Voltage Drop on Ground

[Steelhead]

I come across a subpanel where the grounds and the neutral land on the same bus and the main bonding jumper is installed. Obviously this setup is incorrect and needs to be corrected. There was no EGC run to the panel so I decided to check the integrity of the wire method and found between a hot conductor and the locknut/fitting thru which the subpanel feeds run there is a measurement of 114V whereas between a hot conductor and the neutral there is a measurement of 123V. I can't see 95% of the wiring method because it's inside the walls. Does anyone think that the wiring method, with the 9 volt voltage drop, is not up to par to serve as the EGC?

Thanks

 

[SG-1]

Steelhead, conduit can serve as the EGC. One loose fitting can destroy its integrity. There should be no volt drop on the EGC unless current is flowing over it. The only current supposed to flow on an EGC is fault current.

What type of meter are you measuring the volt drop with ?

 

[kwired]

Steelhead, conduit can serve as the EGC. One loose fitting can destroy its integrity. There should be no volt drop on the EGC unless current is flowing over it. The only current supposed to flow on an EGC is fault current.

What type of meter are you measuring the volt drop with ?

I agree, if there is a metallic raceway and the EGC's are all landed on the neutral bus then the metallic raceway is a parallel conductor to the neutral. Current has been dividing through the neutral conductor and all other metallic paths since the day this was installed.

 

[Steelhead]

Steelhead, conduit can serve as the EGC. One loose fitting can destroy its integrity. There should be no volt drop on the EGC unless current is flowing over it. The only current supposed to flow on an EGC is fault current.

What type of meter are you measuring the volt drop with ?

I measured the voltage with a multimeter. It looks like there may not be the best connection somewhere along the way. Just wanted an opinion as to what you guys would do if you came across the same scenario. Rewiring may be the only option, but it will be expensive and intrusive.

Thanks

 

[Steelhead]

I agree, if there is a metallic raceway and the EGC's are all landed on the neutral bus then the metallic raceway is a parallel conductor to the neutral. Current has been dividing through the neutral conductor and all other metallic paths since the day this was installed.

Yeah, that's what worries me.:blink:

 

[ron]

Is the feeder to the subpanel run underground or is very long where the ungrounded conductors may be upsized due to volt drop or temperature limitations?

If you do supplement with an EGC, the grounding conductor would get upsized in proportion to the upsizing of the ungrounded conductors.

 

[gar]

130604-1005 EDT

Steelhead:

I don't really understand your first post, but the following is a means to troubleshoot the problem.

You need to establish a useful voltage reference point. Most likely this is at a main panel where earthing ground, EGC, and neutral all terminate at one reliable low resistance point. This means under load conditions that there is next to zero voltage difference between any combination of the three wires. This reference point becomes the reference for all voltage measurements. Connect a fourth wire to this point that is one test lead input to your voltmeter. You can use a #18 wire from a roll of wire, or extension cords connected together.

Use a good multimeter with millivolt resolution. I like the Fluke 27.

From this reference point to any location on any EGC there should be very little voltage. If you connect a 1500 watt 120 V heater, this is about 10 A, from a 120 V source to the any point on an EGC there should not be a large voltage difference. This will be a function of the resistance of the EGC. 100 ft of #12 copper is about 0.15 ohms, and the voltage at 10 A would be about 1.5 V. If you suspect EGC problems, then use a much lower source voltage, 6 V maybe, to generate the test current. A filament transformer with 5 to 10 A capacity and appropriate resistor can provide the test current.

Obviously you don't apply a test current if there is noticeable voltage on the EGC. You simply go looking at various points on the EGC to isolate where is the origin of the voltage.

Neutral voltages will be a function of neutral resistance and neutral current.

The voltage change on hot wires will be the wire resistance times the current change in the wire. So changing a known load here can be helpful.

.

 

[Steelhead]

130604-1005 EDT

Steelhead:

I don't really understand your first post, but the following is a means to troubleshoot the problem.

You need to establish a useful voltage reference point. Most likely this is at a main panel where earthing ground, EGC, and neutral all terminate at one reliable low resistance point. This means under load conditions that there is next to zero voltage difference between any combination of the three wires. This reference point becomes the reference for all voltage measurements. Connect a fourth wire to this point that is one test lead input to your voltmeter. You can use a #18 wire from a roll of wire, or extension cords connected together.

Use a good multimeter with millivolt resolution. I like the Fluke 27.

From this reference point to any location on any EGC there should be very little voltage. If you connect a 1500 watt 120 V heater, this is about 10 A, from a 120 V source to the any point on an EGC there should not be a large voltage difference. This will be a function of the resistance of the EGC. 100 ft of #12 copper is about 0.15 ohms, and the voltage at 10 A would be about 1.5 V. If you suspect EGC problems, then use a much lower source voltage, 6 V maybe, to generate the test current. A filament transformer with 5 to 10 A capacity and appropriate resistor can provide the test current.

Obviously you don't apply a test current if there is noticeable voltage on the EGC. You simply go looking at various points on the EGC to isolate where is the origin of the voltage.

Neutral voltages will be a function of neutral resistance and neutral current.

The voltage change on hot wires will be the wire resistance times the current change in the wire. So changing a known load here can be helpful.

.

Sorry if my original post was unclear.

What I was trying to get across was that the subpanel I encountered had a parallel path for neutral current. The MBJ was installed and all the EGCs landed on the neutral bus. I didn't see any EGC run to the panel with the feeders so I wanted to see if the wiring method could serve as the EGC. I used my multimeter to see what reading I would get between the ungrounded conductors and the fitting of the metallic wiring method. I got a reading of 114V. I then measured between the ungrounded conductors and the neutral conductor to get a baseline measurement or a measurement that I would consider the reference measurement. I got a reading of 123V between the ungrounded conductors and the neutral conductor. There was a difference of 9V between the two measurements. My thought was that there may not be a good connection somewhere in the wiring method. My question was does anyone think the wiring method could serve as a sufficient EGC?

Update: Apparently, with all the wires in this panel, I didn't notice a bare EGC run with the feeder conductors and attached to the back of the panel. I installed an equipment grounding bar, terminated all the EGCs to the grounding bar and removed the MBJ. All should be good now. No parallel path for neutral current. I get a reading of 123V between the ungrounded conductors and the new grounding bar I installed.

Thanks for all your help!

 

[don_resqcapt19]

I don't understand how you can have the conduit bonded to the neutral conductor at the sub panel and have a different voltage reading between the hot and neutral and between the hot and the conduit at that sub panel. If they were actually bonded together at that point, they cannot have a different reading.

 

[Steelhead]

I don't understand how you can have the conduit bonded to the neutral conductor at the sub panel and have a different voltage reading between the hot and neutral and between the hot and the conduit at that sub panel. If they were actually bonded together at that point, they cannot have a different reading.

The fitting wasn't correctly bonded to the panel and I had removed the MBJ before taking measurements. I fixed the poorly bonded fitting.

 

[don_resqcapt19]

The fitting wasn't correctly bonded to the panel and I had removed the MBJ before taking measurements. I fixed the poorly bonded fitting.

Even with that added information, unless you have a lot of current on the conduit, I still can't see how you get a different voltage. Even a very high resistance connection will show the same voltage when using a typical meter unles there is current flow on the conduit. If the testing was done with the bonding removed, there should not be any current flow on the conduit.