I was showing a friend where to install a subpanel and we got to discussing the need to keep the neutrals and grounds separate in the subpanel. I know this is the way it should be done and the way I always do it but I could not give him a good explanation as to why.
It was explained to me a long time ago and it made sense at that time but now I am at a loss as to a good layman’s reason beyond “because that’s the way it is done”.
Some help is welcome.
Thanks.
Replies
One purpose of the Equipement Ground Conductor, aka - "the ground wire" is to insure that any "exposed metal" in the house is at the same electrical potential.
That include the metal cases of appliacne and tools, water piping, and furacne ducts in most house.
Anytime that you have current flowing in a wire you have a voltage drop and thus anything referenced to that wire will measure different voltage than another wire without that current or a different current.
In cases of a sub-panel you can have a large amount of currently flowiing in the "neutral" connection. And you will have voltage drop. And if that neutral is also used for the ground and you have bad connection you can have upto 120 volts on that "ground".
Thus you might have a kitchen with a refigerator feed off a sub-panel , but the facuetts connected to the water pipe that is referenced to the main panel ground. And if there is a problem you can have a large voltage difference between the refigerator case and the faucet.
Bill: NOW, I finally understand. The explanation kinda has its inconsistancies. Consider a sub panel a mere 30 ft from the main panel. All the discussion is probably purely academic - you don't have resistance in that length wire that is even measurable w/ a standard ohmmeter. But - consider the sub panel in my well house, 200 ft away, being fed by a 50 amp 220V ckt. Now it all becomes real, and a serious problem. Through voltage drop over that distance w/ nearly 50 amps flowing through it, you can get significant drop. Since the Code is generally written in blood, it cannot make those kinds of fine distinctions, and it has to apply equally to the dimensional extremes in building. I am about to put in a sub panel in my shop to feed an electrical on demand water heater. It will be 60 ft from the svc panel, and will be fed through a 125 A breaker. I can now see where that could become dangerous. Ditto for the bldg ground.Thanks.DonDon Reinhard
The Glass Masterworks
"If it scratches, I etch it!"
Thanks to all for the info.
In this particular instance the subpanel is right next to the main. Just ran out of spaces and the Seimens panel won't allow "twins" for the breakers.
Not to get too picky - if there is a new panel being installed with 2 "neutral" bars would it be ideal to use one for all of the neutrals and one for all of the grounds?
The main panel in question has this provision but the neutrals/grounds are intermixed on each bar.
On another note - an electrician said you can hook up 2 circuits to each breaker. He said he does not like to have 2 wires on the breaker but rather a pigtail with the single wire under the breaker screw. He is the one who installed the 200A main panel and is licensed.
This just seems "wrong" to me. Maybe it is okay but I would rather not do it. Any thoughts?
Thanks.
CC: Wrong guy to ask these questions. I are an injuneer, not an electrician. I'd get a new electrician, however. I don't think it is good practice to put two ckts on one breaker, even if allowed, & I don't think it is. Also, again, even if allowed, In my humble opinion, joining two wires within a panel is also poor practice. Also, The two "Neutral" bars are connected by a jumper, so electrically, they are the same - but, again, good technique & practice is to put all grounds on one & all neutrals on the other. That way, if ever you have to make it a 4 wire entry (Hot, Hot, Neutral,Ground) all you have to do is take off the jumper.Inside a panel, neatness counts!DonDon Reinhard
The Glass Masterworks
"If it scratches, I etch it!"
CC: A couple random thoughts, based on Bill Hartmann's discussion of the reason for keeping grounds & neutrals separate in a sub panel. It is obvious why when you have great distances between the main panel & sub panel - but if, as in your case, the sub is adjacent to the main, those ground loops cannot exist, for the same reason a bird can land on a bare high tension cross country line w/o getting fried. The difference in potential along the wire in the short distance between its feet cannot become gret enough to cause any current to flow through the bird. Same should be true for your sub panel & main.But - As I said, the rules are written in blood, & they are hard pressed to make exceptions - code would become so complex it would be unmanageable.I worked for years w/ the DoD Explosive Safety Manual - also written in blood. Lots of it, as a matter of fact. There were a lot of things that made you scratch your head, but again, the writers had to be consistent, or more blood would flow. We had enough problems w/ the regs as written with people doing absolutely dumbrectum things & getting killed as a result. The best you can do is understand the reason why, no matter how obtuse, and drive on w/ the regs.DonDon Reinhard
The Glass Masterworks
"If it scratches, I etch it!"
Don, you said: "CC: A couple random thoughts, based on Bill Hartmann's discussion of the reason for keeping grounds & neutrals separate in a sub panel. It is obvious why when you have great distances between the main panel & sub panel - but if, as in your case, the sub is adjacent to the main, those ground loops cannot exist,"Oh contraire! IF you have two wires in parallel, the current will divide between the two wires based upon their resisance. Now, with the subpanel so close the wire is say 2 feet long and the resistance is negligable. But, the current flowing still divides between the two wires, so if you connect the ground and neutral at both the main panel and the subpanel, the GROUND wire now carries half the neutral current! Big ground loop! If the neutral current to the load is 30 amps, the current in the ground wire is 15 amps and current in the neutral wire is 15 amps.I ran into this a while back and it really lit a light in my brain. The clamp on ammeter read 25 amps on a ground wire, yet I could touch both grounded objects and not get shocked because the #4 wire was very short in length. But there was 25 amps flowing in the ground wire! Obviously not safe. And the ground loop was causing equipment problems, hence why I was troubleshooting the sub panel to begin with.
Frank DuVal You can never make something foolproof because fools are so ingenious.
Frank: Dang!! you are right. Score another one for the electrician over the injuneer! You are also correct about fools - they can screw up anything! Especially if what they are messing around with is inherently dangerous.DonDon Reinhard
The Glass Masterworks
"If it scratches, I etch it!"
On a main panel with the neutral bus bonded to the ground bus it really does not matter which wire lands on which bus. And different electricans have different styles.However, one advantage of keeping them isolated is if sometimes in the future there is a an upgrade or meter is moved or the like and there is a new panel or disconnect up stream then this is already setup to be a sub-panel without moving a lot of wires. But that is not that common.Most breakers are only rated for one wire. However, IIRC, the Square D QO, Homeline, and Cuttler Hammer CH series is rated for two wires in the 15 and 20 (and maybe 30) amp sizes.Look at the side of the breakers and you will see molded into it the size of wire and the number of wires allowed.Likewise look on the panel label and it will indicate the size and number of wires that you can put into a bus bar hole.But the code only allows for ONE wire in a hole with a neutral. If I am not mistaken it used to read so that you could only have one NEUTRAL wire and locally I have see that it is common to put the ground wire and neutral in one hole. I don't have the latest book, but I think that has been reworded to prevent that.You can have several grounds in the same hole if the label allows that.
just look him in the eye and say" If you dont it will pop the little sh$% out of you"
Never though I'd agree with you BB...
but this aint a pol thread!!!
Best reason ever...
"Cause it will SAVE your sorry/dumb a$$!!!"
Mr. T.
There's a steering-wheel in me pants and it's driving me nuts!!!
Another reason is to avoid what are called "ground loops." As described, current flowing through the neutral will pull the neutral "up" above ground to a certain voltage loosely related to the voltage drop in the wire (yeah, the neutral is grounded at the service panel, but there is no such thing as a perfect conductor, so by the time you get all the way to a subpanel, it's pretty much guaranteed the neutral will not be at ground potential -- its internal resistance will allow it to be pulled above ground -- the formula is the trusty/rusty E=IxR or in other words the voltage will equal the current, in amps, times the internal resistance of the wire, in ohms -- it's usually very low, but it is enough to be aware of as the NEC does). If you connect neutral to ground at a subpanel, the voltage in the neutral above ground will flow in the ground wire (in a loop from neutral to ground -- the ground loop). That can be dangerous, but another problem is it will screw up sensitive electronic equipment big time (like big rolling stripes in your TV picture; who knows what it would do to a computer).
As Bill already said, yet another problem is the ground wire back to the service panel has it's own internal resistance. So now the ground, at the subpanel, is pulled by the neutral, improperly connected in the subpanel, above the real "ground" reference at the service panel. Touch the service panel "ground" and the subpanel "ground" at the same time, and that voltage now flows through you. It's usually a very small voltage -- usually. I wouldn't stake anyone's life on that though. Long ago in a previous career when chasing power quality problems on an industrial site, we measured over 70 volts from a metal fence in the ground to the "ground" at a main panel, with basically "unlimited" current available (like many hundreds of amps). Place drew so much power we could measure the ground loop in the (apparently undersized) entrance cables. Probably could have killed someone.
Edit: dug out the code book -- 250.6(A) (can't have "objectionable current" -- ground loop is "objectionable current"), see also 250.6(B), 250.4(A).
Edited 2/25/2006 1:57 am ET by philarenewal
Edited 2/25/2006 1:30 pm ET by philarenewal