Disconnect Melting

JoelHunter1102 said:
Strathead,
per your itemized list below:


1. We have put monitoring equipment on and tracked amperage for a couple weeks (maybe a month) at a time. It never pulled over 72 amps on any leg. Voltage never dropped or went haywire.
2. We are now on our 4th disconnect (numbers 3 & 4 are 200A HD disconnects for the 100A load). So not sure how high resistance would be a recurring issue across 3 disconnects at the same point.
3. High surrounding temperature is legit, it is above a gun range, outside of all HVAC cooled areas. But wouldn't that cause the issue to happen on more than just the B leg every time?
4. So you suggest maybe upsizing the conduit and wire to give more area for the conductors to breathe?
5. Not sure what would cause high heat to be focused on the line side of the middle leg of a disconnect.... in the picture i was sent it doesnt look like our conduit penetration is directly above.... which i now have the ability to attach pictures. so please see attached!


Thanks all!
Click to expand...
replace the fuses next time, they look like they been around longer than your overheating problems have been around. Probably overheated the tab of the fuse when the first switch failed, that metal will never be the same once it been overheated and will take out the next fuse holder when it heats up again.
 
  • Like
Reactions: J2H
Is there some kind of reason they are using the disco for an on/off switch? Maybe a contactor with a conveniently mounted control would stop this. You should put up a trail cam aimed at the disco
 
Correlation does not mean causation. OK, it was the B-phase each time. That does not rule out an installation error on B-phase both times. One test that can be done after the next repair is to test the the connections using a "Ductor". It puts 20 some amps though a connection and measures the resistance under that load. Another method is to measure the milli-volt drop across connections while under full load.
 
The fuse is not seated correctly because of my coworker messing with it while removing the disconnect. He is telling me that they are always fully seated every time we replace the disconnect, and we have replaced the fuses each time we have replaced the disconnect also.

If we get a chance to swing by the place again (before another melting incident) I will have him take a picture of it as it is currently installed, maybe that will help shed some light.

Thanks!
 
Pics Pics Pics.....Did you guys take pics of all the "ruined" disconnects? It would be good to compare them. Mentioned already but that sure looks internal to me. Pics of conductors before replacement of disconnect to prove otherwise.
 
What is the load?
Is this disconnect being operated daily?
Is the switch being used as an ON-OFF switch for lighting with ballasts?
You said the loading was about 68A. What is the startup/inrush? How long does it last?
Do the fuses often blow before the switch melts? Are they replaced by the owner/users?
 
Check the weatherhead. I notice that the pattern inside the supply lugs is different, and wonder if water is entering one of the supply wires at the weatherhead, wicking down, and damaging only one of the disconnect legs. Of course the different pattern could be caused by the differential heating, but we are looking for something that remains constant between the different disconnects....

-Jonathan
 
Good morning again guys,
Ill try to summarize the answers from all the recent questions:

Loads: 208v/3ph, when we measured load was somewhere around high 60's pretty evenly across all three phases, without any real fluctuation. The motor is a large ventilation style fan to pull the smoke out of the gun range.

There is no weatherhead, this is all indoors.

No we have not been taking pics until recently. Didn't think it was something significant until the 2nd disconnect burned up the same way as the first.

Yes it is being operated daily. No, the disconnect is not being used as the on/off means (unless the customer is doing this on their own and lying to us when we have asked them about that).
I do not have an answer about the in-rush load and duration; last time we monitored was several years ago before we installed the 200A disconnect, and after that we thought this was resolved, so we have put a lot of those specific details out of our memories at this point.
The fuses have never blown, and have only been replaced at the same time we replaced the disconnects (again, unless the customer has replaced fuses without telling us, and lying to us when we asked them).

As for startup loads, there is a VFD on the load side of the disconnect, between the disco and the fan motor. Which should be controlling the ramp up (at least as it pertains to load going to the motor).

Thanks!
 
What do you use to measure phase current? Do you know its frequency response or true-rms capabilities?

The VFD introduces a new wrinkle, because of its input rectifier. It may be the case that the RMS current is the same on all three phases, but that B phase is carrying higher peak current from a distorted waveform. Is there any chance you could put an oscilloscope with current probes on the circuit?
 
ptonsparky said:
Any chance that VFD may be operating under single phase conditions on occasion? That would significantly increase the current on A&B. Still makes me wonder why B only.
Click to expand...
Baldor drives (at least the 15H and 18H of the early 2000s) had a current limiting resistor to lower the inrush as the drive powered up, which was then bypassed either through a relay or contactor (depending on size) once the DC bus came up to spec.
 
Our go-to measuring device is eGauge with CT's. Which (if you are not familiar) allows us to send the load data in real time to a website where we can watch it, and also keep it on for months at a time, then be able to go back and look through months of data.

I do not know the specifics in terms of frequency response, or true-rms capabilities. But that might be worth me looking into to find out.

The outside company we hired to do the testing did use an oscilloscope as well as a handful of other tests, all of which apparently led them to tell us the disconnect was the problem. They then suggested we upsize it to the 200A HD disconnect...and here we are several years later.
However, i wouldnt be opposed to us buying our own oscilloscope for our crew, do you have a brand/model you recommend?

From what i have learned in my career is that there is always a chance that something may be programmed to do something unexpected haha. So while i dont believe the VFD may be operating under single phase conditions, i think it is absolutely worth looking into the next time we are there.

However, my coworker just told me (i am usually texting him to get these answers so i can reply to you guys in as timely a manner as i can :) ) that originally the soft start that was hooked up to the fan motor would occasionally throw an error code and stop working. This led the fan contractor to ask us to install the VFD that we are now using.
So i assume that the problem is hopefully not with the VFD, but whatever was causing the soft start to fail is likely the culprit?

I really do apologize for not having all the information readily available, as i have mentioned quite a bit now, i wasnt the actual installer, and the last time i had any thought about this was at least 3 years ago. So its all coming back into memory as my coworker and i discuss this and retrace steps.

Thanks!
 
Maybe a
ptonsparky said:
Any chance that VFD may be operating under single phase conditions on occasion? That would significantly increase the current on A&B. Still makes me wonder why B only.
Click to expand...

It could also be that the peak L-L voltage across A-C is sufficiently smaller than the A-B and B-C peak voltages that the rectifier diodes do not conduct when the A-C voltage waveform is at its peak. That would make it operate as a 4-pulse rectifier instead of the normal 6-pulse. Lines A and C would then each conduct current during 2 of the 4 pulses per cycle, but line B would conduct current during all 4 of the pulses. That would cause significantly more current on line B.
 
JoelHunter1102 said:
However, my coworker just told me (i am usually texting him to get these answers so i can reply to you guys in as timely a manner as i can :) ) that originally the soft start that was hooked up to the fan motor would occasionally throw an error code and stop working. This led the fan contractor to ask us to install the VFD that we are now using.
So i assume that the problem is hopefully not with the VFD, but whatever was causing the soft start to fail is likely the culprit?
Click to expand...
Entirely possible. Another shot in the dark here, but has anyone meggered the run? Perhaps there's a high-impedance ground fault somewhere along the line. I also like synchro's theory that the CA voltage may be a good bit lower than AB and BC. I've seen those voltage imbalances coming from the utility on our BMS fairly often (though usually it's been one higher rather than two lower).

Also, the range should be able to start/stop the exhaust fan using the VFD's control panel (or even a basic switch connected to a digital input, if the drive has one) rather than using the disconnect as the switch. While it probably won't solve the overheating problem, it will drop the wear and tear on the disconnect down to near-zero.
 
Don't worry that you haven't posted everything that might be relevant. You have a very interesting puzzle, and whatever is going on is something unexpected.

I am not familiar with the eGauge system, but it claims to be able to log and report harmonic distortion or harmonics. One of the things that I'd investigate is if harmonic current is different on the different phases. You might have to monitor for an extended period of time.

Right now I think everyone will be guessing and grasping at straws, because this really is a strange one. It may simply be that you've won the lottery, the same failure happening several times in a row, just by pure coincidence. Or it may be that there is some subtle common failure mode.

Right now I'm wondering if you can put a power quality monitor and an IR camera (and maybe a few temperature probes) on the thing, and report back the next time it fails :(
 
@JoelHunter1102
From my experience, the problem is the style of disconnect you are using, one with a rotary bar operator with double sliding contacts on each pole. The fact that they switch it frequently doesn't help it much. The spring/lever action to engage all six contacts at one time in unison, leads to not all of the contacts fully engaging properly. each one of the six contacts has spring pressure on each blade half, to keep the switch contacts tight to each other. After several operations the rotary bar does not have enough force to fully engage all 12 sliding contact surfaces. 3 poles. 4 sliding contacts per pole.

I had several of that style go bad on different jobs when mildly loaded and switched frequently. I learned to not purchase disconnects of that style construction. if I remember correctly GE and Siemens were making that style back when I was having those problems.

After the switch got some normal use and you turned it on, you could take a screwdriver tip and push on the rotary contact blades and get them to fully engage the rest of the way, the rotating bar turned just a little bit further to fully engage the 12 contact surfaces.

the older style single hinged knife blade construction or the even older clamp-matic ITE/Siemens style were the ones I used new or old, no rotary bars and no more problems. Once one of the blade contact surfaces heats up and galls a bit, the grease dries up and the condition gets worse quickly.
 
You must log in or register to reply here.
Top