Shower shocked!

[quinn77]

I went on an emergency call today. H O was getting jolted when taking a shower and touching cold/hot handles. I took a reading between drain and handles( with water in tub ) and got 25V. Went outside to water main entrance and found new UG pvc water pipe to new well, where 1" steel piping used to be) Well not to my surprise, the rest of the steel water piping for the house was not bonded to GEC. Bonded the pipe and no longer had the 25V reading. I also searched round the attic and found 3 homeruns running most the lenth of the cold and hot steel pipes, within 1/2" or so. Would this be a case of induced voltage with no place to go? I found numerous loose connections on the neutral bus, tightened all neutral connections at panel. The panel is FPE originally installed in '68 with original ungrounded romex. She had a previous electrician install devices/circuitry around 8 weeks ago and I found his work to be sub-par...eventually he tied it in to existing ungrounded circuitry. At one switch location in house I read 30 volts between ground and neutral. I suggested that we disassemble the wiring starting with the most previous work first, since she earlier told me that the shower shock had been progressivly getting worse. Any words on what might be going on here?

 

[nakulak]

turn off all ckts (starting with well) one at a time until voltage disappears, troubleshoot, repeat if necessary. if all ckts off and still get voltage then troubleshoot service.

 

[kwired]

Is drain piping metallic system and if so is it also bonded - this is often easily overlooked as something required to be bonded but if the voltage on the water system should happen to rise for some reason there will be no potential between faucet and drain if they are bonded together.

 

[gar]

100102-2033 EST

First, search for any plastic piping within the home.

If any non-conductive elements exist, then make sure bonding exists around all the non-conductive regions. Bond between hot and cold around the hot water heater. I believe you will find that gas lines are required to be bonded to the rest of the conductive piping systems. I am not a code person so the code people have to instruct you on these details.

Second, make sure there is good bonding between neutral, EGC, and a grounding electrode.

Third, apply power, then using a high impedance digital meter, such as Fluke 27 or 87, measure the voltage between all faucets, drains, conductive tubs and sinks, and EGCs. There should be negligible voltage difference. The EGC bus in the main panel should be your reference. Also measure from the EGC bus to a screwdriver in the earth outside.

A typical reading in my house is about 30 MV between an arbitrary EGC in an outlet and a water pipe. Some of this is high frequency noise, and other noise is 60 Hz from RFI filters to the EGC.

With all these voltages low you have an equipotential system between these components.

Heating ducts or furnace pipes also have to be bonded and checked.

If there is substantial leakage current from some source to any of these components you may be able to use a sensitive clamp-on current probe to detect it.

I suspect the current sensed in the shower was probably less than 1 MA, maybe less than 1/10 MA.

Could there be induced currents from long parallel runs? Maybe. But with proper bonding to make a low impedance equipotential bonded system small currents are not a problem.

.

 

[Psychojohn]

Is it an electric hot water heater? If so, possibly a faulted element.

 

[LarryFine]

At one switch location in house I read 30 volts between ground and neutral.

This suggests a floating/unbonded EGC. Which is hot, the neutral or the EGC? The faucet or the drain? Compare the line-to-neutral voltages.

Remember, when you have a voltage between points which should be common, it helps to know which one is the energized (relative to earth) one.

 

[gar]

100102-2333 EST

If all the conductive elements in the shower are bonded together and to EGC, this includes wire mesh supporting the tile, then there should be no potential difference in the shower. This implies that there is a reasonable length of conductive pipes leading to and from the shower.

Look at the resistivity of typical tap water or even very conductive water, and you will see that leakage current will be shunted to the pipes before getting into the shower.

You treat this as a transmission line with maybe 150 ohms in series with 150 ohms shunt to ground for every centimeter of conductive pipe length. Assume 10 ft of pipe, then the number of stages is about 300. Thus, voltage attenuation is about 0.5^300 . A large amount of attenuation. The assumption is 1 sq-cm of cross section in the pipe.

Try running an experiment with 10 ft of 1/2" pipe, about 1.2 sq-cm. Put an insulated electrode at bottom end of pipe. Fill with tap water. Connect pipe to the neutral, and apply 120 V to the insulated electrode. With a sensitive voltmeter at the top end measure from the water to the pipe.

.

 

[gar]

100103-1810 EST

I have performed the following experiments to provide some verification of my analysis in the previous post.

First a CAUTION. What voltage drop profile exists inside a conductive pipe has no bearing on that in a non-conductive pipe.

Equipment for the experiment.
A 120 V 60 Hz supply with neutral grounded at the main panel, 20 A breaker.
You might want a smaller fuse in series.
Fluke 27 on AC volts range. No filtering.
The Fluke leads were twisted to reduce noise pickup.
23" piece of 1/2" rigid copper tubing mounted vertically.
A 10/32 screw in a piece of Nylon inserted into the lower end of the tube.
When testing 120 V is directly applied to the screw electrode.
The tip of the electrode is 21.5" from the top of the tube.
A piece of Belden 8521 (#16) is used as the voltage probe into the water.
The wire has marks to indicate the distance from the electrode tip.

Current to the electrode was not measured.

Two different samples of water were used. Filtered water with a resistivity of K*4500 ohm-cm. Tap water with 140 ohm-cm.

The voltage measurement is between the voltage probe tip, clean cut end of the Belden wire, in the water and the copper tube.

At 100 MV input the meter current is 0.01 microamp, and 10 MV 0.001 microamp.

Filtered water with 20 times lower resistance than distilled water.
1.5" 53 MV attenuation 1/2272
3.0" 6 MV attenuation 1/20000
Residual seemed about 6 MV.

Tap water with 642 times lower resistance than distilled water.
1.5" 225 MV attenuation 1/533 and lots of gas bubbles
3.0" 105 MV attenuation 1/1142
6.0" 16 MV attenuation 1/7500
12" 3 MV attenuation 1/40000
20" 3 MV

There are obviously differences between the experimental results and my theoretical assumption of a cascaded resistive divider. But note how rapidly the current flowing in the water drops off.

At 3" from the injection point the current is less than 0.01 microamp.

.

 

[ghostbuster]

Also,make sure her well pump ground is not "picking up" stray utility primary ground/neutral current.

 

[LARRY POTTER]

possible solution

possible solution

we ran in to this at a state park everything was bonded grounded and all connected it turned out we had an open grounded conductor between the pole and the main distibution panel

 

[AV ELECTRIC]

Well if this only started happening after the other electrician completed his work I would have to lean to something he did or didn't do Ungrounded system did he ground the new wiring or did he scab onto a water pipe somewhere tied his neutral and ground together in a receptical box . Did the home owner call him back to repair if he refused or the owner didnt trust him you may be wright about being sub par and you are going in the wright direction checking his work thoroughly . It sounds like a bonding issue .

 

[quinn77]

Well if this only started happening after the other electrician completed his work I would have to lean to something he did or didn't do Ungrounded system did he ground the new wiring or did he scab onto a water pipe somewhere tied his neutral and ground together in a receptical box . Did the home owner call him back to repair if he refused or the owner didnt trust him you may be wright about being sub par and you are going in the wright direction checking his work thoroughly . It sounds like a bonding issue .

Yes, I believe starting with "his" work first is best. The other guy installed a new gfi in bathroom, exaust fan on timer, and a switch. Switch and fan/timer were just replaced, gfi was added, and he assured the gfi was grounding. When I found that he tied the 14/2 w ground from gfi into existing circuit in attic and had clipped his new ground and tied in to existing non grounding circuit ( with open splice...no j box ) is when I determined the previous electrician was sub-par. 14 awg has been illegal in that jurisdiction for quit some time, and the quality of his work in general tells me something he did in that house prob has something to do with the voltage im finding on the grounds in house. Ground tied to nuetral is my target. But in a 2200 sq ft house that can be tedious, and selling that to this homeowner who happens to be an engineer for DOW chemical is tough, especially when her electrical engineer friends are telling her its the POCO transformers...which Ive already verified it isn't.

quinn

 

[gaffer]

I've had this once in the past ,it wound up being a screw that had been installed in the past by a jb that finally made contact through the insulation, good luck.

 

[gar]

100104-2208 EST

quinn77:

Your Dow engineer is obviously not an electrical engineer since she is getting information from electrical engineering friends. Her electrical background is probably quite limited. This second hand advice is clearly not very good, they do not have sufficient information.

What kind of meter are you using for your measurements? I suggest a good digital with 10 megohms input impedance and resolution of 0.1 V at 120 V. A 10 V reading will correspond to 1 microampere. A Simpson 260 (5000 ohms per volt sensitivity on AC) will require 200 microamperes to read 10 V on the 10 V range.

You need to determine if there is an earth ground reference. In other words is there a grounding electrode system and in turn is it connected to the neutral bus and ground bus in the main panel. If one does not exist, then start by creating one.

To study problems in the house use the main panel neutral or ground bus as your reference. Obviously there should be no potential difference between these.

Get a long wire that will allow you to make voltage measurements anywhere in the house relative to the main panel ground. Such a wire can be the EGC wire in an extension cord.

Now you can look for non-small voltages on pipes, ducts, fixtures, equipment, EGCs, and neutrals. Under load neutrals might be couple volts. You will learn what is reasonable as you make different measurements. EGCs probably not over 0.1 to 0.2 V, but this could depend upon RF noise that might be present. For example carrier current signalling, X10 stuff or the TED energy monitor.

Measurements like these will avoid a lot of unnecessary guessing.

.

 

[quinn77]

100104-2208 EST

quinn77:

Your Dow engineer is obviously not an electrical engineer since she is getting information from electrical engineering friends. Her electrical background is probably quite limited. This second hand advice is clearly not very good, they do not have sufficient information.

What kind of meter are you using for your measurements? I suggest a good digital with 10 megohms input impedance and resolution of 0.1 V at 120 V. A 10 V reading will correspond to 1 microampere. A Simpson 260 (5000 ohms per volt sensitivity on AC) will require 200 microamperes to read 10 V on the 10 V range.

You need to determine if there is an earth ground reference. In other words is there a grounding electrode system and in turn is it connected to the neutral bus and ground bus in the main panel. If one does not exist, then start by creating one.

To study problems in the house use the main panel neutral or ground bus as your reference. Obviously there should be no potential difference between these.

Get a long wire that will allow you to make voltage measurements anywhere in the house relative to the main panel ground. Such a wire can be the EGC wire in an extension cord.

Now you can look for non-small voltages on pipes, ducts, fixtures, equipment, EGCs, and neutrals. Under load neutrals might be couple volts. You will learn what is reasonable as you make different measurements. EGCs probably not over 0.1 to 0.2 V, but this could depend upon RF noise that might be present. For example carrier current signalling, X10 stuff or the TED energy monitor.

Measurements like these will avoid a lot of unnecessary guessing.

.

Like I said this was an emergency call for a shocking shower, and at $150.00 per hour ( 6:00 pm on new years day-after hour call rate ) she only gave me 1 hour to find the problem, I check my service neutral by loading up the service and taking readings at line side of main. at 130amps I only had a voltage-drop of 6v. I then drug out the extension cord and did exactly what you had suggested and found the areas of the house effected. When I got into the attic is when I found grounding NM tied in to non grounding, tons of cut grounds, loose neutrals, wires stripped with side cutter ( or teeth ) broken inside the wirenuts...just too much to handle in 1 hour by myself. This service is original to the house...1968...FPE 200 amp with a neutral bus, GEC terminated properly to it. Only 1 circuit has been added to the house since, however an additional 1500 sq ft has been added. Its in this addition where the problem originates. PS...Im a service tech, mainly light industrial, motors and controls, water/wastewater/lift-station, the only meters I carry are the Fluke 1589 and a handheld scope.....and yes I bonded water main to make her showers comfortable.

 

[quinn77]

I've had this once in the past ,it wound up being a screw that had been installed in the past by a jb that finally made contact through the insulation, good luck.

had the same thing happen quite a bit, this is built in an area with no AHJ, so any "cut-rate" yahoo can come out and do "electrical work" no prob. this time there are many more variables than "usual". always tricky, always tedious.

quinn

 

[Article 90.1]

homeowner who happens to be an engineer

quinn

Wait a minute, back up, that is the problem. There was no "other electrician" the "other electrician"is the engineer-homeowner. I bet you can get a confession out of them.

 

[quinn77]

Wait a minute, back up, that is the problem. There was no "other electrician" the "other electrician"is the engineer-homeowner. I bet you can get a confession out of them.

yep, prob used me as the problem solver, then called the hubby and supervised the repair...I love Home Depot/ Lowes how to clinics....they keep us in business.