Block of 200+ flats, eleven single channel hot water timers blown this week alone.

[GBDamo]

Got called to "a flat in Manchester has no hot water" at 16:30 today and to be frank I was a little miffed to have to attend.

So arrives at site, spends 20mins finding parking, 30 mins finding the site contact, getting keys and then being informed that it was not one flat but eleven.

Have no idea what the supply is, big I'd guess, nor did I have access to ant distribution equipment, just a list of flat numbers.

So I go to the first one. The 16A MCB was off and was informed that "it goes bang if you try to turn it back on" which I took his word for.

So I worked back from the tank, element OK, thermostat OK, wiring OK however the timer was visibly burnt out.

Did a random sample of the other ten and of the four I checked out all had the same burnt out timer.

Due to the time nothing more was done and an assumption (I know?) made that all had blown.

The only other thing I did was to check the local voltages and all were 236-7V and steady, for the brief time I observed.

These flats were all over, not just one floor so unlikely to be from a single phase or distribution board.

The tennants all made the same complaints of flickering lights and one long recent power outage.

None of the staff had any knowledge.

----es me off that there is nobody on site with any ownership, or understanding of the buildings and their systems, there are two blocks approx 400 flats for christ sake.

I'm half tempted to just replace the controllers and note there may be underlying issues that need further investigations.

How would you guys approach this?

 

[GBDamo]

Here's some photos @Lucien Nunes and others.

The terminals and cables are clean as a whistle.

This is the culprit

Capacitor(?) has blown.

 

[plugsandsparks]

They are rated at 16A. The hot water cylinders are on a dedicated circuit off their own 16A MCB, isolated via a 20A double pole switch, length is a couple of meters.

Couldn't see a rating on the element but looking into that now.

Edit to add,

The elements are 3kW but looking at the plumbing I think they're on a central hot water system as there seems to be a coil on the cylinders.

I'm back there tonight two replace a couple and grab some more info.

Would definitely clamp it , you may get over 14 Amps if lucky...lol but is stuff that good nowadays. BTW i only use Timeguard products for this type of stuff as fed up changing crap out after a few years

 

[freddo]

Looks like a metal oxide varistor connected across the supply, If several blew at once this is a sure sign of overvoltage on the supply. A MOV is normally high resistance, but it's resistance drops as the voltage increases to stop voltage spikes. A sustained overvoltage such as from a supply fault will quickly overheat and destroy an MOV.

 

[Yozza19]

Looks like a metal oxide varistor connected across the supply, If several blew at once this is a sure sign of overvoltage on the supply. A MOV is normally high resistance, but it's resistance drops as the voltage increases to stop voltage spikes. A sustained overvoltage such as from a supply fault will quickly overheat and destroy an MOV.

Would definitely put my money on a floating neutral as some stage some Dno have a search function on there site to advise if there have been any incidents in the locality

 

[Lucien Nunes]

Yup, it sure does look like an MOV that wasn't big enough absorb the available energy in the overvoltage event. When that happens, it can initiate an arc that lasts until the fuse blows, which can dissipate so much more energy that it makes it hard to identify whether the trigger event was an impulse or a prolonged excursion above clamping voltage.

Assuming the building is served by conventional urban underground distribution the latter seems the most likely and I would say the likelihood of a cable fault causing both the timer damage and the flickering / outage events was very high. Other items in the properies with MOVs (and even SPDs) might also have been compromised, but if the MOVs in the timers had a slightly lower clamping voltage then they might have all gone phut before many of the remainder showed signs of distress.

The sub might be specific to the estate or building, in which case any information you can get from the DNO about recent faults might nail it.

 

[GBDamo]

Thanks guys.

Getting stuck into some tinnys now so will pick this back up tomorrow.

Cheers

 

[GBDamo]

My thinking of what we have so far.

The catastrophic failure or the MOV (thanks) in the controller would point to a large over voltage situation.

It is most probably not environmental but a lost Neutral.

Without further information this could be anywhere internal or external.

So I was thinking to approach it on two fronts, to contact the DNO and report a possible Neutral fault whilst surveying the site, reported faults and drawings to see if a pattern can pinpoint the location.

So in a normal fault free situation the installation should look like.

If the fault is in the below position then I should only be hearing of power outages in flats 7, 8 & 9 but not over voltages or the MOV failures I'm seeing.

If the fault is in the below position then flats 4, 5 & 6 and flats 7, 8 & 9 will see voltage fluctuations between 0V and 400V depending on their respective loads, this situation could cause the MOV failures.

And finally, in the below image the fault would be evident across all phases and all flats with under and over voltages. Whilst there are sections of the internal installation that could contain the fault; in this scenario the fault could also be external.

Ii actuality I have already informed the client of the liklihood of a serious fault existing and that it may be internal or external and they will be picking it up from here.

I'm quite disappointed really as its quite interesting to see how far you can fault find on paper if you have enough information.

 

[Lucien Nunes]

Your explanation of the possible scenarios is correct but the configuration in pic.3 is unlikely to exist in practice. Each single phase service will be derived from a 3-phase busbar, with neutrals normally serving one or three phases but seldom two.

If the fault was a cable strike, the neutral may have not merely been broken but also shorted or arced over to a line. A variety of different voltages may have been impressed on the loads in quick succession as the fault developed.

 

[LukeD]

I am loving this thread...some very clever replies

 

[marconi]

I don’t think the burned out disc is an MOV. The timer uses a transformer less way of producing low voltage dc to operate the electronics. I reckon one can see a bridge rectifier - the white block component, and a disc capacitor in series with a resistor of suitable power rating together forming a mains to low voltage dropper. The placement of the resistor and the disc capacitor makes me think their right hand ends are connected to Lin and Nin respectively and their left hand ends to the ac inputs of the bridge rectifier which are normally the inner two leads for this style of rectifier. There is a smoothing capacitor to the left of the bridge rectifier. I reckon the gold rimmed cylindrical component above the relay is a battery for keep alive in the event of power cut and accumulator to handle short duration current surges when the relay turns on.

 

[Strima]

I don’t think the burned out disc is an MOV. The timer uses a transformer less way of producing low voltage dc to operate the electronics. I reckon one can see a bridge rectifier - the white block component, and a disc capacitor in series with a resistor of suitable power rating together forming a mains to low voltage dropper. The placement of the resistor and the disc capacitor makes me think their right hand ends are connected to Lin and Nin respectively and their left hand ends to the ac inputs of the bridge rectifier which are normally the inner two leads for this style of rectifier. There is a smoothing capacitor to the left of the bridge rectifier. I reckon the gold rimmed cylindrical component above the relay is a battery for keep alive in the event of power cut and accumulator to handle short duration current surges when the relay turns on.

A picture of the underside of the PCB would be handy.

 

[GBDamo]

A picture of the underside of the PCB would be handy.

Was dreading that request, damn thing was riveted in.

 

[Lucien Nunes]

The white block will be the dropper capacitor, the bridge rec will be SMD. But yes, let's see some pics of the PCB

 

[marconi]

I was wrong ?

 

[GBDamo]

I was wrong ?

Damn, was hoping my initial guess was a stroke of genius.?

 

[marconi]

I’ll add this to my long list of other things I have got wrong. Yours humbly……

 

[Mike Johnson]

Don't think it's been asked, but did all eleven flats loose their timers at the same time, or was it cumulative over a period.

 

[marconi]

Don't think it's been asked, but did all eleven flats loose their timers at the same time, or was it cumulative over a period.

Following MJ’s theme - MOVs do degrade each time they react to over voltages. It is not unknown for them to eventually suffer thermal runaway which is when their quiescent current rises over time, the MOV warms up, its resistance reduces, it passes more current, it warms up even more…until it cannot dissipate the heat quickly enough to its surroundings and it destructs. All of which can happen in the absence of any overvoltage.

 

[GBDamo]

Don't think it's been asked, but did all eleven flats loose their timers at the same time, or was it cumulative over a period.

That is a good question but impossible to get a definitive answer as we don't know when they "went" only when the tenant discovered and reported them.

The first was reported on Tuesday, I was called on Thursday and by then there were eleven.

All we've got is the times they were reported.

But judging by the crazy disco lights I witnessed the issue is still ongoing and more will go.

 

[Lucien Nunes]

@marconi I'll let you off if you can find a SIL through-hole bridge rec in a white package made in the last 25 years.?

We can now confirm from the PCB pics that the failed part is across the mains input. In theory it could have been a small filter capacitor but:
a) The part shape and burst pattern are more consistent with an MOV
b) A cap of those dimensions would typically be a Y-class part of a few nF which is not appropriate there.
c) An MOV is not a bad idea where a capacitive dropper is tapped off the supply to a relay-switched 16A load of unspecified characteristics.

What it will hopefully have been doing, before it popped, was shaving off transients likely to damage the dielectric of the X2 dropper cap, that would otherwise cause a progressive loss of capacitance to the self-healing process.

But judging by the crazy disco lights I witnessed the issue is still ongoing and more will go.

Somebody needs to switch this lot off now.

 

[GBDamo]

Somebody needs to switch this lot off now.

Twice I have told them it needs attention, verbally and in writing via the job card.

I will chase it up on Monday as the lass who I was dealing with didn't convince me she understood the potential gravity of the situation.

 

[UNG]

Just curious if you did any more voltage checks when you went back although I think to catch any major voltage peaks or dips you would probably need to put some logging equipment in across all three phases


Had a similar problem to this at a business premises I used to work from many years ago although no equipment was damaged the voltage would very often dip to 170 - 185v on one phase around 8:15 - 9:00 in a morning and around 17:30 - 18:30 in an evening but not everyday, after a few weeks we got the DNO involved who put a voltage logger in for a week who told us there was no problem found, a week later 4 DNO vans appeared one afternoon and dug up the link box at the end of the street about 50m away, talking to the DNO lads they said that the box was close to failing due to thermal damage and was being replaced with a resin joint, The voltage problem disappeared after that

 

[Dave OCD]

Could it all have been caused by a one off surge as the power was restored following the reported mains failure across the entire block ?

 

[roper_ni]

They are rated at 16A. The hot water cylinders are on a dedicated circuit off their own 16A MCB, isolated via a 20A double pole switch, length is a couple of meters.

Couldn't see a rating on the element but looking into that now.

Edit to add,

The elements are 3kW but looking at the plumbing I think they're on a central hot water system as there seems to be a coil on the cylinders.

I'm back there tonight two replace a couple and grab some more info.

Time to get the Megger out and 1000v dc @ the heater could be cable or heater.

 

[roper_ni]

Time to get the Megger out and 1000v dc @ the heater could be cable or heater.

Sorry wrong post