*Shallow end, being shocked when grabbing railing*

[JamesW]

I would use the oscilloscope wherever you are testing for voltage so that you can see what the waveform and frequency are.

I would also check the current in AC and DC to see what that shows.

A signal generator would also be helpful to verify which line is responsible for the stray current.

 

[newdude]

would use the oscilloscope wherever you are testing for voltage so that you can see what the waveform and frequency are

I totally agree that there is a scientist somewhere that can decipher it. The hard hat, the hard hats 1st level manager/engineer or the electrician needed to fix it won’t have a clue.

 

[mas985]

The voltage is coming from someplace on the ground bar at the main panel in the house. When I disconnect the ground wire coming from the main panel, everything goes away.

That doesn't mean the voltage is coming from the house side. That just means you are completing the circuit back to ground. Potential is relative. The house ground is connected to neutral which in turn is connected to the power co and every other home so I would consider that to be the reference potential. Everything else should be measured relative to that.

We know that the the deck and the railing are not bonded together properly otherwise there would be no shock. If the railing is bonded properly and the deck has an elevated potential to the house ground/neutral, then when the railing bond is connected to the house ground, there is a potential difference created between the deck and the railing which causes the shock. If you break the bond between the house ground and the railing, the railing is now insulated and cannot pass current to anything so no shock will occur.

More that likely, there is voltage source near the pool that is causing the issue. Is the pool close to your neighbors and do they have landscape lighting?

I think I remember reading that when you turned off the power at your house through the main breaker, the shock was still there but when the neighborhood went dark, it went away. This is likely caused by one of your neighbors.

 

[newdude]

I think I remember reading that when you turned off the power at your house through the main breaker, the shock was still there but when the neighborhood went dark, it went away. This is likely caused by one of your neighbors.

Again, to me, this reeeeeeeally points to one of the other utilities. When the street went dark, so did they. To cross pollinate from the neighbor is step 2 if the street suspect is ruled out.

 

[NowintersinAZ]

The deal about electricity is it will follow all paths to ground. NOT just the easiest or lowest resistance path but all paths to ground. You will obviously get more on the one with the least resistance and less on the one with more resistance.

This is why it could be coming from anywhere or everywhere. I'm hoping on one source and my bets are on an outside source either the neutral, a utility, or inside source at the solar panel system.

I would say the next step is lift the neutral continuity test the boxes to make sure they are not connected in anyway. I am going to guess there is a metal fitting between them where the wires come into the main panel.

If all is clear.
Test for voltage at the pool.
If still present go around the house lifting one at a time the bonds at the utilities.
If still present look into the solar panel being the issue.

I'm going to shy away from a stray voltage from a neighbor, street or any other outside source running through the earth to your house. Reason is when you lift the EGC from the house to the timer it goes away. If it was traveling through the earth it would still be present when lifting the EGC from the house. Your equipotential plane has way more contact with the earth than the silly ground rod you have at the panel.

jm.02

 

[newdude]

I'm going to shy away from a stray voltage from a neighbor, street or any other outside source running through the earth to your house. Reason is when you lift the EGC from the house to the timer it goes away. If it was traveling through the earth it would still be present when lifting the EGC from the house

I think it’s stray voltage from the street traveling in on a utility wire to the house, back feeding the house and ultimately the pool through the common main ground.

2 ways to easily check this is to pull the street coax / phone wires *and* their grounds and test the pool.

or pull the main ground / MGN at the ground rod and see if the ground round is still hot. If it is, the source is another feed / utility sharing the ground rod. If the ground rod isn’t hot, The ground wire / MGN to the panel will be hot and the source is proven inside (although still possibly ‘outside’ like the solar panels or a neighbor). So a better way to say it is to prove it onsite or offsite by one of the 2 ways listed above. If the voltage is ‘onsite’, then pull each of the nuetrals / grounds at the hot bus bar. Test each individual wire pulled and also the bar itself at each step. If it’s only one wire the bar stops being hot when that wire is pulled. If it’s multiples the bar stays hot and the pulled wire is also hot and OP keeps going to prove the 2 or 3 sources.

10 minutes tops to pull the main ground or utilities and we have a direction to zero in on.

 

[mas985]

I'm going to shy away from a stray voltage from a neighbor, street or any other outside source running through the earth to your house. Reason is when you lift the EGC from the house to the timer it goes away. If it was traveling through the earth it would still be present when lifting the EGC from the house. Your equipotential plane has way more contact with the earth than the silly ground rod you have at the panel.

By code, grounding rod must be 8 foot deep to ensure a good ground contact. Surface contact with the ground is a very poor ground unless the soil is soaked. The grounding rod is the true reference. However, potential can be different between two grounding rods in different locations which is one of the reasons why the neutrals are tied to the grounding rods at each house. This makes sure the reference potential for each house in the distribution network is the same as well as for the distribution node itself. There is one power reference for the neighborhood and that is neutral/ground. The power company came out and should have checked both ground and neutral so I think we can say those are ok.

Separating the ground from the bond simply broke the circuit loop from the railing to the deck and isolated one or the other (my guess is the railing was isolated). It doesn't mean that there is anything wrong with the house wiring although any of the scenarios that have been mentioned are possible.

Certainly any underground wire with the insulation compromised, could be an issue. This would include utility, landscaping and even the hot sides of the mains but the power company should be able to detect that last one.

The next course of action, check with the power company to see if there's a load issue. They came out, performed a load test, and everything checked out. The house grounding rod is secure and reads 0V. We've taken out breakers, disconnected all of the wires for the pool motor, timer, outdoor outlets, checked the GFCI's. Nothing, absolutely freaking nothing is leading us to the source.

To me, this indicates that the source is not at the house.

 

[JamesW]

The power company came out and should have checked both ground and neutral so I think we can say those are ok.

This would include utility, landscaping and even the hot sides of the mains but the power company should be able to detect that last one.

The power company really does not always try too hard to figure things out.

They usually just send out the lowest level person to see if there are any obvious problems just to clear the service call.

They check to see if the incoming voltage is in range and they look for sparks or visible damage.

If the voltage is in range and there are no obvious problems like sparks or visible damage, the service person marks the service ticket as all clear, no problems found and the ticket is closed.

Unless you really push for a qualified person, then you cannot expect that the person sent out will do anything worthwhile.

 

[newdude]

Guys. I’m feeling largely ignored here. While then theoretical discussion is fascinating, it’s doing bupkiss to prove anything.

If my feebile 1st level hard hat brain is seeing this wrong with my plan of attack, please say so and stop being polite. I’ll go back to lurking and let the adults continue to speak, with ZERO hard feelings.

 

[mas985]

I would agree that the PC will not troubleshoot the stray voltage but I know they have the equipment to test for proper neutral/ground wiring.

Also, @Newdude, I don't think anyone was ignoring you but your method of testing could lead to a false positive and it is a bit ambiguous. Here are some of the issues that I see:

or pull the main ground / MGN at the ground rod and see if the ground round is still hot.

Ground rod hot relative to what? Voltage is measured between two points, if you are measuring the voltage between the disconnected ground wire and the grounding rod, then that can lead to a voltage difference because of the local ground voltage conditions. The point of the ground wire to the ground rod is to force the ground and neutral to be exactly the same potential. So even if the ground potential is different than neutral, that does necessarily mean that is causing the problem. It just means that there is a local potential difference at the grounding rod location. The ground potential at the pool could be the same or it could be different. At best it leads to an indicator of a problem but is still inconclusive.

If the voltage is ‘onsite’, then pull each of the nuetrals / grounds at the hot bus bar. Test each individual wire pulled and also the bar itself at each step. If it’s only one wire the bar stops being hot when that wire is pulled. If it’s multiples the bar stays hot and the pulled wire is also hot and OP keeps going to prove the 2 or 3 sources.

Again, voltage is measured between two points so what are the two points for each of your measurements? If you just disconnect a neutral from a circuit and the circuit is closed (e.g. the light switch is on), then the voltage from that disconnect neutral to the neutral bar will read 120v because there is continuity between the hot wire and the neutral through the light bulb. You must disconnect both the hot and neutral wiring in the breaker box on each side of the circuit so both are isolated from the mains power. But even if there was a short between a utility wire (phone line, cable, etc), then 120v would also feedback into the utility line and basically blow up anything connected to it (e.g. phone, cable modem) so you should know this right away. I am just having a hard time understanding how this scenario plays out.

I would put in a ground rod at the pool and measure the potential difference between that and the house grounding rod.

 

[JamesW]

I would agree that the PC will not troubleshoot the stray voltage but I know they have the equipment to test for proper neutral/ground wiring.

Unless you specifically request that they check something, it's unlikely that they actually checked anything, especially if it requires any significant time or effort.

Do we know exactly what the power company has done so far?

 

[JamesW]

I think that it would be helpful to measure the current on each hot leg, the neutral and the wire going to the ground rod all at the same time to see what each shows.

The current on on one hot leg should equal the sum of the current on the other hot leg plus the neutral.

The current should be steady for the test.

If the current is fluctuating, you would need a test instrument that could do the math based on an instantaneous reading from both hot wires and the neutral at the same time.

The current on the wire going to the ground rod should be zero or close to zero.

I also think that a signal generator would be helpful to put on each line to see if the signal shows up on the oscilloscope.

 

[newdude]

Ground rod hot relative to what? Voltage is measured between two points, if you are measuring the voltage between the disconnected ground wire and the grounding rod, then that can lead to a voltage difference because of the local ground voltage conditions

Ground rod to actual ground (earth). And ground wire to actual ground.

At best it leads to an indicator of a problem but is still inconclusive.

True but if we get an idea which direction it’s feeding we can proceed further. For example, a false positive with voltage at the ground rod (while disconnected), But zero volts coming from the panel would let us focus outside.

so you should know this right away. I am just having a hard time understanding how this scenario plays out.

I’m not looking for 120v. I’m looking for inductive voltage from a utility that isn’t bonded properly at the street sending stray current to the house. Or similar bleed off voltage from something in/on the home.

 

[newdude]

Further thoughts :

OP pulls the main ground at the panel and the trouble goes away. But it didn’t go away. He simply broke the path of travel with no knowledge of the direction of the travel.

inside at the panel he lost the ability to test for still present voltage as the ground is disconnected. He would need another ground, say a water pipe, which is not bonded to the electric as OP found. (Also providing there is no pex anywhere). So with a new ground source OP could test properly for voltage at the bus bar on the panel. If present we have proved it ‘internal’. If not we have proved it external and the next step is to test everything connected or defacto connected to the main ground rod outside using the earth as the ground side of the meter.

It can still be tricky here with the exponential equalizing and whatnot but we have a better direction to continue here.

 

[NowintersinAZ]

Guys. I’m feeling largely ignored here. While then theoretical discussion is fascinating, it’s doing bupkiss to prove anything.

If my feebile 1st level hard hat brain is seeing this wrong with my plan of attack, please say so and stop being polite. I’ll go back to lurking and let the adults continue to speak, with ZERO hard feelings.

Without knowing where or what is causing the voltage and without all of the tests mentioned we are just a think tank. Any idea could or can lead to a possible solution.

Testing a ground rod for voltage to earth with no wire attached to it will produce no voltage. The only way it could is if the entire property is a voltage gradient (which is possible) and two test leads are far enough apart to pick up on that.

Pulling one EGC or neutral is fine but we still don't have any idea or proof if it's generated inside or outside the property. I believe finding that out is a top priority. After that is figured out, then then the other tests can or might lead to the final chapter of the novel.

 

[newdude]

Testing a ground rod for voltage to earth with no wire attached to it will produce no voltage

The CATV and Phones should be on it also. Or connected to the meter which is connected to the ground rod.

 

[mas985]

Ground rod to actual ground (earth). And ground wire to actual ground.

Ground rod to actual ground (earth)? These are the same thing or should be. The ground rod establishes the ground reference to the local area. How do you measure the potential between a grounding rod to earth without another grounding rod? Grounding rods are deep enough or at least should be to be the established reference. That and the neutral should be the reference against all other voltages.

 

[NowintersinAZ]

The CATV and Phones should be on it also. Or connected to the meter which is connected to the ground rod.

I have never seen a ground rod connected to a meter. That doesn't mean it doesn't happen or exist. The grounding electrode conductor runs from the ground rod to the ground bar in the main disconnect/panel. Then in the main panel the bonding jumper attaches the neutral bar and the ground bar together. With either a green screw or a strap and this is the only place current codes allows the two to be joined together.

A ground rod shall be 8' in contact with the earth and have 25 ohms or less. If it has more than 25 ohms (which in 99.999% of the cases it will) you are required to add a second ground rod at least 6' from the first., but never more than two ground rods no matter the ohm reading. The code realizes that a ground rod is a very poor ground and is not used to save lives.

 

[NowintersinAZ]

The CATV and Phones should be on it also.

Yes, but in the main disconnect or panel.

That's why when I posted to lift the neutral then walk around the house and lift the bonds from the utilities one at a time, then testing. This is all a process of elimination. Things need to be eliminated with testing to narrow down the possibilities.

I say main disconnect because that may or may not be in the panel in some cases. Typical suburban residence will have a meter on the outside of the house then piped to the panel inside the house. In this instance the main breaker is the main disconnect as it is the first disconnect on the property. In the case of a farm you might have a 200 amp disconnect on a pole in the yard that may run to multiple structures on the property. In this case the 200 amp disconnect on the pole is the first disconnect on the property and that's where the ground rod will be located with the grounding electrode conductor running to the ground bar and main bonding jumper tying the neutral and the ground bar together. All other panels on the property are now sub panels and need to have EGCs and neutrals separated.

 

[newdude]

Again, I’m uniquely qualified here in that I find stray voltage for a living. In places / directions where it doesn’t belong. AC, DC, or inductive from either without being a hard cross sending full voltage down any leg. The power company and electricians generally have it easier being one directional 99.99999% of the time and this is why you’ve stumped them both. Also, with all of the above testing, steers my opinions to missing/ bad bonds on one or all of the utilities somewhere down the street, sending inductive voltage to the house using the cable or phone drops as a leg.

We only have one way of zeroing in and that’s to cut it and test both ways. Go further back down the bad leg and sectionalize again. Eventually, Even if dealing with a miles long run, you get there.

Sometimes it’s shot all over and voltage is coming from both sides but each/all of those problems get located by doing the same sectionalizing.