panel fire

[My Old Tools]

I had an interesting and eventful Sunday afternoon at my old shop. I was swapping out some cabinets in the garage apartment. It was hot and humid so I flipped on the heat pump AC. After 10-15 minutes of cooling it flipped the main breaker on the 200A SquareD homeline panel. I went to the panel and reset the main. I had a whole house surge suppressor (15 years old or so, maybe Intermec but I need to check, small metal box with 4 leads) connected to the first breaker (220v). Smoke started boiling out of the metal surge box and it burst into flames. I flipped the main back off, put out the flames, and disconnected the surge leads. After things cooled for a few minutes I reset the main. Everything worked just fine, AC cooled the rest of the day, no after effects other than the burnt up surge box. There have been no changes in that panel in a couple of years. I was not jacking with wiring at all.

 

[sberry]

Another "safety" gizmo strikes again.

 

[Norcal]

What size breaker was used for the surge "protector"?

 

[sberry]

Good catch.

 

[rlitman]

What size breaker was used for the surge "protector"?

Interesting question, but I'm not so sure that it would make a difference.
And since you can even buy an SPD that is rated to be connected to the meter un-fused, it says to me that the problem is in the SPD, and not in the breaker.

The natural way that the MOV components in an SPD fail is into a shorted state. They are supposed to have fuses that take the MOVs out of the circuit before "bad" things happen, and when the fuses open, the status light on the SPD should go out (because you're no longer protected), alerting you to the failure.

Still, the shorting of an MOV can be pretty violent when the panel has sufficient energy available (in a residential setting, that's especially true in a 200A panel). Since the SPD is connected to the panel bus via a low impedance path (by design), the inrush current during the short event can be enormous. In theory, the fuse in the SPD should blow first, preventing a breaker trip, but clearly that did not work, since the main tripped.

Square D's Homeline trip curves show that the minimum current required to trip your 200A main breaker on the magnetic ("instantaneous", though technically this can take as long as 1 cycle to trip) mode would be 1800A, though it could easily be upwards of 4000A. So, at a minimum, 1800A (and probably a lot more) flowed through that SPD in the instant before the main tripped. As I said, that's a lot of energy available in your 200A panel. Enough to blow up glass fuses and flash over inside the SPD, which can create conductive carbon trails inside bypass the broken fuse that lead to the fire.

A quality SPD should be designed to contain this "violence". It should have ceramic-bodied sand-filled fuses (the same advice goes for a multimeter used on such a panel) that should interrupt the short circuit current safely. Clearly the SPD in question was not a good match for a 200A panel. The SPDs connected to the 2000A (480V) panels at my office are potted in epoxy in their molded cases, and then enclosed in steel. Of course that level of protection is not necessary for the OP, but I think he should seek out a better SPD as a replacement. And follow the instructions carefully. Especially with regards to the circuit breaker's interrupt rating.

Also, I would consider discarding the breaker supplying the SPD. The fact that the main tripped before it is disconcerting. I'd have been more comfortable if both tripped at the same time, but that does not appear to sound like how things happened.

 

[sberry]

Most home class ones I have seen are designed for 30A connection.

 

[My Old Tools]

Yep, 30A breaker. And no, best I can remember it did not trip. I'll swap that one out. And as I said, that surge protector was at least 15 years old. I was just really surprised to see flames coming out of it's steel can. After a bit of Googling, it's not uncommon.

 

[rlitman]

Yeah, they can fail by letting the smoke out. They've gotten better nowadays though.
Just keep an eye on the SCCR and make sure it's at least as high as the rating of your panel's main.

 

[William Payne]

I would like a surge protector for my garage to protect my welders. I have been looking into it and to be honest the ones I see that wire in look gimicky. I have seen din rail ones that go inside your panel which I may look at. I need to do more research.

 

[Fixin'Stuff]

Eaton has made a couple of panel surge protectors that look very similar. One, the CHSPUltra requires that you hook it to its own 15 amp breaker. But the CHSPT2Ultra says that it should be hooked to a 50 amp breaker for maximum surge protection. I can see where having the model designed for 15 amps would lets its magic smoke out before tripping a 50 amp breaker, but it sure as heck should have tripped that smaller breaker first, unless your panel is overloaded and running so close to the max that the 200 amp breaker will allow that it tripped that one first.

You don't happen to have one of those Federal Pacific breaker panels, do you?? They're infamous for breakers that won't trip.

 

[csp]

You don't happen to have one of those Federal Pacific breaker panels, do you??

OP stated it's a Square D Homeline panel. My first thought reading the title of the thread was wondering if it was Federal Pacific.

 

[rlitman]

... I can see where having the model designed for 15 amps would lets its magic smoke out before tripping a 50 amp breaker, but it sure as heck should have tripped that smaller breaker first, unless your panel is overloaded and running so close to the max that the 200 amp breaker will allow that it tripped that one first.

You don't happen to have one of those Federal Pacific breaker panels, do you?? They're infamous for breakers that won't trip.

Circuit breaker "coordination" is a tricky topic. Remember that the current that trips a breaker in the magnetic region may be 20x the rated current. And the same current is seen by the breaker closest to the short, as the breakers upstream of the short.

I'll give an example from my real-life experience. At work, we saw a piece of equipment (a static switch) connected to a PDU cause a "bolted short" that tripped the 20A breaker in the PDU, the 30A breaker feeding the PDU, AND the 225A breaker feeding that. The 1000A breaker in that panel did not trip.

The low impedance of the circuit allowed for the short circuit instantaneous current to exceed 900A (though I am not sure by how much), and all three circuit breakers in a line saw that same same current. All three cleared the fault within a cycle or so. The 1000A upstream breaker is microprocessor controlled and would not have tripped for at least 5 cycles, so that main was never in danger of tripping, so long as the downstream breakers were capable of clearing the fault (this was an EXCELLENT design).

It does not take an overloaded panel for a short to trip the main. Remember that this SPD's connection to the panel bus is of very low impedance. So the short circuit current will be enormous. Also, remember that the trip curve for the OP's Homeline panel only kicks in above 9x the breaker rating. Any base-line load on the panel is insignificant compared to the short circuit current.

That being said, I can think of two reasons why the SPD's 30A breaker did not trip initially.
1) It was defective (to the tune of FPE ****).
2) It's trip curve was such that even when seeing currents that should trip it in the magnetic mode, it was too slow to react. With a short circuit current sufficient to trip the main, these two breakers were like gunslingers in a duel. Whoever shot first interrupted the current, and the slower breaker then no longer had the ability to trip. Even in ideal conditions, this sort of breaker can take from 1 to 1.5 cycles to trip. So it is quite possible that it was going to trip, but the tripped main prevented it by dumb luck.
Either way, I'd feel more comfortable taking that breaker out of service (the main will be fine).

 

[sberry]

I would like a surge protector for my garage to protect my welders. I have been looking into it and to be honest the ones I see that wire in look gimicky. I have seen din rail ones that go inside your panel which I may look at. I need to do more research.

You do not need it for welders.

 

[sberry]

I was dicking and testing a fixture, put a plug in without glasses on and had a strand of wire in a 16 cord, bigger than a hair but not by much poke out under an old plug. Can still see the arc strike on a cover plate, knocked over the 15 or 20 it was on as well as 60 up stream but didn't trip the 200 main.

 

[William Payne]

You do not need it for welders.

I have met people who have lost boards in welders due to lightning strikes. Don't know about other surges. My welders are mostly transformer but would rather have it and never need it then need it and never have it.

 

[rlitman]

You do not need it for welders.

Unless he's talking about an inverter welder.

 

[William Payne]

Unless he's talking about an inverter welder.

Only one of mine is inverter but yeah. I figure minimise the risk.

 

[wyliesdiesels]

OP stated it's a Square D Homeline panel. My first thought reading the title of the thread was wondering if it was Federal Pacific.

Same here.

Or zinsco.

 

[Norcal]

Same here.

Or zinsco.

Except for the fact that they have not been made in decades, Zinsco bolt-on breakers are not as bad as the plug-in variety and they do trip when a fault or overload occurs. As much as I dislike Zinsco had to say that, what is funny is that Zinsco/Sylvania used FPE breakers in their gear requiring anything over 225 amperes or over 240 volts.

 

[bjcouche]

I'll add that these surge protectors usually fail after they attempt to suppress a power surge. As an EE I understand breaker coordination as previously mentioned and the fact that it doesn't always work as intended, especially with magnetic trip breakers. I too have a whole house surge protector, CH I believe, and it's on a 15A breaker. I don't know how likely, but it's theoretically possible that by tripping the main 200A breaker, it _saved_ other equipment connected to your panel. If the surge came from upstream of the main breaker, and the 30A surge protector breaker had tripped instead, then the surge would have been allowed to continue to the rest of the panel after the surge protector sacrificed it's life.
Too often, these surge protectors burst into flames at the end of their life. You'd think they would be required to be made out of flameproof materials. I've seen the results of several power strips (some of expensive name brands), that caught on fire. Makes me wonder, how many houses burn to the ground because people (myself included) were trying to protect their electronics from power surges.

 

[William Payne]

Is anyone running the din rail ones that become part of the electrical panel? Like Schneider electric or other brands.

 

[sberry]

Only one of mine is inverter but yeah. I figure minimise the risk.

I have never seen it or heard a reliable acount of this happening.

 

[sberry]

Number 1 thing is to make sure grounding is in place, addi g more atuff without this may give a warm feeling but doesn't do much.

 

[Norcal]

Is anyone running the din rail ones that become part of the electrical panel? Like Schneider electric or other brands.

The DIN rail breakers are listed as supplemental protection, so the only place they are used is in control panels. Residential loadcenters are too cheap to even use those goofy panels anyway, too small.

 

[William Payne]

The DIN rail breakers are listed as supplemental protection, so the only place they are used is in control panels. Residential loadcenters are too cheap to even use those goofy panels anyway, too small.

Oh ok must be different down here. Din rail is everywhere down here. go to the electrical store and ask for a distribution board (what we call electrical panel) it will be din rail.

 

[rlitman]

... I don't know how likely, but it's theoretically possible that by tripping the main 200A breaker, it _saved_ other equipment connected to your panel. If the surge came from upstream of the main breaker, and the 30A surge protector breaker had tripped instead, then the surge would have been allowed to continue to the rest of the panel after the surge protector sacrificed it's life.
Too often, these surge protectors burst into flames at the end of their life...

No, a breaker couldn't have protected from a surge. The inertia of the breaker's moving parts is far too great to allow a breaker to trip anywhere near fast enough.

Look at it this way. The timing of surges is measured in picoseconds (like the timing of the voltage rise in a TDR). The timing of a breaker opening is measured in milliseconds. The difference in their intervals is something like seven orders of magnitude.

Let's assume that the actual surge had enough energy to trip the breaker (it did not). The surge would have had time to pass through the breaker and enter the SPD. It would also continue on the bus past the SPD to the far end of the bus, and reflect back. That reflection would have passed through the main breaker in the opposite direction (and also been partly absorbed by the SPD) well before the breaker could possibly have had time to open. By then, the remnants of the surge will already have propagated to all of the taps on the bus. On a side note, it is because of this reflection that it is recommended that the SPD be tapped furthest from the main on the bus.

So, here's what most likely happened:

The surge passed through the main breaker. lt entered the SPD, and as soon as the voltage rose above the clamping voltage, the MOVs within formed a dead short, which effectively drops the voltage to beneath the clamping voltage. This releases the energy of the surge as heat, but because the duration of the surge is so short, the energy is limited. So far so good, this is normal operation, and the MOV should re-open after the surge passes.

When the MOV fails however, it stays closed, shunting the system voltage across it, and releasing far more energy until the fuses or breakers upstream of it open. The way it is SUPPOSED to work, is that the internal fuses interrupt the current and take the shorted MOV out of the circuit (on high voltage applications, this uses an explosive charge). But if the fuses fail, then the circuit breaker will take its place. However, as demonstrated by the OP, if you re-close the breaker with the fuses in a failed state, the short across the MOVs will start a fire within the SPD. IMNSHO, this is either an inexcusable major manufacturing design flaw in using insufficient fuses to match the SCCR of the panel, OR an engineering mistake by the installation of an SPD that was insufficient for the panel.

Is anyone running the din rail ones that become part of the electrical panel? Like Schneider electric or other brands.

The form factor of an SPD has nothing to do with its effectiveness. A DIN rail SPD would make sense if you had a DIN rail panel. I cannot imagine fitting one inside any residential panel, and in anything other than a DIN rail enclosure, the status LED would be blocked. Use the form factor that fits your purposes best.

Oh ok must be different down here. Din rail is everywhere down here. go to the electrical store and ask for a distribution board (what we call electrical panel) it will be din rail.

Yep, it's a regional thing. Like I said, if you have a DIN rail panel, then a DIN rail SPD will be a good fit.

I have never seen it or heard a reliable acount of this happening.

The most recent encounter I personally had with a surge damaging an inverter, was in the VFD for a motor in a residential elevator (I promptly installed a whole house SPD in that panel after that event). Surges have the potential to cause damage. Whether or not you consider accounts of surges causing damage to be reliable is another story.

 

[bjcouche]

OK, question then,
Are you saying that the SPD clamps the surge voltage to the clamping voltage for as long as it is capable of doing so until the total energy exceeds it's rating and it fails by blowing it's internal fuses? This is how I understand their operation. If we both agree on that, then the only question is whether the SPD can clamp the voltage for long enough for the main breaker to trip. I didn't do the math, but I know the joule rating of the MOV's is small compared to the amount of energy required to trip a breaker on short circuit.
This is why I always use fast blow current limiting fuses on 120V industrial control circuits, so that when somebody shorts a wire out, the arc flash looks more like static electricity, instead of blowing the end of your wire, screwdriver, etc. off waiting for a magnetic breaker to trip.

Brian

 

[sberry]

I say this because while it may have happened to some inverter equipment somewhere in service I never seen or hear of it happening to a welder especially unlikely with the fuggin switch off Edison. This is no personal offense to the guy that said it because it just comes out but I been in this biz a fair while and like welding on cars I never seen it happen but an amateur knows several cases where it occurs on a regular basis.
I have also seen thousands of posts on Miller and Hobart and cant recall a case of a machine hit by lightening or any recommendation from the people designed it warning us we should have panel protection for this reason.

 

[William Payne]

OK, question then,
Are you saying that the SPD clamps the surge voltage to the clamping voltage for as long as it is capable of doing so until the total energy exceeds it's rating and it fails by blowing it's internal fuses? This is how I understand their operation. If we both agree on that, then the only question is whether the SPD can clamp the voltage for long enough for the main breaker to trip. I didn't do the math, but I know the joule rating of the MOV's is small compared to the amount of energy required to trip a breaker on short circuit.
This is why I always use fast blow current limiting fuses on 120V industrial control circuits, so that when somebody shorts a wire out, the arc flash looks more like static electricity, instead of blowing the end of your wire, screwdriver, etc. off waiting for a magnetic breaker to trip.

Brian

I say this because while it may have happened to some inverter equipment somewhere in service I never seen or hear of it happening to a welder especially unlikely with the fuggin switch off Edison. This is no personal offense to the guy that said it because it just comes out but I been in this biz a fair while and like welding on cars I never seen it happen but an amateur knows several cases where it occurs on a regular basis.
I have also seen thousands of posts on Miller and Hobart and cant recall a case of a machine hit by lightening or any recommendation from the people designed it warning us we should have panel protection for this reason.

It's that once in a million thing. Will probably never happen and never need it. I've never seen it happen but have met people who have. For me the thought of it even happening is enough for me set something up just in case of that one in a million thing.

Im used to working in industrial environments where you think about this stuff.

 

[westom]

If we both agree on that, then the only question is whether the SPD can clamp the voltage for long enough for the main breaker to trip.

Breaker is only for a condition that must not happen in any protector - catastrophic MOV failure. Breaker is never to disconnect a surge. A breaker is a secondary protection device - to protect human life - not hardware. If MOVs fail in an unacceptable manner - catastrophically, it is designed to contain that failure long enough for a breaker to trip.

MOVs must degrade - not fail catastrophically as so many assume (because so many plug-in protectors are grossly undersized and fail catastrophically). Should that rare failure happen, a breaker (any properly selected breaker) cuts off power long before flames occur. Something was done incorrectly either by the manufacturer or by an installer. Since that event must never happen - even though it is too common with plug-in protectors.

Why are protectors installed? To protection from something that might occur once every seven years. A number that can vary in states and even in a same town. A number manipulated by local conditions such as utility wire routing and geology beneath the shop. Since that protection is so cheap and so effective, well, it is already installed for free (as required) on telephone, dish, and TV cable installations. Unfortunately we still do not require it on AC electric service. So a property owner must take initiative.

 

[sberry]

Sounds like the safety equipment simply failed and shorted out without faulting thew breaker it was hooked to, what happened first is a guess. Maybe nothing, no external event.

 

[rlitman]

OK, question then,
Are you saying that the SPD clamps the surge voltage to the clamping voltage for as long as it is capable of doing so until the total energy exceeds it's rating and it fails by blowing it's internal fuses?...

Close enough that I'd say we agree.
To be pedantic: the oxide layer in the MOV has a breakdown voltage that is somewhat analogous to the reverse breakdown voltage of a diode (think of a high voltage zener, except that MOVs do not have a directional bias). Below the clamping voltage, the MOV has a high resistance that leaks very little current. Above it, it has a low resistance that acts like a shunt.
The oxide layer in the MOV is damaged (effectively consumed) by energetic surges, leaving it with lessening reserve to absorb energy as it takes hits. Eventually a fatal surge will short it, and the fuses blow. That might be in one awful surge, or after many small ones.

... then the only question is whether the SPD can clamp the voltage for long enough for the main breaker to trip. I didn't do the math, but I know the joule rating of the MOV's is small compared to the amount of energy required to trip a breaker on short circuit.
This is why I always use fast blow current limiting fuses on 120V industrial control circuits, so that when somebody shorts a wire out, the arc flash looks more like static electricity, instead of blowing the end of your wire, screwdriver, etc. off waiting for a magnetic breaker to trip.

Yes, yes, and yes.
Look, energy is the area under the curve. So with a surge's super short duration, while the current may be the full short circuit current capacity of the supply, and the wattage may be staggeringly high, the total energy is actually fairly low. Stretch that time out to the time it takes a breaker to trip, and the energy starts to get rather large. But surges don't last that long.

Yes a fast blow fuse has the ability to limit the energy in an arc flash much better than a breaker. Breakers are all well and good, but fuses have the ability to act more quickly, and good fuses can interrupt greater currents.

I say this because while it may have happened to some inverter equipment somewhere in service I never seen or hear of it happening to a welder especially unlikely with the fuggin switch off Edison...

Yes, a REAL off switch is indeed the best protection. But even that only works if it happens to be switched off.

I'm glad that both my Miller welder and my Hypertherm plasma cutter use real off switches that cut all power to the machine when switched off (they are both in fact circuit breaker type switches with side contacts that allow the machine to trip them off). But I can't say that all manufacturer's provide that sort of quality. My TV's power supply is certainly still connected to the grid, even when it is apparently off. So is any modern computer (since the inception of ATX and WOL). I will freely admit that I have never seen a welder damaged by a surge. But that does not mean that the switching power supplies found in inverter welders are inherently any more resistant to surges than any the IDENTICAL equipment found in appliances with other labels. My guess is that for every hour that a welder is switched on, the sum of the rest of the electronics on your electrical meter's load may see 2000 hours of power.

Whether or not YOUR electrical service is likely to experience damaging surges is also quite variable. A large panel close to the utility transformer may see a lot more surge energy than a small panel far away (due to large differences in the impedance of the feed).

 

[zmaxmotorsports]

I wouldn't call it a panel fire,makes it sound like the panel failed.

 

[sberry]

Back in the day when modems were mostly internal there were a lot of problems, I had one. Blew up a couple 3 or 4 protectors till it dawned on me to read the code section and come to the conclusion that my phone wasn't grounded properly. We have a small company and they don't know fuk.
I did quiz a satalite installer a while back and the guy was really sharp and fussy. But the phone guys learned hand me down in a month and because they don't "understand it" they just do it how they feel like it pretty much.
I have seen some damage, mostly to connected electronics and older puters, grounded and hooked to ungrounded phones where it didn't blow the suppressors on their side. The one I connected to it is grounded, it takes the hit. Once I got with the code guru and he splained in further detail what was happening and I ground it right the problem went away. I had one really violent hit not long after took out the company sup, never replaced another add on unit, never unplug a thing.
I have seen a couple hits in other places, one where it "took out the well" but what it really did was take a hit on an ungrounded metal sheet shed, went all the way thru the circuit, ended up down hole with the service conductors on a metal case well. Ground rods rotted off the service, no casing connection to the service, best place must have been the submersible well pump.
I had one hit, I dont think the original installer knew what he was doing along with the electrician. When I re did the thing I quizzed the code guru and he says ground it to the casing and skip the rods. I welded a lug to it.
As to the op, didn't look back and proof read here but sounds like the thing simply caught on fire. Should have tripped the breaker it was connected to.

 

[sberry]

I'm glad that both my Miller welder and my Hypertherm plasma cutter use real off switches that cut all power to the machine when switched off (they are both in fact circuit breaker type switches with side contacts that allow the machine to trip them off). But I can't say that all manufacturer's provide that sort of quality.

The code calls for all ungrounded conductors to be disconnected in a welder switch. Yes, your TV is another matter.

Whether or not YOUR electrical service is likely to experience damaging surges is also quite variable. A large panel close to the utility transformer may see a lot more surge energy than a small panel far away (due to large differences in the impedance of the feed).

Where are these surges coming from, the poco get a wild hair up its *** and spike the voltage a 100 out of the blue just for the hell of it or is it coming from lightening strikes?
Anything truly sensitive would likely call for insulated grounding. But welding machines are designed to be very tolerant, they run inverters off of **** generators. Most problems not related to storms but some little **** transistor or logic chip on a board goes out for who the fug knows?

 

[sberry]

Edison I am not but feel free and I take this from explanation above,, laymans terms. Above clamping voltage it is a shunt to grounding electrode system and if it has an internal fault needs to be connected to a breaker small enough it will short circuit, either line to line or line to ground? The better the ground the better it works and to a faulty ground ,ight as well not be there as in the case where the original phone sup I had was simply passing it thru.

 

[rlitman]

...As to the op, didn't look back and proof read here but sounds like the thing simply caught on fire. Should have tripped the breaker it was connected to.

You'd think, but breakers aren't designed specifically to prevent fires. The Homeline breakers do not have brains. Personally, I wonder if an AFCI would have tripped in this case.

In the OP's case, the main breaker tripped, because the short circuit current across the failed device (when it flashed over internally, turning the innards temporarily into a fully conductive plasma) was excessive. The fact that the breaker the device was directly connected to did not trip can be explained. The catastrophic event that destroyed the SPD left it internally a hunk of carbon that now acts as a resistor when the main was re-closed. Since it was not designed to dissipate heat with any sustained wattage, it caught fire. But it is perfectly reasonable to believe that the carbon resistor it turned into might catch fire without carrying enough current to trip any breakers. Think about the heat that a heat gun can put out. If you stop the fan on a heat gun, it too will quickly catch fire without tripping a breaker. Now imagine three times that wattage (this was fed by 240V on a 30A breaker) packed into a smaller case. It could easily catch fire without tripping, since it was no longer a dead short.

The code calls for all ungrounded conductors to be disconnected in a welder switch. Yes, your TV is another matter.
Where are these surges coming from, the poco get a wild hair up its *** and spike the voltage a 100 out of the blue just for the hell of it or is it coming from lightening strikes?
Anything truly sensitive would likely call for insulated grounding. But welding machines are designed to be very tolerant, they run inverters off of **** generators. Most problems not related to storms but some little **** transistor or logic chip on a board goes out for who the fug knows?

The code that I am familiar with applies to a disconnect switch feeding a hard-wired welder. I haven't seen one of those in years. Every welder I've used in the past several decades was connected with a plug that serves as a disconnecting means. The switch INSIDE the welder has to meet UL standards. I'm not familiar with those requirements, but I've seen plenty of imported welders that are not UL certified (not that I'd plug such a device in in my shop).

Lightning is one source of surges. But "internal" sources are more common.
You should look here:
http://www.nemasurge.org/history/

 

[sberry]

Good links, I read some. I sat thru a couple lectures with some eggheads a long time ago.

 

[westom]

Lightning is one source of surges. But "internal" sources are more common.

Depends on what numbers exist to define a surge. Internally generated surges are single digit or ten volt transients. Those occur hourly or daily. How many clocks and GFCIs do you replace hourly or daily? None. Because those surges do not do damage. Those are best called noise.

Surges that do damage are created by lightning, linemen errors, tree rodents, utility equipment failures, stray cars, utility switching, and other external events. Surges with numbers high enough to overwhelm protection inside every appliance have higher numbers.

Protectors have a let-through voltage. For 120 volt service, that my be 330 volts. That means a protector ignores everything until voltage well exceeds 330 volts. If anything inside is creating voltages that high, then a first item damaged is that item creating that voltage - if a 'whole house' protector does not exist to make it irrelevant.

Transients high enough to be seen and reacted to by a protector will create well over 330 volts if not connected to earth by that protector. As explained previously, an effective protector really does not do protection. Best protectors absorb less energy. Instead, hundreds of thousands of joules must dissipate harmlessly in earth. If that connection to earth is not low impedance (ie less than 10 feet), then a protector cannot connect energy to where it must dissipate.

Internal appliances do not create surges that a tens or hundreds of thousands of joules. Protectors are not even designed for 'surges' that occur hourly or daily. Protectors are for transients that overwhelm protection inside all appliances. Those typically are generated outside building - not by appliances. Protection from them is always about how that energy is connected to and harmlessly absorbed by earth ground.

A protector is only a connecting device to and as effective as its earth ground. Protection is defined by the item that harmlessly absorbs hundreds of thousands of joules.