I have a question concerning bonding swimming pool grid . The nec says bond all metal parts on the grid, but it also says you don’t have to bond to service equipment, or the pool panel , or the grounding electrodes. So if the grid is not tied to the service how do you clear a fault? Say I drop 2 wire cord in the water , that’s plugged into a gfci. How will the gfci trip if the water bond does not tie to the service anywhere? I realize that when you bond the pool pump motor that the ground wire from the panel to the motor will tie the bonding grid with the ground at the service. But if you have a double insulated motor, how do the 2 get tied together? I always thought that the number 8 wire from the bonding grid needed to be connected to the ground in the panel but I was told not so . What is your thoughts?
Swimming pool bonding.
- Thread starter Rgw 01
- Start date
- Status
- Location
- Chapel Hill, NC
- Occupation
- Retired Electrical Contractor
The grid is not there to clear a fault. It is there to keep all metal objects and the earth at the same potential. Because of stray voltages you can often have a few volts running in the earth. The equipotential bonding keeps all of the pool area at the same potential. You can't get shocked if everything is at the same voltage
Dale001289
Senior Member
- Location
- Georgia
Fault at the motor will typically run via the branch circuit EGC to the motor - not the equipotential grid.
Last edited:
david
Senior Member
- Location
- Pennsylvania
You are correct in your assumption that the grid is required to bond to a min of a 12 awg equipment ground
"Where there is no connection between the swimming pool bonding grid and the equipment grounding system for the premises, this bonding conductor shall be connected to the equipment grounding conductor of the motor circuit. "
in other words if the # 8 green bond ties the grid to the insulated 12 for the lighting branch circuit than you are good. If however there is no place a 12 insulated equipment ground bonds to the grid than you must bond the equipment ground of the branch circuit for the double insulated pump circuit to the grid
Exception: Metal parts of listed equipment incorporating an approved system of double insulation shall not be bonded.
(a) Double-Insulated Water Pump Motors. Where a double-insulated water pump motor is installed under the provisions of this rule, a solid 8 AWG copper conductor of sufficient length to make a bonding connection to a replacement motor shall be extended from the bonding grid to an accessible point in the vicinity of the pool pump motor. Where there is no connection between the swimming pool bonding grid and the equipment grounding system for the premises, this bonding conductor shall be connected to the equipment grounding conductor of the motor circuit.
david
Senior Member
- Location
- Pennsylvania
you extend a min of 12 insulated equipment ground from the pump circuit junction box at the pump location and bond it to the required bonding conductor at the pump location
I new it had to tie back to the equipment ground somehow to be safe. Thanks for the answer.
Thanks for the info.
FionaZuppa
Senior Member
- Location
- AZ
- Occupation
- Part Time Electrician (semi retired, old) - EE retired.
And if the feeder or pump frame loses egc then any fault to the motor case then energizes the whole bonding grid. 240 @ 4mA might be deadly while the gfi just laughs at ya.
depending on where the pump is, what type of water it uses, how often the pump unit gets serviced, the external egc attached to the pump "lug" can become compromised over time. it for this reason i will take a separate bond wire over to the equipment's nearby disco/sub sub panel and attach it to egc there. thus if the pump loses egc your grid does not, etc.
Dale001289
Senior Member
- Location
- Georgia
kwired
Electron manager
- Location
- NE Nebraska
- Occupation
- EC
That is what you want it to do, as long as there are no holes in the grid, users of the pool are not subject to any voltage gradients and don't know there is a problem. Chances are it is still only 120 volts to ground, and chances are the water is conductive enough to easily carry more then 4 mA and will trip the GFCI.
Not impossible for EGC to come loose in the disconnect either, maybe more likely at motor though.
- Location
- Chapel Hill, NC
- Occupation
- Retired Electrical Contractor
hurmpff.. every time I think I know a little something about the Code ....
I thought at one time the code just stated to leave the equipotential bonding conductor long enough to reach a replacement motor that may not be double insulated. Not sure but this rule to connect to the equipment grounding conductor of the motor has been around at least since 2005 NEC
david
Senior Member
- Location
- Pennsylvania
In the past the grid connection to the equipment ground for the premise was typically through the green # 8 bond at a lighting nitch the encapsulated connection in the nitch was a stud that made connection to the bonding grid on the back out side the nitch
Of course with low voltage lighting that # 8 green bond is not required.
Now we have to make that connection at the motor location even if it is double insulated.
Some install the insulated green # 8 bond even when using a low voltage light, but that is rare.
david
Senior Member
- Location
- Pennsylvania
This may have been around the time the international code council and NFPA where coming together in recognizing the NEC as the electrical code for both,
I think around the same time the general purpose rec. and the twist lock motor rec. went from 10 ft for the general purpose and 5 ft. for the twist lock, to 6 ft for both
FionaZuppa
Senior Member
- Location
- AZ
- Occupation
- Part Time Electrician (semi retired, old) - EE retired.
it's not what i would want it to do.....
it's imperative to make sure egc from/to pool stuff is 100%. it's one reason why i advocate that the grid be tied to egc at pump and another location (2 or more separate ties, etc), thus if the bond grid gets energized it will be brought back down to egc value, earth/neutral, and hopefully a ocpd/gfi trips. when you lose the egc the bond grid is ok as long as you dont bridge the ckt near the edge of the grid, etc.
kwired
Electron manager
- Location
- NE Nebraska
- Occupation
- EC
Equipotential bonding isn't about supplementing overcurrent protection. Most of the time you end up with low resistance path back to an EGC anyway, but that isn't the main focus, the focus is keeping current away from pool users, if you have no voltage gradient in/near the pool you have nothing a user can touch that would result in any noticeable current flowing through their body. Yes if the entire network is carrying current the pool user may have a small amount of current but is so small they won't notice.
No different then a bird sitting on a bare high voltage conductor - as long as he can't touch anything of different potential he never knows any different.
FionaZuppa
Senior Member
- Location
- AZ
- Occupation
- Part Time Electrician (semi retired, old) - EE retired.
A bird on a wire that cant touch anything else is not a good example of a pool scenario.
And yes, equipotential everywhere lessons probability of a shock, but why even allow the bond grid to have voltage on it above earth?
And i'll disagree that a bond grid around a pool that is not directly tied to egc has a low ohms path to the egc. Direct tie is zero ohms!
kwired
Electron manager
- Location
- NE Nebraska
- Occupation
- EC
I didn't say it will have low ohms path I said most of the time it likely does. EGC has resistance to it as well, but when when you only need 4-6 mA to make a GFCI trip it can be pretty high in comparison to what is needed to make magnetic trip function of a breaker operate. 6 mA flows when 120 volts finds 20k of resistance. I think most conductive surfaces around a typical pool will have much less then 20k resistance to ground, therefore GFCI's will function regardless of what the situation with EGC' is. And if all those surfaces are bonded together it likely only lessens the resistance, even though you are not intentionally grounding these items the more there are the less resistance to ground there likely will be. That lower resistance isn't the intent of the equipotential bonding, but is a side effect that is still desirable when it comes to supplementing the GFCI protection.
FionaZuppa
Senior Member
- Location
- AZ
- Occupation
- Part Time Electrician (semi retired, old) - EE retired.
we are talking about safety. or in another context, 20k ohms vs zero ohms. with one simple tie i'll take zero ohms.
your 20k is guessing and can fluctuate based solely on weather (rainy for days, 9% RH for months, etc etc). a hard tie is not guessing, etc.
your 20k is guessing and can fluctuate based solely on weather (rainy for days, 9% RH for months, etc etc). a hard tie is not guessing, etc.
kwired
Electron manager
- Location
- NE Nebraska
- Occupation
- EC
All I was saying is 20k is enough to let 6 mA of fault current to flow. I'd bet most of the time the resistance is less then 20k, probably less then 10k quite often.
- Status