Bonding Conduit Stub Ups

Status
Not open for further replies.
A

aperez01

Member
On a recently completed project we were required to install a 4/0 bare copper conductor inside every power, control, and instrumentation ductbank bank to bond every conduit stub up with a grounding clamp. The ductbank was constructed of sch. 40 pvc conduit underground and pvc coated rigid conduit risers/stub ups. We installed an insulated ground conductor in every conduit which was bonded at each termination point (i.e. junction box, panel, equipment, etc.)

Is the additional bonding/grounding clamp a new requirement by NEC or just customer preference?
 
Its a job spec. My personal guess its an old spec, perhaps from a power plant, where the orginal system was ungrounded.
The NEC requires EGC in the raceway or the racway.
But depending on the application, whats spec'd may be appropiate.
 
Can you point me in the direction in the NEC section under Article 250 that references the requirement of EGC in the raceway or the racway.

Thanks. Oh yeah the project was a power plant.
 
Exception

Exception

Check NFPA70 250.86 Exception 2

Exception No. 2: Short sections of metal enclosures or raceways
used to provide support or protection of cable assemblies
from physical damage shall not be required to be
grounded
 
Bonding Conduit Stub-Ups

Bonding Conduit Stub-Ups

Note: this is a very long post, I hope it is useful - right or wrong.

The ductbanks were concrete encased PVC with PVC coated sweeps that transitioned to RGS above ground.

The stubs in question were not technically stubs, but conduits emerging from the underground. A stub is a short projection (Webster) of conduit and is terminated with a ground bushing or similar termination fitting. On the stub ends, the project called out bonding the ground bushings together and to the ductbank 4/0 equipment ground conductor (EGC) as well as the ground bus of the enclosed equipment. There was no connection detail for terminating the EGC at conduits that emerged from the UG ductbank and continued above ground to the loads. As mentioned, all of the pipes had EGC's and were continued from underground to the loads in RGS. Most terminated at the load with LFMC/external bond. Some loads were in boxes with myers hubs, some were pe___heads (old school, motor termination boxes to be PC). Some MTB's cast and some with KO's. The PVC sections were very long in most cases.

The direction was to bond all of the PVC coated rigid risers to the ductbank EGC's at the point of emergence from the UG. The direction referred to this as a bonding jumper for the UG PVC sections (most of which ran through UG pull boxes).

This bonding configuration at the riser is a very common method in industrial installations to insure solid grounding and a direct ground fault current path to the feeder protection. Some may consider it a supplemental ground when there is an internal EGC.

The direction considered the riser bonds a code requirement and not extra work. The contractor disagreed citing the internal EGC, ground bushings and bonds. This is clearly 250-86. And most would agree at first glance. But the EGC bonding is supposed to be permanent and provide a "low" impedance fault path. If there is a fault in the UG PVC Ctd steel elbow or near it, the fault path is from the stub through the RGS, the flex bond (of questionable fault strength), the hub or other connection, the bushing bond to the equipment ground lug and into the EGC and back to the feeder protective device. It is my experience that the flex connections and box connections become lose and/or corroded over time and maintenance disconnection/reconnection. This path is questionable.

The bonded ductbank EGC is a direct, low impedance path back to the feeder protection and more reliable.

So, my opinion is that the EGC connected to the load is intended for for equipment case faults, but not always acceptable for the conduit on the downstream side of the UG PVC.

As for the NEC, the UG ductbank wire is required to meet 250.96 and 300.10. If one thinks that the EGC with the flex, bonding, etc. is sufficient to provide electrical continuity, consider (hypothetically) a small non-metallic Hoffman pull box in the middle of a long (say 800') AG metallic run with internal EGC. By the above rationale, bonding would not be required across the box because of the bonding at the load, 400' away. I don't think an inspector would allow it.

My interpretation of 250.96 and 300.10 is that the RACEWAY has to stand on its own and be permanently electrically continuous and provide the most effective fault current path (250.4(A)(5)). The raceway ends at the end of the flex, so the bonding in the load enclosure is not part of the continuity path. In the above installation, PVC ductbank EGC's are required across the plastic to meet this requirement. Internal ECG's may not always meet the requirements.

Closing notes:
There are several other NEC sections that cover this subject.
"Ductbank EGC" assumes the ductbank is a piece of equipment.
The original use of the term "ductbank bonds" was not an accurate use of the term.
The installation is in a landfill, so the underground is hazardous area - very smelly methane. Seals were installed at all stubs and emergences.
I look forward to comments on this subject.
Yes Andrew, its me.

Thanks for taking the time to read this.
larselect
 
This reminds me of that Randy Travis song, "Diggin' Up Bones." :D

So, essentially, you are confirming that without the additional bonding the installation was code compliant?
 
For the most part the NEC/UL recognizes that the RMC wrenchtight is an effective fault current path, it does not distinguish between hubs or locknut to cabinet connections.

The problem is the LFMC, LFMC is only good for 60 amps fault current for sizes 3/4" and above, having the RMC between the PVC and the LFMC will not provide a safe fault path for the RMC if the fault current is above 60 amps, the EGC inside of the conduit would do nothing either unless the fault were to involve it.

the problem is while a external bond can be installed (250.102(E)) outside the LFMC, it would be limited to 6' any other method will run into trouble with 250.134(B) I have seen conduits racked using 250.136(A) method, but again 250.134(B) will come into play with the EGC routing away from the circuit conductors.

The only thing I see that would work would be junction box's at the changeover to the LFMC which would give an internal bonding point from the EGC's to the RMC's, if bonding of the continuous RMC to loads is required, then bonding locknut's could be used.

From an installer's point of view, the design of trying to bond where the RMC emerges from grade would be almost imposable without junction box's installed to give a code compliant access to the EGC ran with the circuit conductors, the lack of instructions is a failure on the designers part as to how to make this bond, and the lack of questioning this missing information is the failure on the installers part.

P.S. Welcome to Mike Holts site, but remember please see the rules and while we can help with discussions like this, as long as it doesn't look like somthing going to court.
 
Last edited:
Ductbank Bond Wire Connections

Ductbank Bond Wire Connections

This is very long, but I think it will be enlightening and invite comments.

The stubs in question are technically not stubs, they are above ground extensions from underground. A stub is a short conduit projection (Webster). Here?s the basics:

Underground concrete encased PVC ductbanks with PVC CTD RGSC 5? plus sweeps at the ends. 4/0 CU for length of ductbank embedded in concrete. Insulated ground bushings are specified on conduit stubs in equipment enclosures, bonded to ductbank bond wire, each other and equipment grounding bus/terminals. There is no detail for connecting the ductbank bond wire at the above ground extensions to RGSC. All conduits have internal EGC?s. The project is generation in a landfill and underground is considered hazardous due to very smelly methane. Seals are installed at all stubs and extensions. Most loads connected via LFMC with bond jumpers.

The direction was given to bond the ductbank bond wires to all of the risers just below the seals in order to provide bonding jumpers across the PVC and bond the conduit downstream of the PVC as required by code. This is a very common industrial method of terminating ductbank bond wires.

The contractor disagreed with the direction and considered it extra work stating that the conduit was bonded through the EGC, ground bushings and flex bond jumpers. This is clearly in accordance with 250.86. At first glance one would agree with the contractor. But, the installation doesn?t necessarily meet 250.96 and 300.10. The bonding must provide raceway continuity and an effective fault current path in accordance with 250.4(A)(5). Here?s the path for a fault in the conduit just downstream of the PVC: through the RGSC, the flex bond (if it can handle the fault), the hub/bushing threads, the hub/bus bond to the equipment ground lug and through the EGC to the protective device. It is my experience that the connections at pe?heads (old school, motor termination box is pc) and similar equipment hook-ups become corroded over time and maintenance may loosen the connections from disconnecting and connecting. The impedance of this connection is questionable and may not be less than 1 ohm. Another problem is that the EGC connection on some equipment is not necessarily permanent because it is disconnected for maintenance.

Consider a hypothetical 800? steel conduit run with an internal ground and a small plastic pull box in the middle. By the above rationale, the box does not require a bond across it. But the fault path is up to 1200? long as opposed to 400?. I don?t think the inspector will buy it.

Section 250.96 requires the metal raceways be bonded where necessary to insure electrical continuity and fault current capacity regardless of supplementary EGC?s (internal). By this section, the steel ends of the UG raceways must be bonded together across the plastic ? the raceway must stand on its own as a ground conductor. 300.10 further requires that the bonding create a continuous electrical conductor. Although the internal EGC back through the flex bond, etc. is an electrical circuit, it does not make the raceway a continuous electrical conductor.

Notes:
There are several other NEC sections that deal with this issue. I think 250.96 and 300.10 are most important.
I look forward to comments.
Yes Aperez, it is me.
 
Sorry George. Althouh an old, abused electrical guy, I'm a newbie.
I posted my stuff twice by accident.

I contend that the ductbank bonds connected to the conduit risers is a code requirement. What say you?
 
Status
Not open for further replies.
Top