New 24ft resin AGP installed over a month ago with a Hayward AquaTrol SWG. Last week, when standing on the ground next to the pool, I reached in the water and felt a mild electrical shock. Through the process of elimination, I discovered this was most likely being created by the SWG because it was only after the SWG was unplugged from it's power source that the shock was eliminated. I contacted my local pool supplier from whom I purchased this and he contacted Hayward and they say it can not be their system. He agreed to let me install a new cell to see if this eliminated this. I installed the new cell today and this did not fix the problem. The system is grounded using the external ground terminals to its own grounding plate that was installed after this issue was discovered. I haven't measured the level of voltage but I can't see this being much over 5V. Has anyone experienced this before? What else can I do to eliminate this? I'm at a loss and considering returning to chlorine. The system is connected to a GFCI receptacle.
Hayward AquaTrol SWG creating electric shock in pool
- Thread starter Dan24ft
- Start date
Well I am no expert but I am good at troubleshooting and have owned saltwater aquariums where shock is a real concern
First off, Dont swim until this is resolved. A GFI is not a fail proof to being electrocuted, as you now know since you got shocked on a GFI circuit!. The only way the GFI is going to protect you is if the circuit "leaks" current to ground. The GFI looks at current in and out of each phase of power and if one is lower than the other it trips because some of the current is "leaking" to ground. If the "leak" is through you, you will be protected, however if the current is flowing through you from one leg and back into the other the GFI isn't going to do anything.
. I would first check the GFI... and I wouldn't even trust the test button. I would actually ground the danged thing out to see if it trips... If you aren't comfortable doing that, then just replace it or buy a tester gfci tester | eBay
Then I would re measure the voltage and double check that it does in fact only happen when the SWG is plugged in. Next I would double check the grounding, Put an ohm meter on the grounding plate of the SWG and the other end to a known good ground, you should have continuity.
You definitely have a very strange issue going on... If everything is properly bonded and the GFI works, the GFI should trip if you are getting shocked while standing on the cement and touching the water.
When you unplug the unit do you mean you are shutting off the power at the GFI or just disconnecting it or are you shutting off power BEFORE the GFI?
First off, Dont swim until this is resolved. A GFI is not a fail proof to being electrocuted, as you now know since you got shocked on a GFI circuit!. The only way the GFI is going to protect you is if the circuit "leaks" current to ground. The GFI looks at current in and out of each phase of power and if one is lower than the other it trips because some of the current is "leaking" to ground. If the "leak" is through you, you will be protected, however if the current is flowing through you from one leg and back into the other the GFI isn't going to do anything.
. I would first check the GFI... and I wouldn't even trust the test button. I would actually ground the danged thing out to see if it trips... If you aren't comfortable doing that, then just replace it or buy a tester gfci tester | eBay
Then I would re measure the voltage and double check that it does in fact only happen when the SWG is plugged in. Next I would double check the grounding, Put an ohm meter on the grounding plate of the SWG and the other end to a known good ground, you should have continuity.
You definitely have a very strange issue going on... If everything is properly bonded and the GFI works, the GFI should trip if you are getting shocked while standing on the cement and touching the water.
When you unplug the unit do you mean you are shutting off the power at the GFI or just disconnecting it or are you shutting off power BEFORE the GFI?
Is the pool and all equipment bonded to a properly buried bonding loop? Since it's a resin pool the water may have to be specifically bonded.
If everything is wired and connected correctly, it should not be able to produce enough potential for you to feel.
If everything is wired and connected correctly, it should not be able to produce enough potential for you to feel.
Welcome to TFP!
I know Canadian rules for pool electrical are different from US rules, but what exactly is a grounding plate?
Is you pool water, pump etc. connected by a bonding wire which goes to a ground rod?
I know Canadian rules for pool electrical are different from US rules, but what exactly is a grounding plate?
Is you pool water, pump etc. connected by a bonding wire which goes to a ground rod?
John, a ground plate serves the same purpose as a ground rod. It probably offers the same surface area as a rod, but you don't have to pound it in 8ft deep. It's about 18" by 10" by about 1/4" thick galvanized plate with ground lug. See following link for example (http://www.homedepot.ca/product/galvanized-ground-plate-c-w-connector/910158). The only items that I was able to bond to the plate was my heat pump and the SWG cabinet. I don't believe I have a ground lug on my pump which is cord connected so grounded this way; it's all plastic enclosure so I believe this would be considered as being double insulated.
Since I wrote this post, I believe I have found the problem. Since my pool is a resin type (the only metal part is the aluminum wall), it can not itself be physically bonded. I started reading online and found a reference to the NEC stating the pool water needed to be grounded or bonded by way of an intentional bonding means. I went out and measured the voltage in my pool water with a meter and was measuring 10VDC (5VAC). I took a piece of 14ga wire and grounded one end to my ground plate and stuck the other end in the water and this dissipated the unwanted voltage. The SWG must dissipate a voltage in the water and a normal metal pool that is bonded would never experience this as it would get dissipated. I have found a few items online that are UL listed for this purpose and my local pool equipment supplier is looking into getting something for me. Hopefully this will address my situation. In the meantime, I make sure the SWG is unplugged when there is someone wanting to go in the pool.
I read that your pump doesn't seem to be able to be bonded...I think the reason most people don't have this problem is because their pump is bonded and thus dissipates the voltage there...sounds like you have a path forward anyway.
That's half the equation. Bonding safely eliminates voltage gradients in and around the pool, but you want to try to identify the source of the voltage. Even though a metal pool is bonded, the water normally doesn't come in contact with the metal structure.
Is this the pump you have? The document references an external bonding connection, so if it is I'd take a look:
http://www.pentairpool.com/pdfs/WhisperFloOM.pdf
http://www.pentairpool.com/pdfs/WhisperFloOM.pdf
No, I have a 1HP Hayward Turbo-Flo II pump. I looked on the pump and there is no bonding lug. I checked the manual and they make no reference to needing to bond the pump, unlike other Hayward pumps such as the Max-Flow.
Oh I'm confident the source of the voltage is the Salt Water Generator. My local supplier was even nice enough to let me try a new cell to make sure the cell wasn't the issue. As soon as I unplug the SWG, the voltage in the water drops to zero.
I would suspect there is more to it than that. Maybe a problem with the wiring or even stray current in the ground around the pool. Even if bonding hides the problem, that isn't a good solution. Kind of like driving your car with no brakes because you've got air bags.
I have a GFI tester and tested both GFI receptacles and both work fine (I have a 20A GFI receptacle for my pump and a 15A GFI receptacle for my SWG). When I say unplugged, I mean I am unplugging the SWG from the GFI receptacle; the SWG is cord connected. Would you suggest I try measuring the residual voltage in the water with the SWG plugged in but the breaker turned off to see if I get the same result? This would eliminate an issue with the feed from the distribution panel I guess. As for measuring the ground continuity, I already did this several time and I have continuity. I just went outside again and measured from the new ground plate to the ground on both GFI receptacles and I have perfect continuity. Even measured from the ground plate to the heat pump disconnect switch enclosure and I get the same result.
Is it possible that since my pump is double insulated, meaning that none of the internal parts of the pump in contact with the water are grounded, or can be, that I am experiencing this issue? Section 680.26(C) of the NEC 2008 does state that an intentional electrical bond of a minimum conductive surface area of 9 square inches be installed and in contact with the swimming pool water. When I ground the water, my issue is eliminated. I have found a few items that are on the market to bond water, the one I am considering one manufactured by Perma-Cast; http://permacastonline.com/pdf/pbbrochure.pdf
Dan24,
Glad to see you are onto the problem.. what was funny is that before I posted my original post I had also suggested grounding out the water but I took out those instructions for fear of being lambasted here for telling yout o basically throw one end of an extension cord into the water and touch the other end to a ground source... hahah...
Good job finding the problem.
Glad to see you are onto the problem.. what was funny is that before I posted my original post I had also suggested grounding out the water but I took out those instructions for fear of being lambasted here for telling yout o basically throw one end of an extension cord into the water and touch the other end to a ground source... hahah...
Good job finding the problem.
Almost seems too simple now that I think about it. I have no idea what else it could be.