ECN Forum Forums Code Related Discussion NEC & other Code issues Do you bond your boxes?

Pete brought up bonding in another thread. 250-148(a) requires the EGC to be connected to the box, however, in Nebraska the AHJ does not have a unanimous opinion that this is a good thing, so most inspectors are not enforcing it. We always run a EGC, even where we would be allowed to use the raceway as the ground.

The company opinion is that bonding at the box creates parallel paths for the ground and that's a bad thing. Bottom line is we don't do it because it takes more time and the AHJ doesn't enforce it.

What do you think?

GJ

When working with alternating current, you are dealing with inductive reactance instead of resistance. The formula for calculating inductive reactances in parallel is the same as parallel resistance & if memory serves me correctly, the reactance of the parallel paths is lower than the lowest reactance. This is a good thing since higher fault current will flow and this will operate the OCPD faster.

Even though the EGC isn't required when you're using a grounded metal raceway, the fact that you are installing it means you have to comply with all the rules, including 250-148. If you don't comply, you are leaving yourself quite a liability exposure.

Tom

Why is a parallel fault return path a "bad thing"??? The more current that flows, the quicker the fault is cleared.
don(resqcapt19)

Parallel grounded conductors are bad...
Parallel equipment grounding conductors aren't a big deal...Am I right?

i.e. Iso grounded recepts for electronic equipment are not only OK but are desireable.

But say if you use a bonding screw in the disco and use a bonding screw in the load center and you have used 4 conductor wire between the disco and load center, then the EGC will carry current as the GC does.

This is how I understand it, I'm asking if I have all this right rather than answering your questions.

I always run an EGC because I've seen too many loose screws on EMT Conduit couplers and fittings to be assured that the ground is continuous and has the ampacity required for a dead short ground fault. The resistance of steel seems to be significant to me in that especially long runs, the steel would not have the ampacity required to trip an OCPD in the event of a dead short ground fault, or it seems more likely to become a problem anyway without an EGC.

Clear as mud?

-Virgil
"I say the dumb things, so you don't have to!" (all in fun)

Virgil,

As a rule of thumb (as much as I hate them) you can run about 225 feet of EMT before the reactance of the conduit becomes a problem

If you have a recent copy of "Soares Book On Grounding" see table 11 in the back.

Parallel grounded conductors are great, did you ever try to run anything bigger than 500kcmil?

The sin is when you have a grounded conductor and a grounding conductor in parallel.

I'm with you on the loose screws & fittings. Which is the reason I too pull in an equipment ground, and connect it to every box it passes through.

Tom

[This message has been edited by Tom (edited 05-07-2001).]

Tom;
Ok, table eleven,

"Max length of steel conduit or tubing that may safely be used as an equipment grounding conductor"

I'm kinda in the mud along with Virgil here, can you expand a bit more ???

OK, let me get the mud glasses out.

Take #12 wire in a 1/2" EMT. The 500% column is what you need to get the breaker to trip quickly (how quickly depends on the trip curve of the breaker). So we need to make 100 amps flow to get the breaker to trip (this could take 1 second based on the curve I'm looking at). Looking at the EMT column, if I'm not sloppy about tightening the screws, a run of 246 feet is the suggested maximum leangth. As noted at the bottom of the table, this is for 120 volts. At 277 volts, I'm sure the run could be longer.

I've heard that if you can make 10x current flow through a breaker, most brands will now be in the instantaneous trip part of the curve (another of those blasted rules of thumb).

Tom

Why do you say this?
They are fine, terrific, and specified for avant-garde installations where equipment can be sensitive, namely, computers and such electronics.
"200% neutrals" we call them.

They are also fine so long as they are all headed to the same place with sufficient ampacity to get the current there.

True depending on just how isolated they are.
The grounds still must come together at the point of service.

Correct.
That would be a faulty installation if the neutral and ground were bonded downstream from the point of service.

I like that for a lot of reasons.

If a ground is run in addition to the metallic raceway, i believe that is critical to ensure that raceway joints are tight. If a coupling comes apart and a fault occurs, the non-continuous raceway enclosing the grounding conductor will act like an inductive choke and limit the fault current flow, possibly to the point of not allowing OCPD operation, or maybe burning up the EGC? Also due to the skin effect, with a EGC installed, probably 90% of any fault current will flow on the raceway anyway. At least so I believe.

I don't think the choke effect is a problem when both sides of the circuit are in the same raceway. In a grounding electrode conductor raceway you only have one conductor and the choke effect applies. In a conduit system with the circuit conductors you have both the source of the fault current and the return path in the same raceway.
Don(resqcapt19)