[Mr.Natural]

I'm well aware of the reasons that neutral should only be bonded to ground at the main disconnect panel, so I was surprised to see the default configuration in a 125 amp sub-panel that I'm installing.

The panel comes with a bonding screw attached from the neutral bar to the case, but no bonding screw from the ground bar to the case. It seems to me that it should be just the opposite, since some of the bare ground wires running to the ground bar will certainly come in contact with the case, resulting in an unintentional bond from ground to case (and then from case to neutral). So shouldn't the bond screw be removed from the neutral bar and installed on the ground bar? Or am I missing something?

 

[Jim Port]

In a service panel there is no difference between the bars. Typically there will be a bridge between the two bars effectively making them one bar.

 

[concrete_joe]

but Jim, its a sub panel away from the main panel. the sub panels should have isolated neutral bar.

 

[rjniles]

but Jim, its a sub panel away from the main panel. the sub panels should have isolated neutral bar.

The manufacture does not know how you will use the panel and sets it up to be a main panel. You remove the neutral bonding screw to use as a sub. The ground bar is bonded to the tub in all cases.

 

[Jim Port]

If the sub is fed by a 3 wire feeder and there is no other metallic path between the buildings the neutral is again bonded like a service.

 

[Oso954]

You may not have a ground bar. They are often sold separately.

You may need to purchase the ground bar(s) and mount them in the panel for use as a sub-panel.

 

[concrete_joe]

The manufacture does not know how you will use the panel and sets it up to be a main panel. You remove the neutral bonding screw to use as a sub. The ground bar is bonded to the tub in all cases.

my Eaton sub panel came with neutral bar isolated, had a long green bonding screw in a separate bag.

so i guess its always best to look over the panel and instructions before installing.....:thumbsup:

 

[ritelec]

The manufacture does not know how you will use the panel and sets it up to be a main panel. You remove the neutral bonding screw to use as a sub. The ground bar is bonded to the tub in all cases.

Tub ??:laughing:

Unless you've created an isolated ground.. (which is doubtful), then there would be two.. one bonded to the tub thumbsup and one floating..

 

[rjniles]

You have never heard the metal chassis of a breaker panel called a tub?

 

[dmxtothemax]

You have never heard the metal chassis of a breaker panel called a tub?

NO !
have not heard that term at all ?
They say you learn something new everyday !

 

[Know A Little]

You have never heard the metal chassis of a breaker panel called a tub?

44 years in the trade, NOPE never heard it called that.

 

[jbfan]

You have never heard the metal chassis of a breaker panel called a tub?

Yep, just not much.

 

[Stubbie]

I have !!!! it's a rather confusing term though .. CAN makes much more sense ... :jester:

 

[Mr.Natural]

From what I've read in the past, my personal experiences, and the comments to my original post (thanks, guys), my confusion between what the NEC says and why it says it is ever-growing. I'll try to be succinct.

The panel I'm installing (GE TLM1212CUP) came with both a neutral bar and a ground bar installed. I have an identical one elsewhere that's configured properly (ground bar bonded to the case (tub, can, whatever), and the neutral bar floating. So apparently, this latest one was just assembled incorrectly. I'll switch the bonding screw from the neutral bar to the ground bar.

My understanding of the NEC is that the bond between neutral and ground should only occur at the main breaker panel - never within any device, and never within a sub-panel. In either of the two latter situations, a fault in the neutral line upstream of the device or panel would result in the ground line carrying the return current, with the possibility of it exhibiting a raised potential on all connected grounded equipment, posing a shock hazard. If a sub-panel had neutral bonded to ground and the ground wire in the feed line were of a smaller gauge than the main conductors, then an upstream fault in the neutral line could result in the ground line carrying the return current with the possibility of overload and risk of fire. Therefore, the only place neutral should be bonded to ground (if at all) is at the main breaker panel.

So here's my final state of confusion. If neutral is bonded to ground at the main breaker panel, and if a fault occurs upstream in the neutral line (e.g., at the pole), then the earth ground line will carry the return current, again with the possibility of raised potential on all connected grounded equipment and a resulting shock hazard. In addition, if the earth ground line is of a smaller gauge than the main conductors, an overload could occur with the risk of fire. So this suggests to me that neutral should never be bonded to ground, regardless of what the NEC says.

On the other hand, if neutral is not bonded to ground anywhere, an upstream fault in the neutral line would result in an imbalance in the two power legs (the degree of imbalance would depend, of course, on the power demand on each leg), with the possibility of equipment damage due to over- or under-voltage. So the choice seems to be whether to bow to safety (no bond) or equipment protection (bond).

As I said - am I missing something?

 

[rjniles]

It would be very unsafe to not have a neutral to ground connection. If a ground fault were to occur, the fault current would only flow to the ground rod (or other grounding medium). The resistance to the ground is often high enough that enough current would not flow to trip the breaker.

Example: If the resistance to ground were 10 ohms and the supply voltage is 120 volts, during a ground fault 12 amps would flow and not be enough to trip the breaker.

With a neutral to ground bond at the first disconnect, the resistance is well under 1 ohm and enough current would flow to clear the fault. Example: resistance is 0.5 ohms and supply is 120 volts, fault current would be 240 amps.

 

[Oso954]

So apparently, this latest one was just assembled incorrectly.

If I really believed that my panel was assembled incorrectly, I would take it back.

I googled your panel number. (GE TLM1212CUP) The only hit on it is this thread, which may indicate a typo in the number you provided.

 

[concrete_joe]

It would be very unsafe to not have a neutral to ground connection. If a ground fault were to occur, the fault current would only flow to the ground rod (or other grounding medium). The resistance to the ground is often high enough that enough current would not flow to trip the breaker.

Example: If the resistance to ground were 10 ohms and the supply voltage is 120 volts, during a ground fault 12 amps would flow and not be enough to trip the breaker.

With a neutral to ground bond at the first disconnect, the resistance is well under 1 ohm and enough current would flow to clear the fault. Example: resistance is 0.5 ohms and supply is 120 volts, fault current would be 240 amps.

so, the poco tie of neut & gnd can vary, but if poco N lifts i am guessing the gnd is no longer there either, thus things just wont work because current cannot flow to the N bar, thus any single pole circuit will be "live" on both hot and N wires. having N and gnd tied in the main service panel is needed.

 

[Daniel Holzman]

To the OPS. I think some of your confusion on the topic of bonding neutral to ground is due to your terminology. You use the terms "upstream" and "downstream" several times in your post, but they really have no meaning in an AC system, since current reverses 60 times per second. So I really don't know what exactly you are referring to when you discuss the "upstream" neutral.

Something else to consider. The neutral is bonded to earth ground at the power pole by the utility company. This is done so that the voltage on the neutral is essentially earth potential, typically taken to be zero volts. So at least in principal, touching one hand to the neutral and one to earth should not result in a shock, since they are both in theory at the same potential. In practice, it is certainly common for the neutral to have a voltage on it, usually small, but not zero. This can occur for a variety of reasons, in particular since neutral carries current, and has some resistance, there will be voltage drop along the neutral, so the neutral at the device will be at higher potential (voltage) at the device than at the panel, and will be at higher potential at the panel than it is at the power pole. So grabbing the neutral is never a good idea, but generally the voltage level on the neutral is going to be low.

If there is an open neutral between the device and the panel, your device will not function, since there is no return path for current. If for some reason you bonded the neutral to the equipment grounding conductor (EGC) at the device, well then current would flow along the EGC, back to the panel, and back to the power pole via the (presumably intact) neutral between the panel and the power pole. If the neutral between the power pole and the panel was open, then current would flow between the device and the grounding rod via the EGC, but since the grounding rod typically has relatively high resistance (25 ohms is pretty common), you would have a series circuit connecting the device to ground via the grounding rod, and likely your device would not function properly, if at all, since the combined resistance would be above the resistance of the device alone, hence current would be low, perhaps too low to correctly run the device.

There would certainly be potential for shock if you grabbed the EGC under these conditions, since the EGC would be at something close to 120V, depending on the ratio of resistance between the EGC and the device. And unfortunately the breaker would not trip under these conditions. So clearly you don't want to bond the EGC and the neutral at the device. As for bonding them at the panel, that is for lightning protection only, certainly it is not to allow an alternate path to ground in the event the POCO neutral is lost.

 

[Mr.Natural]

If I really believed that my panel was assembled incorrectly, I would take it back.

I googled your panel number. (GE TLM1212CUP) The only hit on it is this thread, which may indicate a typo in the number you provided.

Yup, you're right; it's TLM1212CCUP.

 

[Mr.Natural]

To the OPS. I think some of your confusion on the topic of bonding neutral to ground is due to your terminology. You use the terms "upstream" and "downstream" several times in your post, but they really have no meaning in an AC system, since current reverses 60 times per second. So I really don't know what exactly you are referring to when you discuss the "upstream" neutral.

...

"Upstream" refers to the source of the power; "downstream" refers to the load. So "upstream" from the main panel, for instance, would be the incoming line, transformer, etc.; "downstream" from the main panel would be house wiring, device, etc.