Big surge protector needed

You could have added that it is worth seeing what happens at around the

17 minutes mark.

Presumably it shows what happens when the HV primary voltage of a distribution system ends up on the the LV supply lines going to a house.

I'm amazed that the house wasn't actually on fire. Don't people bond metal roofs to a ground rod in the US?

-- Sue

| snipped-for-privacy@ipal.net wrote: |> Try this on your surge protector: |>

|> | You could have added that it is worth seeing what happens at around the | 17 minutes mark. | | Presumably it shows what happens when the HV primary voltage of a | distribution system ends up on the the LV supply lines going to a house.

It could be just MV.

| I'm amazed that the house wasn't actually on fire. Don't people bond | metal roofs to a ground rod in the US?

I don't that it would help in this situation. There's too much available current.

Awesome. It looks like a cherry picker shorted against the utility wires. Not a surge - the duration is _way_ too long. Kiss your surge protector goodbye. In the video, they almost have to kiss the house goodbye!

Ed

On Thu, 24 Apr 2008 18:48:17 GMT ehsjr wrote: | snipped-for-privacy@ipal.net wrote: |> Try this on your surge protector: |>

|> | | Awesome. | It looks like a cherry picker shorted against the utility wires. | Not a surge - the duration is _way_ too long. Kiss your surge | protector goodbye. In the video, they almost have to kiss the | house goodbye!

At least three houses I could see were affected. A barely audible comment suggested all the houses in "the main street" were affected (presumably within the range of commonly connected LV circuits). I wonder if any telephone or cable wires were affected.

I could see in some parts of the video some pulled down wires in places apart from where the cherry picker was. Maybe the linemen were there to do repairs and the line was re-energized somehow (recloser closing back again and getting stuck).

How many people here think that the houses affected, assuming there was minimal fire damage, at least have to have all the wiring replaced (at the cost of the power company if it was their fault, or by insurance)?

Now for theory:

Sure, surge protectors you buy in Wal-mart would be destroyed by something like this. But, what would it take to at least protect a device like a computer?

One thing the plug-in surge protectors are touted to do is maintain the same voltage level on all wires once the voltage reaches the level where the MOVs would conduct. Now in this case, the available current and the long duration would just make those MOVs join the air pollution. So it would take something a LOT bigger. If there was something that could survive it, then in theory the computer would see the voltage rise (and drop and rise the other way and drop, 60 times a second). But as long as all connections to the computer rise and drop together, there is no voltage _difference_ and only charging current would flow. Lightning surges can be a lot higher voltage than MV distribution lines. So it should be a matter of making something that can hold up to the distribution voltage and the crossover current long enough for some really large fuses rated for such voltages to burn out. I'm thinking along the lines of a "whole house" type of protector. It might not be practical or worth the rare risk like this, but it is a challenging thought project.

:-) Nice turn of phrase!

Don't know. The assumption we are making is that massive energy reaches the point of use protector undiminished. But the only thing that counts is the energy that actually does reach the point of use protector, not the huge energy we see outdoors. I would also assume that massive energy reaches the protector so we agree on that - but it's still an assumption.

Well, at the homeowner level it would be treating the symptom, not the failure. The fix would have to involve the distribution system. The answer to the "cherry picker event" is underground distribution - which would bring a different set of possible failure scenarios that might somehow short the cables.

Ed

Christmas lights from Hell!

That happened to the summer cottage my family has, about 20 years ago. One winter, the ant-weakened top of a pole broke off, allowing the MV distribution wires (7200V I believe) to contact the 120V/240V to the cottages. Two of them burned to the ground. (it was winter and the roads were snowbound, no way the fire dept could get in there) Our cottage was fed from the same transformer, but my father always throws the main breaker for the winter. That certainly saved it from a similar fate. There's also a big surge protector on the breaker box but it was safe behind the main breaker.

|> -- |> |WARNING: Due to extreme spam, I no longer see any articles originating |> from | |> | Google Groups. If you want your postings to be seen by more |> readers | |> | you will need to find a different place to post on Usenet. |> | |> | Phil Howard KA9WGN (email for humans: first name in lower case at |> ipal.net) | | | | | You know, you can create a news account using a ficticious e-mail address, | then include your real e-mail address on a case-by-cace basis by entering it | as at dot com. This will allow anybody that wants to | reply to you to do so, while keeping the spambots at bay. Then you can see | news messages from the unwashed fools on Google Groups. Even an unwashed | fool has a valuable thing to say from time to time.

The spam issue isn't email spam in this case ... it's usenet spam. Google readers can see my posts just fine. Spammers scraping Google Groups can get my email address.

The issue is spam in the usenet postings. Google's CAPTCHA got cracked and Google seems to not be taking steps to counter that. The spam volume rose dramatically every week until I decided to block it. I see it as Google putting the posts by legitimate users of their service at risk, not me.

| Spambots know how to ignore "nospam" buried inside an email address, | especially when you set it off with punctuation. You could filter spam | better with philnewsnospam at ipal dot net.

Actually, the modified email address is a real mailbox. It gets about as much spam as my main mailbox, which is seen in a lot more places. I used to generate dyanmically random email addresses (within my domain) in the signature here. Back when I first started doing that, one of those got spammed within 7 minutes of the posting it appeared on.

------------------ Surge protectors aren't designed for long term situations like this. They are intended to handle switching or other surges which are in the order of less than a millisecond- but not line frequency currents. To be able to conduct for 17 minutes would likely result in a prohibitive cost (more than the house). Since it is apparent that circuit breakers on the MV(?) supply didn't operate, the current must have been relatively low (or something was horribly wrong with the fusing/ protection of the primaries. Assuming 5A at

7200V, there is only 36KVA involved but in 17 minutes the energy is about 37 megajoules. I'm not sure that even the surge arrestors used on EHV lines could handle that energy. I do know that their 60Hz withstand at anything over about 120% of rated voltage is in the order of seconds- not minutes.

| Surge protectors aren't designed for long term situations like this. They | are intended to handle switching or other surges which are in the order of | less than a millisecond- but not line frequency currents. To be able to | conduct for 17 minutes would likely result in a prohibitive cost (more than | the house). Since it is apparent that circuit breakers on the MV(?) supply | didn't operate, the current must have been relatively low (or something was | horribly wrong with the fusing/ protection of the primaries. Assuming 5A at | 7200V, there is only 36KVA involved but in 17 minutes the energy is about 37 | megajoules. I'm not sure that even the surge arrestors used on EHV lines | could handle that energy. I do know that their 60Hz withstand at anything | over about 120% of rated voltage is in the order of seconds- not minutes.

What would you call (a term to use to identify) protection for a situation like this?

We really don't know why the MV breakers/reclosers did not operate, or came back on (there were apparently time intervals where power was off).

Protection against this I think can be done. But it most certainly would not be in the form of long term fault current conduction if the currents are high.

If the current is LOW, such as the 5A you suggest, then the system impedance must be rather high. Maybe too high. How much voltage drop would a street of homes see at their 240V service when each draws 100A at each of 5 homes? That's 120kW total, 16.66A at 7200V. So the MV system impedance needs to be about to deliver that 16.66A without an intolerable voltage drop. Since some of that drop would happen in the transformer, the distribution can only be a portion of it. But even if we assume a 5% drop in voltage, the impedance has to be no more than 21.6 ohms, which would result in 333A in a distribution fault. That's a severe voltage drop. I suspect the system impedance would be lower.

So I think we cannot rule out some kind of failure upstream. It happens.

I believe the arcs in the video are much more than 5A. Remember, the power is a function of the amps times the voltage _drop_ of the arc itself, not the system voltage. If the system is so high an impedance that a fault would be only 5A, then more power is dissipated in the system than at the fault.

A protection device for this would first need to be able to handle the full distribution voltage. It would then have 200A fuses (whatever the amperage of the service drop is). These would have to be MV rated fuses. So this is not cheap. But it's not the price of a house, either. Then after the fuses is a device that would conduct well at medium voltage and do so long enough for the fuses to blow. It would have to conduct between the wires.

The risk here is that the MV phase was contacting the LV wires in common. That is, all 3 wires were at the same potential, and the voltage at the house was only relative to ground. It would take a very good earthing system to be able to get the current high enough to burn the fuses. So there are cases where this still won't work.

Homes in rural areas that are more spread out would typically be served by a single transformer. So then there would be no pole to pole LV wires exposed to potential MV droppings. By setting in a separate pole away from the MV lines, and bringing one MV branch over to one side of it, and run the LV drop from the other side of that pole, it would at least minimize the chance of such a mess as this video shows, in these rural situations.

The better solution is to go underground, even if the MV lines stay above ground. Using pad transformers (protected against vehicle impacts) would be safer, IMHO.

As for a failure upstream, that can depend on what failed, and how the various devices like a recloser are actually designed. I don't know how they really work inside. So I'm guessing maybe a recloser would have three current sensor transformers. Now what if one shorted out on one phase and that happened to be the phase that fell? If it faults to neutral, it would not be detected. If it happens to hit another phase, then the fault current on the other phase would be detected and the recloser opens. After a while it closes again and the whole thing starts over. Can the recloser completely reset altogether if current flows normally (from its point of view, being blind to one bad phase)?

I pulled out the 5A figure just to show the power and energy levels at a low current level. Now consider a current as you indicate (arc impedance ignored) and multiply my numbers accordingly.

However: Your calculation assumes that the transformer and line impedance is purely resistive but this is not true. The HV/MV transformer would have an X/R ratio of possibly 5 to 8 so that for a 5% drop the resistance would be of the order of 15 ohms and reactance 75 ohms for a total impedance of 76.5 ohms. Using these figures with a current of 16.7A 1.0 pf, at 7200V results in a drop of 4.9%. Now consider short circuit conditions- the current will be nearly completely reactive (generally resistance is ignored in fault studies) and will have a magnitude of 99A. Actually it would be less because ground impedance is a factor. Note that the fusing would have to be set to trip quickly (at most a second or so at, say, 60A (would have to be co-ordinated with the MV system fault level at the point under consideration, which presents problems on a 100 or

200A service as well as logistical problems so, in my opinion in house protection is impractical.

If, however, you assume that the main transformer has 5% impedance based on

7200V 150MVA (phase) then the fault current would be in the range you give (but a lot lower voltage drop under normal load- something less than 1%). THen 3 MVA fast blow fuses as you suggest might work as long as clearances between the incoming line was such that there was no danger of flashover.

So- we are considering an unknow scenario because we don't have the facts to do a proper analysis-

Your suggestions regarding arrangements of MV and LV lines are likely more practical and could be cheap or expensive depending on the load density.

We really don't know whether there was a 3 phase MV line in this region-- one phase and neutral may have been used- this is not uncommon. The reclosers are likely single phase- 3 shots and it stays open. If 3 phase, I would expect interlocking. You are absolutely correct in that something upstream went wrong and it took an unconscionable time to do something about it.

| I pulled out the 5A figure just to show the power and energy levels at a low | current level. Now consider a current as you indicate (arc impedance | ignored) and multiply my numbers accordingly.

Multiply what numbers? The arc impedance will dictate the current given a voltage. If you have 5 amps, you have far more impedance on a 7200 volt system than an arc will give you.

| However: Your calculation assumes that the transformer and line impedance is | purely resistive but this is not true. The HV/MV transformer would have an | X/R ratio of possibly 5 to 8 so that for a 5% drop the resistance would be | of the order of 15 ohms and reactance 75 ohms for a total impedance of 76.5 | ohms. Using these figures with a current of 16.7A 1.0 pf, at 7200V results | in a drop of 4.9%. Now consider short circuit conditions- the current will | be nearly completely reactive (generally resistance is ignored in fault | studies) and will have a magnitude of 99A. Actually it would be less because | ground impedance is a factor. | Note that the fusing would have to be set to trip quickly (at most a second | or so at, say, 60A (would have to be co-ordinated with the MV system fault | level at the point under consideration, which presents problems on a 100 or | 200A service as well as logistical problems so, in my opinion in house | protection is impractical. | | If, however, you assume that the main transformer has 5% impedance based on | 7200V 150MVA (phase) then the fault current would be in the range you give | (but a lot lower voltage drop under normal load- something less than 1%). | THen 3 MVA fast blow fuses as you suggest might work as long as clearances | between the incoming line was such that there was no danger of flashover.

There could be flashover on the service drop triplex. But if the protection is somehow installed between it and the house, it could protect the house.

| So- we are considering an unknow scenario because we don't have the facts to | do a proper analysis-

Probably.

I'm quite convinced these arcs were a LOT more than 5 amps. I've seen real faults that get cleared fairly fast, but they had enough time to do things on the scale of what happened in the video. These I saw would not be 5 amps since that's only 36 kW max. If the system could not deliver more than 5 amps, it can't provide much service to a neighborhood of many homes.

I once did see a downed MV line that was just sizzling and smoking in the grass. I could easily believe that one was 5 amps or less, and earth impedance was high. At the time I came upon it, a policeman was there telling people to stay back. But he was about 10 feet from it himself. I had to talk HIM into getting further away. If he were to see this video, he might realize the danger he exposed himself to. I'd already seen half a dozen such events myself by then, so I already knew. Since then I've even seen (online) the gory results of just the arc blast. These are things you and I and linemen know to stay well away from.

| Your suggestions regarding arrangements of MV and LV lines are likely more | practical and could be cheap or expensive depending on the load density.

It would certainly be an issue in a compact residential area. There, the way to go (and new installations do this for other reasons) is underground and using pad transformers.

I'm looking at building a house in a rural location, so I would have room. The question I want to figure a good answer for is just how far from the house I want to have the step from MV to LV. Too close and I have MV over or under more of my land. Too far and I get more voltage drop on the LV service. If I could get 480 volts (remote possibility in a rural area), that might work out better.

| We really don't know whether there was a 3 phase MV line in this region-- | one phase and neutral may have been used- this is not uncommon. The | reclosers are likely single phase- 3 shots and it stays open. If 3 phase, I | would expect interlocking. You are absolutely correct in that something | upstream went wrong and it took an unconscionable time to do something about | it.

My hypothesis is the recloser was acting sometimes. That's based on the way the tape was edited. If I had 20 minutes of tape with about 4 minutes worth of "fun arcs" to show on YouTube, I'd edit it down and indicated the times. That seems to be what was done (this did come from YouTube). So I believe it quite likely that the missing times on the tape are times without faults. That may be because the wires were not touching (it was windy at times as seen in the trees, so that can change). But maybe the lines were dead at the times of no arcs. Why would a lineman go up in a bucket if he believed a damaged line was still energized? It may have been a combination of recloser activity and lines not always contacting.

OTOH, the linemen should have made sure the circuit was shut off _and_ grounded on the downstream side before working on it. During a big storm that knocked out a lot of power in my area, I could tell when driving back home from a trip into town to eat at a restaurant, that the linemen were finally in my area working on it when I noticed grounding cables had been attached just beyond the last point where fuse cutoffs were on the line leading to where I lived. Power was back on 45 minutes after that.

In any case, I'm trying to guess a hypothesis of why a recloser might work sometimes, and just leave fault current flowing at other times. If I knew more about how they were designed, I could probably guess better. But even then it is a guess, as we don't know what really happened in this event.

Still, my guess seems to fit. That is, these are three phase lines (they look like it) and the CT used by the recloser was defective (or shorted) on the one phase that was the primary culprit. When that phase arced to ground, the recloser didn't see it. When it crossed over to another phase (infrequent event) then it saw it.

What does a recloser, that is supposed to lockout after 3 tries, do if it only had to open 1 or 2 times, and the fault was cleared? Does it reset the count after some time? Or is the next event in a few months going to lock out immediately? If there is a timer that resets the count after some amount of time with no observable faults, then maybe it was resetting for the times between phase crossover faults in this video? Again, just a guess.

BTW, in the video, a couple shots of the lines give the appearance that at least one pole has been pulled partly down sideways in a way that would have pull on the lines downstream and snapped one or more. I definitely saw at least one broken wire.

Another thing I notice is that the point of arcing were moving along the line. It's hard to see as the guy operating the camera was moving toward the arcs as they moved back away from him. In one scene a row of small bushes were well ahead. In another he had come just past them.

True. It is typically for short duration transients. However that is exactly why a friend with a 'whole house' protector suffered something equivalent without damage while others suffered damage to plug-in protectors and to appliances attached to plug-in protectors.

In this case, a line of 33K or 69K dropped on what may be the 4000 volt distribution primary. The resulting voltage literally exploded meters right out of the meter pans. Pieces of utility meters (hundreds) were found 30+ feet away. Friend with a 'whole house' protector suffered tripped circuit breakers, an exploded utility meter, and no household electronics damage (except two GFCIs).

In a brochure from Intermatic long ago, they show pictures of a sale manager's Intermatic 'whole house' protector in Ft Lauderdale just after the hurricane (Andrew?) passed through. The primary fell upon his secondary (240 volts) service. Intermatic was blackened. It shunted long enough for breakers to disconnect the fault. He also suffered no household damage.

-------- Scale up to the levels you are considering. The 5A was just a number drawn out of a hat to indicate that even at a low level, for the duration involved, the energy is such that a practical MOV isn't going to hack it.

-----------------

That is possible. However, it likely couldn't be done on a "one size fits all" as the fuse rating would depend on the impedance of the upstream transformer and the line between, on the basis of the small probability that something like this incident would happen. Next would be to design cars to survive head on collisions with semi's as the probability of such collisions is definitely higher.

-------------

----------- It appears that you did not read what I wrote after you first questioned the

5A figure. I do agree -see the second part of what I wrote. The first part (paragraph starting with However) is based on 5% drop with normal loading and pointing out that your calculation , on the basis of a unity pf load, doesn't deal with the actual fault conditions. The next paragraph (If, however....) considers what is more realistic in considering an ideal source behind the HV/LV transformer and assuming 5% impedance (actually a bit low considering lines). This will result in a "bolted" fault current of 417A and a fuse able to handle 3000KVA (single phase -more if 3 phase). Will such fusing do the job? It depends on other factors such as ground potentials that may occur.

------------

I have met, after the fact, such a situation- really a longer story involving a day starting with a lot of whiskey and a broken guy wire flipping over a 7200V single phase line and an attempt to clear the wire from ground- it blew out his rubber boots at the ankles, and ended up on the guy's back, burning holes in his flesh but not knocking him out and not being detected by the recloser. This went on for over 15 minutes- every now and then he would try to get up but this - he lived but was essentially a basket case.

-----------

-------- Right and the cable drop to the underground system is on a pole at the edge of my lot. It could have been across the street where the 3 phase line is but then the developer would have to pay for the extra run under the road (and patching the road). It is not a problem.

----------------

-------- A friend of mine dealt with this in another way - one MV span into his (wooded) property and the rest underground at 120/240V. He did all the UG wiring and dealt with the appropriate wire sizing. No problems and no eyesores. More money on copper in this case bt it was worth it.

-----------

---------- That's possible. As far as a lineman in a bucket- the purpose of the bucket is to allow live line work. In addition, a lineman who wants to collect his pension should never assume a line is dead. It should be grounded on both sides of the work area (particularly where otherlines are physically in parallel. There are standard safety procedures to be followed.

------- That would be something that one could expect.

A primary line dropping on secondaries will rapidly destroy MOVs. UL (US) requires the MOVs be disconnected when they overheat and fail.

Plug-in suppressors may connect the protected load across the MOV so it is disconnected with the failing MOV, or across the incoming line. If connected across the MOV a plug-in suppressor may (or for crossed primary wire may not) provide protection. A few plug-in suppressors disconnect on overvoltage. A UPS may provide protection.

Long ago Byte magazine columnist Jerry Pournelle had a 16kV primary wire cross. A computer connected to a UPS continued to function through the event. Some other equipment was not so fortunate. (This was well before UL required thermal disconnects.)

At about 6kV there should be are-over from bus to panel enclosure (US). That may provide slight protection, but may also start a fire in the house.

As usual, Bud disputes reality.

Effective protectors are not supposed to provide this protection. But being effective, properly sized 'whole house' protectors did more than what is expected.

What sometimes happens to Bud's protectors are located in some of the most fire dangerous places? A problem that occurs so often as to recently get more attention from fire departments and fire marshals. Scary pictures - more reality that Bud must deny to convert every discussion into his personal pissing contest:

or Bud will lie: proclaim every protectors was before UL1449. Every protector in those scary pictures is after UL1449 was created. Another problem with protectors so grossly undersized as to depend only on the fuse. The fuse is suppose to be another layer of protection. So why so many 'scary pictures and resulting fires?

Will that thermal fuse stop 4000 or 13000 volts? Of course not. And yet the thermal fuse alone must keep fire from desktop papers and carpets.

Let's see. The 'whole house' protector tripping the circuit breaker will also disconnect (protect) electronics. A fuse inside a plug-in protector only disconnects the MOV leaving everything on that circuit connected to 4000 or 13,000 volts. What kind of protection is that? Protection that Bud recommends. Protection that may also creates 'scary pictures' where fire is most hazardous. Why no damage where a 'whole house' protector was installed? 'Whole house' protector provided protection long enough for circuit breakers to trip or meter to explode off the building.

Bud disputes what actually occurred. Primary shorting to secondary with so much energy as to explode hundreds of meters; no damage to appliances protected by only one 'whole house' protector. Others using plug-in protectors had failed plug-in protectors and damaged appliances. Bud must dispute these actual events.

| Scale up to the levels you are considering. The 5A was just a number drawn | out of a hat to indicate that even at a low level, for the duration | involved, the energy is such that a practical MOV isn't going to hack it.

If an MOV cannot conduct 5A for a long sustained time, then sure, it is not a good solution. The time frame doesn't need to be as long as the video was showing faults taking place. It only needs to be as long as necessary to blow the fuses. Of course, if you have fuses intended for a 200A service, getting them to blow is a function of getting enough energy dissipated in them, over some time frame, to burn out the fuse element. That might be

400A for a long time or 4000A for a very short time. You need to make the conductive element last LONGER than the fuse, obviously. But that is all. We can assume a requirement that if the fuses blow and need to be replaced, the conductive element also needs to be replaced because it will have taken enough damage that it cannot do the job the next time.

Clearly, a simple MOV like you find inside a plug-in surge protector is not going to accomplish that. I don't know if an MOV can be scaled up enough for one, or even several in parallel, to do the job. I'm leary of parallel because I can envision them not all working at the same time, and the high current blowing them up in cascade too quickly.

I'm thinking more along the lines of a sacrificial arc gap combined with current limited MOVs.

And maybe a transformer after that point.

|> There could be flashover on the service drop triplex. But if the |> protection |> is somehow installed between it and the house, it could protect the house. | ----------- | That is possible. However, it likely couldn't be done on a "one size fits | all" as the fuse rating would depend on the impedance of the upstream | transformer and the line between, on the basis of the small probability that | something like this incident would happen. Next would be to design cars to | survive head on collisions with semi's as the probability of such | collisions is definitely higher.

The transformer that is stepping the 7200+ volts down to 240 would not be a part of that equation, since the issue is voltage bypassing it. But that is probably an oversimplification of the range of possible scenarios.

|> I'm looking at building a house in a rural location, so I would have room. |> The question I want to figure a good answer for is just how far from the |> house I want to have the step from MV to LV. Too close and I have MV over |> or under more of my land. Too far and I get more voltage drop on the LV |> service. If I could get 480 volts (remote possibility in a rural area), |> that might work out better. | -------- | A friend of mine dealt with this in another way - one MV span into his | (wooded) property and the rest underground at 120/240V. He did all the UG | wiring and dealt with the appropriate wire sizing. No problems and no | eyesores. More money on copper in this case bt it was worth it.

I want to minimize the amount of MV on my land. If I can't do maintenance on it myself, I want as little of it there as possible, preferrably none. But I'm not talking about miles of distance. It would just be hundreds of feet or meters. My calculations using Gerald's voltage drop calculator indicate that for the shorter distances I'm thinking of, 240V all the way would be OK. If I can get 480V or 600V, that distance without MV would be even more. Then if the distance is any longer, I will have to have MV. Then it is a matter of whether the advantage of putting the MV underground is worth the extra cost.

| That's possible. As far as a lineman in a bucket- the purpose of the bucket | is to allow live line work. In addition, a lineman who wants to collect his | pension should never assume a line is dead. It should be grounded on both | sides of the work area (particularly where otherlines are physically in | parallel. There are standard safety procedures to be followed.

And either the line the guy was approach in the bucket was NOT grounded, or the grounding came loose. It sure seems to me that more than one or two things went wrong at the same time in that event. Possibly, he was trying to ground it in that trip up the bucket, or maybe pull fuses too close to the fault area?

| As to what the sequence of "failures" was or why they occurred- is a matter | of conjecture.

Unfortunately, that's about all we can do, here.

But given that I know about multiple incidents of MV crossing over to LV for various reasons, I have an interest in explore all the possible approaches to protection and prevention against this, above and beyond what is already done. I do know the whole scenario is possible well before any linemen show up to even try to work on it. I've seen lightning hit and break a MV line and said line drop to ground and arc there briefly before being cut out. It could have just as easily come down across LV lines right then and there.

I once saw multiple events like this during an ice storm. I already had _everything_ in my apartment unplugged (that I could unplug) just in case.

| At about 6kV there should be are-over from bus to panel enclosure (US). | That may provide slight protection, but may also start a fire in the house.

Mount the panel on a large (expensive) copper sheet that is multiply grounded. You could run the grounding wires from the panel along this sheet and bolt or weld them to the sheet every few inches along the way. Just keep as much that is burnable away from the panel as possible.

Apparently w__ already converted it into a pissing contest.

w_ can't understand his own hanford link. It is about "some older model" power strips and says overheating was fixed with a revision to UL1449 that required thermal disconnects. That was 1998. There is no reason to believe, from any of the links, that there is a problem with suppressors produced under the UL standard that has been in effect since

1998. None of the links even says a damaged suppressor was UL listed.

A service panel suppressor, on failure of it?s MOVs, may only open the UL required internal thermal disconnects. It may trip the circuit breaker for the branch circuits it is connected to. Or it may trip the main circuit breaker.

As my post clearly said, the protected load may be connected across the MOVs and be disconnected with the MOVs when they overheat, fail, and are disconnected. A good suppressor will be connected that way. There is extensive discussion of this at an IEEE guide on surges and surge protection:

I made no recommendations.

I said nothing about whether w_?s event happened. But I would like to see an independent news article. Must have made it into the news.

In any case, it is anecdotal science.

Neither plug-in or service panel suppressors are intended to protect against crossed power lines.

Bud is here to make things nasty. Bud follows me everywhere to post nasty and to post propaganda. Provided are numerous safety problems that are most problematic when primary shorts to secondary. Since Bud cannot contradict even two examples, he picks one. Then Bud forgets that UL1449 was created 28 Aug 1985 - not in 1998. Bud would have us believe all those burned protectors were before 1985. Otherwise he must admit the safety threat created by plug-in protectors. An honest Bud would have noted all plug-in protectors met UL1449 and still created scary pictures. Most 'scary pictures' are protector after 1998. But then Bud could not insult and would have to admit safety problems with a product he promotes for.

'Whole house' protectors (not intended for that primary to secondary fault), in multiple locations, resulted in no electronics damage whereas neighbors without 'whole house' protectors suffered both plug- in protector damage and electronics damage. To avoid that reality, Bud must insult. Bud is so dishonest as to not even admit he promotes plug-in protectors.

Fire marshal and fire departments both note the problem. Protectors depend only on emergency safety backup circuit (thermal fuse) to stop potential house fires. Why would anyone put a dangerous device adjacent to paper piles or on a rug? A danger that is greatest when primary wire falls on secondary wire since that thermal fuse is not intended to open (block) 4000 or 13,000 volts.

The worst type protector is one located on or adjacent to flammable materials. Instead, Bud pretends 'scary pictures' do not exist - especially:

or Why pay tens of times more money per protected appliances for something that only increases the fire risk? Because it results in increased profits. 'Whole house' protector has a history of protecting from primary to secondary faults AND are not located where the fire risk is highest. Bud must insult and post half truths to deny this human safety problem.

Household wires are only rated for 600 volts. Primary may be 4000 or 13,000 volts. How does that plug-in protector also protect building wires? Just another reason why one 'whole house' protector is better suited to making that fault less problematic. Not only did the 'whole house' proector protect electronics. It also protected all interior wires - again without creating those 'scary pictures'.