intermittant tripping of breaker running pump through VFD

[h20byemf]

I am running a 7.5 hp 260v 3 phase pump motor using a 20hp vfd with single phase 260v input. (the location does not have 3 phase power available) The unit has run without problems for several years until 3 months ago when I started having trouble with the 100 amp main breaker tripping. The first time it happened, I took off the inspection plate on the motor and noted that one of the leads was showing signs of overheating, so I re-terminated it, checked the motor for shorts to ground, there weren't any.

I assumed that the connection at the lead had either heated up from resistance or grounded out inside the terminal box. So I powered it back up and things worked well for another month or so. Then I came back one day, and noted that one of the slow-blow 60 amp fuses had blown, and there was damage to the disconnect between the vfd and the subpanel. Again I looked for signs of overheating, shorts, etc. and found no concerning signs. So I connected the VFD directly to the 50 amp subpanel breaker.

This was the beginning of the current (not a pun) problem: The motor runs for 1-2 hours, then trips either the 100amp main or the 50 amp subpanel breaker. Each time, the VFD will show a fault on the either "DC bus low voltage" , or more often: "Line Loss." I have tested the motor insulation and each phase to ground is uniformly 55M ohm for a 5 minute 500v test. I am concerned that the motor may be at fault as it is not an inverter grade motor. It is an f grade of insulation, which when I spoke to a motor tech at Baldor said it would be worth trying to see how well it works.

I'm having a hard time figuring out what I need to correct. There is a very significant short going on, sufficient to trip the circuit breaker, but it does not persist. When I reset the breaker, the motor runs again for another hour. I have checked the voltages and amperages of the 3 phases, and the 3 legs show close to equal amperages ( +/- 15 amps)lamp-on, as do the 2 legs of the single phase (+/- 25 amps). Suggestions on where to go from here in correcting this problem?

 

[ptonsparky]

Check for poor connections would be my first suggestion. Overheated wires are the result of poor connections if properly sized to begin with. The poor connection can be internal to the CB.

Use a thermal scanning device or a voltmeter to check those connections.

I hope you forgot a decimal point to the left of your amperage, .15 & .25, otherwise you have a major problem.

 

[LarryFine]

In no particular order:

There could be a problem with the 1ph supply to the VFD.

Being a pump, its power demand could have risen over time.

 

[h20byemf]

amperage values

amperage values

No Ptonsparky, those are actual values of amperage from my clamp-on meter. the FLA on the motor nameplate is 18.8, and I'm running it at 50 hz through the VFD.

Yes Larry Fine, I have been inspecting, but just ordered an IR thermometer to more accurately look for hot spots.

 

[h20byemf]

ptonsparky, I looked at this again and perhaps the way that I stated it was ambiguous. What I meant to say was: the 3 phase amperage, read on the clamp ammeter at the motor leads, operating at 50 hz, is around 15 amps. It does fluctuate a bit on all 3 legs. meaning the amperage dances back and fourth between 14.2 to 16.3. All three legs are within the same range.

 

[ptonsparky]

ptonsparky, I looked at this again and perhaps the way that I stated it was ambiguous. What I meant to say was: the 3 phase amperage, read on the clamp ammeter at the motor leads, operating at 50 hz, is around 15 amps. It does fluctuate a bit on all 3 legs. meaning the amperage dances back and fourth between 14.2 to 16.3. All three legs are within the same range.

:thumbsup: Not you, it was me.

 

[StarCat]

On the Subject of Tripping breakers.....

On the Subject of Tripping breakers.....

Many years ago I had a Mystery trip happening on an RTU in Allen, Texas at a retail store. It was something like a 7.5 Ton Lennox GCS series running 480V 3-PH supply. So I am babysitting as we call it and looking for a reason for this system to be tripping the main, and fortunately on this day I was able to catch and zero in on the location. For whatever reason I had the unit running with the blower panel off, which is where the power entrance was brought into the cabinet through the roof curb. I think in this case it was run with 1-1/4" conduit. I was able to detect a flash down that conduit run when the fault happened which must have been fairly close to the roof on the vertical run of the pipe. There was obviously a wire in the pipe with compromised insulation that was arcing intermittently.
The Electrician that was called on to replace the wire run, refused to accept what I told him.
Interestingly enough, I was able to get a " repeat " of the ground fault with him as a witness, where he then realized that he had to replace a run of about 150 Ft.
Maybe he got the fact on that day that not all HVAC guys are lacking or predictable with respect to Electrical analysis.
The Problem was stopped cold once the new wire was in.

 

[petersonra]

I am running a 7.5 hp 260v 3 phase pump motor using a 20hp vfd with single phase 260v input. (the location does not have 3 phase power available) The unit has run without problems for several years until 3 months ago when I started having trouble with the 100 amp main breaker tripping. The first time it happened, I took off the inspection plate on the motor and noted that one of the leads was showing signs of overheating, so I re-terminated it, checked the motor for shorts to ground, there weren't any.

I assumed that the connection at the lead had either heated up from resistance or grounded out inside the terminal box. So I powered it back up and things worked well for another month or so. Then I came back one day, and noted that one of the slow-blow 60 amp fuses had blown, and there was damage to the disconnect between the vfd and the subpanel. Again I looked for signs of overheating, shorts, etc. and found no concerning signs. So I connected the VFD directly to the 50 amp subpanel breaker.

This was the beginning of the current (not a pun) problem: The motor runs for 1-2 hours, then trips either the 100amp main or the 50 amp subpanel breaker. Each time, the VFD will show a fault on the either "DC bus low voltage" , or more often: "Line Loss." I have tested the motor insulation and each phase to ground is uniformly 55M ohm for a 5 minute 500v test. I am concerned that the motor may be at fault as it is not an inverter grade motor. It is an f grade of insulation, which when I spoke to a motor tech at Baldor said it would be worth trying to see how well it works.

I'm having a hard time figuring out what I need to correct. There is a very significant short going on, sufficient to trip the circuit breaker, but it does not persist. When I reset the breaker, the motor runs again for another hour. I have checked the voltages and amperages of the 3 phases, and the 3 legs show close to equal amperages ( +/- 15 amps)lamp-on, as do the 2 legs of the single phase (+/- 25 amps). Suggestions on where to go from here in correcting this problem?

1. It is highly unlikely that something on the motor side could trip a breaker on the line side of the VFD if they are sized properly.

2. Because both the 100 Amp feeder breaker and the 50 Amp branch breaker have tripped, it suggests the fault is in the circuit downstream of the 50 Amp breaker.

I would be looking for an intermittent ground fault of some kind there. I think the first step would be to disconnect the wires at both ends and megger the two wires to see if they are damaged in some way. It is not a perfect test, as sometimes meggering can miss insulation breakages, but it is a good first start.

 

[Jraef]

We are assuming the CB and fused disconnect are on the LINE side of the VFD and that there is no Line Reactor ahead of the VFD. I'm also going to assume you meant 240V, not 260V...

Also, is there anything else between the line side and the VFD besides the breaker? A contactor or motor starter for example? If not, does anyone cycle the input power to the VFD a lot? If so, does the tripping happen at that time?

As to the fault record, that's what you will see when the breaker trips, Line Loss and/or Low DC Bus Volts, so that in and of itself isn't unusual. The real issue is, why are the breakers tripping?

Assuming all the other things I asked are negative, i.e. nothing in between the sub-panel and the VFD, nobody is cycling the power to the VFD and the fault happens randomly, not just when powered up; here is my SWAG on what's happening.

A 20HP VFD is likely rated at around 54A and in boost mode can draw upward of 150% of that for a second or two, so the drive is capable of pulling 84A in that circumstance and that may be enough to trip the 50A breaker. That however is not likely enough to cause the 100A breaker to trip, it's too short of a time frame. But when a drive initially powers up, the current draw to charge the capacitors is upward of the Available Fault Current, limited only by the impedance of the circuit ahead of the VFD terminals! So the drive COULD theoretically pull 1000+ amps.

Because of this, all VFDs have what's called a "pre-charge circuit" that limits that capacitor charging current, typically to that 150% of the drive rating. That pre-charge circuit however could interfere with getting full voltage to the motor, so it has a timer associated with it that shorts it out of the circuit after about 1 second, which is all it takes to charge the caps. The problem comes in when there are line side transients. Diodes, like those used in the bridge rectifier, conduct when the sine wave of voltage exceeds a certain level, called the "forward conduction threshold". When the line voltage dips below the forward conduction threshold of the diodes because of a ringing transient event, they cease to conduct when they should be. If the VFD is running the motor at that moment, the motor is pulling energy out of the caps but one or more of the diodes is not contributing to replace it. So when the transient is over, the next diodes in sequence that can conduct again will conduct into the "empty" caps and the caps will pull that ultra high current again. Only this time, because the drive was ALREADY powered up, the pre-charge circuit is still shorted out so there is no current limiting effect. That high steep pulse of current draw is what will trip your breakers. The solution to this is to install a Line Reactor ahead of the VFD. A line reactor adds impedance to the circuit, plus it adds an "inductive time constant" meaning that current and/or voltage cannot change any faster than a certain rate, so it slows down the rate of rise or fall of the transients, so they have less of an effect on the diode bridge. That's the main value of adding line reactors to VFDs, it not really about "harmonics" as most people think. In a single phase feed, it's even MORE important because you only have 4 diodes conducting, not 6, so a transient on one side of the 240V feed has more of an effect.

Side issue: the NEC, article 430.122, requires that your conductors FEEDING the VFD must be rated for 125% of the VFD input current. The input current on that 20HP VFD is likely somewhere around 50A, so your conductors must be rated for a minimum of 62.5A. I would have used a 70A feeder for that.

 

[MrJLH]

I am running a 7.5 hp 260v 3 phase pump motor using a 20hp vfd with single phase 260v input. (the location does not have 3 phase power available) The unit has run without problems for several years until 3 months ago when I started having trouble with the 100 amp main breaker tripping. The first time it happened, I took off the inspection plate on the motor and noted that one of the leads was showing signs of overheating, so I re-terminated it, checked the motor for shorts to ground, there weren't any.

I assumed that the connection at the lead had either heated up from resistance or grounded out inside the terminal box. So I powered it back up and things worked well for another month or so. Then I came back one day, and noted that one of the slow-blow 60 amp fuses had blown, and there was damage to the disconnect between the vfd and the subpanel. Again I looked for signs of overheating, shorts, etc. and found no concerning signs. So I connected the VFD directly to the 50 amp subpanel breaker.

This was the beginning of the current (not a pun) problem: The motor runs for 1-2 hours, then trips either the 100amp main or the 50 amp subpanel breaker. Each time, the VFD will show a fault on the either "DC bus low voltage" , or more often: "Line Loss." I have tested the motor insulation and each phase to ground is uniformly 55M ohm for a 5 minute 500v test. I am concerned that the motor may be at fault as it is not an inverter grade motor. It is an f grade of insulation, which when I spoke to a motor tech at Baldor said it would be worth trying to see how well it works.

I'm having a hard time figuring out what I need to correct. There is a very significant short going on, sufficient to trip the circuit breaker, but it does not persist. When I reset the breaker, the motor runs again for another hour. I have checked the voltages and amperages of the 3 phases, and the 3 legs show close to equal amperages ( +/- 15 amps)lamp-on, as do the 2 legs of the single phase (+/- 25 amps). Suggestions on where to go from here in correcting this problem?

Is this by chance a parallel pump system to a common header?
Was the pump recently rebuilt?
What kind of pump?

 

[h20byemf]

Is this by chance a parallel pump system to a common header?
Was the pump recently rebuilt?
What kind of pump?

Thanks for your inquiry. It is the main pump on an RO unit, no header. The pump is a Grundfos CRN vertical multistage , No it hasn't been rebuilt. Seems to be performing same as previous, no pressure or volume fluctuations.

 

[h20byemf]

We are assuming the CB and fused disconnect are on the LINE side of the VFD and that there is no Line Reactor ahead of the VFD. I'm also going to assume you meant 240V, not 260V...

Also, is there anything else between the line side and the VFD besides the breaker? A contactor or motor starter for example? If not, does anyone cycle the input power to the VFD a lot? If so, does the tripping happen at that time?

As to the fault record, that's what you will see when the breaker trips, Line Loss and/or Low DC Bus Volts, so that in and of itself isn't unusual. The real issue is, why are the breakers tripping?

Assuming all the other things I asked are negative, i.e. nothing in between the sub-panel and the VFD, nobody is cycling the power to the VFD and the fault happens randomly, not just when powered up; here is my SWAG on what's happening.

A 20HP VFD is likely rated at around 54A and in boost mode can draw upward of 150% of that for a second or two, so the drive is capable of pulling 84A in that circumstance and that may be enough to trip the 50A breaker. That however is not likely enough to cause the 100A breaker to trip, it's too short of a time frame. But when a drive initially powers up, the current draw to charge the capacitors is upward of the Available Fault Current, limited only by the impedance of the circuit ahead of the VFD terminals! So the drive COULD theoretically pull 1000+ amps.

Because of this, all VFDs have what's called a "pre-charge circuit" that limits that capacitor charging current, typically to that 150% of the drive rating. That pre-charge circuit however could interfere with getting full voltage to the motor, so it has a timer associated with it that shorts it out of the circuit after about 1 second, which is all it takes to charge the caps. The problem comes in when there are line side transients. Diodes, like those used in the bridge rectifier, conduct when the sine wave of voltage exceeds a certain level, called the "forward conduction threshold". When the line voltage dips below the forward conduction threshold of the diodes because of a ringing transient event, they cease to conduct when they should be. If the VFD is running the motor at that moment, the motor is pulling energy out of the caps but one or more of the diodes is not contributing to replace it. So when the transient is over, the next diodes in sequence that can conduct again will conduct into the "empty" caps and the caps will pull that ultra high current again. Only this time, because the drive was ALREADY powered up, the pre-charge circuit is still shorted out so there is no current limiting effect. That high steep pulse of current draw is what will trip your breakers. The solution to this is to install a Line Reactor ahead of the VFD. A line reactor adds impedance to the circuit, plus it adds an "inductive time constant" meaning that current and/or voltage cannot change any faster than a certain rate, so it slows down the rate of rise or fall of the transients, so they have less of an effect on the diode bridge. That's the main value of adding line reactors to VFDs, it not really about "harmonics" as most people think. In a single phase feed, it's even MORE important because you only have 4 diodes conducting, not 6, so a transient on one side of the 240V feed has more of an effect.

Side issue: the NEC, article 430.122, requires that your conductors FEEDING the VFD must be rated for 125% of the VFD input current. The input current on that 20HP VFD is likely somewhere around 50A, so your conductors must be rated for a minimum of 62.5A. I would have used a 70A feeder for that.

Thanks for the detailed response. It took me a while to work through it, but I think I now understand your analysis, and it makes a lot of sense. (especially the "ringing" transients) I plan to install a line reactor as the next step. There are a few other issues with this system however, so I’m going to mention them as well.

You asked about other devices between line and VFD. There aren't any starters or other devices between the sub panel and the VFD except the fused disconnect, which has now been removed.

There are 2 other pumps on the sub panel, a 3hp single phase and a 1 hp 3 phase, also running on a VFD at constant pressure. The 1 hp is always running when the 7.5 is, as it is the booster to the RO unit, and the 7.5 is the main pump for the RO. The 3hp is usually on, and I haven’t been able to correlate the tripping of breakers with it being off or on.

The disconnect that was removed was a 60 amp HVAC model with 60 amp slow blow fuses. I came over one day and noted that the the plastic part of the pullout was melted. I assumed the fuses were blown, but when I checked them today they still have continuity. One side of the metal pullout contact was fried. Probably was a mis-match to use the slo-blow fuses on the economy hvac disconnect?

Speaking of fried, there is a 100 amp breaker at 200 amp service that feeds the sub panel. I inspected it today and found a loose screw connection at the CB. It had signs of arcing and charring of plastic where the CB plugs into the panel, only on one leg. No other signs of arcing or hot spots were found.

The feeder from service to subpanel is #4 THHN, but the feeder from the SUB to the VFD is only #8 THHN, definitely not 70A.

Since I replaced the 100 amp CB yesterday (and moved it to a new spot on the panel), the system has been running fine. I’m planning on installing a line reactor next.

Motor leads: I had used NEC ampacity chart (similar to NEC310.15(B)(16)) to determine sizing for the #8 motor leads mentioned above. Do you recommend up sizing them as well to try to prevent future problems?

Breaker connection: In the past I have only used antioxidant paste on aluminum connections. Would you recommend using that on my CB connections as well?

Sizing of the line reactor: I’m not sure what will be the controlling variable here. As you say, the 20hp VFD could draw in excess of 50 amps, but my normal draw is 24-25 amps single phase, running the 18.8 FLA 3phase motor at 50 HZ. The literature indicates that the reactor should be sized for the load to be effective, and that the TCI-KDR series filters can tolerate loads of 200%. Do you have any suggestions regarding this?

Also, I am currently installing a 7.5hp 460v submersible pump using a 10kVA step-up transformer in an autotransformer configuration to feed single phase 480v into a VFD. of course there will be a dv/dt output filter. thoughts on applicability of a line reactor in this configuration? new thread?

 

[h20byemf]

We are assuming the CB and fused disconnect are on the LINE side of the VFD and that there is no Line Reactor ahead of the VFD. I'm also going to assume you meant 240V, not 260V...

Also, is there anything else between the line side and the VFD besides the breaker? A contactor or motor starter for example? If not, does anyone cycle the input power to the VFD a lot? If so, does the tripping happen at that time?

As to the fault record, that's what you will see when the breaker trips, Line Loss and/or Low DC Bus Volts, so that in and of itself isn't unusual. The real issue is, why are the breakers tripping?

Assuming all the other things I asked are negative, i.e. nothing in between the sub-panel and the VFD, nobody is cycling the power to the VFD and the fault happens randomly, not just when powered up; here is my SWAG on what's happening.

A 20HP VFD is likely rated at around 54A and in boost mode can draw upward of 150% of that for a second or two, so the drive is capable of pulling 84A in that circumstance and that may be enough to trip the 50A breaker. That however is not likely enough to cause the 100A breaker to trip, it's too short of a time frame. But when a drive initially powers up, the current draw to charge the capacitors is upward of the Available Fault Current, limited only by the impedance of the circuit ahead of the VFD terminals! So the drive COULD theoretically pull 1000+ amps.

Because of this, all VFDs have what's called a "pre-charge circuit" that limits that capacitor charging current, typically to that 150% of the drive rating. That pre-charge circuit however could interfere with getting full voltage to the motor, so it has a timer associated with it that shorts it out of the circuit after about 1 second, which is all it takes to charge the caps. The problem comes in when there are line side transients. Diodes, like those used in the bridge rectifier, conduct when the sine wave of voltage exceeds a certain level, called the "forward conduction threshold". When the line voltage dips below the forward conduction threshold of the diodes because of a ringing transient event, they cease to conduct when they should be. If the VFD is running the motor at that moment, the motor is pulling energy out of the caps but one or more of the diodes is not contributing to replace it. So when the transient is over, the next diodes in sequence that can conduct again will conduct into the "empty" caps and the caps will pull that ultra high current again. Only this time, because the drive was ALREADY powered up, the pre-charge circuit is still shorted out so there is no current limiting effect. That high steep pulse of current draw is what will trip your breakers. The solution to this is to install a Line Reactor ahead of the VFD. A line reactor adds impedance to the circuit, plus it adds an "inductive time constant" meaning that current and/or voltage cannot change any faster than a certain rate, so it slows down the rate of rise or fall of the transients, so they have less of an effect on the diode bridge. That's the main value of adding line reactors to VFDs, it not really about "harmonics" as most people think. In a single phase feed, it's even MORE important because you only have 4 diodes conducting, not 6, so a transient on one side of the 240V feed has more of an effect.

Side issue: the NEC, article 430.122, requires that your conductors FEEDING the VFD must be rated for 125% of the VFD input current. The input current on that 20HP VFD is likely somewhere around 50A, so your conductors must be rated for a minimum of 62.5A. I would have used a 70A feeder for that.

Thanks for the detailed response. It took me a while to work through it, but I think I now understand your analysis, and it makes a lot of sense. I plan to install a line reactor as the next step. There are a few other issues with this system however, so I’m going to mention them as well.

You asked about other devices between line and VFD. There aren't any starters or other devices between the sub panel and the VFD except the fused disconnect, which has now been removed.

There are 2 other pumps on the sub panel, a 3hp single phase and a 1 hp 3 phase, also running on a VFD at constant pressure. The 1 hp is always running when the 7.5 is, as it is the booster to the RO unit, and the 7.5 is the main pump for the RO. The 3hp is usually on, and I haven’t been able to correlate the tripping of breakers with it being off or on.

The disconnect that was removed was a 60 amp HVAC model with 60 amp slow blow fuses. I came over one day and noted that the the plastic part of the pullout was melted. I assumed the fuses were blown, but when I checked them today they still have continuity. One side of the metal pullout contact was fried. Probably was a mis-match to use the slo-blow fuses on the economy hvac disconnect?

Speaking of fried, there is a 100 amp breaker at 200 amp service that feeds the sub panel. I inspected it today and found a loose screw connection at the CB. It had signs of arcing and charring of plastic where the CB plugs into the panel, only on one leg. No other signs of arcing or hot spots were found.

The feeder from service to subpanel is #4 THHN, but the feeder from the SUB to the VFD is only #8 THHN, definitely not 70A.

Since I replaced the 100 amp CB yesterday (and moved it to a new spot on the panel), the system has been running fine. I’m planning on installing a line reactor next.

[h=1]Motor leads: I had used NEC ampacity chart (similar to NEC310.15(B)(16))[/h]to determine sizing for the #8 motor leads mentioned above. Do you recommend up sizing them as well to try to prevent future problems?

Breaker connection: In the past I have only used antioxidant paste on aluminum connections. Would you recommend using that on my CB connections as well?

Sizing of the line reactor: I’m not sure what will be the controlling variable here. As you say, the 20hp VFD could draw in excess of 50 amps, but my normal draw is 24-25 amps single phase, running the 18.8 FLA 3phase motor at 50 HZ. The literature indicates that the reactor should be sized for the load to be effective, and that the TCI-KDR series filters can tolerate loads of 200%. Do you have any suggestions regarding this?