Open Neutral Readings

[PhaseShift]

I am aware that with an open neutral on a 120/240V service the voltage across the loads will be determined by whatever the resistance of the loads are set up as a voltage divider across the total 240V from the panel. These loads can see much greater than 120V when the neutral is disconnected.

My question is, I know when when a neutral is lost the L-N voltage readings will indicate the voltage on the loads set up by the voltage divider, however what will the L-G readings be for these loads? Will these readings be determined by where the measurement is taking place? Since the ground is bonded to the neutral should the L-G readings still be the correct L-N readings?

 

[charlie b]

It depends on where the neutral connection was lost, and where you are taking your L-G and L-N readings. If, for example, the utility's neutral connection to the service is good, and if the N-G bond at the service is good, and if the neutral wire serving a set of loads is disconnected internal to the service panel, then any L-N readings taken near the loads (i.e., far from the point at which the neutral was lost) will not be the same as L-G readings taken at the same location.

 

[e57]

I'll also add that a neutral connection can be resistive +/or intermittent and only do so under load. Making these types of connection loss on a neutral harder to find.

In order to find these elusive types of failure - it may be necessary to HEAVILY load the circuit on ONLY one leg or EQUALLY with a disposable load - regular or halogen lamps or an old heater....

Once the connection gets hot, you can change each leg voltage to N or G by moving load from one leg to another. If the loose connection is not totally obvious - as it will be normal on one side of the bad connection, and all screwy on the other - sometimes it can be hard to find it in a panel because you might not measure with your leads in the right place, or the connection only weakens under a very heavy load - you can find it under load with a IR thermometer. The bad connection will be hotter than the area around it...

It is also best to draw out a one-line or line diagram to assist in visual referance to help troubleshoot. By writing down readings and where they were taken you can narrow down the possibilities.

 

[gar]

090825-0807 EST

PhaseShift:

You have indicated you are an electrical engineer. So the following should be natural for you.

Draw an equivalent circuit as follows:

A center tapped transformer secondary, 120-0-120 V. Connect the center tap to a ground rod at the transformer, and to a neutral wire to the main panel. Two hot wires to the main panel. Obviously there is a meter in the path. Maybe an outside disconnect, and/or maybe a main breaker or fuses in the panel as a disconnect. There is an earth ground at the main panel from the neutral and EGC bus. Some locations there may be slight variations on some of these details.

Neglect inductance and assume resistance for the transformer secondary and all wiring and devices.


Assume 0.03 ohms for each half of the secondary (120*0.05/200=0.03). If you know more about the transformer, then a better estimate is possible.

Assume 0.01 ohms for each hot and neutral wire. For your real application you should make an estimated calculation of the resistance. Neglect the meter, main breaker, and main panel bus bars.

Assume 25 ohms from transformer center tap to the main panel neutral via an earth ground path. This could be much higher or lower. You need information on your soil type and other factors to make a better estimate, or make measurements.

Assume equal size wires for hot, neutral, and EGC on any branch circuit. #14 wire is 2.53 ohms/1000 ft at room temperature. And #12 is 1.59 ohms/1000 ft.

Consider a 50 ft run of #12. The resistance is 0.05*1.59 = 0.08 ohms.

Often times you do not have an open neutral, but rather one with a higher than normal resistance and likely varying. Sometimes it may be arcing.

If a 12 A 120 test (10 ohms, a 1500 W electric heater) is done on the above assumptions and all wires and connections are good, then the change in voltage at the end of the 50 ft branch circuit would be about (0.03+(2*0.01)+(2*0.08))*12 = 0.21*12 = 2.5 V. The neutral voltage drop from transformer center tap to main panel is about 0.01*12 = 0.12 V (120 MV). The neutral drop from the main panel to the end of the 50 ft branch circuit is 0.08*12 = 0.96 V.

If there was a truly an open neutral wire from the transformer to the main panel, then that path now follows the ground path and the neutral resistance is 25 ohms from the above assumption. Assuming no load on the other phase, then the 12 A load (10 ohms) in series with 25 ohms produces a voltage across the load of (10/35)*120 = 35 V.

These neutral drops can be directly measured with some long test leads. An extension cord is a useful test lead extender. At the transformer a direct connection to the ground rod is a fairly good point to reference the transformer center tap. Interior to the house any EGC may be a good reference to the main panel neutral bus provided there is no current flowing in the EGC.

Knowing what the voltage drops on a good circuit are allows you to more easily locate the problem on the circuit when it is bad. In some cases it may be difficult to isolate the problem without sufficient instrumentation of the correct type.

.