Fused vs. non-fused disconnect

[petersonra]

I'm gonna disagree with that statement, and I use the following chart as evidence:

I also refer back to 240.86(A) which allows an engineer to permit the use of fuses for existing installations where the fault current exceeds the rating of a breaker. A fuse can be used to permit such an installation provided the fuse operates fast enough to avoid the dynamic impedance of an opening breaker.

My understanding is the only way this can be determined is by testing, and not by using some chart.

 

[mayanees]

My understanding is the only way this can be determined is by testing, and not by using some chart.

My understanding is that the let-through curves were superseded by the let-through charts which were developed through testing. If I'm charged with pass/fail determination for a seemingly underrated device in an existing installation, then I use these tables to pass the device. I also look at the breaker and fuse operating characteristics on a TCC for verification that the fuse is cleared before the breaker starts to open.

 

[petersonra]

My understanding is that the let-through curves were superseded by the let-through charts which were developed through testing. If I'm charged with pass/fail determination for a seemingly underrated device in an existing installation, then I use these tables to pass the device. I also look at the breaker and fuse operating characteristics on a TCC for verification that the fuse is cleared before the breaker starts to open.

I think you might want to take a closer look at your understanding. Do any of the software packages that calculate available SCC use this method?

 

[mayanees]

I think you might want to take a closer look at your understanding. Do any of the software packages that calculate available SCC use this method?

This is what I go by:
NEC 240.86 Series Ratings.
Where a circuit breaker is used on a circuit having an available fault current higher than the marked interrupting rating by being connected on the load side of an approved overcurrent protective device having a higher rating, the circuit breaker shall meet the requirements specified in 240.86(A) or (B), and (C).
(A) Selected Under Engineering Supervision in Existing Installations.
The series rated combination devices shall be selected by a licensed professional engineer engaged primarily in the design or maintenance of electrical installations. The selection shall be documented and stamped by the professional engineer. This documentation shall be available to those authorized to design, install, inspect, maintain, and operate the system. This series combination rating, including identification of the upstream device, shall be field marked on the end use equipment.
For calculated applications, the engineer shall ensure that the downstream circuit breaker(s) that are part of the series combination remain passive during the interruption period of the line side fully rated, current-limiting device.

 

[petersonra]

This is what I go by:
NEC 240.86 Series Ratings.
Where a circuit breaker is used on a circuit having an available fault current higher than the marked interrupting rating by being connected on the load side of an approved overcurrent protective device having a higher rating, the circuit breaker shall meet the requirements specified in 240.86(A) or (B), and (C).
(A) Selected Under Engineering Supervision in Existing Installations.
The series rated combination devices shall be selected by a licensed professional engineer engaged primarily in the design or maintenance of electrical installations. The selection shall be documented and stamped by the professional engineer. This documentation shall be available to those authorized to design, install, inspect, maintain, and operate the system. This series combination rating, including identification of the upstream device, shall be field marked on the end use equipment.
For calculated applications, the engineer shall ensure that the downstream circuit breaker(s) that are part of the series combination remain passive during the interruption period of the line side fully rated, current-limiting device.

As I understand it, UL has determined that the only way to do this is by testing.

 

[mayanees]

As I understand it, UL has determined that the only way to do this is by testing.

That's why there's NEC 240.86 Series Ratings.
My first attempt to pass the equipment in a short-circuit evaluation is with 240.86(B) (B) Tested Combinations.
The combination of line-side overcurrent device and load-side circuit breaker(s) is tested and marked on the end use equipment, such as switchboards and panelboards.
But if I can't find that it's been tested, then I scrutinize it using 240.86(A) Selected Under Engineering Supervision in Existing Installations.
When we do a Power Study/Arc Flash analysis the equipment is evaluated for short-circuit withstand and if the software fails anything then it takes allot longer to do the analysis because we exhaust all options to pass something - because the software only reports the basics.

 

[petersonra]

That's why there's NEC 240.86 Series Ratings.
My first attempt to pass the equipment in a short-circuit evaluation is with 240.86(B) (B) Tested Combinations.
The combination of line-side overcurrent device and load-side circuit breaker(s) is tested and marked on the end use equipment, such as switchboards and panelboards.
But if I can't find that it's been tested, then I scrutinize it using 240.86(A) Selected Under Engineering Supervision in Existing Installations.
When we do a Power Study/Arc Flash analysis the equipment is evaluated for short-circuit withstand and if the software fails anything then it takes allot longer to do the analysis because we exhaust all options to pass something - because the software only reports the basics.

I don't believe you can get a series rating except by testing. At least so UL claims. Maybe you have some way of doing it that UL doesn't know about.

 

[mayanees]

I don't believe you can get a series rating except by testing. At least so UL claims. Maybe you have some way of doing it that UL doesn't know about.

No, that's not it. UL isn't paid to test all of the possible combinations. It's that I'm an engineer who practices engineering and is paid to make interpretations based on available codes and engineering principles. It's that I'm not gonna make an owner pay money to replace equipment if there's a chance that I can make determination that the application is safe. I've cited the reason a number of times, and you still make a nasty remark about me knowing more than UL so it's obvious that I'm not gonna change your mind, not that I even care to.
I present the information for the good of the forum so other folks can be aware of the existence of 240.86 and how an engineer looks at the application.

 

[petersonra]

If this was a legitimate means of dealing with the problem, don't you think the fuse manufacturers would be pushing it like hell rather than paying UL to test series combinations?

 

[electrofelon]

You cannot use the let through current of a fuse in calculating the available short circuit current.

I disagree. I see nothing in the code that prohibits this, especially if you look at the definition of current limiting fuse in article 240. Of course if we are talking about two OCPD's in series, the NEC says the combination must be tested or engineer approved. If we are talking about a downstream passive device such as switch, I don't see why the ASCC can't be determined by the uostream fuse characteristics. If the NEC doesn't want this, perhaps they should provide a good definition of available short circuit current and which methods they allow and which methods they don't. As it is, I am seeing nothing in the NEC that points to this being unacceptable again especially considering the definition they provide of current limiting fuse.

 

[nhee2]

If this was a legitimate means of dealing with the problem, don't you think the fuse manufacturers would be pushing it like hell rather than paying UL to test series combinations?

My understanding / experience both in project work and in forums like this one - the method that Mayanees describes is acceptable and code compliant, but not all that frequent, due to the unwillingness of PEs to seal/stamp as mentioned in post #20. But - maybe it is more frequently applied in older installations.

 

[petersonra]

My understanding / experience both in project work and in forums like this one - the method that Mayanees describes is acceptable and code compliant, but not all that frequent, due to the unwillingness of PEs to seal/stamp as mentioned in post #20. But - maybe it is more frequently applied in older installations.

it is only allowed at all in existing installations. it cannot be used in new work.

 

[Jraef]

My understanding / experience both in project work and in forums like this one - the method that Mayanees describes is acceptable and code compliant, but not all that frequent, due to the unwillingness of PEs to seal/stamp as mentioned in post #20. But - maybe it is more frequently applied in older installations.

That’s how I have experienced it. I first came across this issue when a client ordered a bunch of MCCs built to 42kA and after the transformers arrived and were installed under a separate contract, we calculated the fault current at 44kA. Most of the buckets in the MCC were all 65kAIC or greater so it was really only the MCC bus bracing that was the issue, but it was an insurmountable one from the standpoint of changing that in the field. We looked for a PE to sign off on putting CL fuses ahead of each MCC, couldn’t find one that would sign off on that. This has come up again msny times when people allow OEMs to send control panels that have only the “courtesy” SCCR of 5kA that UL allows without testing, which makes it almost impossible to install in most industrial settings. People try to just use CL fuses and discover that it needs a PE’s stamp, then can’t find a PE willing to do that. So even if it is technically allowed for new (depending on your interpretation) installations or only applies to existing, from a practical standpoint I’ve never seen it actually happen.

 

[topgone]

That’s how I have experienced it. I first came across this issue when a client ordered a bunch of MCCs built to 42kA and after the transformers arrived and were installed under a separate contract, we calculated the fault current at 44kA. Most of the buckets in the MCC were all 65kAIC or greater so it was really only the MCC bus bracing that was the issue, but it was an insurmountable one from the standpoint of changing that in the field. We looked for a PE to sign off on putting CL fuses ahead of each MCC, couldn’t find one that would sign off on that. This has come up again msny times when people allow OEMs to send control panels that have only the “courtesy” SCCR of 5kA that UL allows without testing, which makes it almost impossible to install in most industrial settings. People try to just use CL fuses and discover that it needs a PE’s stamp, then can’t find a PE willing to do that. So even if it is technically allowed for new (depending on your interpretation) installations or only applies to existing, from a practical standpoint I’ve never seen it actually happen.

I was in that very situation when the company wanted me sign and seal something done by another and only to see the ratings of the starters and the bus bracings well below the available fault current! I did a lot of convincing them that things they wanted is just not possible. We have to let the vendor change all those that they brought with them. It's important that the project was well specified and we had the advantage at that time.

 

[petersonra]

Ironically, many of the small control panels that have the so-called courtesy short circuit current rating could have a 65K rating if the control panel manufacturer would just put a label on the control panel saying it would be 65K with (for instance) 100 amp max type j fuses upstream.

As often as not what I have seen of these type of panels the issue is that they put a non-fuse disconnect switch in to save money on fuses. UL 508a does not require any kind of overcurrent protection device in the main and it's a relatively expensive thing to add. But The standard also allows such fuses to be located outside of the control panel and supplied externally if it is marked that way.

 

[Jraef]

Ironically, many of the small control panels that have the so-called courtesy short circuit current rating could have a 65K rating if the control panel manufacturer would just put a label on the control panel saying it would be 65K with (for instance) 100 amp max type j fuses upstream.

As often as not what I have seen of these type of panels the issue is that they put a non-fuse disconnect switch in to save money on fuses. UL 508a does not require any kind of overcurrent protection device in the main and it's a relatively expensive thing to add. But The standard also allows such fuses to be located outside of the control panel and supplied externally if it is marked that way.

Yes, but only IF the panel has been fully vetted by the builder to have a higher SCCR when used behind those specific fuses, which would mean the label on the panel needs to show that higher SCCR. If they get lazy and just put on the 5kA label, that can’t be changed in the field.

And I totally agree that it is not at all difficult to attain a higher SCCR on a panel, most of the major component mfrs now provide very easy to use charts or spreadsheets laying out what works with what to attain it. It just takes a couple of extra steps to document it and it might mean not selecting the cheapest version of every part regardless of manufacturer. Yet almost every week, I see panels arrive (usually OEM provided) with 5kA labels on them and end users asking how they can deal with that. Nobody likes the answer to that question.

 

[GoldDigger]

As I understand it, UL has determined that the only way to do this is by testing.

As I see it, you are correct if the device that has the low current rating includes any sort of OCPD. But if the device is passive (for example a simple unfused switch) then there are no downstream OCPDs that are part of the series combination. The code may still require a PE stamp, but I do not see the specific requirement for testing in this case.
I think that the difference between AIC and SCCR comes into play in this situation. A passive component like a switch does not have an AIC, But a panel, intended to contain OCPDs will have both an AIC and an SCCR which is conditioned on the breakers installed in that panel.
(I may not be using the language perfectly correctly, but I still think I have a point with respect to the unfused disconnect.

 

[petersonra]

As I see it, you are correct if the device that has the low current rating includes any sort of OCPD. But if the device is passive (for example a simple unfused switch) then there are no downstream OCPDs that are part of the series combination. The code may still require a PE stamp, but I do not see the specific requirement for testing in this case.
I think that the difference between AIC and SCCR comes into play in this situation. A passive component like a switch does not have an AIC, But a panel, intended to contain OCPDs will have both an AIC and an SCCR which is conditioned on the breakers installed in that panel.
(I may not be using the language perfectly correctly, but I still think I have a point with respect to the unfused disconnect.

I thought we were talking about circuit breakers.

I am inclined to agree that there ought to be some procedure to allow things other than OCPDs to gain a higher SCCR by using a current limiting fuse.

In fact, UL508a has such a process, but it is a lot less useful than you might think. It basically mostly works on 30 A or less fuses. Above 30 Amps I have not found the provision to be real helpful because the let through current is too high to do any good.