Solution for Overdutied Industrial Control Panels

[EC Dan]

Hi all,

I’ve inherited a set of 480 VAC industrial control panels that I’ve determined to be overdutied based on the previous engineer’s available fault current analysis (analysis performed after panel design and installation). For some reason as part of that analysis he did not also identify that the panels were overdutied as a result of the AFC. In these cases, the AFC is about 9.9 kA and the panels have elements that render it with the base 5 kA SCCR. These panels contain a number of motor control circuits, each with (in order from line to load) a manual motor starter, door-mounted disconnect for LOTO, PowerFlex VFD, Phoenix Contact 3-level motor terminal block, and motor. The door-mounted disconnect and VFD are the limiting factors. I’ve read every thread I can get my hands on in this forum about SCCR and have learned a tremendous amount, so I believe I can see what my options are, and I’d like some advice. As I can see it, I can either attempt to reduce the AFC to below 5 kA with an isolation transformer or replace the components in the panel that do not give me series combinations to get me above 10 kA. Both options have pros and cons:

• Isolation transformer: this would be the easiest to implement, in that it’s a single component to add, and would give me benefits of protecting the line from any harmonics from the VFDs, however I believe I’d be cutting it too close to the non-instantaneous trip region of the circuit breaker. The CB is 100 A and is comfortably in the range of instant trip with an arcing current of 6.5 kA (based on 9.9 kA bolted fault current), however unless I can very narrowly target 4-5 kA bolted fault current (3-4 kA arcing current), I may be too close for comfort to the non-instantaneous trip portion of the TCC, which would vastly increase my incident energy.
• Replace components: both the VFD and door-mounted disconnect have improved SCCR with class CC/J fuses, however there is not room for both the MMS and fuse holders, so I would need something like a Bussmann fuse block with disconnect switch to replace the MMS entirely and maintain the ability to disconnect individual motor circuits. I lose the thermal trip part of the MMS in that case, but as each motor is on a VFD I think the VFD would handle that fault protection. I wanted to avoid fuses but as I’ve learned it’s very difficult to achieve higher than 5 kA without fuses and the series combinations they enable. I’m also not seeing very many options for fuse holders with disconnect switches (at least of the DIN-rail mounted variety for use in a panel), which leads me to believe they are not widely used, which makes me second guess their application here.

I’d appreciate any feedback about my evaluation of options and whether I’m overlooking anything or options I haven’t considered. I'd be happy to provide any clarifying details and information.

 

[zbang]

There are times where adding length to the feeder adds enough impedance to reduce the AFC, that's an option to consider (check the actual numbers).

If you don't go the transformer route- remove the manual starters, set the VFD's as overload protection, and install the appropriate fuses. Assuming a single feeder to the panel, one set of fuses should cover the AFC for the the entire panel, so the door-mounted disconnects would still be usable for LOTO. (I'd consider removing the manual starters anyway, they're another point of failure.)
Remember to handle single-phasing when a fuse blows.
What's controlling the VFDs?

Most difficult thing might be convincing the moneybags to pay for the changes.

 

[EC Dan]

There are times where adding length to the feeder adds enough impedance to reduce the AFC, that's an option to consider (check the actual numbers).

If you don't go the transformer route- remove the manual starters, set the VFD's as overload protection, and install the appropriate fuses. Assuming a single feeder to the panel, one set of fuses should cover the AFC for the the entire panel, so the door-mounted disconnects would still be usable for LOTO. (I'd consider removing the manual starters anyway, they're another point of failure.)
Remember to handle single-phasing when a fuse blows.
What's controlling the VFDs?

Most difficult thing might be convincing the moneybags to pay for the changes.

I'll look at the extra conductor option. Initially I had dismissed it since I don't have much space to store the extra runs (the control panels are typically very close to the distribution panels), however it could be that the foot print is similar to that of an isolation transformer of the appropriate size.

To your comment on using one set of fuses, I've been wondering about that. When an OCPD is series tested with a component to increases it's SCCR, I had been assuming the increased rating only holds for a single OCPD protecting a single component. Are you saying I can use a single set of fuses to provide the SCCR increase to every downstream device that is series tested with that class of fuse (assuming the fuse is sized appropriately for the collection of components)?

The VFDs are all controlled by a PLC via Ethernet IP.

 

[jim dungar]

Are you saying I can use a single set of fuses to provide the SCCR increase to every downstream device that is series tested with that class of fuse (assuming the fuse is sized appropriately for the collection of components)?

Maybe.
There are different rules for determining series rating of over current protective devices in series than there are for groups of components internal to industrial control panels.

 

[petersonra]

Keep in mind that if you can use CL fuses to fix your problem, they do not have to be physically contained in the enclosure. They can be in the feeder going to the box if the numbers work out.

Without knowing which PF drive it is, it is hard to help you with that but many of the AB drives match up with an AB MMS with a combined rating much higher than the default.

 

[EC Dan]

Keep in mind that if you can use CL fuses to fix your problem, they do not have to be physically contained in the enclosure. They can be in the feeder going to the box if the numbers work out.

Without knowing which PF drive it is, it is hard to help you with that but many of the AB drives match up with an AB MMS with a combined rating much higher than the default.

Most of the drives are 3 HP PowerFlex 525. You're right that AB sells an MMS (140M) that gives the series combination a SCCR of 65 kA, however the door-mounted disconnect (Mersen M163DM) would still be left in the wind with it's default SCCR. Both get a higher SCCR with CC, T, or J fuses though.

Regarding putting fuses on the feeder to protect the whole panel, does that work when the drive and disconnect specify maximum fuse sizes in order to get the higher SCCR? For example, the disconnect gets a 10 kA SCCR when used in series with a maximum fuse size of 30 A, which obviously is too small for a whole panel but could conceivably protect three 3 HP drives (7.5 FLA x 1.25 x 3 < 30 A).

 

[petersonra]

The switch you mentioned is only a 20 A rated switch. If you put a 30 A class CC fuse in the feeder to the panel it would give the switch a 10 kA SCCR.

Without knowing exactly what you have, it is really hard to help you.There is also the problem that the UL procedure for determining SCCR requires (in a legalistic way) the industrial control panel be UL listed to UL508a. I doubt you are able to do that yourself.

 

[EC Dan]

Are you saying I'm unable to apply series combination SCCR changes outside the context of a UL panel shop? All these panels were built in-house and were never assigned an SCCR. My main goal, besides obviously preventing components from exploding in the event of a short circuit, is to cover our base in the event of an inspection. My understanding was that a panel needs to have both the available fault current and SCCR labeled on it. The latter is currently missing and a cursory inspection of the components would show it to be overdutied.

I'm happy to provide more details. Let's focus on one panel. It is fed from a 100 A circuit breaker into a main disconnect and a distribution block. Those components are all suitably rated. The distribution block splits the power to eight individual motor circuits. All eight circuits have, in order, an LSI MMS (MMS-32H-13A), Mersen disconnect (M163DM), and PowerFlex 525 3 HP. My thought was to replace the LSI MMS with a Bussmann fuse holder with disconnect (CCP2-3-30CC).

 

[petersonra]

Are you saying I'm unable to apply series combination SCCR changes outside the context of a UL panel shop? All these panels were built in-house and were never assigned an SCCR. My main goal, besides obviously preventing components from exploding in the event of a short circuit, is to cover our base in the event of an inspection. My understanding was that a panel needs to have both the available fault current and SCCR labeled on it. The latter is currently missing and a cursory inspection of the components would show it to be overdutied.

I'm happy to provide more details. Let's focus on one panel. It is fed from a 100 A circuit breaker into a main disconnect and a distribution block. Those components are all suitably rated. The distribution block splits the power to eight individual motor circuits. All eight circuits have, in order, an LSI MMS (MMS-32H-13A), Mersen disconnect (M163DM), and PowerFlex 525 3 HP. My thought was to replace the LSI MMS with a Bussmann fuse holder with disconnect (CCP2-3-30CC).

What I am saying is the UL procedure you are using to determine the SCCR of the panel only applies to a UL listed control panel. Note what it says about the requirements in the supplement you use to determine the SCCR.



If I was redesigning the thing, I would start the fix by getting rid of the oddball LSI MMS and replacing them with a class CC fuse holder so each of those power circuits is no longer at issue and the Mersen switch now is good up to 10 kA, which is adequate. By the way, MMS are not listed for use on 480 V delta systems. If you have a 480 V delta power system, the MMS is not an appropriate choice anyway.

The distribution block and main switch are likely problematic and would need to be fixed somehow. The simplest solution is probably to replace the main switch with some kind of MCCB and add a 100 A class J fuse block in front of the distribution block that is probably only rated for 10 kA without CL fuses. Many such distribution blocks have a pretty high SCCR when fed from a CL fuse. If not, you might have to change out the distribution block for one that has a better rating. Incidentally, in MOST cases, distribution blocks are limited to a single wire per terminal and many times the SCCR is dependant on a minimum wire size. For instance, the higher SCCR may not apply unless the wire size is at least #10.

Unfortunately, the design of these panels does not lend itself to an easy fix and maybe your best bet is to find some way to reduce the SCC the panel sees rather than reworking the panel. Before I went any farther, I would be taking a second look at the available SCC at the terminals to the panels to make sure they really are what the previous guy said they are. He might have been wrong in his calculations.

 

[EC Dan]

What I am saying is the UL procedure you are using to determine the SCCR of the panel only applies to a UL listed control panel. Note what it says about the requirements in the supplement you use to determine the SCCR.

View attachment 2557074

If I was redesigning the thing, I would start the fix by getting rid of the oddball LSI MMS and replacing them with a class CC fuse holder so each of those power circuits is no longer at issue and the Mersen switch now is good up to 10 kA, which is adequate. By the way, MMS are not listed for use on 480 V delta systems. If you have a 480 V delta power system, the MMS is not an appropriate choice anyway.

The distribution block and main switch are likely problematic and would need to be fixed somehow. The simplest solution is probably to replace the main switch with some kind of MCCB and add a 100 A class J fuse block in front of the distribution block that is probably only rated for 10 kA without CL fuses. Many such distribution blocks have a pretty high SCCR when fed from a CL fuse. If not, you might have to change out the distribution block for one that has a better rating. Incidentally, in MOST cases, distribution blocks are limited to a single wire per terminal and many times the SCCR is dependant on a minimum wire size. For instance, the higher SCCR may not apply unless the wire size is at least #10.

Unfortunately, the design of these panels does not lend itself to an easy fix and maybe your best bet is to find some way to reduce the SCC the panel sees rather than reworking the panel. Before I went any farther, I would be taking a second look at the available SCC at the terminals to the panels to make sure they really are what the previous guy said they are. He might have been wrong in his calculations.

You're right about the main disconnect being a problem. The 100 kA rating that I thought it had turns out to be contingent on using 100 A class J fuses, so it has the same default SCCR as the door mount disconnects. I would need to do as you suggested and add the 100 A class J fuses upstream of both the dist block and the switch. At this point I think re-designing the whole panel is not a great option. The extra length of conductors may be my best bet, since I should be able to calculate the length of the conductors to nudge me just below 5 kA (after I re-evaluate the AFC).

One thing I still don't understand very well is if the panel were designed correctly from the beginning and used 100 A class J fuses on the feeder, wouldn't that affect the AIC of the breaker and/or fuse since there's also a 100 A circuit breaker? As I understand it, as one opens it affects the clearing time of the other which affects the maximum let-through current and invalidates the SCCR. At what point is this not considered a problem with series OCPD? Just to be clear I don't plan on implementing this since I believe this is beyond my engineering ability right now, I'm just trying to understand. Is it also the case that in order to get any decent SCCR, fuses are pretty much need to be part of the solution since it seems most manufacturers only do series combination ratings with fuses?

 

[petersonra]

One thing I still don't understand very well is if the panel were designed correctly from the beginning and used 100 A class J fuses on the feeder, wouldn't that affect the AIC of the breaker and/or fuse since there's also a 100 A circuit breaker? As I understand it, as one opens it affects the clearing time of the other which affects the maximum let-through current and invalidates the SCCR. At what point is this not considered a problem with series OCPD? Just to be clear I don't plan on implementing this since I believe this is beyond my engineering ability right now, I'm just trying to understand. Is it also the case that in order to get any decent SCCR, fuses are pretty much need to be part of the solution since it seems most manufacturers only do series combination ratings with fuses?

It is not real hard to get 65 kA SCCR with MCCBs. You just have to select components correctly. Pretty much can't use power distribution blocks. That is not as big a deal as you might think. Most MCCBs these days can have multi-lugs on the load side so you can hook up 6 wires to each pole. Often I add feeder breakers off the main breaker with multi-lugs and use them for power distribution. It helps meet the tap rules too.
I have been known to put two sets of CL fuses in series so that I can claim one set is in the feeder circuit and get credit for the let thru rating of the fuse. Why UL thinks it matters if the CL fuse is in the feeder circuit versus a branch circuit escapes me, but that is the rule and I just follow it.
The AIC rating of a CB or fuse is not affected by having multiple OCPDs in series. What can happen is you can get a higher AIC rating for a CB if it is in series with a CL fuse that it has been tested with.

 

[EC Dan]

The peak let-through MUST be based on UL 508A Table 4.2 correct? Even if the manufacturer has current-limitation charts that have more detailed information, such as peak-let through at lower SCC? I ask because Eaton shows that with certain 100 A Class J fuses at 10 kA (close to my SCC), the let-through current would be 2 kA, which means I wouldn't need to change anything inside the panel and the panel SCCR becomes the AIC rating of the fuse, but I believe this is a discredited methodology for determining SCCR, correct?

Edit: Actually the 2 kA is RMS let-through, not peak. Peak would still put me at 7 kA.

 

[paulengr]

Hi all,

I’ve inherited a set of 480 VAC industrial control panels that I’ve determined to be overdutied based on the previous engineer’s available fault current analysis (analysis performed after panel design and installation). For some reason as part of that analysis he did not also identify that the panels were overdutied as a result of the AFC. In these cases, the AFC is about 9.9 kA and the panels have elements that render it with the base 5 kA SCCR. These panels contain a number of motor control circuits, each with (in order from line to load) a manual motor starter, door-mounted disconnect for LOTO, PowerFlex VFD, Phoenix Contact 3-level motor terminal block, and motor. The door-mounted disconnect and VFD are the limiting factors. I’ve read every thread I can get my hands on in this forum about SCCR and have learned a tremendous amount, so I believe I can see what my options are, and I’d like some advice. As I can see it, I can either attempt to reduce the AFC to below 5 kA with an isolation transformer or replace the components in the panel that do not give me series combinations to get me above 10 kA. Both options have pros and cons:

• Isolation transformer: this would be the easiest to implement, in that it’s a single component to add, and would give me benefits of protecting the line from any harmonics from the VFDs, however I believe I’d be cutting it too close to the non-instantaneous trip region of the circuit breaker. The CB is 100 A and is comfortably in the range of instant trip with an arcing current of 6.5 kA (based on 9.9 kA bolted fault current), however unless I can very narrowly target 4-5 kA bolted fault current (3-4 kA arcing current), I may be too close for comfort to the non-instantaneous trip portion of the TCC, which would vastly increase my incident energy.
• Replace components: both the VFD and door-mounted disconnect have improved SCCR with class CC/J fuses, however there is not room for both the MMS and fuse holders, so I would need something like a Bussmann fuse block with disconnect switch to replace the MMS entirely and maintain the ability to disconnect individual motor circuits. I lose the thermal trip part of the MMS in that case, but as each motor is on a VFD I think the VFD would handle that fault protection. I wanted to avoid fuses but as I’ve learned it’s very difficult to achieve higher than 5 kA without fuses and the series combinations they enable. I’m also not seeing very many options for fuse holders with disconnect switches (at least of the DIN-rail mounted variety for use in a panel), which leads me to believe they are not widely used, which makes me second guess their application here.

I’d appreciate any feedback about my evaluation of options and whether I’m overlooking anything or options I haven’t considered. I'd be happy to provide any clarifying details and information.

Ok so a lot going on here. Your isolation transformer info is partly right. It does increase line impedance thus lowering SCCR. The tripping issue is also true. Any overcurrent device is seeing it through the transformer. Realistically trying to do this is usually not worth it. Just put a fuse block on the secondary side and be done with it.

What is not true is how much it will improve your harmonics. This is one of the biggest lies going. A delta-wye transformer cancels triplen harmonics (google this). Guess which harmonics 6 pulse VFDs do not produce? You get some benefit from an LR filtering effect but it’s small. Plus why buy the transformer? A line reactor gives you exactly the same thing directly at a better price point. The only reason to use a transformer is if you need to create a neutral reference or if you need to change voltages (step up/down). Otherwise use the line reactor.

The CC/J fuse thing is another partial myth. As far as current reduction it’s usually tiny. Maybe you can current limit by 50%, maybe not. If you are tripping or otherwise using it, the I ssue is with series ratings you need tested combinations. If there is no combination listed in the VFD manual this is not an option. Same story with the MMS.

As far as door mounted combo fuse/disconnect blocks the best one out there is the Fuserbloc from Socomec. I’ll warn you up front there is a bewildering variety of options with them and part numbers that defy logic.

Another option you didn’t mention is semiconductor fuses. Again tested combination is the key but if available these ultra fast fuses can easily protect VFDs. If nothing in the manuals again you have a pig in a poke because unless you are directly reducing AIC and not just tripping faster, it doesn’t fix the issue.

 

[paulengr]

Also realized you are attempting to protect motor with MMS overload. Simply put this doesn’t work. You are preserving nothing.

If you try to protect from the line side, you are seeing the drive as a load. The motor is isolated. So for instance if I run the motor at 15 Hz with a corresponding 1/4 of rated Voltage but 400% current (assuming VFD is big enough) the overload relay in the MMS never sees this and is unaware you are cooking the motor. Power draw from the VFD is the same (P=1.732 x V x I x of) so the overload relay is utterly useless except when V to the motor is at line voltage. This assumes efficiency is 100%. At less than 100% at best it might trip prematurely running at 60 Hz.

With one exception ALWAYS use the VFD to perform overload protection. The one exception is with a VFD driving multiple parallel motors. In this case you have to put overload relays, one per motor, on the VFD output, or use thermal overload switches or similar motor protection.

 

[paulengr]

It is not real hard to get 65 kA SCCR with MCCBs. You just have to select components correctly. Pretty much can't use power distribution blocks. That is not as big a deal as you might think. Most MCCBs these days can have multi-lugs on the load side so you can hook up 6 wires to each pole. Often I add feeder breakers off the main breaker with multi-lugs and use them for power distribution. It helps meet the tap rules too.
I have been known to put two sets of CL fuses in series so that I can claim one set is in the feeder circuit and get credit for the let thru rating of the fuse. Why UL thinks it matters if the CL fuse is in the feeder circuit versus a branch circuit escapes me, but that is the rule and I just follow it.
The AIC rating of a CB or fuse is not affected by having multiple OCPDs in series. What can happen is you can get a higher AIC rating for a CB if it is in series with a CL fuse that it has been tested with.

Since when are power distribution blocks useless?

https://cdn.automationdirect.com/static/specs/edisonpb.pdf

200 kA is pretty high for most applications. Just have to watch out for UL 1953 Listed vs 1059 Recognized and maintain spacing which shouldn’t be a problem.

 

[petersonra]

Only if you are willing to put in fuses. A lot of us are not allowed fuses.

 

[EC Dan]

Only if you are willing to put in fuses. A lot of us are not allowed fuses.

I'm curious why you wouldn't be allowed to use fuses.

The previous guy who designed these panels really seems to not have understood the fuse requirement in order to achieve the high SCCR rating. I *almost* don't blame him, since on first glance when looking at these items (such as Edison PDBs, which we use here) is that they just intrinsically have the high SCCR, but it's clear upon reviewing the technical specifications that they are all contingent on fuses for their high SCCR values.

 

[petersonra]

I'm curious why you wouldn't be allowed to use fuses.

The previous guy who designed these panels really seems to not have understood the fuse requirement in order to achieve the high SCCR rating. I *almost* don't blame him, since on first glance when looking at these items (such as Edison PDBs, which we use here) is that they just intrinsically have the high SCCR, but it's clear upon reviewing the technical specifications that they are all contingent on fuses for their high SCCR values.

A lot of our customers demand very high uptime. You blow a fuse and you don't have one in house at 3:00 in the morning it may be a day or two before you can get one.

The other thing is that in larger sizes fuses cost more than circuit breakers and take up a lot more space.

Note that in the post I referred to MCCB, or molded case circuit breakers. I think it a response to something somebody mentioned about needing to use fuses to get high short circuit current ratings.

 

[EC Dan]

A lot of our customers demand very high uptime. You blow a fuse and you don't have one in house at 3:00 in the morning it may be a day or two before you can get one.

The other thing is that in larger sizes fuses cost more than circuit breakers and take up a lot more space.

Note that in the post I referred to MCCB, or molded case circuit breakers. I think it a response to something somebody mentioned about needing to use fuses to get high short circuit current ratings.

Do you have any recommendations for a series of current-limiting panel-mountable MCCBs that have multi-lug kit accessories? I know that's pretty broad without giving more specs. These sound like these would be useful for future panel builds since we also require high uptime.

Also, about determining let-through current for current-limiting feeder CBs and fuses: is it required to use the UL table with the default listings for the purposes of determining SCCR or can the manufacturer's current-limitation charts be used? It would seem to me based on the wording in UL the tables must be used, but I was reading through an applications note from Rockwell that seems to indicate the manufacturer's charts may be used.

 

[petersonra]

Do you have any recommendations for a series of current-limiting panel-mountable MCCBs that have multi-lug kit accessories? I know that's pretty broad without giving more specs. These sound like these would be useful for future panel builds since we also require high uptime.

Also, about determining let-through current for current-limiting feeder CBs and fuses: is it required to use the UL table with the default listings for the purposes of determining SCCR or can the manufacturer's current-limitation charts be used? It would seem to me based on the wording in UL the tables must be used, but I was reading through an applications note from Rockwell that seems to indicate the manufacturer's charts may be used.

I have not used a CL CB, at least that I can recall. Just never needed one. Most of the time I can readily get to 65 kA SCCR without CL fuses or CBs, although it is sometimes, but not always, cheaper with CL fuses.

I think you have to use the UL charts for fuses but can use the manufacturers charts for CBs.

We use mostly AB MCCBs and they have 6 hole lugs for all the frame sizes except frame M. Sometimes have to be a little careful on wire size as only the G frame multilugs go down to 14 AWG, so sometimes I have to use larger wire than I otherwise would.

I have considerable experience with Siemens breakers too and I think all or most of them have the multi-lugs as well. Also pretty sure I have used Schneider MCCBs with multi-lugs.