MWBC & neutral current

[sw_ross]

If the neutral current on a MWBC is only the imbalance of current from L1 & L2, what would be a cause for when the numbers don't match up?

In my case L1 = 10-11 amps and L2 = 15 amps. my neutral current is 9-10 amps.

This is a single phase panel that is fed by a 3-phase 120/208 system

OOPS! I think I just answered my own question with that last statement!

 

[LarryFine]

You did. 120 deg. vs 180 deg.

 

[infinity]

If the neutral current on a MWBC is only the imbalance of current from L1 & L2, what would be a cause for when the numbers don't match up?

In my case L1 = 10-11 amps and L2 = 15 amps. my neutral current is 9-10 amps.

This is a single phase panel that is fed by a 3-phase 120/208 system

OOPS! I think I just answered my own question with that last statement!

3Ø and 1Ø are not the same formulas when calculating the neutral current on a 3 wire MWBC. For 3Ø with equal current on the two phase conductors the neutral current would be about the same as the phase current.

 

[sw_ross]

Thanks for the MH visual! I have seen that before and it totally makes sense.
I was trying figure out why I had a larger than expected neutral current, then as I typed out that last statement in #1 it totally clicked in my brain!

 

[synchro]

If the neutral current on a MWBC is only the imbalance of current from L1 & L2, what would be a cause for when the numbers don't match up?

In my case L1 = 10-11 amps and L2 = 15 amps. my neutral current is 9-10 amps.

This is a single phase panel that is fed by a 3-phase 120/208 system

OOPS! I think I just answered my own question with that last statement!

If there were only L1-N and L2-N loads, then with L1 = 10-11 amps and L2 = 15 amps the neutral current should be closer to 13 amps instead of the measured 9-10 amps. So it appears that you have some L1-L2 loading as well as L-N, because L-L loads would not be adding to the neutral current.

 

[infinity]

Thanks for the MH visual! I have seen that before and it totally makes sense.
I was trying figure out why I had a larger than expected neutral current, then as I typed out that last statement in #1 it totally clicked in my brain!

That graphic didn't directly address your question but it contained the formula which makes it helpful.

 

[sw_ross]

If there were only L1-N and L2-N loads, then with L1 = 10-11 amps and L2 = 15 amps the neutral current should be closer to 13 amps instead of the measured 9-10 amps. So it appears that you have some L1-L2 loading as well as L-N, because L-L loads would not be adding to the neutral current.

So I looked at the circuit again today to double check my numbers and play with the formula.

L1 = 11.4
L2 = 14.9

Using the formula my neutral current should be 13.5 amps.

It was actually 8.9 amps.

This circuit is for street lights. As far as I'm aware (based on info from the maintenance guy in charge of the street lights) all the lights are setup for 120v.

 

[gar]

191116-2135 EST

sw_ross:

Formulas are just mathematical descriptions, and may or may not represent the real world.

A number of tests you can do to see what is going on.

1. Put line 1, 2 and neutral thru one current transformer. What is the result?
2. Put line 1, and 2 thru a current transformer. How does this current reading compare to the neutral reading?
3. Look at each current waveform, and record each on an oscilloscope. Compare the sum of line 1 and 2 with the neutral.

.

 

[synchro]

191116-2135 EST

sw_ross:

Formulas are just mathematical descriptions, and may or may not represent the real world.

A number of tests you can do to see what is going on.

1. Put line 1, 2 and neutral thru one current transformer. What is the result?
2. Put line 1, and 2 thru a current transformer. How does this current reading compare to the neutral reading?
3. Look at each current waveform, and record each on an oscilloscope. Compare the sum of line 1 and 2 with the neutral.

.

I was just thinking about proposing tests like your #1 and #2 , gar
For sw_ross, when you make these current measurements keep the load end of all the wires on the same side of the current transformer.

 

[ramsy]

So I looked at the circuit again today to double check my numbers and play with the formula.

L1 = 11.4
L2 = 14.9

Using the formula my neutral current should be 13.5 amps.

It was actually 8.9 amps

Another way to get N = 8.9A

L1 = 11.4
L2 = 14.9
L3 = 4.78

Maybe field tech missed L3.

 

[synchro]

Another possibility is that the measured neutral current is reduced because some of the neutral current is flowing through the EGC or other ground path.
This is happening because:
1. there's a connection or fault between the neutral conductor and ground further down toward the loads,
or
2. one or more of the loads is connected to the EGC instead of neutral

 

[sw_ross]

Another possibility is that the measured neutral current is reduced because some of the neutral current is flowing through the EGC or other ground path.
This is happening because:
1. there's a connection or fault between the neutral conductor and ground further down toward the loads,
or
2. one or more of the loads is connected to the EGC instead of neutral

This is something I've been wondering about....

This circuit, like I said, is a string of street lights in our small town. The wire that daisy-chains from light to light is direct burial and was probably installed before 1980 (just a guess). The fixtures are definitely from the early 80's. The circuit is on a 2-pole 50 amp breaker. These lights have been serviced as needed by the town maintenance guy (i.e. not an electrician).

Periodically this circuit trips. It has happened more frequently this fall. I'd say it might've happened 4 times since August. We keep looking for a pattern related to occurrence such as ground moisture, etc. but there doesn't seem to be a common denominator. It happened again Thursday night. Previously the last trip was around the first half of October.

Yesterday, based off a previous suggestion by Larry Fine, I installed a different breaker that has the handle-tie removed (it didn't say "common internal trip"). I'm hoping that whenever it next trips I'll be able to see which leg trips to help narrow down the issue.

The town is at "year end money", so they don't have money to throw at this issue. I'm hoping if I can identify the issue I can inform the city council and they can work a fix into the budget for year 2020.

 

[ptonsparky]

You need to open the conductors from pole to pole and check them for ground faults. Check each set of wires going up to the fixtures as well. IDR if you have said in the past if an EG had been included in that install but it was common enough back then to just use rods at each pole. Not a good method as you know.

You could turn the lights on and check voltage from each pole to a screw driver stuck in the ground about 3' away. Use your gloves. Look for dead dried up toads kissing the pole. Not joking.

 

[synchro]

You need to open the conductors from pole to pole and check them for ground faults. Check each set of wires going up to the fixtures as well.

If there's evidence of a ground fault from measurements at the panel and there are lot's of lights in the chain, to speed things up a binary (aka half-interval) search approach could be used. Start in the middle of the chain and check if the fault is located upstream or downstream. Then move in the direction of the fault and pick the light pole in the middle of the remaining part of the chain to make further measurements, and so on until you isolate the fault. This may be obvious or common knowledge, but it's a reminder in any case.

It's not as sensitive as opening the wires and measuring resistance, but if a ground fault is large enough you may be able to identify it by putting your clamp around both hot and neutral wires as gar mentioned to see if there's any measurable current that would constitute "leakage" downstream. This is what a GFCI does when it measures the amount of imbalance in downstream load current to determine if there's a ground fault. You should also check the wires to the fixture as ptonsparky said by putting the clamp around these wires.

The breaker may or may not be tripping from a ground fault. But if a ground fault is present then the wires could be sufficiently damaged or deteriorated in the same location to cause both L-G and L-N faults, especially with repeated 50A breaker tripping occurring.

 

[Russs57]

I deal with this stuff a lot after hurricanes. Have a lot of boxes in the ground with a mixture of 277 and 480 (single phase) lighting loads. I find it is worth the price to add fuse holders at each fixture. IIRC I use littlefuse LEB series. Just don't put them in backwards

I find a megger invaluable.

 

[kwired]

This is something I've been wondering about....

This circuit, like I said, is a string of street lights in our small town. The wire that daisy-chains from light to light is direct burial and was probably installed before 1980 (just a guess). The fixtures are definitely from the early 80's. The circuit is on a 2-pole 50 amp breaker. These lights have been serviced as needed by the town maintenance guy (i.e. not an electrician).

Periodically this circuit trips. It has happened more frequently this fall. I'd say it might've happened 4 times since August. We keep looking for a pattern related to occurrence such as ground moisture, etc. but there doesn't seem to be a common denominator. It happened again Thursday night. Previously the last trip was around the first half of October.

Yesterday, based off a previous suggestion by Larry Fine, I installed a different breaker that has the handle-tie removed (it didn't say "common internal trip"). I'm hoping that whenever it next trips I'll be able to see which leg trips to help narrow down the issue.

The town is at "year end money", so they don't have money to throw at this issue. I'm hoping if I can identify the issue I can inform the city council and they can work a fix into the budget for year 2020.

Is there a separate EGC run to the poles? If the bonded the poles (and maybe even a ground rod) to the neutral conductor then you have some earth return current, how much will depend on just how good any electrodes (naturally occurring or made electrodes) actually are.

If you have non linear loads (particularly LED's or even electronic ballast driven lamps) that throws neutral current off some on a wye system with the harmonic currents being additive in the neutral. You also need a true RMS meter to get more accurate reading of that current.

 

[synchro]

So I looked at the circuit again today to double check my numbers and play with the formula.

L1 = 11.4
L2 = 14.9

Using the formula my neutral current should be 13.5 amps.

It was actually 8.9 amps.

This circuit is for street lights. As far as I'm aware (based on info from the maintenance guy in charge of the street lights) all the lights are setup for 120v.

You mentioned in your post about this issue last month that there are two MWBCs with lights on them, and that one of them is having the breaker tripping problem. Have you measured the currents on the other MWBC that's working fine to see how the currents compare? In particular, does the measured neutral current on the well-behaved MWBC match the calculated current better than the MWBC with the problem does? If the problem-free circuit does match better, then maybe the discrepancy you noted above is because there's an open circuit in the neutral conductor further down the chain, and so there's no neutral current being drawn after that point. Since the measured 8.9A is 66% of the expected 13.5A, perhaps the last third of the chain has the neutral disconnected, just as a rough guess.
This assumes that the lights would still operate with 2 in series across 208V L-L and with no connection to neutral. I think it's possible that they would still operate, but maybe they could have excess inrush current if they don't share voltage evenly all of the time? I think it's also possible that this could cause prolonged startups and therefore a longer duration for the inrush current even if it doesn't increase the peak current value. And then that could cause a "thermal" trip on the breaker.
Just speculation from this end but something to consider.

 

[sw_ross]

You mentioned in your post about this issue last month that there are two MWBCs with lights on them, and that one of them is having the breaker tripping problem. Have you measured the currents on the other MWBC that's working fine to see how the currents compare? In particular, does the measured neutral current on the well-behaved MWBC match the calculated current better than the MWBC with the problem does? If the problem-free circuit does match better, then maybe the discrepancy you noted above is because there's an open circuit in the neutral conductor further down the chain, and so there's no neutral current being drawn after that point. Since the measured 8.9A is 66% of the expected 13.5A, perhaps the last third of the chain has the neutral disconnected, just as a rough guess.
This assumes that the lights would still operate with 2 in series across 208V L-L and with no connection to neutral. I think it's possible that they would still operate, but maybe they could have excess inrush current if they don't share voltage evenly all of the time? I think it's also possible that this could cause prolonged startups and therefore a longer duration for the inrush current even if it doesn't increase the peak current value. And then that could cause a "thermal" trip on the breaker.
Just speculation from this end but something to consider.

I did look at the current draw on the other MWBC that is fed by this panel,

L1 = 17.7
L2 = 13.1
N = 15.6

That's almost right on where it should be using the formula.
This circuit doesn't seem to have issues.

The last time I fired up the circuit that was having the issue, using my ammeter I was going between the legs to see about an inrush issue but there was not more than a 2-3 amps of additional current during startup. I'm not saying it's not an inrush issue, I just didn't see it at that time.

I know there's 2 fixtures in the circuit that have lamps that cycle on/off , so assume the lamp is bad. I told the maintenance guy those lamps should be changed out, but I don't think something as simple as that would contribute to the issue.

 

[sw_ross]

Is there a separate EGC run to the poles? If the bonded the poles (and maybe even a ground rod) to the neutral conductor then you have some earth return current, how much will depend on just how good any electrodes (naturally occurring or made electrodes) actually are.

If you have non linear loads (particularly LED's or even electronic ballast driven lamps) that throws neutral current off some on a wye system with the harmonic currents being additive in the neutral. You also need a true RMS meter to get more accurate reading of that current.

Yes, there is a separate EGC run with the circuit. I don't think the maintenance guy totally understands the difference between the grounded conductor and the EGC.

 

[kwired]

Yes, there is a separate EGC run with the circuit. I don't think the maintenance guy totally understands the difference between the grounded conductor and the EGC.

If you disconnect N and G from supply (so there is no main bonding jumper connected) then you should not have any continuity between them out to the load(s). If you do have continuity they probably are bonded together somewhere and should not be.

Can also clamp the EGC and check to see if majority of current you think is missing from the neutral is flowing on the EGC.