Available fault current approximation

[Isaiah]

Is there a ‘quick and easy’ rule of thumb way to determine available fault current for a (power) Panelboard without doing complex calculations or using Etap?


Sent from my iPhone using Tapatalk

 

[jim dungar]

There are many online calculators you can use. Almost every breaker or fuse manufacturer offers one.

It helps if you know the starting fault current value.

 

[Isaiah]

There are many online calculators you can use. Almost every breaker or fuse manufacturer offers one.

It helps if you know the starting fault current value.

Thanks Jim
The upstream CB is 800AF/AT and is connected to an 85kAIC switchgear bus rated 4000A. The panel itself is MLO with 800A bus. There are 48 circuits but nearly 60% and unused spares or spaces. The panel is very lightly loaded with mostly HVAC, small dry type lighting transformers and a few miscellaneous 3 phase loads. We haven’t purchased the Panelboards (there are several actually) and the distributor wants to know the AIC rating. I have a gut feeling it should be around 35kAIC but without knowing I may have to go up to 65kAIC which of course will raise the price and lead time significantly.
I’m going to look for the online calculator you mentioned. Do you have a recommendation?


Sent from my iPhone using Tapatalk

 

[tom baker]

i used to use charts in Firms Fast Finder.
for software, start with Bussman.

 

[petersonra]

Thanks Jim
The upstream CB is 800AF/AT and is connected to an 85kAIC switchgear bus rated 4000A. The panel itself is MLO with 800A bus. There are 48 circuits but nearly 60% and unused spares or spaces. The panel is very lightly loaded with mostly HVAC, small dry type lighting transformers and a few miscellaneous 3 phase loads. We haven’t purchased the Panelboards (there are several actually) and the distributor wants to know the AIC rating. I have a gut feeling it should be around 35kAIC but without knowing I may have to go up to 65kAIC which of course will raise the price and lead time significantly.
I’m going to look for the online calculator you mentioned. Do you have a recommendation?


Sent from my iPhone using Tapatalk

It does not matter what the rating of the panel board upstream is or how lightly or heavily loaded it is. You need to know what the available short circuit current is upstream and then work your way downstream to where the panel boards you are going to be installing are located and figure out what the available short circuit current will be at that point. The only way to do this is to have a known value at some point and work your way down stream until you get to your panel boards.

Usually the utility will tell you what the available short circuit current is at the terminals of their transformer that supplies the building. That is where you probably have to start.

At that point it is mostly an exercise in ohm's law to determine what the short circuit current will be at any point downstream.

 

[ramsy]

At that point it is mostly an exercise in ohm's law to determine what the short circuit current will be at any point downstream.

Ideal's newer SureTest models have built-in ASCC functions, where the tester generates a calibrated load internally.

ASCC = E/Z, leg voltage(E) over leg impedance(Z). Z=VD/I. VD= dropped volts and I= SureTest's ampere pulse.

Voltages => 300v will destroy this tester, but weather 480/277Y or lower system, L-L may also destroy this tester.

Attempting a less accurate separate clamp-metered load & voltage can be attempted for L-L readings, using the same formula.

 

[Isaiah]

Ideal's newer SureTest models have built-in ASCC functions, where the tester generates a calibrated load internally.

ASCC = E/Z, leg voltage(E) over leg impedance(Z). Z=VD/I. VD= dropped volts and I= SureTest's ampere pulse.

Voltages => 300v will destroy this tester, but weather 480/277Y or lower system, L-L may also destroy this tester.

Attempting a less accurate separate clamp-metered load & voltage can be attempted for L-L readings, using the same formula.

Great info thanks to all


Sent from my iPhone using Tapatalk

 

[Hv&Lv]

Is there a ‘quick and easy’ rule of thumb way to determine available fault current for a (power) Panelboard without doing complex calculations or using Etap?


Sent from my iPhone using Tapatalk

Point to point.
Or simpler is V/Z.

Please don’t take the simple route here..
You can use simple and your numbers will be inflated way out of proportion..
Get a good calculator program, (there are several P2P calculators out there)
You won’t regret it later

 

[electrofelon]

Thanks Jim
The upstream CB is 800AF/AT and is connected to an 85kAIC switchgear bus rated 4000A. The panel itself is MLO with 800A bus. There are 48 circuits but nearly 60% and unused spares or spaces. The panel is very lightly loaded with mostly HVAC, small dry type lighting transformers and a few miscellaneous 3 phase loads. We haven’t purchased the Panelboards (there are several actually) and the distributor wants to know the AIC rating. I have a gut feeling it should be around 35kAIC but without knowing I may have to go up to 65kAIC which of course will raise the price and lead time significantly.
I’m going to look for the online calculator you mentioned. Do you have a recommendation?


Sent from my iPhone using Tapatalk

You can probably use series ratings and not have to worry about calculating the actual fault current. However with an 800 amp upstream device you will probably need 22 kaic branches, not regular 10kaic. I recently actually priced a factory panel board with both and the 22s were only about three bucks more per Pole.

All that said, I personally would like to know about what the AFC is.

 

[Tulsa Electrician]

Is there a tag in the trans for Z.
How far away are going to run the feeder from the switchboard.
Let say you have trans, 3500 kva @ 5.7% Z so you start at 73,860 approx! 3500 x 1000= /480/1.733. 1/Z as a multiplayer. Take available amps times multiplayer. 4210x17.54385964= 73859.64912
Your at 4000 and 85k.
You'll need wire, lenght, sets motor loads etc. You will need FLA X4 for motors. It pays to one line it.

As far as quick stuff on lower kva you can use trans as quick check. They larger the system the more care should be taken.
One reason we set a lot of transformer. Use the above example for a 75 kva @ 5% Z for a 120/208 secondary.
Then one motor of 50 FLA.
208 amps
1/.05= 20
20 x 208 = 4160
50 x4= 200
200+4160= 4360
Quick with out any secondary conductor etc. Well below 10k
Now if was to come in with 80% of 10k I would be looking a little closer at install. In that case if the trans was close to new main breaker panel I would do full calc. If in ele closet with a longer feeder it would be better. It would lower the available.
One additional item to think about it is that it take fault current to open ocds during a ground fault. The available will impact time. So branch will need to be looked at with a low available.
With the 4160 the trip time will be pretty quick base on a 10 k time curve and good execution of work.
Sorry if a little long.
A good engineer can design a safe install.

 

[Isaiah]

Is there a tag in the trans for Z.
How far away are going to run the feeder from the switchboard.
Let say you have trans, 3500 kva @ 5.7% Z so you start at 73,860 approx! 3500 x 1000= /480/1.733. 1/Z as a multiplayer. Take available amps times multiplayer. 4210x17.54385964= 73859.64912
Your at 4000 and 85k.
You'll need wire, lenght, sets motor loads etc. You will need FLA X4 for motors. It pays to one line it.

As far as quick stuff on lower kva you can use trans as quick check. They larger the system the more care should be taken.
One reason we set a lot of transformer. Use the above example for a 75 kva @ 5% Z for a 120/208 secondary.
Then one motor of 50 FLA.
208 amps
1/.05= 20
20 x 208 = 4160
50 x4= 200
200+4160= 4360
Quick with out any secondary conductor etc. Well below 10k
Now if was to come in with 80% of 10k I would be looking a little closer at install. In that case if the trans was close to new main breaker panel I would do full calc. If in ele closet with a longer feeder it would be better. It would lower the available.
One additional item to think about it is that it take fault current to open ocds during a ground fault. The available will impact time. So branch will need to be looked at with a low available.
With the 4160 the trip time will be pretty quick base on a 10 k time curve and good execution of work.
Sorry if a little long.
A good engineer can design a safe install.

Lots of great Info here thanks Tulsa


Sent from my iPhone using Tapatalk

 

[electrofelon]

You should ask the power company for the AFC at the secondary terminals of their transformer or the service point whichever they provide. In my experience the value they give will be 50% to double what the infinite primary fault current figure actually is. There is of course some Prudence to assuming a higher value to account for possibilities like transformer upgrades in the future which may be of larger size and/or lower impedance, but many times if I can see the transformer data I will compute it myself and make my own judgment based on the circumstances rather than use a potentially bloated figure they give.

AFC can vary widely. On one extreme, you can have the POCO come up with only needing a small transformer Bank relative to your service size and a long service drop or lateral. And you'll easily be below 10K at the service equipment. On the other hand you could have an oversized transformer bank for various reasons. One building I did had a 1000 amp service but a 3*166 bank (about 1300A) likely because it used to be a printing company and they had some very large motors in there. Another job, the POCO engineer called for a 3x100 bank they put in 3x166's because they are out of 100s at that time and didn't want to delay my project. Then impedance can vary widely too. The first 166 case I mentioned above, those were submersible units with very low a 500 KVA pad would typically be two and a half times that.

So long story short, it's pretty tough and inadvisable to come up with rules of thumb if you can't see what the power company is serving you with and aren't familiar with the calcs and equipment a bit.

 

[Isaiah]

You should ask the power company for the AFC at the secondary terminals of their transformer or the service point whichever they provide. In my experience the value they give will be 50% to double what the infinite primary fault current figure actually is. There is of course some Prudence to assuming a higher value to account for possibilities like transformer upgrades in the future which may be of larger size and/or lower impedance, but many times if I can see the transformer data I will compute it myself and make my own judgment based on the circumstances rather than use a potentially bloated figure they give.

AFC can vary widely. On one extreme, you can have the POCO come up with only needing a small transformer Bank relative to your service size and a long service drop or lateral. And you'll easily be below 10K at the service equipment. On the other hand you could have an oversized transformer bank for various reasons. One building I did had a 1000 amp service but a 3*166 bank (about 1300A) likely because it used to be a printing company and they had some very large motors in there. Another job, the POCO engineer called for a 3x100 bank they put in 3x166's because they are out of 100s at that time and didn't want to delay my project. Then impedance can vary widely too. The first 166 case I mentioned above, those were submersible units with very low a 500 KVA pad would typically be two and a half times that.

So long story short, it's pretty tough and inadvisable to come up with rules of thumb if you can't see what the power company is serving you with and aren't familiar with the calcs and equipment a bit.

Agreed completely. I will start my research tomorrow with the upstream utility transformer


Sent from my iPhone using Tapatalk

 

[electrofelon]

I just noticed the impedance value of those 166 submersible units didn't make it into the post. It was 1.7%

 

[paulengr]

Is there a ‘quick and easy’ rule of thumb way to determine available fault current for a (power) Panelboard without doing complex calculations or using Etap?


Sent from my iPhone using Tapatalk

Yes. Follow the IEEE ANSI procedure! It was created at a time when calculators did not exist and computers took up a large office building.

Ok so first and least accurate method that I see used everywhere: infinite bus. Let’s assume all transformer losses are reactive (almost entirely true) and the utility bus is very stiff, much bigger than the transformer (sometimes true), and that there are no reactive loads, and all cables have no impedance. So really we just have one reactance. So we just calculate the transformer theoretical maximum short circuit current at the secondary terminals. KVA x 1000 / (1 or 1.732) / V / (%Z / 100) is short circuit current.

Ok but cabling makes a huge difference!! No problem. Look at the Cooper Bussman point by point method. It is a couple tables and formulas and napkin math. Slide rule optional!! It significantly reduces the value from infinite bus even with a little bus bar. And we can go through multiple cables and transformers in series (point by point).

The only thing point by point misses is DC transients. Those can be ignored with UL 1077 equipment (switchboards, switchgear) where everything has a 30 cycle rating unless you have some big synchronous machines with long time constants. So check for that. This is why the IEEE 1584 arc flash spread sheet uses point by point.

Ok, so still not good enough? You can use the full ANSI method. Get the reactance data for each cable. Get motor data. Throw out anything 25 HP or less (transient collapses in under 1 cycle). Add all motor HP from 30 to 150 HP and treat as one motor. Calculate transient reactance (current) from the ANSI tables. Calculate the resulting system using Ohms Law since everything is reactance only. I’m glossing over details but the procedure is really that simple. This is exactly what ETAP does. I’ve also got another simple one from a Soares book but it’s not in PDF or anything like that.

So you want resistance too? Same as above but use the full R+jX values. This takes longer and since resistance has very little impact on short circuit (1%) nobody does it.

This gets you within 10%. Need more accuracy? Then screw ETAP. You need one of the various EMTP programs and a lot of personal abuse! This does time series level analysis.

Remember NEC does not specify which method to use. This is where engineering judgement applies. For short circuit purposes our concern is whether or not SCCR is an issue. So if we do infinite bus or even point by point with just the feeder and we are well under SCCR, with no big synchronous machines or motors and X/R is reasonable, we are done. No need to go further. If it is way over we also have a result. But if it’s close, we need to use another method or just buy the bigger breaker.

With arc flash it’s a much worse problem. The math works in reverse. So if you overestimate short circuit current you will underestimate arc flash, by s lot. So arc flash calculations are best done with the most complicated model, full R+jX. ETAP, Easypower, SKM, etc. do this automatically. The short circuit report will be in ANSI but the arc flash uses the full model plus an adjustment on arcing current specified by the 1584 model,

 

[Isaiah]

Yes. Follow the IEEE ANSI procedure! It was created at a time when calculators did not exist and computers took up a large office building.

Ok so first and least accurate method that I see used everywhere: infinite bus. Let’s assume all transformer losses are reactive (almost entirely true) and the utility bus is very stiff, much bigger than the transformer (sometimes true), and that there are no reactive loads, and all cables have no impedance. So really we just have one reactance. So we just calculate the transformer theoretical maximum short circuit current at the secondary terminals. KVA x 1000 / (1 or 1.732) / V / (%Z / 100) is short circuit current.

Ok but cabling makes a huge difference!! No problem. Look at the Cooper Bussman point by point method. It is a couple tables and formulas and napkin math. Slide rule optional!! It significantly reduces the value from infinite bus even with a little bus bar. And we can go through multiple cables and transformers in series (point by point).

The only thing point by point misses is DC transients. Those can be ignored with UL 1077 equipment (switchboards, switchgear) where everything has a 30 cycle rating unless you have some big synchronous machines with long time constants. So check for that. This is why the IEEE 1584 arc flash spread sheet uses point by point.

Ok, so still not good enough? You can use the full ANSI method. Get the reactance data for each cable. Get motor data. Throw out anything 25 HP or less (transient collapses in under 1 cycle). Add all motor HP from 30 to 150 HP and treat as one motor. Calculate transient reactance (current) from the ANSI tables. Calculate the resulting system using Ohms Law since everything is reactance only. I’m glossing over details but the procedure is really that simple. This is exactly what ETAP does. I’ve also got another simple one from a Soares book but it’s not in PDF or anything like that.

So you want resistance too? Same as above but use the full R+jX values. This takes longer and since resistance has very little impact on short circuit (1%) nobody does it.

This gets you within 10%. Need more accuracy? Then screw ETAP. You need one of the various EMTP programs and a lot of personal abuse! This does time series level analysis.

Remember NEC does not specify which method to use. This is where engineering judgement applies. For short circuit purposes our concern is whether or not SCCR is an issue. So if we do infinite bus or even point by point with just the feeder and we are well under SCCR, with no big synchronous machines or motors and X/R is reasonable, we are done. No need to go further. If it is way over we also have a result. But if it’s close, we need to use another method or just buy the bigger breaker.

With arc flash it’s a much worse problem. The math works in reverse. So if you overestimate short circuit current you will underestimate arc flash, by s lot. So arc flash calculations are best done with the most complicated model, full R+jX. ETAP, Easypower, SKM, etc. do this automatically. The short circuit report will be in ANSI but the arc flash uses the full model plus an adjustment on arcing current specified by the 1584 model,

Paul...You really nailed it!
This thread gives me a great start!


Sent from my iPhone using Tapatalk

 

[kwired]

Thanks Jim
The upstream CB is 800AF/AT and is connected to an 85kAIC switchgear bus rated 4000A. The panel itself is MLO with 800A bus. There are 48 circuits but nearly 60% and unused spares or spaces. The panel is very lightly loaded with mostly HVAC, small dry type lighting transformers and a few miscellaneous 3 phase loads. We haven’t purchased the Panelboards (there are several actually) and the distributor wants to know the AIC rating. I have a gut feeling it should be around 35kAIC but without knowing I may have to go up to 65kAIC which of course will raise the price and lead time significantly.
I’m going to look for the online calculator you mentioned. Do you have a recommendation?


Sent from my iPhone using Tapatalk

I guess question mostly been answered, but what I was going to say is the size of gear and AIC rating of that gear doesn't necessarily mean much to you.

First thing that matters is size and impedance of the source. That sets maximum available current at terminals of that source, it will lessen as you put more conductor between source and point you are determining available fault current at because the conductors will have a resistance. Every feeder between your point and the source will often be even smaller conductor and add even more resistance. So you could have something near that 85kA available at the main gear, but only have 2kA available at some remote panelboard feeding a limited load with say only a 30 amp feeder supplying it.

If you don't know the impedance of the source hopefully you at least know it's KVA rating and can plug in maybe a best and worst case possibility for impedance and see where that puts you for available current at your point of interest, which is fine for determining withstand rating of equipment, but not necessarily going to be all that good for determining incident energy for arc flash calculations, those you do need more accurate values of what is actually there, along with clearing time information on OCPD's to have much accuracy on incident energy.

 

[Isaiah]

I guess question mostly been answered, but what I was going to say is the size of gear and AIC rating of that gear doesn't necessarily mean much to you.

First thing that matters is size and impedance of the source. That sets maximum available current at terminals of that source, it will lessen as you put more conductor between source and point you are determining available fault current at because the conductors will have a resistance. Every feeder between your point and the source will often be even smaller conductor and add even more resistance. So you could have something near that 85kA available at the main gear, but only have 2kA available at some remote panelboard feeding a limited load with say only a 30 amp feeder supplying it.

If you don't know the impedance of the source hopefully you at least know it's KVA rating and can plug in maybe a best and worst case possibility for impedance and see where that puts you for available current at your point of interest, which is fine for determining withstand rating of equipment, but not necessarily going to be all that good for determining incident energy for arc flash calculations, those you do need more accurate values of what is actually there, along with clearing time information on OCPD's to have much accuracy on incident energy.

I agree. I’m thinking the fault current will be really low at the smaller 208/120V Panelboards downstream. If the breakers were undersized in terms of AIC would they be destroyed under fault conditions?


Sent from my iPhone using Tapatalk

 

[kwired]

I agree. I’m thinking the fault current will be really low at the smaller 208/120V Panelboards downstream. If the breakers were undersized in terms of AIC would they be destroyed under fault conditions?


Sent from my iPhone using Tapatalk

no, because the available fault current would be low at that point in the system.

 

[electrofelon]

I agree. I’m thinking the fault current will be really low at the smaller 208/120V Panelboards downstream. If the breakers were undersized in terms of AIC would they be destroyed under fault conditions?


Sent from my iPhone using Tapatalk

no, because the available fault current would be low at that point in the system.

keep in mind it takes a longer distance than you probably think to get under 10k where you have 65K upstream. For a 200A feeder its going to be around 150 feet. OP has an 800A which is 4 times the wire so it going to take a lot of length.