Definition of Subpanel Ground/Neutral Isolation

[Davey4000]

When adding a subpanel I know you are supposed to isolate the ground and neutral in the subpanel. My question is what exactly is the “electrical” definition of an isolated ground and neutral in the subpanel?

In the main panel the grounding bar is bonded to the box. Neutrals (white) and grounds (green) are terminated to this grounding bar. Okay, now I run 4 wires from the main panel to the subpanel. The white wire goes to the neutral lug in the subpanel and the green wire goes to the box. The neutral lug and the box are not bonded (no ground strap or screw from the neutral lugs to the box). Now my question is this…since the white and green originate from the same place in the main panel does this not defeat the effort to isolate them in the subpanel? In the subpanel you can put an Ohm meter across the neutral lug and the box and get a reading, as opposed to being open. Is this still considered being “isolated”?

 

[ishiboo]

When adding a subpanel I know you are supposed to isolate the ground and neutral in the subpanel. My question is what exactly is the “electrical” definition of an isolated ground and neutral in the subpanel?

In the main panel the grounding bar is bonded to the box. Neutrals (white) and grounds (green) are terminated to this grounding bar. Okay, now I run 4 wires from the main panel to the subpanel. The white wire goes to the neutral lug in the subpanel and the green wire goes to the box. The neutral lug and the box are not bonded (no ground strap or screw from the neutral lugs to the box). Now my question is this…since the white and green originate from the same place in the main panel does this not defeat the effort to isolate them in the subpanel? In the subpanel you can put an Ohm meter across the neutral lug and the box and get a reading, as opposed to being open. Is this still considered being “isolated”?

Yes. The goal is to have the two isolated in every point EXCEPT for the main panel So even though they are connected, there are reasons for the two different wires/functions and they should only be bonded at the main panel. The sub panel cannot connect the two.

 

[Davey4000]

Excellent! Thank you very much for the response.

 

[mrb]

the reason for this is current flows on the neutral, and does not flow on the ground unless there is a fault (short). If N+G are connected together at the subpanel, you will have current flowing on all available paths (such as a water pipe between the house and shop)

 

[soj]

This comes up quite often, and it seems to confuse many people. One way to think about it is compare it to a 110 circuit. No one seems to question having a separate neutral and a ground wire to a wall outlet, yet the two wires connect to the same buss bar in the service entrance panel. It's because, as was said earlier, one is a current carrying conductor, the other is a grounding conductor. Connecting the grounds and neutrals together in a sub panel would be the same as connecting the ground and neutral together on the back of a receptacle.

I don't know if that will help anyone else, but once I thought about it that way, it made more sense to me.

 

[hillbilly1]

Parralel paths are one reason, but the main reason is safety, if for some reason the neutral is disconnected or has a bad connection, everything grounded by it will become "Live" to ground due to the loads that use the neutral will try to find an alternate path back, which could be you! Code considers the neutral wire a "hot wire" once it leaves the main service disconnect. This is done for practicality, as the utilities would have to add a wire from the meter back to the transformer to completly remove the chance of a poor connection causing an electocution hazard.

 

[fefarms]

With nothing connected to your subpanel, the neutral bar should be isolated from the case of the subpanel. A tester should show "open" or "infinite" resistance. If it does not, you have not properly prepared the panel for use as a subpanel, There is usually a green bonding screw on the neutral bar that can optionally connect the neutral bar to the panel. It is to be screwed in for a main panel and backed out for a subpanel.

With wires connected to the neutral bar and the ground bar, and properly connected at the main panel end, the panel would show a low resistance path between the case and the neutral, because the tester current flows over the ground and neutral wires back to the main panel and the common bonding point.

 

[Davey4000]

With wires connected to the neutral bar and the ground bar, and properly connected at the main panel end, the panel would show a low resistance path between the case and the neutral, because the tester current flows over the ground and neutral wires back to the main panel and the common bonding point.

Thanks FE. Here are my readings at the subpanel (voltages rounded):

L1-L2 = 240V
L1-N = 120V
L2-N = 120V
L1-G = 120V
L2-G = 120V
L1 or L2 to EMT as it leaves the subpanel = 120V
N-G = 0V

Ohm meter across N and G bars = Continuity just like holding the probes together.

Sound okay?

 

[ishiboo]

Thanks FE. Here are my readings at the subpanel (voltages rounded):

L1-L2 = 240V
L1-N = 120V
L2-N = 120V
L1-G = 120V
L2-G = 120V
L1 or L2 to EMT as it leaves the subpanel = 120V
N-G = 0V

Ohm meter across N and G bars = Continuity just like holding the probes together.

Sound okay?

Sure, these are all correct, but that won't tell you whether the neutral and ground are bonded in more than one place - the readings would be the same

To answer FE's question, you needed to test the sub panel prior to wiring it, making sure that there was infinite resistance between the neutral and ground busses.

 

[dougmac]

Ok ... I drew a picture to help make it a bit more understandable.

I hope this helps!!

 

[soj]

To test now, turn off power at the feed breaker in the house, remove the feeder ground and neutral wire from each buss. Now you should have continuity between the ground buss and the sub panel housing and open between the neutral buss and the sub panel housing and open between each buss. If you get a reading between the neutral buss and the housing, look for a green screw through the buss or a strap from the side of the buss to the housing. Either one should be removed for a sub panel.

 

[Davey4000]

Although I didn't test the subpanel before wiring it I am sure it is correct. The neutral bar does not have a strap to the box. A green screw was supplied to bond the neutral bar to the box and it is still in the plastic bag it came in. I had to purchase a separate bar and add it for the grounds. I was confused over why the neutal and ground in the subpanel would not be open when checking it with an Ohm meter. Now I understand that what we're doing is not providing a return path from the subpanel on the grounding conductor. Makes sense.

 

[ishiboo]

Although I didn't test the subpanel before wiring it I am sure it is correct. The neutral bar does not have a strap to the box. A green screw was supplied to bond the neutral bar to the box and it is still in the plastic bag it came in. I had to purchase a separate bar and add it for the grounds. I was confused over why the neutal and ground in the subpanel would not be open when checking it with an Ohm meter. Now I understand that what we're doing is not providing a return path from the subpanel on the grounding conductor. Makes sense.

Bingo. It's confusing because schematically at first glance, there does not appear to be a difference... but there is in practice

Sounds like you're in good shape!

 

[CreekRat]

Great thread and great picture dougmac.

This topic has always confused me and the discussion and especially the picture really helped out.

 

[mrb]

Ok ... I drew a picture to help make it a bit more understandable.

I hope this helps!!

can you also add pictures showing the path for fault current? Do one with everything properly installed, and another that shows what will happen with a ground rod at the detached subpanel and no EGC back to the main.

 

[MrMark]

To test now, turn off power at the feed breaker in the house, remove the feeder ground and neutral wire from each buss. Now you should have continuity between the ground buss and the sub panel housing and open between the neutral buss and the sub panel housing and open between each buss. If you get a reading between the neutral buss and the housing, look for a green screw through the buss or a strap from the side of the buss to the housing. Either one should be removed for a sub panel.

You can do this test by just disconnecting the ground OR the neutral. I don't see why you need to take both neutral and ground off?

 

[dougmac]

can you also add pictures showing the path for fault current? Do one with everything properly installed, and another that shows what will happen with a ground rod at the detached subpanel and no EGC back to the main.

Let me try to explain this way......

1st you need to understand how the grounding system works:

So you can see from this illustration of the drill that the line voltage (L1) goes to the switch. The neutral is connected to the other side of the motor. The case of the drill is bonded to to the ground (EGC).

If I have a fault where my neutral is shorted to the case of the drill I would not get shocked because it is all ready connected to the neutral. If I am touching two neutrals together electricity does not flow.

Also any other electrical device in my shop that could be energized is connected to the ground (EGC) if it is properly wired. I should be able to theoretically touch the neutral and any other grounded device and not get shocked. That is why it is important to that EVERYTHING that is required to be grounded needs to be bonded to the ground circuit.

If I have a fault where line voltage (L1) shorted to the case the current will travel back to the neutral on the ground circuit and trip the breaker to clear the fault preventing my **** from getting shocked by the energized drill case.

Once again if EVERYTHING that is required to be grounded is grounded, any short to the frame of any device should trip the breaker preventing it from being energized.

For the system to work there must be a low resistance path back to the neutral. That "low resistance path" is the ground conductor.



So ... ground rods.....

A ground rod is intended to dissipate static, switching surges, and lightning. It does not provide a low resistance path back to the neutral and is not designed to clear a fault. The only path it has back to the main panel is through the earth which is not a "low resistance". So the ground rods do not function as a path for the equipment grounds.

I hope this helps to better explain it!
Doug

 

[mrb]

ok take your drawing with the ground rods, connect your faulted drill to it and show the path the fault current takes. Then take the same drawing with the EGC to the subpanel removed, and show the problem. That would reallly help out the people who dont grasp this concept and think the ground rod is a safety device and if its 'grounded' with a rod (or the required two rods so its double grounded) theyre ok.

 

[dougmac]

If the sub is wired properly with a ground wire (EGC) and I have a line to ground fault in my drill the current would travel down the ground wire (yellow/red arrows) directly back to the neutral. This would be a direct short from line to neutral and trip the breaker preventing the drill case from being energized.

If I did not have the ground wire back to the main panel and only had ground rods. The only path back would be through the earth. The earth is not conductive enough to draw enough amperage to trip the breaker. The drill case would become energized. With the drill case wired directly to the ground system, the entire ground system becomes energized.

ALL OF THE OTHER DEVICES THAT ARE BONDED TO THE GROUND SYSTEM WOULD ALSO BECOME ENERGIZED. THIS CREATES A VERY DANGEROUS SITUATION

..... is that what you are looking for?

 

[MrMark]

GREAT DRAWINGS! Are you showing a current divider where it is not a complete short of line to ground, just a partial? I ask this because you are dividing current back on the neutral and the ground.

The only modification I would do is to take off the yellow arrows up off the first ground rod or make the arrows tiny to represent that only a tiny amount of current is flowing down the ground rod at the subpanel and through the earth and up the main panel ground rod and then back up the neutral to the transformer.

 

[danski0224]

And this thread illustrates why it is best to call someone that knows what they are doing, rather than cobble something together from box store parts.