High current on neutral with a balanced load

[Fred B]

I've seen most focus to be on the single pole loads being the posible source of the unbalanced return on neutral, that would be the most likely, but i haven't seen any mention of line/line difference on the 2 pole loads (other than well pump that seemed to be balanced from OP). The current on neutral beyond the NEV referenced, may not be an issue, just unbalanced loads, or from faulty or unbalanced loads on 2 poles appliance (electric stoves will carry unbalanced loads from single pole clocks and lights). Might be chasing ghosts in solving reason for light flicker and computer damage that started the OP. As previously mentioned most likely is surge current and or faulty neutral is the cause. (High resistance on neutral to utility not necessarily open neutral, could be responsible for some of the readings, caused by loose or corroded connections, this also showed as variable readings from moment to moment) I have seen this many times especially in barried cable splice. Even had one at utility ground transformer with barried lines neutral connection was loose, this was found even after poco said their system was fine, they only responded to look, after I presented documentation of metered readings.

 

[jhardy13]

You guys are way more educated than I am, so ... if I make you laugh, don't insult me too hard.

It seems to me that the OP has gotten way off track. While it's intriguing to see all the different wave forms, the original problem seems to have either disappeared or has been put on the back burner.

If the power company has a deteriorating neutral in their distribution system, there can be current flowing through their neutral to a very well grounded house panel because the earth is a parallel path for the distribution system neutral.

Jhardy13, did you ever check for a deteriorating neutral coming into your service by checking voltages between L1/N, L2/N, and L1/L2 while unloaded and then again loading just one L?

Let me share this story with you. The PoCo service man showed up at my neighbor's house one day. Being the inquisitive guy I am, I asked my neighbor what was going on. They said they were having some light dimming and flickering problems. The tech checked the voltages and said everything is fine. Well me being me, ... I asked if he could hold on for a minute (he was packing up to leave). I had my neighbor turn on a hair dryer while I was watching voltages and one L went up while the other went down when measured to N and L/L was stable. I told the tech that they had a neutral going bad and everything is not okay. He argued a little but went to his truck and called his supervisor.

He came back with what he called "the beast". Plugged it into the meter socket, did some testing, then told my neighbor they had a neutral problem and a digging crew would be out to find the problem. It turned out there was a splice (that was deteriorating) near the street and they were able to repair it that day.

I had a marina one time that had a voltage between the waterway and the ground of the electrical system. The only way to get rid of that voltage was to disconnect the PoCo neutral. This told me the PoCo neutral had enough VD that a couple of volts developed between the waterway and the PoCo neutral.

Electricity takes all paths.

So, being the dummy I am, I would check voltages while loading one L, and I would check for current on the PoCo neutral with all breakers off. If there is current on the PoCo neutral, then I would try to determine if it's going to ground through the GEC or if it's going to ground through the well connection.

Flickering lights and computer damage are typical signs of a deteriorating neutral. I think this has been covered already, but since this thread keeps going I thought I might repeat some of these things. Smart people sometimes overthink things. Sometimes.

The original problem is still in full swing. Our neutral current is still on average 3-5 amps higher than the difference between both hot legs. My goal in acquiring all of these measurements is to to determine all of the possible ways this could be happening. If their is voltage on my neutral, it seems logical to me that this could be causing excess current to flow through the neutral which would explain the high readings. I am no expert on any of this however. Just speculation. Your neutral theory is interesting and it may be as simple as that. Thanks for taking the time to provide your thoughts.

 

[jhardy13]

I've seen most focus to be on the single pole loads being the posible source of the unbalanced return on neutral, that would be the most likely, but i haven't seen any mention of line/line difference on the 2 pole loads (other than well pump that seemed to be balanced from OP). The current on neutral beyond the NEV referenced, may not be an issue, just unbalanced loads, or from faulty or unbalanced loads on 2 poles appliance (electric stoves will carry unbalanced loads from single pole clocks and lights). Might be chasing ghosts in solving reason for light flicker and computer damage that started the OP. As previously mentioned most likely is surge current and or faulty neutral is the cause. (High resistance on neutral to utility not necessarily open neutral, could be responsible for some of the readings, caused by loose or corroded connections, this also showed as variable readings from moment to moment) I have seen this many times especially in barried cable splice. Even had one at utility ground transformer with barried lines neutral connection was loose, this was found even after poco said their system was fine, they only responded to look, after I presented documentation of metered readings.

It should be impossible to have more amps on the neutral than what my home is drawing unless the additional current is coming from an external source along our mains neutral. For example, if I am drawing 20 amps on L1 and 5amps on L2 there should be 15amps on the neutral. In my case, if I am drawing 20amps on L1 and 5amps on L2, my neutral shows an amperage between 18 to 23. The 15amps are expected because the loads are not balanced. But where is the additional 3 to 8 amps coming from? That is the root of the issue here. I've posted this picture before, but in case you haven't seen it, this is what I am talking about. I want to know where this extra current is coming from. It has proven to be very difficult to solve. Everyone has told me it can't be coming from an external source because when I turn my 200amp breaker off, the neutral current drops to 0.2amps. I am convinced at this point that it must be coming from an external source. Nothing else makes sense.

 

[gar]

200926-1254 EDT

jhardy13:

Going back to your first post. You are listed as an industrial enginering student. That is good, but does not really indicate your background. You may have studied both high school and college physics, but I almost doubt it.

I can now verify that you are dealing with a single phase source with a center tapped secondary. This means you have a two phase output when viewing the transformer secondary with the center tap as a reference. The two phases are 180 degrees apart. This verification comes from your waveform display of both phase voltages in a recent post. Those voltages were 180 degrees apart, thus from a center tapped secondary, and not two phases of a 3 phaase systyem.

With this type of source it is theoretically possible to load the two phases so that the neutral current is 2 times the current of one phase. From a practical perspective it is not quite possible, but could be approached. However, it is a very unlikely loading. This could come from a capacitive load on one phase (very close to 90 degrees is possible), and from an inductive load on the other phase (90 not possible with practical inductors). But this type of loading would be extremely unlikely, and would produce a sign wave current. You have an extremely nonlinear neutral current.

More later.

.

 

[jhardy13]

200926-1254 EDT

jhardy13:

Going back to your first post. You are listed as an industrial enginering student. That is good, but does not really indicate your background. You may have studied both high school and college physics, but I almost doubt it.

I can now verify that you are dealing with a single phase source with a center tapped secondary. This means you have a two phase output when viewing the transformer secondary with the center tap as a reference. The two phases are 180 degrees apart. This verification comes from your waveform display of both phase voltages in a recent post. Those voltages were 180 degrees apart, thus from a center tapped secondary, and not two phases of a 3 phaase systyem.

With this type of source it is theoretically possible to load the two phases so that the neutral current is 2 times the current of one phase. From a practical perspective it is not quite possible, but could be approached. However, it is a very unlikely loading. This could come from a capacitive load on one phase (very close to 90 degrees is possible), and from an inductive load on the other phase (90 not possible with practical inductors). But this type of loading would be extremely unlikely, and would produce a sign wave current. You have an extremely nonlinear neutral current.

More later.

.

I am young and still have a lot to learn. I acknowledge that. Thank you for your input.

 

[oldsparky52]

Everyone has told me it can't be coming from an external source because when I turn my 200amp breaker off, the neutral current drops to 0.2amps. I am convinced at this point that it must be coming from an external source. Nothing else makes sense.

How can it be coming from an external source when it goes away after you turn off your main breaker?

Have you done one line loading test and checked the voltages? Your original problem was the flickering lights and computer damage. That has deteriorating neutral conductor written all over it.

 

[jhardy13]

How can it be coming from an external source when it goes away after you turn off your main breaker?

Have you done one line loading test and checked the voltages? Your original problem was the flickering lights and computer damage. That has deteriorating neutral conductor written all over it.

I agree with you about the possibility of a deteriorating neutral conductor. I will do the one line loading when I get a chance and will post the results. What I still don't fully understand is how the neutral current can be 8 amps greater than the difference between the two hot leg amperages. If it is possible, that is where my understanding is lacking. I want to understand how that can happen.

 

[oldsparky52]

What I still don't fully understand is how the neutral current can be 8 amps greater than the difference between the two hot leg amperages. If it is possible, that is where my understanding is lacking. I want to understand how that can happen.

I agree with your puzzlement. IMO you should throw out all of your testing and mindset and start the testing all over again. You may have more than one problem, or one problem and one interesting condition that is not a problem.

 

[synchro]

... What I still don't fully understand is how the neutral current can be 8 amps greater than the difference between the two hot leg amperages. If it is possible, that is where my understanding is lacking. I want to understand how that can happen.

I believe the maximum RMS neutral current in a 120V/240V three wire system can be as much as √2 times the RMS line current when the line currents are equal.

Consider the somewhat extreme example where the two line currents are pulses that have no overlap in their time waveforms (in other words whenever the magnitude of one line current is > 0, the other line current is zero). Then the squares of their instantaneous currents vs. time also do not overlap. And so the mean square of the sum of the two line current waveforms is equal to the sum of the mean squared values of each line current waveform (because the cross product of the two waveforms is zero). Therefore if the line currents have equal RMS (and therefore MS) values, then the mean square of the neutral current would be 2 times the mean square of the line current. Hence the RMS (Root Mean Square) neutral current would be √2 times the RMS line current.

But I don't see how even this most extreme case of current waveforms would be damaging appliances. You would need a significant overvoltage for that to happen.
I agree with oldsparky52 that a deteriorating neutral conductor is the most likely cause of the symptoms you are experiencing.

 

[jhardy13]

I believe the maximum RMS neutral current in a 120V/240V three wire system can be as much as √2 times the RMS line current when the line currents are equal.

Consider the somewhat extreme example where the two line currents are pulses that have no overlap in their time waveforms (in other words whenever the magnitude of one line current is > 0, the other line current is zero). Then the squares of their instantaneous currents vs. time also do not overlap. And so the mean square of the sum of the two line current waveforms is equal to the sum of the mean squared values of each line current waveform (because the cross product of the two waveforms is zero). Therefore if the line currents have equal RMS (and therefore MS) values, then the mean square of the neutral current would be 2 times the mean square of the line current. Hence the RMS (Root Mean Square) neutral current would be √2 times the RMS line current.

But I don't see how even this most extreme case of current waveforms would be damaging appliances. You would need a significant overvoltage for that to happen.
I agree with oldsparky52 that a deteriorating neutral conductor is the most likely cause of the symptoms you are experiencing.

That is a good explanation, thanks. I have one more question. Why is there a voltage imbalance when measuring from an isolated reference to L1 and L2 when there is no imbalance when measuring from the neutral? I doubt that is related to this, but out of curiosity, what causes this? You can see it in the pictures a couple posts back.

 

[GoldDigger]

The changes in voltage between L1and remote earth (isolated ground) and between L2 and remote earth while L1 to N and L2 to N remain unchanged can only be explained one way: the voltage between N and remote earth is changing.

The stable L1-N and L2-N voltages suggest that there is not a problem with the resistance between the main panel and the POCO transformer center tap.
But the local ground and the neutral are bonded together at the main disconnect and also at the POCO transformer. The local ground electrodes are moving away from remote earth. Remote earth is assumed to be at a constant potential. The best explanation for the offset is that either because of current flow through the ground electrodes or because of current flow or because of a very substantial current flowing through the earth.
When you see the unbalanced current in N, there may be a current through the GES which makes the currents balance.
But that does not tell me where to go in trying to understand it or change it.

Sent from my Pixel 4a using Tapatalk

 

[gar]

200926-2-19 EDT

jhardy13:

To respond to synchro's comment. The reason you can theoretically get twice the current in the neutral is:
1. For a sine wave applied voltage capactive current is a sine wave and leads the applied voltage by 90 degrees.
2. For a sine wave applied voltage inductive current is a sine wave and lags the applied voltage by 90 degrees.

Assuming both currents are equal in magnitude, then, if one voltage lags the other voltage by 180 degrees, then the neutral currents will sum together producing a sine wave current that is double the value of one of the line currents.

Back to jhardy12. In highschool physics I was first introduced to a formal procedure for performing an experiment. This consisted of:
1. Definition of the problem to be studied.
2. Equipment to be used.
3. Procedure for performing the experiment.
4. Results of the experiment.
5. Conclusions.

Before this I was certainly exposed to this procedure, even in first grade, but not in such a formal way.

You have only partially done this in most of your posts. You have not followed a consistent logical approach to your problem. One thing you need to do is simplify your experiment.

From some of your responses to my questions I now know for sure that you have a single phase source with a center tapped secondary. But I don't know if those voltages were loaded or not. I assume unloaded. The waveforms are good and what I might expect, but not as good as the 1960s. Now, if you put resistive loading of 10 A or so on one phase at a time and, then on the other phase what do the voltage and current waveforms look like.

For some reference:
I have a 50 kVA pole transformer that feeds my neighbor and me from a delta primary source.
In my yard between two screwdrivers about 12 ft apart I seldom see more than about 100 mV.

At my main panel a loading of about 10 A on one 120 V phase produces about 0.4 V increase in voltage on the opposite phase, and 0.8 V drop on said loaded phase.

More later.

.

 

[synchro]

200926-2-19 EDT

To respond to synchro's comment. The reason you can theoretically get twice the current in the neutral is:
1. For a sine wave applied voltage capactive current is a sine wave and leads the applied voltage by 90 degrees.
2. For a sine wave applied voltage inductive current is a sine wave and lags the applied voltage by 90 degrees.

Assuming both currents are equal in magnitude, then, if one voltage lags the other voltage by 180 degrees, then the neutral currents will sum together producing a sine wave current that is double the value of one of the line currents.

gar, I understand what you are saying about summing the two reactive currents at the neutral which will add in phase because 0° + 90° = 180° - 90°.

 

[ptonsparky]

I wish my knee was good, COVID was a nonissue, and Joplin were closer. I need a road trip

 

[gar]

200927-1217 EDT

jhardy13:

There is a reasonably orderable path that you can follow to try to determine the cause(s) of your weird current waveforms.

You need to understand the instruments that you are using, and possible limitations.

A Pico ADC 216 I have has low resistance DC conductivity from the BNC input shells to the chassis, and EGC of the computer AC power cord to which the Pico is connected. I suggest using a battery powered computer on battery with no AC power connection for your measurements.

A current transformer or Hall device should be used for current measurement. This provides for DC isolation in that path. A suitable Hall device can provide DC current measurement with DC isolation, but these are noisy.

The circuit you are trying to measure is hard to simplify. Roughly speaking:
1. You have a single phase transformer with a center tapped secondary.
2. The center tap is earthed (grounded) at the transformer.
3. Transformer primary is likely from one phase of a wye source. Wye neutral is likely earthed here also, and many other places.
4. Three distribution wires run from the transformer to your main panel. neutral and two hots.
5. Some place close to the main panel a grounding wire goes ftom neutral to an earthing means, likely one or two ground rods.
6. This earthing path from transformer to main panel is in parallel with the neutral wire of the same path, and is likely a much higher impedance than a good neutral wire. Thus, won't carry much neutral current.

With this distribution source in good condition, and sized appropriately the transformer voltages should be moderately unchanged at the main panel from what they are at the transformer. Thus, knowing what changes happen to the main panel voltages with loading of the main panel is important.

More later.
.

 

[gar]

200927-1670 EDT

jhardt13:

What I have described above pretty much describes my home, except:
1. My grounding rod at the house is about 150 ft of 1.25" copper tubing to the street, water line.
2. My primary supply is delta. No ground wire between power company transformers. Only the earth between the various power company transformer ground rods, home ground rods, and various pipes.
3. Made some neutral, and water pipe current measurements. The results were unexpected, and require study.

More later.

.

 

[gar]

200827-2050 EDT

jhardy13:

Continuing from last post.

I took a scope to my main panel, internal to the scope I synced to the line frequency, and I used a Fluke Hall device current probe on its 20 A range. The line sync provides me with timing information without looking at the line voltage.

Without my about 10 A 120 V unbalanced resistive load (space heater) I had some messy currents on both neutral and the lead to the water pipe. The two waveforms were not the same. There appears to be a capacitive current component at 60 Hz, and higher frequency components. I don't want to turn off most loads in my home tonight.

With the unbalanced 10 A resistive load the neutral current waveform had a major in phase component from the resistive load. If I had a 100 A resistive load it would get closer to a sine wave. I have many different nonlinear loads in my home.

When one guesses at the ground path resistance from pole transformer ground rod to the home grounding electrode you can see why there is likely little neutral current in this path.

More later.

.

 

[oldsparky52]

3. Made some neutral, and water pipe current measurements. The results were unexpected, and require study.

When you can, would you elaborate on this? Thanks.

 

[gar]

200928-1319 EDT

oldsparky52:

I will get back with comments sometime later. I believe I am seeing capacitive current from pole transformer primary to secondary going out on my water pipe ground.

200929-1640 EST

My experimental setup now consists of:
1. An AC outlet at the main panel from a relatively unloaded 120 V circuit from which my scope is powered.
2. A battery powered Fluke Y8100 Hall device current probe. This is old, but good. DC to many kHz. But I am using scope on AC to strip off DC component. Using lowest range, 2 V out equals 20 A in. Saturates somewhat above 20 A. Can clamp around 0000 cable.
3. Scope is Rigol DS2072A with lots of memory.
3A. Horizontal is 2 mS/major div, and vertical is 10 mV/major div.
3B. Sync is internal from AC line voltage. Zero crossing is at center of screen.
3C. Waveform is average of 8 samples per point for displayed waveform. Quite good to reduce fluctuations.

Most of the time I am down in the 10 to 20 mV range for the waveform peak. Waveform varies quite a bit from time to time. I have not yet turned off my loads. Others should start looking at what is occurring at their homes. Not so sure I am seeing capacitive current. Lots to learn.

More later.

.

 

[gar]

200029-2344 EDT

The 8 sample averaging geatly enhances the ability to see a stable waveform. Thus, real load changes are more apparernt.

Just now I was seeing an approximate square wave current, peak 10 mV or 100 mA, with 1/2 cycle centered horizontially (meaning capacitive). Putting my 10 A 120 V space heater on the panel added some resistive current.

.