MOV fire risk

[miegapele]

I would like to ask what is general consensus, about MOVS in products and fire risk from them.

I stumbled upon this post there MOV exploded even without triggering fuse, and fuses are quite frequently used with MOVS in surge protection strips, like for example here .
However, it seems MOVs now appear in regular products and even without a thermal fuse. example IKEA here (but not in older generation here ), Samsung 65W here , even cheap Chinese smart plugs have them. However, Apple teardowns I have seen, seems to still omit them.

So, that's the current trends regarding use of MOVs in such devices, is it OK from the fire perspective?

[DavidAlfa]

Of course Apple sees easy money there, they will just argue it was a mains surge...
Please buy a new expensive thing from us!

[jonpaul]

Bonjour,

See GE original MOV handbook.

Limited life and fire risk are known.

In 1983 we had a 20 mm GE MOV on a wall plug transient protectopr.

About 20m away a 12 KW cinema lamp had a fault and very high transients were on the 240B mains.

My engineer in the adjacent lab shouts:

"JON!! THERE IS FIRE!"

The MOV exploded and flamed dqwn towars the floor.

The carpet caught fire.

Since that day I havwe never used a MOV.

Jon

[Stray Electron]

   I found a heap of very nice and well-made COMMERCIAL quality surge suppressors in some surplus a while back.  They were mounted in steel boxes with a lot more room than necessary and the boxes were then filled with dry sand.  Presumably to absorb any blast from exploding MOVS. I just grabbed one out of the pile and I don't see a PN on any of the MOVs but they're in groups of 5 each and there are 7 groups of them and each MOV is larger in o.d. than an American quarter. There are five large wires (probably #8 gauge) and three small wires (for monitoring) coming out of each one and they appear to have been made for use on a 3 phase Y-connected circuit. 

  MOV tend to fail shorted so they all need to be fused IMO and packing them with sand and in a sturdy flame proof enclosure is a GOOD idea. These a bit bit weird in they way that they're fused, each incoming power lead goes through a 1 Ohm 1% resister (probably a fusible resistor) and then has a small 22? gauge bare wire that connects to a tent shaped insulator with 9 other small wires attached and each of the other wires connects to a group of MOVs. So it looks like the 1 Ohm resistors and the small bare wires are intended to serve as fuses.   I picked up about 60 of these of various sizes and we tested a bunch of them and they seems to be designed to operate on 120, 240, 277 and 480 volt circuits and both Delta and Y configurations.

[shapirus]

MOV tend to fail shorted so they all need to be fused IMO

Shorting the circuit is their actual purpose, is it not?

1) the MOV begins to conduct once voltage across its terminals exceeds a certain value;
2) current flowing through the fuse installed before the MOV increases rapidly and blows the fuse;
3) the MOV is rated for a specific amount of energy that it can absorb before blowing up (which determines its size -- or, rather, its size determines the energy);
4) if the fuse blows before that energy is exceeded, then the circuit designer used proper values and every part has done its job properly. Otherwise, it's bad design.

Am I wrong?

[joeqsmith]

https://www.littelfuse.com/media?resourcetype=datasheets&itemid=bd475732-1071-4352-b8aa-f78b0007eb05&filename=tmov-itmov
https://www.tdk-electronics.tdk.com/inf/70/db/var/SIOV_Housed_T14_20series.pdf

[shapirus]

https://www.littelfuse.com/media?resourcetype=datasheets&itemid=bd475732-1071-4352-b8aa-f78b0007eb05&filename=tmov-itmov
https://www.tdk-electronics.tdk.com/inf/70/db/var/SIOV_Housed_T14_20series.pdf

Basically an additional safeguard against the main fuse failing to do its job quickly enough. Good thing.

[madires]

Or alternatively MOV plus thermal fuse in heat-shrink tubing.

[coppice]

Or alternatively MOV plus thermal fuse in heat-shrink tubing.

Those heat shrink sleeves are remarkably good at holding the bits together when a MOV pops. Without that sharp bits can punch through a thin plastic case.

[shapirus]

Or alternatively MOV plus thermal fuse in heat-shrink tubing.

Considering the thermal resistance between the two, when they are separate parts, and the speed of the process, having them both in the same package (of course still with the heat shrink cover) gives a better chance of them working as expected.

[dietert1]

As far as i remember MOVs also suffer from aging but i can't say whether it is just time or number of surges. Aging tends to increase leakage currents and the thermal fuse is meant to disconnect the circuit when it starts to consume power, heating up to some level. Of course the thermal fuse is far to slow in case of a nearby lightning strike. MOVs also have a significant capacitance, which may help for EMI suppression.

Regards, Dieter

[nctnico]

Or alternatively MOV plus thermal fuse in heat-shrink tubing.

Those heat shrink sleeves are remarkably good at holding the bits together when a MOV pops. Without that sharp bits can punch through a thin plastic case.

Agreed. I have seen MOVs being covered by heat shrink tubing in many devices.

[coppercone2]

I seen mov utterly detonate and disappear but they never seemed to cause a fire risk. Maybe its 240v vs 120.

I also have alot of them around, i think they split up surges or something like that. maybe if there was just one little mov in the house it would be worse

[Siwastaja]

Shorting the circuit is their actual purpose, is it not?

1) the MOV begins to conduct once voltage across its terminals exceeds a certain value;
2) current flowing through the fuse installed before the MOV increases rapidly and blows the fuse;
3) the MOV is rated for a specific amount of energy that it can absorb before blowing up (which determines its size -- or, rather, its size determines the energy);
4) if the fuse blows before that energy is exceeded, then the circuit designer used proper values and every part has done its job properly. Otherwise, it's bad design.

Am I wrong?

You are correct, just missing one detail. Point 4, fuses "reset" nearly indefinitely as they cool down, hence the energy you refer to in point 3 is an integral which is also assumed to reset: as you explain, if the MOV and fuse are dimensioned together correctly, if the fuse does not blow, then the MOV does not blow up either, both cool down and reset for the next cycle. Fair enough, this is correct.

But MOVs have additional energy limitation, which is non-resetting over the lifetime. Every clamped peak, even benign ones which are not even close to exceeding said repetitive limit (and thus not even close to blowing the fuse), wears out the MOV, until it one day fails short from such small event.

This of course does not change the fact that a correctly dimensioned fuse is needed to protect the MOV from fire, and usually, the components surrounding the MOV also need some protection, at least a physical gap.

[madires]

The thermal fuses commonly used together with MOVs are fuses (stay open after blown) and not thermal switches.

[Kleinstein]

MOVs get increasing leakage when getting old. So they may slowly run hotter and hotter and thus can overheat without blowing a normal mains fuse. The MOV is usually not flamable (after all it is metal oxide), but the hot PCB and neighboring parts may catch fire.

There are MOVs that aready come with a thermal fuse, e.g. Little fuse TMOV series.

[SeanB]

I have had one or three Zinc oxide MOV units, NOS GE ones, in the red dipped package, die short circuit while in storage, but in general with 230VAC mains they will all ineviatably fail as short, so best is to make sure they are fused correctly. But as all of those I have used failed safely, small one blowing apart, bigger ones drawing enough current to trip a breaker rather rapidly, often with no apparent damage to the MOV, other than it having a low varying DC resistance, in the order of 10's of ohms, with 1V applied from a multimeter, as opposed to being well over 10M normally. Seen also units with fuses in the path to the MOV units, either as PC board traces, or a proper ceramic fuse soldered into the board, or for the one, with a 20A ATO automotive fuse soldered into a slot in the board. Last one made by Clearline, guess they were happy using a 20A 32VDC rated fuse well past the warranted ratings.

I have plenty in use, often simply mounted in a connection block to an external LED light, to keep the surges down, to keep the lamp life as long as possible, even if the lights are rated for 3kV surge, extra always helps. they still all fail after a while, some soon, some after a year or two, and a few that still work after 4 years.

[NiHaoMike]

But MOVs have additional energy limitation, which is non-resetting over the lifetime. Every clamped peak, even benign ones which are not even close to exceeding said repetitive limit (and thus not even close to blowing the fuse), wears out the MOV, until it one day fails short from such small event.

It helps to have a large film capacitor, on the order of 5-10uF, in parallel to shunt the small switching transients. That, of course, can only be applied to the L-N or L-L MOVs, using one to ground would cause too much leakage current.

[CosteC]

Nobody mentioned that while in EU nominal voltage is 230 VAC, the realistic voltages in in many places of network reach 250 V due to photovoltaic and other prosumers trying to sell energy. If MOV is rated for 250 VAC it will leak a lot and age quickly. I am not sure how this phenomenon widespread in USA with its 120 VAC system.

Other aspect is common issue - fuses rated "as high as possible" not "sensible". Lack of inrush current protection forces high fuse ratings, this does not work well when MOV will activate or there is failure of some primary side semiconductor. As usual cheap designs done quickly are not good designs and may caught fire.

[T3sl4co1l]

Nobody mentioned that while in EU nominal voltage is 230 VAC, the realistic voltages in in many places of network reach 250 V due to photovoltaic and other prosumers trying to sell energy. If MOV is rated for 250 VAC it will leak a lot and age quickly. I am not sure how this phenomenon widespread in USA with its 120 VAC system.

Very rarely, it can be much worse: since the supply is a 240VCT transformer, if the neutral comes loose for some reason (say, improper, faulty, or damaged wiring), more load on one side than the other will shift the neutral voltage, browning out one side and overvolting the other.  UL 1449 for example covers this.

Tim

[iJoseph2]

In the UK fitting MOV's to the mains incoming supply will become more and more common since the regulations have been updated. For example ..

https://hager.com/uk/support/regulations-18th-edition/surge-protection

18th Edi­tion Re­quire­ments

The 18th edi­tion BS 7671 now stip­u­lates

Pro­tec­tion against tran­si­ent over­voltages shall be provided where the con­se­quences caused by an over­voltage could

(i) res­ult in ser­i­ous in­jury to, or loss of, hu­man life
(ii) failure of a safety service, as defined in Part 2
(iii) significant financial or data loss

For all other cases, protection against transient overvoltages shall be provided unless the owner of the

installation declares it is not required due to any loss or damage being tolerable and they accept the risk of

damage to equipment and any consequential loss.

[CosteC]

Nobody mentioned that while in EU nominal voltage is 230 VAC, the realistic voltages in in many places of network reach 250 V due to photovoltaic and other prosumers trying to sell energy. If MOV is rated for 250 VAC it will leak a lot and age quickly. I am not sure how this phenomenon widespread in USA with its 120 VAC system.

Very rarely, it can be much worse: since the supply is a 240VCT transformer, if the neutral comes loose for some reason (say, improper, faulty, or damaged wiring), more load on one side than the other will shift the neutral voltage, browning out one side and overvolting the other.  UL 1449 for example covers this.

240 VAC is UK peculiarity Still, the lost neutral is surely fault condition, unfortunately it may last hours. Good fuse shall disconnect equipment, MOV usually does not survives it, and this is expected behaviour for commercial equipment.

Again, fuse selection is critical, thermally protected MOV will help, but not be enough in case of so long overvoltage - other components will not survive.

[AVGresponding]

Nobody mentioned that while in EU nominal voltage is 230 VAC, the realistic voltages in in many places of network reach 250 V due to photovoltaic and other prosumers trying to sell energy. If MOV is rated for 250 VAC it will leak a lot and age quickly. I am not sure how this phenomenon widespread in USA with its 120 VAC system.

Very rarely, it can be much worse: since the supply is a 240VCT transformer, if the neutral comes loose for some reason (say, improper, faulty, or damaged wiring), more load on one side than the other will shift the neutral voltage, browning out one side and overvolting the other.  UL 1449 for example covers this.

240 VAC is UK peculiarity Still, the lost neutral is surely fault condition, unfortunately it may last hours. Good fuse shall disconnect equipment, MOV usually does not survives it, and this is expected behaviour for commercial equipment.

Again, fuse selection is critical, thermally protected MOV will help, but not be enough in case of so long overvoltage - other components will not survive.

In this case, the 240V is not a UK peculiarity; the quoted poster is in the US and was referring to the 240V centre-tapped neutral found in domestic supplies there. AKA split phase, they use both phases where they need more power, ie a tumble dryer, for example.

[CosteC]

In this case, the 240V is not a UK peculiarity; the quoted poster is in the US and was referring to the 240V centre-tapped neutral found in domestic supplies there. AKA split phase, they use both phases where they need more power, ie a tumble dryer, for example.

True, yet USA split phase or two phase system is even more peculiar. Still lost neutral at source may generate even more exothermic effects...
You are right. I am forgetting about USA "solutions" mea culpa.

[coppercone2]

well it seems that I read about more MOV and RIFA explosions from europe. I don't think the US power grid is that bad of an idea. Split phase reserved for crazy devices that belong in the garage.

You pay a bit extra in copper weight but in general the failures are less spectacular.

and I noticed way more brown outs / surges in europe then in the USA. They have less local transformer in europe and there is more line effects that damage devices.


and its nice to be able to service most domestic things with a much less hazardous voltage.

[CosteC]

well it seems that I read about more MOV and RIFA explosions from europe. I don't think the US power grid is that bad of an idea. Split phase reserved for crazy devices that belong in the garage.

You pay a bit extra in copper weight but in general the failures are less spectacular.

and I noticed way more brown outs / surges in europe then in the USA. They have less local transformer in europe and there is more line effects that damage devices.


and its nice to be able to service most domestic things with a much less hazardous voltage.

230 V is more deadly when touched - true. RCDs however reduce this risk significantly.
120 V causes significantly more fires, this is hard to solve "electronically" and easier kills entire family. Power distribution network also is significantly more tricky and 11 kV is omni-present in suburban areas. European 15-20 kV is less exposed.
I however does not know any comprehensive statistics. Planet opts for 230 V rather than 110 V however. Possibly only due to reason of 3 phase 230V being more efficient.

Getting back to MOV issue, today I think USA standardisation is ahead in this matter with more restricted demand to protect against MOV overheating and fire. Still problem remains in good testing of product. Cheap imports are cheap imports with questionable design quality.

[coppercone2]

where is this risk of fires increased with 120V ?? who made these stats.

[coppice]

where is this risk of fires increased with 120V ?? who made these stats.

Arc fault fires seem to be a much bigger issue in the US than in Europe. I don't know is that relates to the tradeoff between voltage and current, or if US fittings are generally worse than European ones.

[Gyro]

That's probably the wirenuts and other crappy wiring practices, weak socket contacts etc. Not to mention all the general stuff pulling twice the current.

[SiliconWizard]

Yes, twice the current is like not negligible.

[coppice]

Yes, twice the current is like not negligible.

Its questionable if they have twice the current. The weak load that's possible from a US socket means they don't have most of the high consumption appliances, like 3kW kettles and 2.5kW hair dryers that are normal in 220-240V mains countries.

[coppercone2]

how much faster do you need that water boiled god damn

[AVGresponding]

how much faster do you need that water boiled god damn

Hey, when I want a cup of tea, I want it in the next few minutes, not the next day!   

[CosteC]

where is this risk of fires increased with 120V ?? who made these stats.

https://www.usfa.fema.gov/downloads/pdf/statistics/v19i8.pdf
https://www.ishn.com/articles/95801-electrical-malfunctions-a-top-cause-of-home-fires-in-us

I suppose higher currents, toxic relationship with wirenuts, widespread of wood in housing construction might have something to do with it.
Also USA sockets are famously horrible and famously widespread. More famous is only tendency of manufacturers to use thinner mains cables or make the out of aluminium.

I believe in 230 V systems mains impedance is of lesser importance, so it is easier to rupture a fuse if MOV of anything else go sideways.

[jfiresto]

... The weak load that's possible from a US socket means they don't have most of the high consumption appliances, like 3kW kettles and 2.5kW hair dryers that are normal in 220-240V mains countries.

At least in these parts, I would say that 3kW is not unusual. A few years back, our local Aldi ended up with a big pile of returned Strix-designed, 3kW electric kettles which were too much for the 10A circuits in many of the flats built the first couple decades after the war. I have since not seen more than ca. 2kW from Aldi Nord.

[5U4GB]

MOVs get increasing leakage when getting old. So they may slowly run hotter and hotter and thus can overheat without blowing a normal mains fuse. The MOV is usually not flamable (after all it is metal oxide), but the hot PCB and neighboring parts may catch fire.

I've always wondered about that, ten-cent MOVs get added to cheap power strips so the sellers can double the price by touting the "surge protection" it offers (the random-trigger incendiary functionality gets thrown in for free), but a while back I scored some NOS Islatrol IE-series DIN-mount hardwired power filters for a fraction of their pretty hefty new price, and a major design element was a "MOV array" which I wouldn't have expected in something that's rated for continuous operation over (at a minimum) its ten-year warranty period.  However I'm not that familiar with the design requirements for these things so maybe they're using them in some way that doesn't cause, or isn't affected by, their degradation over time.

[Siwastaja]

Very rarely, it can be much worse: since the supply is a 240VCT transformer, if the neutral comes loose for some reason (say, improper, faulty, or damaged wiring), more load on one side than the other will shift the neutral voltage, browning out one side and overvolting the other.  UL 1449 for example covers this.

Neutral or PEN faults are, if not common, very not-non-existing in the countryside here, and a regular-ish issue on TN-C-S distribution systems. Most of the electronics, lightbulbs etc. blowing up is least of the problems; dying is worse, much more dangerous is that PE and N are connected together at the distribution box, so with phases present without the PEN conductor, now every "earthed" metal surface is at a possibly high potential, through the low impedance of the connected loads. Instructions how to handle this situation go along the lines, "when you see some of your lightbulbs glowing dim and others very brightly, you are experiencing a neutral fault. Take a long non-conductive object, for example a log of wood, to carefully turn off the main switch in your metal-case distribution box. Don't touch any device or any metal object, or you die. Alternatively, go sit in the middle of the room and sob."

[Siwastaja]

Yes, twice the current is like not negligible.

Its questionable if they have twice the current. The weak load that's possible from a US socket means they don't have most of the high consumption appliances, like 3kW kettles and 2.5kW hair dryers that are normal in 220-240V mains countries.

This. Current is not twice, because all large loads in the US use 240V sockets. The opposite point can be made; German Schuko is officially rated to 16A, but fires are reported with sub-par sockets and plugs over 8A. Here, this is all we have, and 3kW loads must use this. In the US, 3kW loads use the same voltage, but plugs and sockets rated to actually survive 16A, and much more.

We mitigate by avoiding loads >10A which can be plugged in by end user completely; we just call an electrician every time we want to install a trivial thing like electric stove, to make a permanent connection - or do it illegally ourselves.

[Siwastaja]

True, yet USA split phase or two phase system is even more peculiar.

Split phase is NOT two-phase, and it's not that peculiar. There is a reason why the term split phase has been coined as a separate thing from two-phase. Two-phase systems have usually 90 deg phase difference so they produce a rotating vector, while a 180deg phase shift produces nothing useful. Split-phase is just a dual voltage single-phase system. AFAIK, two-phase systems aren't used for distribution anywhere, at least not in large scale, because of the awkward dimensioning of neutral wire for double the phase current; 3-phase distribution avoids that. But we still see two-phase design in stepper motors, some small BLDC fans, and in capacitor-run "1-phase" induction motors.

[Marco]

Two phase is an ancient and irrelevant historical artefact, forcing a differentiation between two phase and dual/split-phase has long become knee jerk pedantry rather than needed for clarity.

[dietert1]

Here in Sao Paulo we have a two phase mains distribution, with two times 120 V at 180° phase. It's a huge system like that. I can use all 230 V appliances i brought from Germany and i can buy and use US products, too. In our house we have two sets of sockets: white for 120 V and red for 240 V. Many modern appliances are "bivolt".

Regards, Dieter

[tooki]

Of course Apple sees easy money there, they will just argue it was a mains surge...
Please buy a new expensive thing from us!

Apple’s chargers are famously spare-no-expense designs made entirely with premium components. If they aren’t using MOVs, there’s a good reason. (And historically, Apple chargers have proven to be very durable.)

There are plenty of reasonable things to criticize Apple about, but power supply quality most certainly is not among them.

[BrokenYugo]

Yes, twice the current is like not negligible.

Its questionable if they have twice the current. The weak load that's possible from a US socket means they don't have most of the high consumption appliances, like 3kW kettles and 2.5kW hair dryers that are normal in 220-240V mains countries.

I would assume the very common 1500 watt 12.5 amp space heaters are the real killer. They're typically the only continuous heavy 120V load in the house. Combine with a loose socket or crappy 16 gauge extension cord (or worse yet 2 on one with a 20A breaker just hanging on), or some illegal flying splice in the wall or attic done with less than reputable wire nuts and there's the fire.

Some power tools and appliances also pull 12+ amps, but aside from the 12 amp leaf blower I can't think of any I use that do so for hours on end like a space heater will.