A contractor wants to have a "clean ground" at a location about 100' from the Service. The panel located 100' from the Service is being fed using a metal raceway as the EGC. Can the contractor just tie wrap an insulated conductor to the exterior of the conduit and use this conductor as his "clean ground"? I said No but need a code reference to back me up.
You can always say raceways are not to be used as "support" of other wiring methods but that is a pretty thin beef. Why isn't he just running it in the pipe? Even back in the olden days when IBM liked IG outlets we let them run with the circuit conductors. I suspect mythology has trumped reason with this guy.
George, The clean ground is still an EGC and must be in the raceway with the other circuit conductors. 300.3(B). The rule cited by Greg also applies. Is he trying to install an illegal "isolated" ground? Don
Use as it is in this location of the code it becomes a subjective word and in my mind it means in the same raceway, cable or trench.
There is a condition for running the grounding conductor separate from the other conductors as defined in 250.130(C) for receptacles and extentions and I would accept a single conductor for that type of installation. For example where it is a NM cable job. But after seeing what you guy/gals have posted I'm convinced that the proper installation would be to have the IG installed in the same wiring method as discribed above.
I agree George. I am just afraid you will get tangled up in the mythology that got this guy's panties in a wad about the "clean ground" in the first place. You may hear the definition of "with" thrown around. You can certainly toss back that the inside of the raceway is a "cleaner" environment than the outside but that is probably beyond the scope of your job. Is this ground going to be protecting the case of metal equipment or is it just for his electronics in an insulated box. That could be the deciding factor about whether it requires the mechanical protection the raceway provides.
BTW one of the arguments I have heard against IG is that the "one way" penetrations through metal cabinets and raceways that are not bonded at both ends actually creates a choke that reduces the effectiveness of the IG. (similar to GECs) IBM decided it was better to bond the hell out of everything. In the 60s we had single point grounding and we had to do a "baseplate ground" <continuity> check, lifting the single point ground and verifying that was the only path to ground. After all the engineering changes were installed that "single point" was getting up to 20 or 30 additional grounding wires (added to fix problems) we had to remove to clear all paths to ground and sanity finaly won out. They stopped thinking star grounding was important. It was about that time that the IG dissapeared from the physical planning manual. In your typical PC you will have lots of DC to AC ground points once you start hooking up I/O.