ProDave said:

If the landscaper damaged a cable that has to be repaired, what is to say he didn't bruise it in another place that is now allowing water in unseen?

 

That was one idea I had when I posted the same thought earlier.  The sticking point with it as the possible cause is the low Ze that's just been measured.  It's hard, but perhaps not impossible, to see how a damaged PEN in the cable could still give a Ze as low as 0.2 ohms.  I guess a cable fault could have somehow fixed itself, or maybe there's a fault condition that changes under load.  Most testers don't measure Ze at a high current; IIRC mine does loop impedance testing at either a steady 4.5 A (in non-RCD mode) or pulses the test current faster than the RCD response time when measuring in RCD mode (so as not to cause a trip). 

 

I've seen dodgy joints that have acted in some odd ways before now, including one in a low current circuit that acted as a crude diode, so it's certainly possible for cable faults to change their behaviour depending on the load current.  Taking this hypothesis one step further, then if there was a phase imbalance at the consumer end, such that a high neutral current was flowing, then maybe a possible PEN fault could only appear when there is a fairly high current.  Damned hard to find out whether this may or may not be the cause without digging the cable up, though.

 

Also, I'm pretty sure SSE now specify wavecon for all buried LV supplies, and that has a copper PEN with aluminium phase conductors.  Copper shouldn't corrode if water gets through the outer sheath, and I have a feeling that's why wavecon uses copper for the PEN. 

[jack]

My electrician's just dropped by for a cup of tea and a chat after finishing up for the day.

 

It seems that the ASHP compressor motor has packed up, with at least one dead short across one of the coils (not sure if more than one). The noise filter has packed it in, and another control board is also suspect. One possible scenario is that the short was initially intermittent, resulting in high speed switching within the coil. The thinking is that this might have caused spikes that have caused the damage.

 

Again, this doesn't seem particularly likely, but it's no less plausible than any other proposal currently on the table.

 

Oh, and it also doesn't explain the >30V fault that was seen on the surge protector. It was, however, established that this protector wasn't installed correctly, as it's much too far from what it's supposed to be protecting (not that this explains the 30V fault light).

 

The next step will be a full house inspection to see whether there are any other issues.

 

Between them, my electrician, the other electrician (works for the company that wired the house), and the ASHP guy have decided that a separate surge protector at the ASHP would be a useful thing to have. I think we'll also install an RCD on the incoming three phase main.

 

Thanks again for all your help with this. It's clearly an unusual situation, as everyone is scratching their heads about how something that measures so well now can have behaved so badly only a couple of weeks ago.

[Jeremy Harris]

If the >30V N to PE was recorded by a surge protector some distance from the supply, then it could well have been a false reading, especially if the surge it measured was just a short spike.  Cable inductance comes into play when there's a catastrophic failure, as seems to be the case with the ASHP motor, as it's unlikely that an over-voltage event on the supply caused an insulation failure in that motor, much more likely to be the other way around, I think.  3 phase motor insulation is rated at over the maximum phase-to-phase voltage, with a fairly high factor of safety (a normal domestic insulation resistance test would be conducted at 500V), so the motor failure may well have been the cause, rather than an effect, of the problem. 

 

Fast, high current, spikes can be generated under insulation breakdown in a high current motor, perhaps as a consequence of over-heating for some reason, that won't trip a breaker, as they typically take around 20ms to operate (the max allowable is usually 30ms).  These short duration spikes may well have resulted in high voltage, short duration, voltage surges, that over-stressed some of the equipment in the house and caused it to then fail.

 

This neatly fits the three hour event timeline, with the motor intermittently shorting internally and then clearing, and with every short clearance event the motor inductance would generate a high voltage spike.  The interference suppressor almost certainly failed as a consequence of the high dV/dT, which again fits with the pattern of events. 

 

An RCD on the 3 phase supply would have almost certainly picked up this failure in the early stages, I think, as it sounds as if there were some high current imbalances in the 3 phase supply, before the catastrophic failure at midnight.

[jack]

26 minutes ago, JSHarris said:

If the >30V N to PE was recorded by a surge protector some distance from the supply, then it could well have been a false reading, especially if the surge it measured was just a short spike. 

 

The fly in this ointment is that surge protector's tech department insists that the warning light doesn't latch. That means that if you see the light, there is an ongoing voltage difference of at least 30V between N and E. I don't know how long that was present for, but it was at least hours. In fact, I think my parents in law said it was still flashing the next day. I can check.

 

It's also possible the tech person has this wrong - it would certainly make a lot more sense if it latched (and make it more plausible that the ASHP motor failing caused the problem).

[Jeremy Harris]

Some do latch for a time, I have a couple feeding my PC etc and they stay on for a time after a surge.  Not sure how long, as I've only ever seen them come on a couple of times, but it's long enough for you to notice that the light's on, so of the order of tens of minutes.  Given that there would probably have been a lot of spikes, as the motor intermittently shorted and then cleared, over a period of about three hours, then it seems possible that it just kept on being regularly triggered.

[ProDave]

>30V N-E is virtually impossible on TNC-S as N and E are one and the same, in many supply heads the N and E terminals are in fact the same big block of brass, in others they are separate blocks of brass with a link between them.

 

Up front 3 phase RCD just sounds like a recipe for nuisance whole house tripping. An individual 3 phase rcd for the heat pump sounds a good idea and rcbo's for things like single phase sockets circuits etc.

 

It is beginning to sound like this is an old direct drive heat pump, rather than inverter driven? The inverter would likely have tripped in the latter.

[jack]

12 minutes ago, JSHarris said:

Some do latch for a time, I have a couple feeding my PC etc and they stay on for a time after a surge.  Not sure how long, as I've only ever seen them come on a couple of times, but it's long enough for you to notice that the light's on, so of the order of tens of minutes.  Given that there would probably have been a lot of spikes, as the motor intermittently shorted and then cleared, over a period of about three hours, then it seems possible that it just kept on being regularly triggered.

 

Latching would seem to make sense, but apparently the tech support guy was adamant that the light only operates in real time.

 

15 minutes ago, ProDave said:

>30V N-E is virtually impossible on TNC-S as N and E are one and the same, in many supply heads the N and E terminals are in fact the same big block of brass, in others they are separate blocks of brass with a link between them.

 

Yes, all the electricians who've seen the installation agree, especially given the low loop impedances that have been measured.

 

18 minutes ago, ProDave said:

It is beginning to sound like this is an old direct drive heat pump, rather than inverter driven? The inverter would likely have tripped in the latter.

 

According to my father in law this morning, it's only 4 years since they moved in (I was convinced it was 5). It's a new build and the ASHP was new when installed. Would Mitsubishi still have been selling non-inverter-driven ASHPs at that time? I'll see if I can get the model number and check.