Neutral Cable Fault 2

[dbran1949]

I write software that tries to detect problems with power distribution.

1. My understanding about residential power is that a single phase of medium voltage 6K VAC is stepped down by a pole top transformer to 240 VAC. The low side winding is center tapped, the tap is neutral.

2. Three wires are brought to the residence measuring from either "hot" to neutral will read 120 VAC.

3. If I monitor the two 120 VAC legs and detect an intermittent drop on one and a corresponding increase on the other can I assume it is a bad connection on the neutral (center tap) back at the transformer?

4. A little clarification. The sensor is sampling RMS voltage once a second. The problem manifests as random spikes in the data lasting 15 or 20 seconds, then smoothing out. And during this time the sum of the two 120 lines is always 240

Thanks

 

[magoo2]

On your step 3, I wouldn't assume that a drop in one leg and an increase in the other would indicate a bad neutral.

Keep it simple. Assume you start with no load on either leg. If you then connect a load to one leg, it will result in a small drop on that leg. You have essentially 240 V across both legs. If one sees a decrease, the other will see an increase. This is pretty normal.

With a bad neutral, first assume it's a high impedance path. If you again start with no load and apply a load to one leg, you get a much higher voltage drop. On the other leg you get a much higher voltage rise. Across the two legs, you still have essentially 240 V.

With a bad neutral, next assume it's completely open. Applying a load to one leg, will result in a very low voltage, close to zero in some cases. The other leg will make up the difference because you still have 240 V across both legs. When this situation happens with loads on both legs, the load impedance determines how much voltage is on each leg. You still get 240 across both legs. In this situation you can rapidly burn up light bulbs or cause breakers to trip on the high leg.

The location of a bad neutral doesn't necessarily occur back at the transformer. It could be anywhere between the transformer and the service panel.

 

[dbran1949]

magoo2
Thanks this is essentially how I understood it works.

The information about the open neutral is interesting.

The real question is:

If I detect the situations you describe which I would summarize as

1. If a "significant" drop in one leg is simultaneously accompanied by an equal increase in the other there is a problem with the neutral.

2. The significant figure is user settable to some percent of nominal.

Can I assume it can only be caused by a neutral problem or could a high impedance fault in one of the hot legs exhibit the same symptoms?

thanks again

 

[jghrist]

I think you are correct that you have a bad neutral connection. Either it is intermittently bad or the load is intermittently unbalanced.

A load that is balanced between the two legs will have no current in the neutral and there will not be a voltage difference. A high impedance fault in one leg would also cause a differential voltage. Anything causing a significant voltage drop in the neutral could cause the voltage difference.

Can you monitor the current as well as the voltage?

 

[magoo2]

I think you summarized it correctly.

If you had a high impedance fault on a hot phase, you'd draw very little current for a single phase load on that leg. This leads to a negligible voltage on that leg, but you would still get 120 V on the other leg.

If you had a completely open hot phase, you'd get no voltage across one leg and 120 V across the other.

Utilities use something called a beast of burden to check for bad neutrals. You'll get a lot of details on what you measure on each leg if you look at:

http://www.swaindistribution.com/tools/Superbeast.htm

 

[dbran1949]

I'll try to do some clarification

1. It's not that I am trying to determine if I have a bad neutral connection. I am trying to write software that will check a very large number of sensors placed on distribution transformers.

2. If I detect the above mentioned data can I reasonably assume it is bad neutral

3. I develop software but have some understanding about electrical matters. I am not sure how an unbalanced load would cause the above mentioned scenario. The simultaneous intermittent decrease and increase on each leg.

4. If for some strange reason someone was running a 120 VAC kiln (I don't think there even is such a thing) on one leg and only a light bulb on the other.

4a. Would the voltages be unbalanced (greater than some small percent of nominal)?

Remember The data I see is showing the spikes at one second frequency over a 15 to 20 second duration

 

[dbran1949]

magoo2
thanks the link is very helpful, I don't think I can get current readings from the sensors

 

[itsmoked]

Please note that your 6kV statement is not always true. I'd bet most are NOT 6kV. Mine used to be 12kV and is now 21kV.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[jghrist]

I am trying to write software that will check a very large number of sensors placed on distribution transformers.

If the sensors measure voltage at the transformer, they will not detect the voltage difference, which occurs at the load, or at least beyond the point where the bad neutral connection is.

 

[dbran1949]

Ok
Little light bulb goes on (jghrist I don't know how to add your text as a quote)

At the pole top transformer my voltage reading is from each leg to neutral so at that point neutral is fine and each leg will read 120

At the load end, or if my sensor's neutral tie point is on the load side of the bad connection is where I'll see the bad reading.

Since I've seen the data the sensor's connection to neutral must have been on the load side of the bad (corroded and loose) neutral connection

Since the connection now acts like a resistor there has to be a voltage drop across it. So (bear with me I am trying to understand this with only a little training on power) The load side of the neutral connector actually appears to have a voltage now that is not zero. So the sensor (meter) now reads 105 volts from 1 hot leg to neutral because it reads the potential between the two.

Kind of understand that but then how does the other hot leg read 135? that really looses me (I know I have gotten off topic here feel free to point me to a AC power for dummies text)

 

[waross]

Example #1;
The conductors have equal impedance.
You connect a load to one 120 volt leg that causes a 5 volt drop in both the line and the neutral.
The voltages are now>
Load voltage = 110 volts
Line to line voltage = 235 volts.
The voltage on the unloaded line = 125 volts.
The voltage drop on the neutral will add to the voltage on the lighter loaded line.
Example #2;
The neutral conductor is under sized or has a high resistance connection.
Now, the voltage drop under load may be 10 volts but 9 volts may be on the neutral and 1 volt may be on the line conductor.
The voltages are now>
Load voltage = 110 volts
Line to line voltage = 239 volts.
The voltage on the unloaded line = 129 volts.
Does this describe your readings?
respectfully