I wish there was a way to check the "voltage drop" across this one... Check out the attached image!!

What we have here is the top of a 12,470 volt fuse holder. It is drawing around 425 amps, feeding a 4.5MW single phase induction furnace. As you can see from the image, it's pushing 400 deg. F. These are spring/camlock design, S&C 400 fuses and holders.

This image may be alarming to some, and I'll be honest, it's alarming to me too when I have a 12kV connection that is that hot. But, I've actually seen quite a few of these in the past few years. Problem is, we can't figure out what's causing it. We just replace them when then get this bad, if they don't melt down first. I have some digital photos I could share as well. It looks like arcing between the fuse holder and the clips. Yet, everything is at the same potential! We suspect that there are times when the current spikes to around 1000 amps, yet somehow it does not blow the fuses?!?!

Furnace_7_Sw_Gear-76040012.tmw (160 Kb, 261 downloads)

Ooops. I'll attach it as a Word Document this time.

Furnace_7_12kV_fuse.doc (240 Kb, 237 downloads)

quote:

Originally posted by Bill Schmitt:
What we have here is the top of a 12,470 volt fuse holder. It is drawing around 425 amps, feeding a 4.5MW single phase induction furnace. As you can see from the image, it's pushing 400 deg. F. These are spring/camlock design, S&C 400 fuses and holders.

I hope you're wearing some serious PPE when you play with these things. Geez. That's a scary bit of business, that.

quote:

We suspect that there are times when the current spikes to around 1000 amps, yet somehow it does not blow the fuses?!?!

Short pulses won't blow the fuse (even if it's fast acting) unless the current is really high. I've had wiring heat up enough to catch on fire and melt into a puddle without the fuse blowing, and the fuse was a fast-acting KTK sized to 'protect the wiring'.

quote:

Originally posted by Metalworker Mike:
I hope you're wearing some serious PPE when you play with these things. Geez. That's a scary bit of business, that.

It's in an enclosure and I'm viewing it through a viewport, so NO I'm not exposed to it. I'm not that crazy!!

Bill,

you are looking through an IR camera view port. If this view port was retrofitted into dated switchgear, don't let the "Arc-proof" rating of the view port lead you to believe that you are safe without appropriately rated PPE. Your level of protection is only as good as the arc blast withstand rating of the old swgr. cell door. The latter is an unknown quantity in most existing installations.

I think the jury is still out on what protection we need when using IR windows and viewports. I wish I had clear cut answers.

Although I do not have hard data to back this hypothesis up, I suggest that many arc flash incidents are triggered by actions associated with opening the panel cover. If we can leave that closed, we minimize the risk of initiating an arc flash.

Of course, IF an arc flash occurs when the viewport (and probably an IR window too) is open and in use, risk of exposure may be greater than when it is closed. We need more testing on this but there are also probably too many variations (location of window, locaton and power of incident, etc) to come to simple conclusions.

I have proposed that the NFPA 70E committee consider these important questions during the current revision cycle.

Aren't the windows and viewports UL listed? Otherwise, wouldn't the installation remove the listing of the cabinet? I was given the understanding that this was a major issue when they were introduced.

John: Yes, this should be considered by the NFPA 70E group. I would be less concerned about the opening (when dealing with blast waves, some of the designs are meant to diffuse the wave) and more concerned with the mounting.

Also, yes, the idea is that when a cabinet is opened some impending problem may be disturbed. There was a recent incident at one of my clients where a thermographer happened to use his camera as the component that caused the arc flash.

You asked about the possible explanation for the arcing. I have wondered about that a little myself in various situations. If it is a making/breaking contact under load, of course pits from arcing can occur during circuit interruption. But I have seen evidence of arcing in contacts which never switch load. To me it indicates final stages of failure. I honestly don't understand how that can happen unless the contact actually begins to open circuit, but here is a passage that talks about arcing "Fundamentals of Electrical Contacts" by Milenko Braunovic´, Valery V. Konchits, and Nikolai K. Myshkin (2006, Taylor and Francis)

• "In a good connection, the temperature of the interface is only slightly higher than the bulk temperature but in a poor connection, the supertemperature increases the bulk temperature and accelerates deterioration of the contact areas, causing higher resistance. The deterioration is cumulative resulting in increasingly higher temperatures and ultimate failure of the connection. In certain circumstances, such as short-circuit conditions, melting of the contact zones can occur even in welldesigned joints, resulting in the formation of bridges of molten metal. As further deterioration occurs, these molten zones coalesce into larger areas in which arcing occurs and the whole joint assembly becomes overheated.
  
• At first sight, it might appear that the creation of welded contacts would improve the connection stability. However, on subsequent cooling, the metal shrinks and cracks because of the internal stresses set up in the process of solidification of these bridges. Oxidation of the contact zones further reduces the number of available electrical conducting paths. Thus, overheating and ultimate mechanical failure of the joint occur.
  
• From the above considerations, it is clear that one of the most important requirements for good connector performance is for the real area of contact to be sufficiently large so that even with initial and long-term deterioration, a reserve of contact area is still available to prevent overheating conditions in the joint. Unfortunately, present knowledge and understanding of various aspects relating to the failure mechanisms of electrical contacts are insufficient to assess with any degree of accuracy just how large the true area of contact should be in any given type of joint."

If I'm remembering right, melting of aluminum doesn't happen until 1000F and most other metals are higher still, so you are nowhere near that in what is visible (which is not to say that runaway failure due to thermal effects of oxidation and spring relaxation are not possible... just to say it's not clear to me whether the expalantion above that mentions melting applies to your situation or not. )

If you have repeating failures, you of course want to look at tension adjustments and lubrication. Possibly call in the manufacturer or at least let him look at the carcasses. Also you mentioned something about 425 Amps on a S&C 400 fuseholder... I'm thinking that might mean it is about 25 amps above its continuous rating? I'm not sure if that's correct, but even if it is, I wouldn't think that alone would explain your problems.

This message has been edited. Last edited by: electricpete,

Bill there is another alternative to monitoring critical items/points with IR without using a port of opening the cabinet. There is a system out there that uses permanently mounted IR sensors, which report back to a modbus data module for 24/7/365 monitoring for trending and alarming.

Even if the IR port is UL listed, it does not mean it keeps the arc flash rating of the cabinet when installed.
From reports I've read, mechanical intrusion into the cabinet while it's off, can disturb the "bad" connections, or something can be left in a "bad" spot, and when the cabinets are re-energized, it happens. The arc flash can also happen when a connection becomes critical. I've had customers tell me of switchgear exploding while in service.

The whole key to heading off catastrophic failures is in finding the problems while they are small. And a small problem may take a couple of months or a few weeks to become a bigger problem.
In much of the metal clad switchgear, doing a complete thermal scan can only be done with the panel opened up and shields removed. With a permanently mounted IR based constant thermal monitoring system, the sensors can be installed within the enclosures of the cabinet, and then provide the 24/7/365 monitoring / alarming with out ever opening the panels again.

I should clarify... I'm using the H.VIR 75 Comet windows and not a "port" to see through.

As for the 425 amps... that was calculated based on our downstream metering of approx. 7700 line amps at 670 volts from the transformer going to the furnace. They probably run real near 400 when at full power, and they occasionally spike much higher (estimated 600+ amps on the 12kV side)when they get a chunk of iron that goes through the loop and the furnace "barks".

Since posting this question, I took another image of the fuse when it was at 491 deg. F. and I'm sure it's actual temperature is much higher than that, inside where the contact is being made. (See attachment)

PLUS, it was finally taken out of service a couple days ago and I have some more pictures of the failed components that I will post tomorrow.

Furn_7_12kV_fuse.doc (254 Kb, 79 downloads)

Here's a picture of the failed fuse. The fuse itself never did "open", but the fiberglass body came apart. The electricians cold smell it, and saw that the end was ready to burn off it was so hot, so they changed it.

Looking at the picture; the spring and fuse (bottom) go inside of the fiberglass body (top). The spring is to put tension on the fuse, when the fuse "opens" the spring pulls it back and indicates (viewed through the fiberglass body)that it is open. There is also a large "snuffler" that is not pictured, that screws to the bottom (left side of this picture) to soak up the arc when the fuse opens.

It's obvious which end was hot.

Hey Bill,

You forgot the picture.
Enquiring minds need to see

Dave

Sorry. The first attachment was WAY too big. I'll try this again.

Yikes!! This pictures still really big, I'll try to make the rest smaller.

This message has been edited. Last edited by: Bill Schmitt,

Here's one of the copper strand that runs inside the spring.

In case you're wondering, the brass cap that is threaded on the top of this fuse was welded to the brass body from the heat, so I had to saw it off, right above where the copper bonds to the brass.

This message has been edited. Last edited by: Bill Schmitt,

And here's a close up of the top end, right side of the brass, where the connection is made with the clip.

Like ePete said earlier, you can see evidence of the so-called microscopic "welding".

quote:

If you have repeating failures, you of course want to look at tension adjustments and lubrication.

I have to agree with E'Pete here. It appears there is loss of contact in part of the connection. Has the spring clip been changed out? It's possible you need new ones on there, especially if the clips have the same type surface distress the fuse head is showing. Perhaps the contact surface area is becoming less and less on the clips each time the anomaly occurs.

Let us know what you find.

Dave

Actually, we replace the fuse clips quite often. Usually they are the first thing changed, because they are cheaper than the fuse holder.

Here's a picture of one that was near failure from a year ago. It's not the actual one from this failure, but is identical.

I find it rather strange that it's almost always the top of the fuse and top fuse clip that get hot, and break down, and fail... rarely the bottom one, even though the connection is identical, exact same clip. I think I've seen 1 in 12 failures on the bottom clip. I know, heat rises... but there has to be another reason.

Oh yeah, we are looking at bolt-in style fuses, but apparently they are having trouble finding something that will retro fit our switchgear. The manufacturer wants use to buy complete new, but our company doesn't want to spend the money. Apparently, they'd rather keep buying fuse clips and fuse holders.

Great photos. Thanks for helping to close the loop on this story. One of the issues with clips, or any spring metal, is that it begins to anneal after a month or so at relatively minor temperatures (100C); at that point it is no longer a "clip" in the way it needs to be and, as such, this kind of heating and damage can begin to occur.

I still strongly suggest that the primary benefit of thermography is to validate asset health and/or locate components that are out of spec. "All" that is required is the right conditions, a camera and an ability to use it correctly. To trend or monitor, though it can be done, is much more difficult as surface temperature has many drivers other than the severity of the problem.

Nice work Bill!

So is the moral of the story to buy better clips or to replace the, periodically as IR indicates?

Is there a better scheme than these types of clips?

Is the manufacturer aware of this issue?

Imagine the plants not using Infrared trying to get to the bottom of this problem.

Terry O

Bill,

Next time you replace a clip and fuse, try to conduct a contact resistance test before initial energization. That way, you will have first-hand information whether the same problem will occur in the future.

Also, basing on the picture of the clip assembly, the spring bar is most probably shunting some of the current load resulting to a gradual change of its characteristic. A good design should not allow current to flow through the springs so that the spring characteristics is not altered.

Regards & God bless!