A neighbor wants to put a 60A subpanel in his detached garage, about 30' from the house. He's doing it himself, has already run a UG 1" conduit with 3 #4, & a #8 EGC.
AHJ said he's required to install another grounding electrode @ the garage, but why?? I don't have a good answer for him.
His reasoning is that if the sub was on the fence, 30' further, it wouldn't be required. I can't argue with that.

Electure,

Perhaps someone with access to the history of the NEC could give you the real "why" behind the code article.

The requirement is in 250-32 and in addition, 2 made electrodes will likely be required since this installation must still comply with 250-56.

why?
i ask that a lot too.

why can i run a 60A single circuit to my shed a mile away to weld with, and not do 2 G-rods?

why can i run a 4-wire 1/4 mile in the same building and not do G-rods?

why should i run a 3-wire 10' to an outbuilding and reground with 2 G-rods in close proximity to the serving buildings G-rods ( read into this, I double-dog dare ya!)

why does a disco rate more 'lighting protection" 30' from a mobiler home , that the home itself?

why is 'voltage gradient' not univerasl?

why?

Reminds me of the philosophy student who got an A+ on the final exam.
The only question on the exam was "Why?"
His answer---"Why not?"

it may as well be philsophical, for lack of rationale.....maybe someone would be kind enough to follow thru on Tom's suggestion.

most inspectors call sheds, detached garages and light poles "seperate structures" Has nothing to do with actual electrical theory.

nesparky,

It's interesting that you mentioned light poles. I've heard this before, is that true?

Do they require Gr rods at every pole location?

Bill

The issue of a light pole being a structure has not be specifically addressed in the NEC. The was a code change in the 93 or 96 code in 225-32 to address this issue as far as a disconnecting means at a light pole. It seams that some inspectors, based on calling a light pole a structure, were requiring a disconnect at each pole (structure). 225-32 Exception #3 was added to say that a tower or pole used as a lighting standard can have its disconnect located elsewhere. Maybe we need the same type of exception in 250-32.
Don(resqcapt19)

Come to find out, the subject of my original Q has been absolutely dissected (in gargantuan proportions)in other BB's, without a final consensus in any of them.
It seems like this really needs to be addressed by the CMP, although I don't really know how they would/could reword it.
Any ideas?


[This message has been edited by electure (edited 07-23-2001).]

elecuture,
i would say that looking at what our nieghbors do would be a good start.

Sparky,

Don't try that here in WV, especially in Virgil's neighborhood.

The reason, I think the inspector is requiring the subpanel to be grounded is so the two buildings will have an equipotential ground. That way if a fault occurs on either of the systems, there is no voltage differential between the two. I have done several industrial facilities in that fashion so the entire facility is at the same ground potential.
Just an opinion.

lol!
yes it would seem WV & VT are made of the same asteroid remants!
I meant our international nieghbors, note the MEN system;
http://www.epanorama.net/documents/groundloop/electrical_wiring.html

bill


yes the inspctors here do require ground rods at lite poles- also almost all specs and drawings will show them- usually with #6 bare tied to ground wire from panel

We don't have to put in ground rods here, but my company bonds the pole to the rebar cage in the pole base with a #6. I don't know that it's required though. I've sure seen a lot without.

I am going to take a stab at this even if I get my head bit off. The difference lies in the panel being in a building which is a structure that is meant to be occupied at least part of the time by people. The reason for the additional electrode at the second structure is to limit the voltage difference between the exposed conductive surfaces in the building and the grounded surfaces in the building during lightning strikes and fault conditions. If the earth and the conductive surfaces that will conduct stray currents to it are held to nearly the same potential during these events the hazard to personnel is reduced. The entire purpose of installing a Grounding Electrode System is to avoid a human being becoming the Grounding Electrode Conductor between the exposed conductive surfaces and ground.

One example of the hazard would be the existence of a fault between a ungrounded conductor and a grounded box. The sweaty home owner is kneeling on his garage floor, using a three wire tool on his kid's soap box derby racer. Without the grounding electrode at the garage the voltage drop over the Equipment Grounding Conductor may raise the potential of the drill in the owners hand several volts above the grounded concrete garage floor. If a tenth of an amp flows between the occupants hand and his knee his heart may go into fibrillation. Unless effective CPR is begun at once and a defibrillator reaches him within six minutes he is a dead man. Not likely you say, I was the first due Emergency Medical Technician on the call. The patient was not discovered immediately and died of his injuries. The current was not high enough to burn either his hand or his knee. We only discovered the source of the shock by trying to turn the lights on in the garage to work on the patient. The breaker had tripped. After the medic unit left with the patient I tried to reclose the breaker. It tripped immediately and in the darkened interior of the garage I could see a visible spark between the drill surface and the garage floor. I asked the engine company to secure the scene and came back with my work van. When the cops were finished with the scene I moved the drill to a wooden work bench. I attached a fluke DMM set to hold the highest voltage reading between the drill housing and a scrap of braided copper lightning protection conductor laying in a puddle of normal saline I poured onto the garage floor. The fluke DMM showed a voltage of 39 volts after I reclosed the breaker and it tripped again. I do not know if a fluke DMM is fast enough to actually measure the full voltage rise on the circuit but the voltage rise I did measure was an indication of the presence of a substantial voltage rise during the fault condition. The voltage rise during a lightning strike would have been far greater. There was a water spigot supplied via galvanized iron piping a few feet from the garage. I took a piece of stranded #10 AWG and ran it between the ground terminal of a cord cap and a ground clamp I installed on the spigot riser. I plugged the cord cap in to a receptacle near the garage door close to the spigot. When I then reclosed the breaker the fluke DMM did not measure any significant rise although it did beep as if the voltage had changed. The breaker that tripped was protecting the lighting circuit. The breaker for the circuit supplying the outlets never tripped. The garage did not have GFI receptacles so I plugged the drill into the portable GFI I use to protect myself and repeated the experiment with and without my improvised Grounding Electrode Conductor. Neither one tripped the plug-in GFI. I then took the heat gun I use for heat shrink tubing on UF splices and wired it between Ungrounded Conductor and Equipment Grounding Conductor on the receptacle circuit. With the switch thrown to the high setting of 2000 watts the voltage across the ungrounded conductor and the Equipment Grounding Conductor dropped to 102 volts verses an unloaded voltage of 123 with the same meter. There was no such voltage drop when the same measurement was taken between the Grounded Circuit Conductor and the Ungrounded Circuit Conductor. The voltage dropped only two volts on the normal circuit. So with nineteen volts of voltage drop attributable to the impedance of the Equipment Grounding Conductor it was not hard to imagine how the current necessary to stop the victims heart might have flowed. I cant be sure that a properly installed Grounding Electrode System would have saved that man and I realize that my post incident investigation was crude at best. The police called in people from the state to do a more formal investigation after I told them what I found. I never heard if they got the same kind of results or not. I was told that the fault was traced to an internal failure in a photo voltaic sensor controlling the garage's outdoor lights the accident having occurred at dusk.

U sure can post tdhorne!
I see we have EMS in common. Kinda gives me rectal pucker to be first in on some of those scenes...
( I take it he did not have a shockable rythum)
anyway, are you claiming that a GEC could have cleared the fault? could be a looong thread here....
what condition was the ECG in?
was the drill double-insulated?

as somewhat applicable to electure's original Q i pose the following;

Why can i install a service lateral in rigid steel, 3 wire
triplex to a dwelling, but not it's outbuildings?

fire at will.....

Mo shockable rhythm. Asystole, AED said "shock not recommended start CPR."

No I think the GEC would have kept the drill case too close to ground potential for the voltage between his hand and his knee to overcome the resistance of his body.

The drill was three wire grounded if it had been double insulated he would not have been in contact with the Equipment grounding conductor and he would not have been killed.
--
Tom

tdhorne.......
U were there, I was not, so i will refrain from pickin' apart your call
I hate it when i'm on the recieving end of such a deal...woulda, coulda, shoulda...

however, you could start a thread here; http://www.emsvillage.com/forums/

and kind of weave your two job decsriptions together, you have the most direct link bettweeen the two as ever was posted!
save the electric eel thread, but that's another story.....

And what if he'd been in the same building, but 1000' away? The extra GEC wouldn't be required or allowed by Code.

[This message has been edited by electure (edited 08-05-2001).]

Without getting into Code specifics, I would put the problem with the detached garage like this:
1. If you pull an EGC to the subpanel, then you treat it like a feeder, and DO NOT drive a ground rod, nor do you bond the neutral at the subpanel.
2. If you leave the EGC to the subpanel out, then, you treat the subpanel like a service. You would drive a ground rod, and bond the neutral at the subpanel.
In example #1 the fault current would travel along the EGC, back to the main panel, through the main bonding jumper, out to the utility transformer, back through the MCB, through the feeder breaker, through the feeder, and trip the branch circuit OCP.
Example #2 would basically be the same except the fault current would travel on the neutral from the subpanel back to the main panel.
electure,
I hope the above makes sense to you. I think if you do some research, you'll find it to be true. Good Luck.

SlamTex
I believe you are mistaken about not "driving a ground rod" when you bring an Equipment Grounding Conductor with the conductors supplying the building. If there is more than one branch circuit to or in the structure you must bond the panel or disconnect enclosure to a Grounding Electrode. That electrode could be any of the ones in 250-50. If you ran an Equipment Grounding Conductor with the supply circuit/s you do not bond the grounded conductor to the enclosure. If there are no metallic pathways of any kind between the buildings you could omit the Equipment Grounding Conductor from the supply circuits or feeder and bond the grounded conductor to the enclosure.

would a phone wire bettween buildings constitute a 'mettalic path"

I think the code says yes but that does not make sense to me. A coaxial cable might serve as a parallel path to the grounded circuit conductor but the telephone wiring is not itself deliberately grounded anywhere outside the telephone central office. The protectors at each building are grounded but if they fail they are designed to fail clear leaving a crude air gap lightning protector in place. The only time the telephone wiring would carry the circuit currents is after a power cross. Given there gauge they would last about a second.

Although SlamTex's interpretation of this section is incorrect, I believe it has more "common sense" merit than the actual Code. I see a big difference between an underground feeder with an Equip Grd and an aerial Triplex to an outbuilding.
A 30' length of #8 copper has a lower resistance than the homeowner's knee to concrete connection. In the case of a hot to box short, it will cause the OCPD to trip.
Lightning in this area (OC, CA) is a rarety.
For us to run a 700'+ feeder in a warehouse building is not at all uncommon.
Why then, is the Grounding Electrode required, even though the buildings might be only 1' apart, on the separate structure?
(As a side thought, would I be called if the ground rods were 2' apart from each other?? curious)

intent & theory do not coexist in 250. it would almost be better if 250 was formed into 2 lists, ground it, don't ground it...

case in point,
if i run a 100A 240V branch circuit to an outbuilding, no OCPD's, a branch circuit, let's pick a nice constant load, a hot h20 heater here....no GEC or G-rods required per 250-32.

there is no rationale to that

Just to set the record straight .....

Article 230 B. Overhead Service-Drop Conductors
230-21. Overhead Supply
Overhead service conductors to a building or other structure (such as a pole) on which a meter or disconnecting means is installed shall be considered as a service drop and installed accordingly.

[This message has been edited by Joe Tedesco (edited 08-09-2001).]

Joe,
that would just lead you to 250, and 250-32.
So if you where to generalize and require a G-rod(s) with this, we could have a situation where 30A overhead rates a GEC, 100 a UG does not....

again, confusion....

where is that 99' ad-hoc committee who's agenda was a user-freindly 250????

Joe, I don't understand your post's relevance.
The requirement in 250-32 doesn't reference to 230. It doesn't differentiate between OH or UG. What about your opinion on 250-32?
(I see now. Bill's post of 7-23-01)

[This message has been edited by electure (edited 08-09-2001).]