odd 400A install, right?

[mjmike]

Was reviewing a 120/240V 1-phase installation that seems odd. On the building exterior, there is 1 set of large service conductors (500MCM possibly) (2 phases and grounded (neutral)) that go thru the meter and enter a non-fused disconnect. Inside the disconnect, the neutral is bonded to 2 grounding electrodes. Leaving the disconnect are 2 sets of 2 phase conductors and neutral with each set going to a different 200A main breaker panel (size looks to be about a 250 MCM). Once you pull a panel cover, you will see the 2 phases and the neutral. It does not appear to be a ground ran to each panel, the conduit is PVC, and there are no ground rods installed at each panel. In each panel, the green bonding screw is installed to bond the neutral bus to ground bus.

I would think the main needs to be fused, would need a ground with each set, and would need to isolate the neutral and ground in each panel. Install is from the mid 90's.

 

[augie47]

A bit odd, but certainly not a unique install.
Over the years I've seen inspectors look at those installs differently.
Taking 230.91 into account some inspectors might consider the outside equipment as "the service disconnect" and require separated neutrals and equipment grounds, Others may consider the outside switch simply a switch and the panels the service equipment and allow the route you saw.

 

[retirede]

Still done around here on rural systems. The meter with breaker on the pole, 2 phase and 1 grounded conductor run to the house. Inside the house is wired as the service disconnect with GEC bonded to neutral.

 

[mjmike]

A bit odd, but certainly not a unique install.
Over the years I've seen inspectors look at those installs differently.
Taking 230.91 into account some inspectors might consider the outside equipment as "the service disconnect" and require separated neutrals and equipment grounds, Others may consider the outside switch simply a switch and the panels the service equipment and allow the route you saw.

230.91 says the overcurrent needs to be integral or immediately adjacent. Lets assume the "immediately adjacent" is stretched and distance is not a factor, now are you saying the outside disconnect is the service disconnect and the 2 remote panels are the "immediately adjacent" overcurrent protection? If so, I can believe that.

Would the 2 panels be considered 2 service disconnects and shouldn't there be a ground of some sorts ran to the 2 panels? If no ground, then should the grounding electrodes go to each of the panels instead of the non-fused disconnect?

Keep in mind, there is no ground wire or grounding electrode going to the 2 panels and the conduit is PVC.

 

[GoldDigger]

230.91 says the overcurrent needs to be integral or immediately adjacent. Lets assume the "immediately adjacent" is stretched and distance is not a factor, now are you saying the outside disconnect is the service disconnect and the 2 remote panels are the "immediately adjacent" overcurrent protection? If so, I can believe that.

Would the 2 panels be considered 2 service disconnects and shouldn't there be a ground of some sorts ran to the 2 panels? If no ground, then should the grounding electrodes go to each of the panels instead of the non-fused disconnect?

Keep in mind, there is no ground wire or grounding electrode going to the 2 panels and the conduit is PVC.

Yes, even if you can justify the consideration of the combination of the outside switch and the inside breakers as being the service disconnect, you would, IMHO, either have to connect the ground electrode system at each of the remote panels or run an EGC from the disconnect to each of the two panels.

It would be a really big stretch to consider that the GEC at the outside disconnect removes the need to bond neutral to GEC or EGC at the inside panels.

 

[augie47]

Yes, even if you can justify the consideration of the combination of the outside switch and the inside breakers as being the service disconnect, you would, IMHO, either have to connect the ground electrode system at each of the remote panels or run an EGC from the disconnect to each of the two panels.

It would be a really big stretch to consider that the GEC at the outside disconnect removes the need to bond neutral to GEC or EGC at the inside panels.

but..
250.24(A)(1)(1) General. The grounding electrode conductor connection shall be made at any accessible point from the load end of the overhead service conductors, service drop, underground service conductors, or service lateral to, including the terminal or bus to which the grounded service conductor is connected at the service disconnecting means.

 

[mjmike]

but..
250.24(A)(1)(1) General. The grounding electrode conductor connection shall be made at any accessible point from the load end of the overhead service conductors, service drop, underground service conductors, or service lateral to, including the terminal or bus to which the grounded service conductor is connected at the service disconnecting means.

Good point, so assuming the 2 panels are the "overcurrent" portion of the service disconnect, the ground does not necessarily need to be present in the 2 panels themselves. The ground can be present and bonded elseware in the system. SO when looking at the panel and only seeing the 2 phases and the neutral, it would be the same as looking at a service panel where the N-G bond is made upstream in the system someplace.

 

[GoldDigger]

Good point, so assuming the 2 panels are the "overcurrent" portion of the service disconnect, the ground does not necessarily need to be present in the 2 panels themselves. The ground can be present and bonded elseware in the system. SO when looking at the panel and only seeing the 2 phases and the neutral, it would be the same as looking at a service panel where the N-G bond is made upstream in the system someplace.

Pretty much. But note that since you have only two hots and a neutral coming in to the two remote panels, there must be (and is) a bond between the ground and neutral bars in the panel to provide an effective fault clearing path for the EGCs that originate in those panels.
There are two variations on the situation with a bond upstream of the service equipment:

1. There is a metallic path (EMT, conduit, etc.) between the bond point upstream and the service panel. You are not allowed an additional bond at the remote panel/service panel. This is not what you have.
2. There is no metallic path and therefore there must be a ground-neutral bond at the service panel(s) too. The only difference is that AFAIK the GEC connection can be either there or upstream also.

 

[infinity]

1. There is a metallic path (EMT, conduit, etc.) between the bond point upstream and the service panel. You are not allowed an additional bond at the remote panel/service panel. This is not what you have.

If the neutral is bonded in the meter enclosure and a metal raceway is run to the service panel the neutral still must also be bonded in the service panel.

 

[GoldDigger]

If the neutral is bonded in the meter enclosure and a metal raceway is run to the service panel the neutral still must also be bonded in the service panel.

:thumbsup:
I was confusing it with the rule for separate buildings.

 

[infinity]

:thumbsup:
I was confusing it with the rule for separate buildings.

Ah that makes more sense.

 

[mjmike]

All makes sense. Thanks.

 

[Smart $]

If the neutral is bonded in the meter enclosure and a metal raceway is run to the service panel the neutral still must also be bonded in the service panel.

Odd way of saying a grounded neutral conductor is always bonded to service equipment enclosures, including the service disconnecting means enclosure through the main bonding jumper.

 

[Smart $]

Good point, so assuming the 2 panels are the "overcurrent" portion of the service disconnect...

Are these panels MCB OR MLO.

If the former, the MCB's are the service disconnecting means (and the non-fused disconnect could be a violation if not required by POCO or locally adopted amendment). If the latter, and there are more than 6 throws in each panel, you have a non-compliant installation (even ignoring the 'immediately adjacent ocp' rule).

 

[mjmike]

Are these panels MCB OR MLO.

If the former, the MCB's are the service disconnecting means (and the non-fused disconnect could be a violation if not required by POCO or locally adopted amendment). If the latter, and there are more than 6 throws in each panel, you have a non-compliant installation (even ignoring the 'immediately adjacent ocp' rule).

The 2 remote panels are MCB. Again, they ran only a neutral, the conduit is PVC and they put the ground bonding screw in at the panels. So the ground and the neutrals at the 2 panels are all connected to the neutral. There are no ground rods at the remote panels.

 

[Smart $]

The 2 remote panels are MCB. Again, they ran only a neutral, the conduit is PVC and they put the ground bonding screw in at the panels. So the ground and the neutrals at the 2 panels are all connected to the neutral. There are no ground rods at the remote panels.

Then the panels (more precisely their MCB's) are the service disconnecting means (SDM)... being the upstream disconnect is non-fused. GEC's do not have to land in the SDM(s) as noted by augie47. The only question I see is whether the non-fused disconnect is compliant.

PS: you did not mention whether or not the disconnect and the panels are on/in the same building or structure. If separate bldg or structure, a grounding electrode system would be required at each.

 

[mjmike]

Then the panels (more precisely their MCB's) are the service disconnecting means (SDM)... being the upstream disconnect is non-fused. GEC's do not have to land in the SDM(s) as noted by augie47. The only question I see is whether the non-fused disconnect is compliant.

PS: you did not mention whether or not the disconnect and the panels are on/in the same building or structure. If separate bldg or structure, a grounding electrode system would be required at each.

They are the same structure. Even if the EGC is established at a different location, you would think there should still be a ground going to the remote panel. Since they put in the bonding screw, the neutral at the remote panel, which can carry current, is also the ground.

 

[GoldDigger]

They are the same structure. Even if the EGC is established at a different location, you would think there should still be a ground going to the remote panel. Since they put in the bonding screw, the neutral at the remote panel, which can carry current, is also the ground.

That is OK in the specific case of service conductors and the service disconnect.

 

[Smart $]

They are the same structure. Even if the EGC is established at a different location, you would think there should still be a ground going to the remote panel. Since they put in the bonding screw, the neutral at the remote panel, which can carry current, is also the ground.

As GD said. It is more important by Code where you establish the EGC system (service disconnecting means) than it is where the system is grounded (service side under NEC, not necessarily in the service disconnecting means enclosure).