sparky question about ground and neutral at panel

[mound]

I had a licensed electrician run a 100amp subpanel into my attached garage workshop last summer.   The house is 60 years old and the main panel is pretty much stuffed full.  I was fascinated to see that the large ground and neutral wires feeding the subpanel were both simply connected together using one of those two-wire lugs (rated for copper and/or aluminum) which was simply bolted to the chassis. It looks like he scraped away the paint, drilled a hole and then just used a bolt and nut to attach it. 

There's another 100amp sub-panel in the house which was installed around 2005 for an addition.  Whoever installed that did the same thing - bolted a lug to the chassis for both ground and neutral.  

Got me going down the rabbit hole learning about ground and neutral only being allowed to bond at the main/first panel.  I learned that you're not supposed to put more than one neutral wire under each screw on the neutral bus. In my case, pretty much every screw on the neutral bus has more than one wire, several have both bare copper and white neutral under the same screw.  Given that it's been like this for decades, I'm not looking to change it, just learning. 

My question is, if neutral carries return current and is bonded directly to the chassis, why doesn't the chassis of the main panel get energized? Does it all just go directly to the ground rod in the earth at that point?

[Roly]

I had a licensed electrician run a 100amp subpanel into my attached garage workshop last summer.   The house is 60 years old and the main panel is pretty much stuffed full.  I was fascinated to see that the large ground and neutral wires feeding the subpanel were both simply connected together using one of those two-wire lugs (rated for copper and/or aluminum) which was simply bolted to the chassis. It looks like he scraped away the paint, drilled a hole and then just used a bolt and nut to attach it. 

There's another 100amp sub-panel in the house which was installed around 2005 for an addition.  Whoever installed that did the same thing - bolted a lug to the chassis for both ground and neutral.  

Got me going down the rabbit hole learning about ground and neutral only being allowed to bond at the main/first panel.  I learned that you're not supposed to put more than one neutral wire under each screw on the neutral bus. In my case, pretty much every screw on the neutral bus has more than one wire, several have both bare copper and white neutral under the same screw.  Given that it's been like this for decades, I'm not looking to change it, just learning. 

My question is, if neutral carries return current and is bonded directly to the chassis, why doesn't the chassis of the main panel get energized? Does it all just go directly to the ground rod in the earth at that point?

You are confusing current (amperage) and potential (voltage).   The neutral current MUST return to the transformer supplying it.   It will take all paths to do this,  the neutral wire, water pipes and the earth.  While there is a small voltage present to do this most of the time it is not noticeable other than pools.   The neutral will be connected with the ground at normally the main panel.   For code a lug must be marked to take more than one conductor.    A lot of services used a bare neutral which was grounded at the panel, was this the bare ground you are talking about ?     For parallel ground paths the current will divide inversely to the resistance of the path but the neutral will carry by far the most current.   Roly

[mound]

You are confusing current (amperage) and potential (voltage).   The neutral current MUST return to the transformer supplying it.   It will take all paths to do this,  the neutral wire, water pipes and the earth.  While there is a small voltage present to do this most of the time it is not noticeable other than pools.   The neutral will be connected with the ground at normally the main panel.   For code a lug must be marked to take more than one conductor.    A lot of services used a bare neutral which was grounded at the panel, was this the bare ground you are talking about ?     For parallel ground paths the current will divide inversely to the resistance of the path but the neutral will carry by far the most current.   Roly

Interesting, thanks. "not noticeable other than pools" - can you elaborate on that? What changes with pools?  The neutral bus (smaller screws where all the #12 and #14 NM ("Romex") type cables from branch circuits in the house land) seems to have the whites and bare coppers all landing together on the same bus, some with the white and bare copper together under the same screw.  The sub-panel used a dual rated lug, with two openings/set-screws and the ground went into one, neutral into another, whole lug bolted to the chassis.

[Roly]

Interesting, thanks. "not noticeable other than pools" - can you elaborate on that? What changes with pools?  The neutral bus (smaller screws where all the #12 and #14 NM ("Romex") type cables from branch circuits in the house land) seems to have the whites and bare coppers all landing together on the same bus, some with the white and bare copper together under the same screw.  The sub-panel used a dual rated lug, with two openings/set-screws and the ground went into one, neutral into another, whole lug bolted to the chassis.

Being wet lowers the electrical resistance of the skin which allows more current to flow which is what you feel. Also the area of contact is much greater if you are in water which also lowers your bodies resistance which increase current.  This is where you get into equipotenial grounding where everything is the same potential even if it is not zero.    If everything is the same you do not notice any voltage,  like a bird on the wire.   Roly

[mound]

Being wet lowers the electrical resistance of the skin which allows more current to flow which is what you feel. Also the area of contact is much greater if you are in water which also lowers your bodies resistance which increase current.  This is where you get into equipotenial grounding where everything is the same potential even if it is not zero.    If everything is the same you do not notice any voltage,  like a bird on the wire.   Roly

Gotcha. I can't say I've ever felt any sort of shock getting in or out of a pool!

[Roly]

Gotcha. I can't say I've ever felt any sort of shock getting in or out of a pool!

Thats good.      Most of the time it is a handrail that does not extend into the water and even though it is grounded it is not at the same potential as the water.  Roly

[mound]

Being wet lowers the electrical resistance of the skin which allows more current to flow which is what you feel. Also the area of contact is much greater if you are in water which also lowers your bodies resistance which increase current.  This is where you get into equipotenial grounding where everything is the same potential even if it is not zero.    If everything is the same you do not notice any voltage,  like a bird on the wire.   Roly

Roly, so speaking of pools, interested your thoughts. I've been working through my own pool at my new house trying to understand what was done with the goal being to replace the gas heater with a heat pump.  I've completed all wiring (supply and ground, all insulated #6 copper) for the heat pump and am confident with all that, but then came to a stopping point as I see no appropriate place to connect the chassis bonding lug to.  

I'll preface to say I believe I have a decent grasp now on the difference between equipotential bonding and grounding. Grounding being the failsafe for a failed neutral or other fault condition to get current back to the source with low resistance/trip breakers, and the equipotential bonding to bring the water and all related/nearby metal components and deck to the same potential. (the whole "bird on a wire" concept)..   Correct me if I'm wrong, but these two systems (equipotential bonding grid, and electrical ground) should not be connected to each other, right? Otherwise couldn't some fault deliver voltage into the water or to the ladder etc?

The pool is at least 30 years old (I've reached out to the town to find out exactly how old and hopefully who built it).. It's inground, surrounded by concrete. The pool itself isn't Gunite or concrete, nor is it a vinyl liner pool. So I'm assuming it's fiberglass. It has plastic coping but metal ladder and handles embedded in metal cups and an underwater light in the shell with a metal ring. 

The gas pool heater's chassis bonding lug is connected to a piece of #8 stranded, insulated copper which disappeared into the ground in a thin metal pipe sticking up out of the concrete. Just this single wire into the concrete.  I was assuming this connected to the bonding grid around the pool and had planned to use #8 bare copper between the new heat pump chassis and this point.  I wanted to test that assumption first. 

So I used my multi-meter to test resistance/continuity between various components and this single copper wire that disappeared into the concrete. I used one half of a 150' extension cord to make a very long lead for the meter. 

I showed infinite resistance (no continuity) between this point and any of the metal cups for the handles and ladder as well as the water and wet concrete deck.  But I did show continuity between this point and the underwater metal light ring as well as to the above-ground metal junction box that feeds the light niche..

hmm.. 

So then out of curiosity, I tested between this point and the grounding bus on the main panel, and ground pins on various accessible outlets around the house. All showed continuity and near zero resistance. 

Apparently this lonesome piece of copper wire disappearing into the concrete floor is obviously not attached to an equipotential bonding grid, it's attached to the main electrical system's ground... somewhere... 

So the chassis of the gas heater, while it looks like it's bonded, seems to be just grounded to everything else in the house.  There is nothing connected to the bonding lug on the pool pump motor.  Both pieces of equipment have dedicated, insulated, EGC's alongside their supply lines and it's all more than 5' from the pool. 

Anyway, long story short, in looking for any evidence of equipotential bonding I see none. Everything is grounded, but there doesn't look to be any equipotential bonding system and this setup has apparently been safe for decades. 

I worry that if I were to bring in an inspector, they'd say "you're adding a new heater, it must be bonded to the equipotential bonding grid, but there is none, so you must tear up all your concrete and replace it to include one and you're not allowed to use the pool until this happens.."   That would simply not be feasible (maybe this fear is irrational). 

The heat pump is wired with two hots and a dedicated EGC.  With nothing attached to its bonding lug, I show continuity between the bonding lug and ground, which I hadn't quite expected if ground and bonding should be isolated from each other.  The heat pump must internally already ground its chassis. So it almost seems like, in the absence of an equipotential bonding grid, the heat pump's bonding lug should simply remain empty.  But if it were present, wouldn't this inherently tie together the bonding grid and the main electrical systems ground?  After all, the old heater's bonding lug was grounded to the main electrical system, and the new heat pump's lug is at this point as well, even with nothing connected to it. 

Does any of this make sense? I've permanently disconnected power to the light niche because it was tripping the breaker. If and when I replace it I'll use a transformer and a 24v light and I also plan to replace the metal ladder and handle with non-metallic versions.   

Love to hear your thoughts!

[Roly]

Here is a list of free publications from Mike Holt ( very respected in the electrical field)    LINK   Look towards bottom of list and it will list pools.   As far as grounding the bonding wire I would think it would need to be tied to the ground ask inspector before actual inspection.   If there is no electrical equipment by the pool  it does not need to be grounded but needs the bonding wire as I understand it.  (NOT A POOL EXPERT BY ANY MEANS)   If the equipment is double insulated it may not have a grounding lug. Toward the end of the pool article it talks about equipotential grounding.   A lot to take in and digest.
He has many other free articles available along with paid videos.   Roly

[mound]

Here is a list of free publications from Mike Holt ( very respected in the electrical field)    LINK   Look towards bottom of list and it will list pools.   As far as grounding the bonding wire I would think it would need to be tied to the ground ask inspector before actual inspection.   If there is no electrical equipment by the pool  it does not need to be grounded but needs the bonding wire as I understand it.  (NOT A POOL EXPERT BY ANY MEANS)   If the equipment is double insulated it may not have a grounding lug. Toward the end of the pool article it talks about equipotential grounding.   A lot to take in and digest.
He has many other free articles available along with paid videos.   Roly

Thanks I've actually been reading that very site already. I guess ultimately the question is, what to bond to when the pool itself has no equipotential bonding grid.

[Snipe Hunter]

Thanks I've actually been reading that very site already. I guess ultimately the question is, what to bond to when the pool itself has no equipotential bonding grid.

The pool should have a bonding grid. Ground and neutral should be separated at the sub panel. Neutral not bonded to the sub panel but bonded at the main panel. Otherwise, you can have a hot ground between the 2 panels