Current limiting fuses to protect equipment

[zbberry]

I found many threads about current limiting fuses, but not for this specific case. Most of what I can find on the current limiting fuse topic involves trying to protect panels or series ratings with CBs.

In our case, we have a Chiller with SCCR of 5000. and an available fault current at 9800, it has been suggested by more than one person/company, that we can solve with with a current limiting fuse (they showed me it would work using the up, over, down method with the fuse they picked). Question 1) Can you utilize current limiting fuses to lower the available fault current at the equipment? Question 2) If we add a CB at the chiller that is series rated (at 10k or higher)with an upstream fuse at the panel, does the chiller see the advantage of

I know we can lower the available fault current with a lot of extra cable (not feasible here) or by adding a 480v-480v transformer.
Question 3)If we can't use the solutions of questions 1 and 2, which have already been brought up in meetings, how do I explain to the average person (owner/maintenance person) why we can't use those solutions and justify the added cost? Is there a clear code reference or other documentation I can point them to?

 

[dkidd]

There is a good article at http://ecmweb.com/design/sccr-liability-trap

An option is to add an isolation transformer to increase the impedance of the supply.

The answer to your first two questions is no.

 

[d0nut]

If your chiller isn't too old, you could check with the manufacturer to see what would be required to raise the SCCR of the chiller. It may be as simple as getting the factory to swap out a couple of components and put on a new label.

 

[mayanees]

I found many threads about current limiting fuses, but not for this specific case. Most of what I can find on the current limiting fuse topic involves trying to protect panels or series ratings with CBs.

In our case, we have a Chiller with SCCR of 5000. and an available fault current at 9800, it has been suggested by more than one person/company, that we can solve with with a current limiting fuse (they showed me it would work using the up, over, down method with the fuse they picked). Question 1) Can you utilize current limiting fuses to lower the available fault current at the equipment? Question 2) If we add a CB at the chiller that is series rated (at 10k or higher)with an upstream fuse at the panel, does the chiller see the advantage of

I know we can lower the available fault current with a lot of extra cable (not feasible here) or by adding a 480v-480v transformer.
Question 3)If we can't use the solutions of questions 1 and 2, which have already been brought up in meetings, how do I explain to the average person (owner/maintenance person) why we can't use those solutions and justify the added cost? Is there a clear code reference or other documentation I can point them to?

Your need is to comply with NEC 240.86 where there can be a listed combination, or for existing systems where a PE certifies that the equipment specified meets the fault current requirements of the system. Application of a current-limiting fuse requires that the PE spec the exact fuses and you label that, etc. But the problem becomes that the downstream device could start opening before the current-limiting (CL) fuse, so the PE must interpret the situation. The up-over-down method you mentioned has been replaced with current let-through tables, and the CL fuse must be operating in the current limiting region of its curve. But you can see from the ecmweb article that was provided it's best to do it from the beginning, but perhaps you are where you are. The easiest thing now is if you can find a fuse or breaker that series-rates with whatever device is in your 5k panel. You can't just replace the main breaker in the panel with a 10kA-rated breaker because it breaks the UL listing on your new equipment.
Good luck with it. Are you a PE? Perhaps this gets your company to pay for you getting your PE license. I think it's doable if you can find the right fuse, but it takes allot of scrutiny of the devices. The downstream element/breaker becomes a dynamic impedance, meaning it's no longer a short-circuit as it attempts to open so the CL fuse operation has to be very fast. Good luck with it.

 

[mayanees]

I looked back and reread what you asked and tried to edit, but no luck... so,
It sounds like you have no CB at the new equipment. You need to be comfortable that if there's a fault in the 5kA-rated gear, that it will stay together while the fuse reduces/clips the current to less than 5kA. I think you could find a fuse that limits the let-through to 5kA or less. But that depends on where it's at in your electrical system relative to the available fault current. Is the system modelled in SKM or some other package? That's what I'd do is model the system back to the Utility source and look for the right CL fuse.
In direct response to your ?s
1) yes, depending on the fuse and the available fault current
2) chiller advantage? no
3) the code reference is 110.9 or 10 which speaks to short-circuit requirements.
The cost associated with doing it this way will probably be in the time spent on the power systems analysis for determination of a CL fuse that works while under Utility operation, or under generator operation. That fuse will operate in less than a quarter cycle so it won't be cheap. And of course it's a one-shot operation.
Good luck

 

[paulengr]

Sounds like your “protection” is a cheap UL Tyoe F device or a supplemental style circuit breaker., or some part is such as a very cheap VFD or soft start that they put in without any short circuit protection and left it up to the buyer. What happens is that in the event of a short it can’t handle the current and either blows apart while attempting to interrupt or welds shut and will never trip. The really ridiculous part is that it’s probably less than $10 price between the “expensive” and “cheap” part. We’ve been seeing one particular chiller manufacturer paying huge amounts of money to have motors pulled, repaired to spec, and reinstalled, on site technician time to swap out parts, you name it. Like when a large hog processing plant buys several 1000+ HP 4160 V chillers only to have them blowing out bearings in 6 weeks because the motor manufacturer put the wrong bearing in the motors. We were happy for the warranty work but that’s crazy.

It is physically trivially easy and quick to go into the panel and swap out the weak components for the right ones. Or if it’s a drive add semiconductor fuses or a line reactor. But right now it’s a Listed assembly so modifications are not allowed unless you “rebuild it” with a new Listing or get PE sign off, or bring your external available short circuit current down. That may be possible with current limiting fuses or breakers but going from 10 kA to 5 kA is probably a bridge too far. That is where line reactors come in. An isolation transformer does the same thing but unless you are changing voltages or need a neutral, it serves no purpose. Line reactors are just 3 coils instead of 6 (cheaper). Have whoever calculated 9800 A plug in 3% line reactors and see where that gets you. Just have them run a motor starting time calculation too so you don’t get in trouble. But the upgrade is so cheap I’d check with the chiller manufacturer. The only problem you will have with them is if they contracted out the panel, you will quickly get push back since their vendor will try to hit them up for some huge engineering bill. If it’s in house they should respond fast.

Keep in mind there is a downside to impedance regardless of how you add it. It will affect starting current on motors so it might cause you starting issues.

The chiller manufacturer should be ashamed of themselves. 5 kA is the DEFAULT MINiMUM for some components (others are 10 kA). Their panel builder clearly does not understand power distribution and dumped it on you. I would never sell you anything that chintzy.

 

[petersonra]

The HVAC industry as a whole is notoriously CHEAP.

They also do not want to change anything because that costs money. I have seen HVAC panels labeled 5kA SCCR that could be upgraded to 65 kA with virtually no material or labor cost increase but it would require time to update maybe a dozen series of drawings and they just do not want to do it.

 

[Jraef]

If your chiller isn't too old, you could check with the manufacturer to see what would be required to raise the SCCR of the chiller. It may be as simple as getting the factory to swap out a couple of components and put on a new label.

Field labeling is not allowed.

The issue here is that you cannot change the RATING of the equipment in the field with the CL fuses because that is what's referred to as providing a "series rating", and the only acceptable ways to do that are outlined in 240.86 (A) and (B). (A) says it can be done by a registered PE, (B) says it can be done via TESTING, which is not a simplistic task, its some that only equipment mfrs can afford when they can amortize that cost across sales of high volumes of components. The PE issue is theoretically possible, but the times I have tried to find a PE willing to risk his professional license on a stupid little panel have all been unsuccessful, so I have no idea what they would charge for that if you find one willing to do it.

If you are stuck, you could remove the panel and send it to a local listed panel shop to be redesigned (and possibly reworked) to meet the criteria of UL 508A Supplement SB, but it probably will require replacement of components with those that have ALREADY been series listed by the manufacturers at higher SCCR levels. All in all this might be the cheapest.

The sad part is, the supplier of the equipment COULD have done this before shipping it, but instead, they chose to be cheap and lazy so they took the "courtesy" rating of 5kA that UL allows for untested combinations. All it takes to attain a higher rating is to pick out components that have already been tested together at higher levels. But that means having to pick the SAME supplier for the separate parts, so they cannot negotiate down the cost as well as if they chose disparate untested components. So they maybe saved a few dollars on parts and dumped a load of hurt onto you.

If you get a chance, inform your buyers that from here on out, they need to require an SCCR that meets or exceeds the AFC!

It's not hard...

 

[winnie]

Isn't a key question weather or not the panel at the chiller contains OCPD? Does the absence or presence of OCPD in the chiller panel change the viability of current limiting fuses?

Jon

 

[ptonsparky]

Yup, I can see it. _____ system all but installed. Mechanically everything in place. EC gets the call.
Ok, have you called the POCO?
Last year, but they haven’t showed yet.
Where is this job?
.....
and I’m supposed to tell them they need an entirely different control panel on something I had no involvement with.

Yup ”It’s not hard.”

 

[petersonra]

Isn't a key question weather or not the panel at the chiller contains OCPD? Does the absence or presence of OCPD in the chiller panel change the viability of current limiting fuses?

Jon

No. You cannot change the rating of a control panel that is listed in the field. you can send it back to the manufacturer or to another manufacturer to be reworked and relabeled but you can't do it yourself. The only thing you can do is reduce the short circuit current that is available at the terminals of the panel. The real problem is stupid people who don't tell the manufacturer what they need upfront and then blame the manufacturer for selling them the cheapest possible thing.

 

[petersonra]

Incidentally for a small control panel that has maybe one motor you can actually have current limiting fuses in the disconnecting switch that don't actually count for reducing the let-through current inside the control panel because it's not in the feeder circuit. I have actually put two sets of fuses in series solely to create a feeder circuit because the only place you can take credit for the current limiting fuses is in a feeder circuit. If it's in a branch circuit it doesn't count.

This problem has been around a long time and anybody competent knows about it in the electrical business. this is like going to a car dealer and saying I want to buy a car I don't care what it is just get me the cheapest possible price. Then after they deliver a white car you decide oh I need a green one and somehow this is the car dealers fault.

 

[winnie]

No. You cannot change the rating of a control panel that is listed in the field. you can send it back to the manufacturer or to another manufacturer to be reworked and relabeled but you can't do it yourself. The only thing you can do is reduce the short circuit current that is available at the terminals of the panel. The real problem is stupid people who don't tell the manufacturer what they need upfront and then blame the manufacturer for selling them the cheapest possible thing.

Let me rephrase my question: Does the use of so called 'current limiting fuses' reduce the short circuit current available at the panel?

I know that the answer is 'NO!' if the panel contains devices such as OCPD which will attempt to open during a fault, because the 'dynamic impedance' of such devices interacts with the 'dynamic impedance' of the 'current limiting fuse'. When a current limiting fuse is in series with an untested OCPD, it might actually not limit current correctly, or the other OCPD might see more fault stresses than expected given the 'fuse limited current'. This is the whole reason that series rating is required.

But if the panel does not contain OCPD, can a current limiting fuse with the 'up over down' method be used to limit the fault current available at the panel? Or is this old technique no longer accepted?

Thanks
Jon

 

[petersonra]

Let me rephrase my question: Does the use of so called 'current limiting fuses' reduce the short circuit current available at the panel?

I know that the answer is 'NO!' if the panel contains devices such as OCPD which will attempt to open during a fault, because the 'dynamic impedance' of such devices interacts with the 'dynamic impedance' of the 'current limiting fuse'. When a current limiting fuse is in series with an untested OCPD, it might actually not limit current correctly, or the other OCPD might see more fault stresses than expected given the 'fuse limited current'. This is the whole reason that series rating is required.

But if the panel does not contain OCPD, can a current limiting fuse with the 'up over down' method be used to limit the fault current available at the panel? Or is this old technique no longer accepted?

Thanks
Jon

This technique was never really accepted.

Incidentally, there's no requirement that the overcurrent protection device be in the control panel. But to take credit for current limiting it has to be in the feeder circuit and if you don't have overcurrent protection in the control panel it's not a feeder circuit so UL 508a will not allow you to take credit for it.

why it matters whether the branch circuit protection is in the feeder circuit or the branch circuit escapes me

 

[Jraef]

Let me rephrase my question: Does the use of so called 'current limiting fuses' reduce the short circuit current available at the panel?

I know that the answer is 'NO!' if the panel contains devices such as OCPD which will attempt to open during a fault, because the 'dynamic impedance' of such devices interacts with the 'dynamic impedance' of the 'current limiting fuse'. When a current limiting fuse is in series with an untested OCPD, it might actually not limit current correctly, or the other OCPD might see more fault stresses than expected given the 'fuse limited current'. This is the whole reason that series rating is required.

But if the panel does not contain OCPD, can a current limiting fuse with the 'up over down' method be used to limit the fault current available at the panel? Or is this old technique no longer accepted?

Thanks
Jon

Sort of.
IF the panel maker used power circuit components that have ALREADY been series listed with a particular fuse, the panel supplier can put a label on the panel that says something to the effect of "Suitable for use on a system capable of xx kA Fault Current when protected by {yyy -yyyy} fuses, where the yy-yyyy is the very SPECIFIC fuse that the components are listed with (or an exact equivalent of another mfr in most cases). But, again, the components inside must have been TESTED AND LISTED behind those specific fuses by the mfr of the components. If that has not been done and that label applied by the panel builder, you cannot do it in the field after the panel has already been listed at 5kA.

 

[electrofelon]

Let me rephrase my question: Does the use of so called 'current limiting fuses' reduce the short circuit current available at the panel?

I know that the answer is 'NO!' if the panel contains devices such as OCPD which will attempt to open during a fault, because the 'dynamic impedance' of such devices interacts with the 'dynamic impedance' of the 'current limiting fuse'. When a current limiting fuse is in series with an untested OCPD, it might actually not limit current correctly, or the other OCPD might see more fault stresses than expected given the 'fuse limited current'. This is the whole reason that series rating is required.

But if the panel does not contain OCPD, can a current limiting fuse with the 'up over down' method be used to limit the fault current available at the panel? Or is this old technique no longer accepted?

Thanks
Jon

We recently had a long thread on this. My claim was that the CL fuse changes/reduces the available SCC just like a transformer or conductor would - its not about series rating or tested combination. Everyone else in the thread pretty much said you can't do that despite the fuse manufacturer saying you can and providing the charts, and the NEC definition of CL fuse seemingly allowing it. I respect the knowledge of all those on here and they certainly know more than me in this topic, but I still haven't really gotten a satisfactory explanation of why a CL fuse can't be used this way.

 

[jim dungar]

...I still haven't really gotten a satisfactory explanation of why a CL fuse can't be used this way.

The fuse CL curves are based on a specific X/R ratio. They may need to be modified based on the actual circuit where they are installed. This could imply that their CL effect is not constant.

 

[paulengr]

Field labeling is not allowed.

The issue here is that you cannot change the RATING of the equipment in the field with the CL fuses because that is what's referred to as providing a "series rating", and the only acceptable ways to do that are outlined in 240.86 (A) and (B). (A) says it can be done by a registered PE, (B) says it can be done via TESTING, which is not a simplistic task, its some that only equipment mfrs can afford when they can amortize that cost across sales of high volumes of components. The PE issue is theoretically possible, but the times I have tried to find a PE willing to risk his professional license on a stupid little panel have all been unsuccessful, so I have no idea what they would charge for that if you find one willing to do it.

If you are stuck, you could remove the panel and send it to a local listed panel shop to be redesigned (and possibly reworked) to meet the criteria of UL 508A Supplement SB, but it probably will require replacement of components with those that have ALREADY been series listed by the manufacturers at higher SCCR levels. All in all this might be the cheapest.

The sad part is, the supplier of the equipment COULD have done this before shipping it, but instead, they chose to be cheap and lazy so they took the "courtesy" rating of 5kA that UL allows for untested combinations. All it takes to attain a higher rating is to pick out components that have already been tested together at higher levels. But that means having to pick the SAME supplier for the separate parts, so they cannot negotiate down the cost as well as if they chose disparate untested components. So they maybe saved a few dollars on parts and dumped a load of hurt onto you.

If you get a chance, inform your buyers that from here on out, they need to require an SCCR that meets or exceeds the AFC!

It's not hard...

Are you kidding me? Every component manufacturer provides dozens of pages of series ratings charts. The chiller manufacturer won’t buy the breaker or whatever the device in the panel is, will.

What you are suggesting is so crazy the manufacturer can’t even build a panel to your specs.

OP: open the door and look at what device has such a crazy low SCCR. It’s probably either a breaker or similar device and will be printed on it.

 

[topgone]

@paulengr,
As far as I understand, what @Jraef meant to say for the poster to see another way of doing things in his future projects! Or you could force the poster to follow your way; but anything that helps should be welcomed! Customers are king nowadays. That said, whatever the specifications manufacturers get from customers will have to be followed, else delivered products will be rejected. That's how I see it, IMHO.

 

[drktmplr12]

I used to work for a panel builder making lift station panels and they are pretty competitive. Anything to get the price down, even if it's $100. It isn't difficult for a 508 shop to build a 65kaic panel, but it does cost a bit more and the panel might be a bit bigger. The cheaper shops would prefer to load it up with the cheapest parts available and make it up later in change orders.

This is easily combated by:

inform your buyers that from here on out, they need to require an SCCR that meets or exceeds the AFC!

It certainly isn't hard if done before the panel is built, but it is quite expensive to upgrade an existing install to 65kaic. UL could also do a field inspection to "verify" listing is valid, but be prepared to pay $$$$ and find other things wrong that they will want to see fixed.