What happened here? Your best guess welcome... Photo1 1

[N1755L]

Hello everyone,

Attached is a photo of a disconnect switch that was installed to disconnect a 600V, 3-phase, 30A water heater tank. The switch is indeed rated for 600V, 30Amps.

http://files.engineering.com/getfil...7-4d6c-43ca-987f-65985df98eba&file=Photo1.JPG

As you can see from the photo, the two outer screws have suffered a considerable amount of meltdown on the downward side facing the plate, and though the plate has spatter on it, it doesn't appear to have any arc burn damage (for as much as I can visually ascertain). Yes, that is a bit of insulation lifted off the left-side wire, but I can't really tell if it was blown off because of arcing, or if it was handling damage (whoever did this installation was not easy on the wire). The next thread, "What happened here? Your best guess welcome... Photo2" shows another angle of that left-side wire. The strange part is that the exposed copper is immaculate where that insulation is nicked, all the strands from the copper there has absolutely no damage at all (see photo2.jpg, on next thread, sorry, can't get the bloody camera to focus exactly on the wire, but let me assure you, there is no damage to the copper).

Another electrical company had been there earlier this morning and changed the elements in the hot water tank 'cause they'd made the assumption that there must be a problem with the elements 'cause it appears from what we've been told that the main fuses for the restaurant had previously blown before the breaker for this circuit had a chance to disengage. The manager says that after the elements had been changed, the circuit had been ok for approx. 15 mins. this morning, then the main fuse had popped again. Apparently, the previous times this "main fuse popping" had happened, the hot water tank would run for a period of time and eventually one of the main fuses would go, like this morning. I'm guessing that a second or third element in the hot water tank kicks in when this happens... will take a closer look at the tank tomorrow when I go install the replacement for that switch (a fused disconnect).

Is it possible for arcing to cause burning like what is seen on the switch screws in this photo, without similar burning at the other end of the arc (the nicked wire)? I can't imagine, I'm fairly certain it could not, but then where else would the arc have been going to on the end opposite the screws on the switch?

The burnt side is the line side, yes.

How did these switch screws sustain so much damage without any further damage being seen/found on either the wires, nor the cover of the box?

What's your best guess? Thanks.

 

[itsmoked]

Yes an arcing fault. Between the outer screws, with minor involvement with the center screw. Two possible causes.

1) Over voltage exceeding 600V from transformer issues, PoCo issues, or lightning. Had a storm lately? Have a big motor nearby?

2) Contamination. Rodent, insect, or water. You did mention a water heater. Can you find any sign of misplaced water? Rust nearby or tracked dust?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[VEBill]

Lightning.

 

[srbelec]

I would say overcurrent. If you had a problem with the elements, combined with poor fuse/breaker co-ordination, you could see this kind of damage. They may have replaced the elements, but did anyone megger them? or at least check the resistance to determine if there is a short?

 

[VEBill]

On the Photo 2 thread, Keith noticed the same thing I see: A pattern on the cover plate where the arcs landed. I'm thinking it would have had to be a common-mode event, and therefore probably lightning. It jumped onto the ground circuit in the switch (lots for everyone). Heading for the water heater because water pipes make good grounds.

Just a guess...

 

[N1755L]

Thanks guys,

Looking at the cover here, I have to say that the marks on the cover most certainly appear to be nothing but the spatter of the molten metal off the screws.

For some of you, I have to point out that no, there absolutely has been no lightning in this area (Toronto, ON, Canada) of late, and as for the location of the 4x4 box into which this switch was mounted, it's was mounted on a VERY DRY concrete wall, approx. 9-10 ft in the air, and the water tank is mounted on this same wall, but at least 2 ft away from this switch box, and there is absolutely no humidity up there (just so we don't go off on the wrong track here).

Remember again, that this is a 600V hot water tank, so one of the elements would have been connected to both of the phases that have the burnt screws on the switch in the photo. I was beginning to think that somehow the arc must have gone from one screw, through the steel plate, and then jumped back to the second screw, but I believe that as soon, immediately when arcing would have occurred from one of the phases to the grounded steel plate, the main fuse or the circuit's circuit breaker would have opened, but nonetheless, there would still be burn arc evidence to see on the metal plate.

I'm now fairly convinced that somehow, the arc might have been directly between the two burnt switch screws shown in the photo, and that the reason they are only burnt on the bottom end of the screws is 'cause the switch's plastic insulating material between the screws is tapered on that bottom end, providing less dielectric resistance on that bottom end. Perhaps closer the proximity of the metal on that said bottom end also helps the arc to form more easily?

But why the voltage spike? How many thousands of volts of spike would have been needed to make possible the arc jumping across from phase to phase in this manner? I measured the phases once we'd repaired and installed a fused disconnect, all three phases were measuring the exact same amperage (17.8 Amps). A fair amount of power on each phase.

Can this voltage spike result from when one of the contactors disconnects this particular element which was connected between these two burnt phases?

What if one of the wires underneath one of these screws had been loose? Could a micro-second of switch from/to varying amounts of resistance as a result of this loose connection make possible and inductive kick as described here?:

http://books.google.com/books?id=_D...maxm_is=0&as_maxy_is=&as_brr=0&as_pt=ALLTYPES

I realize the element should be purely resistive, but the circuit taken as a whole might possibly have a small component of inductance?

Lastly, I can't be certain that this wire came off the one screw because it was loose initially, perhaps it somehow got loose as a result of whatever happened when the arc damaged the screw.

Thanks for reading, I know it's long winded, but then it's not so easy to cover everything and every possibility in exact detail for the purpose of trying to answer all tangibles for the benefit of whoever is going to try to think this one through.

Thanks again, and forgot to sign my name last time.


Richard

 

[rjvh]

Hi richard

I have seen similar burned contacts due to loose wire conections in not maintained switch panels.

if wires getting hot and cool down again it expand and contracts if the conection where not realy tight in the first place it could lead to loosen up the wire in the terminal and the wire contact in terminal will be able to arc.

Robert-Jan

 

[hacksaw]

any photos of the enclosure? grounded or floating neutral and how is the system grounded.

 

[N1755L]

Hi Robert-Jan,

The strange thing is that the arc appears to have happened between the two outer lugs of the switch. I've seen a 100A disconnect with a loose lug burn #3 AWG wire to the point where there were only 1 or 2 strands left, the insulation on that leg was burnt clean off the copper for nearly a foot. They called us in 'cause their lights were dim.

hacksaw,

there's no neutral in the panel from where this circuit originates, only 3-phase 600V. I haven't seen the main electrical room where the transformer is located, I'm hoping to be able to have a look to see how/if the secondary side of the transformer is grounded.

 

[hacksaw]

have seen flash over events in process environments with carbon fines, dusts,metal fines, etc inside the enclosure
looks similar

an image of the box interior would be benefical

 

[OperaHouse]

A great book is Cobine's 1941 "Gaseous Conductors" It notes that the greatest heat is generated in the transition from conductive plasma to metal. The screws melting is a good example. The basic reason EDM and welding work so well. The plasma is actually a good conductor.

 

[N1755L]

Hello OperaHouse,

Thanks for the "heads up", am finding lots of interesting info as a result of your suggestion. A lot of reading to be done yet.

and, Hello hacksaw,

There certainly is no dust of any kind in that environment, but what you're saying makes good sense and any kinds of dust/particles as you mention would have been a good catalyst.

As for a photo of the box, I might not be able to return to the site, since it appears the problem has been resolved as a result of the replacement of the switch for a fusible disconnect. I just can't help but wonder if voltage spikes are not still occurring on that circuit over there.

Just trying to find some sort of an explanation, just for the sake of curiosity, though seems as though I might not ever find out.