Grounding bushings

[tortuga]

250.64(E)(1) requires bonding of both ends for a ferrous raceway containing a grounding electrode conductor to said conductor. Other than that NEC is is silent,

I'd add 314.3 Exception requires a bonding means in a non-metallic box to a metal raceway or cable no matter what end:

Exception No. 1: Where internal bonding means are provided
between all entries, nonmetallic boxes shall be permitted to be
used with metal raceways or metal-armored cables.

 

[SSDriver]

That’s good to know the wire size is not a deciding factor. that’s one of the things I here on the job is if the wire size is over 4/0 you need a grounding bushing. I hear people say a lot of different Prerequisites for grounding bushing’s and I don’t think half of them are right.

There is a size requirement for #4(not 4/0) and larger to have a plastic bushing or smooth bushing (not a bonding bushing). Plastic bushings are also required on all rigid conduit regardless of wire size.

 

[electrofelon]

Ah. Understood. What type of application? Normally in my area of work we have the main bonding jumper taking the ground fault back to the neutral.

Say you have a pullbox between a meter socket and a service disconnect. Let's say the pull box was not directly bonded to the neutral/grounded conductor that runs thru it. Let's say there is RGS conduit between the meter socket and the pull box. You could use a bonding bushing at each end of the RGS to bond that box.

 

[infinity]

But there are situations where a service raceway would be a fault current path and a jumper would be required on both ends.

Do you have an example?

Say you have a pullbox between a meter socket and a service disconnect. Let's say the pull box was not directly bonded to the neutral/grounded conductor that runs thru it. Let's say there is RGS conduit between the meter socket and the pull box. You could use a bonding bushing at each end of the RGS to bond that box.

I would have to give this some thought but I'm leaning towards the PB being required to be bonded to the neutral.

 

[electrofelon]

Do you have an example?

I would have to give this some thought but I'm leaning towards the PB being required to be bonded to the neutral.

I don't see why I can't use an SSBJ to bond the box. But I am having second thoughts on whether a raceway can be used as an SSBJ

 

[kwired]

Let's say I have a 277/480 volt feeder inside EMT conduit, between 2 pull boxes. and at each end the the conduit terminates in a concentric / eccentric KO. do you need a bonding jumper on each end?

If the raceway is the EGC yes, that assures continuity of the path, if you pull a EGC through the raceway you only need to bond raceway in one location, though a grounding bushing on one end is likely most common choice, could be done with say an external bonding jumper to a pipe grounding clamp anywhere on the raceway as well.

 

[infinity]

If the raceway is the EGC yes, that assures continuity of the path, if you pull a EGC through the raceway you only need to bond raceway in one location, though a grounding bushing on one end is likely most common choice, could be done with say an external bonding jumper to a pipe grounding clamp anywhere on the raceway as well.

I've always argued that this is true. The EGC provides the fault return path and the bonding bushing on one end bonds the raceway. Problem is I can't find this anywhere in the NEC.

 

[kwired]

I don't see why I can't use an SSBJ to bond the box. But I am having second thoughts on whether a raceway can be used as an SSBJ

Typical meter socket with metal raceway on supply side to a overhead supply is done that way all the time, only bonding of that raceway is normally via the hub to the meter socket enclosure. And in line with the conversation is only bonded at one end.

With service raceways bonding both ends just makes the raceway a better parallel path, which is not an intended current carrying path.

Meter socket to service disconnect, you need to bond the raceway with something other than a standard locknut, but only need one end to be be that way to successfully bond the raceway.

RMC to service disconnect from underground but change over to PVC once underground - same thing, bond the RMC with something other than a standard locknut, and in one location, you generally will do this at the service disconnect end of the run but certainly could run an external bonding jumper to a pipe bonding clamp on the RMC as well.

 

[Dennis Alwon]

@Hitschamp

I see someone posted the need for no 4 to have a bushing. So the people on the job are either confusing 4 awg with 4/0 or you are mishearing it.​

300.4(G) Insulated Fittings. Where raceways contain 4 AWG or
larger insulated circuit conductors, and these conductors enter
a cabinet, a box, an enclosure, or a raceway, the conductors
shall be protected by an identified fitting providing a smoothly
rounded insulating surface, unless the conductors are separated
from the fitting or raceway by identified insulating material
that is securely fastened in place.
Exception: Where threaded hubs or bosses that are an integral part of a
cabinet, box, enclosure, or raceway provide a smoothly rounded or flared
entry for conductors.
Conduit bushings constructed wholly of insulating material
shall not be used to secure a fitting or raceway. The insulating
fitting or insulating material shall have a temperature rating
not less than the insulation temperature rating of the installed
conductors.

 

[electrofelon]

Typical meter socket with metal raceway on supply side to a overhead supply is done that way all the time, only bonding of that raceway is normally via the hub to the meter socket enclosure. And in line with the conversation is only bonded at one end.

With service raceways bonding both ends just makes the raceway a better parallel path, which is not an intended current carrying path.

Meter socket to service disconnect, you need to bond the raceway with something other than a standard locknut, but only need one end to be be that way to successfully bond the raceway.

RMC to service disconnect from underground but change over to PVC once underground - same thing, bond the RMC with something other than a standard locknut, and in one location, you generally will do this at the service disconnect end of the run but certainly could run an external bonding jumper to a pipe bonding clamp on the RMC as well.

But it seems in those cases you mention it's different. There the conduit is bonded, but it's not being used AS a SSBJ to bond other things in the system. For the situation I mentioned in post #23, the conduit actually is a SSBJ. I'm not dead set on this, I have to think about it some more.....

 

[Grouch1980]

If the raceway is the EGC yes, that assures continuity of the path, if you pull a EGC through the raceway you only need to bond raceway in one location, though a grounding bushing on one end is likely most common choice, could be done with say an external bonding jumper to a pipe grounding clamp anywhere on the raceway as well.

If you pull an EGC wire through the raceway though, wouldn't that EGC wire be considered the alternate / redundant pathway for a ground fault? Meaning the raceway is still the primary pathway... so it's still required to be bonded on both ends, not just one end?

 

[tortuga]

If you pull an EGC wire through the raceway though, wouldn't that EGC wire be considered the alternate / redundant pathway for a ground fault? Meaning the raceway is still the primary pathway... so it's still required to be bonded on both ends, not just one end?

An Equipment Grounding Conductor (EGC) is bonded to the metal boxes in that case. At least the ones containing splices Per 250.148.
A ground bushing is not required for that

 

[kwired]

If you pull an EGC wire through the raceway though, wouldn't that EGC wire be considered the alternate / redundant pathway for a ground fault? Meaning the raceway is still the primary pathway... so it's still required to be bonded on both ends, not just one end?

Yes the green wire assures there is continuity regardless whether there is questionable connections in the raceway, but you are still required to bond the raceway so if there is questionable raceway connections on both ends of the raceway you definitely would need additional bonding method in at least one place to assure the raceway is bonded. The green conductor per NEC minimum requirements is redundant if you have metal raceway other than a limited number of situations, one being in health care locations. Might even need to read between the lines a little there, it says the wiring method itself must qualify as an EGC, plus you need to add the green conductor. If the wiring method itself must qualify as an EGC a little reading between the lines might suggest bonding both ends might be necessary if they are potentially impaired connections?

 

[tortuga]

Here is the flaw in that flow chart, say you have some metal flex or EMT leaving a loadcenter (metallic) going to a plastic J box., even if its a 240V circuit, you run into 314.3, which prohibits non-metallic boxes in general then allows them by exception

 

[packersparky]

But it seems in those cases you mention it's different. There the conduit is bonded, but it's not being used AS a SSBJ to bond other things in the system. For the situation I mentioned in post #23, the conduit actually is a SSBJ. I'm not dead set on this, I have to think about it some more.....

I agree with your hesitance on making a statement on this. 250.30(A)(2) allows a non flexible metallic raceway to serve as the supply supply bonding jumper for a separately derived system. 250.92 is not clear at all as to if it is allowed for a service. That needs to be clarified.

 

[kwired]

Here is the flaw in that flow chart, say you have some metal flex or EMT leaving a loadcenter (metallic) going to a plastic J box., even if its a 240V circuit, you run into 314.3, which prohibits non-metallic boxes in general then allows them by exception


View attachment 2560870

GEC inside >> yes>>>required on both ends - needs to include the fact it must be bonded to the contained GEC

 

[Grouch1980]

Yes the green wire assures there is continuity regardless whether there is questionable connections in the raceway, but you are still required to bond the raceway so if there is questionable raceway connections on both ends of the raceway you definitely would need additional bonding method in at least one place to assure the raceway is bonded. The green conductor per NEC minimum requirements is redundant if you have metal raceway other than a limited number of situations, one being in health care locations. Might even need to read between the lines a little there, it says the wiring method itself must qualify as an EGC, plus you need to add the green conductor. If the wiring method itself must qualify as an EGC a little reading between the lines might suggest bonding both ends might be necessary if they are potentially impaired connections?

Unfortunately I did not follow.

 

[kwired]

Unfortunately I did not follow.

Forget the last part with reference to health care. The general rules are the raceway must be bonded as well as other non current carrying metallic boxes, cabinets, appliance frames, etc. If a raceway has solid metal to metal contact with a box, cabinet, etc. that normally is acceptable bonding method. If that connection is impaired in any way, (on both ends of raceway or anywhere in the run) then additional bonding methods need to be utilized. one point of bonding will be sufficient to facilitate handling fault current should an ungrounded conductor fault to the raceway, won't matter if that point is at one end, both ends, or even somewhere in the middle of the run.

A GEC in ferrous raceway must be bonded at both ends, for different reasons than bonding for EGC purposes. A single conductor in a ferrous container (the raceway) will have magnetic effects on the container when current is flowing this will turn the entire setup into an inductor and will have an impedance (resistance to current flow) that is not normally desired. Bonding both ends to the conductor puts the raceway in parallel with the conductor and lessens the inductive effects.