What would you have done?

Dsg319

Senior Member
Location
West Virginia
Occupation
(Green)Master Electrician
On a project replacing a couple sub panels.
One being the main sub panel that feeds several others.

Replace a federal pacific with a new 200amp GE. All other sub panels are old 3 wire feeders with neutral bonded at the panel.

How ever after replacing this one I noticed the conduit that feeds the sub panel was all emt. So I checked the resistance between the feeder pipe and the unhooked neutral and it was bonded back at the main disco.

So with that being said I decided to just use the EMT as an EGC and isolate the neutral in this panel (although rest of the sub panels fed from it are 3 wire feeders neutral bonded).

Should I have left it alone after being that away after so many years or was what I done correct?

I did however simulate a (bolted) ground fault on the far end of the building and it cleared probably within 3 seconds or less. Granted that was a 14gauge circuit on a 20amp breaker that has to be changed🙄
 

hillbilly1

Senior Member
Location
Atlanta,Ga
Occupation
Field coordinator/ technical support
If those other panels were FP, it’s a wonder you were able to clear the fault at all when you tested it! LOL!
 

Dsg319

Senior Member
Location
West Virginia
Occupation
(Green)Master Electrician
I can see ups and downs both ways. Will probably leave it how it is now as I have lost the screws for the bonding strap😝
 
Location
Tulsa
Occupation
Electrician
Roughly a withstand rating of 182 amps.
#14 bare (4110 CM @ 29.1)
3 seconds
182/15= 12.13333
I would think to long clearing time on bolted fault. Unknown available fault current.

Took it was nm cable
 
Location
Tulsa
Occupation
Electrician
It would the ocd handle size multiplier.
This would fall into the instantaneous range. Which should have cleared much quicker. Agian using three seconds bare conductor. Bolted ground fault. Just some rough math.
Math based on five second withstand rating. Amp squared time, CM /29.1.
Corrected 5 second rating to 3 seconds for allowed
amps before damage to conductor

It stood out to me as to much voltage drop back to the source to open the OCD.
However not knowing the available fault current and using EMT as ground, unknown distance and fitting etc.
So just rough math.
Depending on feeder length from main panel where main bonding jumper is installed and that raceway size and length to the other panels and conductor sizes.

Now that being said you started the sub panel had a bonding screw in it. Bonding the neutral bar to the enclosure. This tells me you a parallel grounded conductor circuit which each has a different impedance. Pipe and wire acting in parallel. Now you bolted a fault to ground on a branch circuit conductor to the EMT raceway at an unknown distance.
So I assumed the longest length for that branch circuit of that sub panel.
My point being when you bolted that fault after you removed that screw you lost one of the intended paths back to the source (grounded conductor). The wire will carry a portion of this fault current back to to he mbj, Now it does not. Based on parallel circuit law. Circuit impedance has increased causing a greater voltage drop on the circuit.
If the sub panel is let's say 100 feet away and your branch raceway is 200' away. You can see how my first thought was vd on fault based on the three seconds so I did rough math. I would look at the raceway assembly and be sure all is good and mechanically connected well. This of course if the building is of age which I bet it is and no egc conductors installed.
I my self would have also removed the bond at subs after looking over the raceway installation.
Now if the sub panel is adjacent to the main panel. I would add egc conductor in Fedder and add ground bars in sub. Then your back where you started before removing sub panel bonding screw. This being code compliant and removing the EMT from the grounded neutral conductor.
Hope this explains what I was thinking with the crude math.
You did say commerical garage. So I will assume gfci protection on receptacles. Or age of building not.

So the screw removal may have effected the low impedance effective ground path. It's a trade off some times. Me I want my ocd to open in time. I cringe every time I hear a singing fault. Any fault for that matter.
Did not mean to be long winded. Just wanted you to see why I did what I did.
 

Dsg319

Senior Member
Location
West Virginia
Occupation
(Green)Master Electrician
It would the ocd handle size multiplier.
This would fall into the instantaneous range. Which should have cleared much quicker. Agian using three seconds bare conductor. Bolted ground fault. Just some rough math.
Math based on five second withstand rating. Amp squared time, CM /29.1.
Corrected 5 second rating to 3 seconds for allowed
amps before damage to conductor

It stood out to me as to much voltage drop back to the source to open the OCD.
However not knowing the available fault current and using EMT as ground, unknown distance and fitting etc.
So just rough math.
Depending on feeder length from main panel where main bonding jumper is installed and that raceway size and length to the other panels and conductor sizes.

Now that being said you started the sub panel had a bonding screw in it. Bonding the neutral bar to the enclosure. This tells me you a parallel grounded conductor circuit which each has a different impedance. Pipe and wire acting in parallel. Now you bolted a fault to ground on a branch circuit conductor to the EMT raceway at an unknown distance.
So I assumed the longest length for that branch circuit of that sub panel.
My point being when you bolted that fault after you removed that screw you lost one of the intended paths back to the source (grounded conductor). The wire will carry a portion of this fault current back to to he mbj, Now it does not. Based on parallel circuit law. Circuit impedance has increased causing a greater voltage drop on the circuit.
If the sub panel is let's say 100 feet away and your branch raceway is 200' away. You can see how my first thought was vd on fault based on the three seconds so I did rough math. I would look at the raceway assembly and be sure all is good and mechanically connected well. This of course if the building is of age which I bet it is and no egc conductors installed.
I my self would have also removed the bond at subs after looking over the raceway installation.
Now if the sub panel is adjacent to the main panel. I would add egc conductor in Fedder and add ground bars in sub. Then your back where you started before removing sub panel bonding screw. This being code compliant and removing the EMT from the grounded neutral conductor.
Hope this explains what I was thinking with the crude math.
You did say commerical garage. So I will assume gfci protection on receptacles. Or age of building not.

So the screw removal may have effected the low impedance effective ground path. It's a trade off some times. Me I want my ocd to open in time. I cringe every time I hear a singing fault. Any fault for that matter.
Did not mean to be long winded. Just wanted you to see why I did what I did.
Nope you had a great explanation. Will probably think on it some on and off this week and maybe change my mind to re-bonding the panel when I get back to it next weekend.

Kind of a double edged sword. Objectionable current/ vs long winded fault current.
 
Top