VFD (Solving DC bus issue)

Learn the NEC with Mike Holt now!
Status
Not open for further replies.

pguerra75

Member
Location
US
I am currently working a project that uses a 40HP 460VAC motor to drive a reciprocating pump. I have done several of these in the past without issue, however my boss decided to through me a curve ball and add a VFD to the mix. The motor SF is 1.15 so I selected a 50HP VFD (GS4-4050). The issue I am having is once the motor gets moving (over 50hz) I get a DC Bus Overvolt alarm. I believe I have traced this to the fact it is a reciprocating pump and the motor regeneration is causing the over volt. Now from what I have read is that there is 3 ways to solve this 1. Add mass (not sure I could sell this one) 2. Using a breaking resistor 3. Buying a drive that monitors shaft load and can back of on excitation to avoid over generation. Option 2 being the most economical I would love to be able to go that route but the Brakes for the drive I currently have are only rated to 10% duty cycle and I am not sure how to determine if my current load would exceed that. Anyone know how to calculate this? Alternatively anyone know of a good VFD that monitors shaft load?

Thanks for the help,
Paul
 

Besoeker3

Senior Member
Location
UK
Occupation
Retired Electrical Engineer
I am currently working a project that uses a 40HP 460VAC motor to drive a reciprocating pump. I have done several of these in the past without issue, however my boss decided to through me a curve ball and add a VFD to the mix. The motor SF is 1.15 so I selected a 50HP VFD (GS4-4050). The issue I am having is once the motor gets moving (over 50hz) I get a DC Bus Overvolt alarm. I believe I have traced this to the fact it is a reciprocating pump and the motor regeneration is causing the over volt. Now from what I have read is that there is 3 ways to solve this 1. Add mass (not sure I could sell this one) 2. Using a breaking resistor 3. Buying a drive that monitors shaft load and can back of on excitation to avoid over generation. Option 2 being the most economical I would love to be able to go that route but the Brakes for the drive I currently have are only rated to 10% duty cycle and I am not sure how to determine if my current load would exceed that. Anyone know how to calculate this? Alternatively anyone know of a good VFD that monitors shaft load?

Thanks for the help,
Paul
Regeneration would cause a rise in DC liknk voltage but I don't see why a reciprocating as opposed to centrifugal pump would cause that. Perhaps the decelerating rate is too great. A flywheel could exacerbate the problem. Slowing the decel rate should be the simplest fix. Restive braking may be an answer but needs more hardware and more cost
 

Besoeker3

Senior Member
Location
UK
Occupation
Retired Electrical Engineer
Extending it would certainly help but if ACTL starters worked for the rest, why does it need a deceleration rate at all? Coast to stop seems appropriate.
Well, yes if it always ran at fixed speed. But then, why use a VSD unless you want to vary the speed? And that could be up or down.
 

Besoeker3

Senior Member
Location
UK
Occupation
Retired Electrical Engineer
The OP was surprised with the VFD where the fixed speed worked before. Changing the speed could certainly be advantageous but is deceleration that important for this setup? More of a question for the OP, than you.
Agree. The rationale for including a VFD has not been explained.
 

winnie

Senior Member
Location
Springfield, MA, USA
Occupation
Electric motor research
When you say 'reciprocating pump' do you mean something like a 'pumpjack' where the beam is rising and falling pretty slowly?

My understanding is that unless these are very well balanced they commonly regenerate over part of the pumping cycle. When connected across the line the motor draws power for part of the cycle (say lifting the beam) and then regenerates to the grid on the other half.

If this is the case, then on a VFD you are regularly going to be regenerating a significant portion of your power consumption.

You can always use a larger braking resistor; resistors are separately rated in ohms and watts, and you simply need a resistor of the proper resistance (based on drive to give proper braking performance) and proper wattage (to handle the continuous regeneration). However once you consider the energy lost to this resistor it is almost certainly not the cheapest approach.

In addition to the options you are considering, you can also look at the following:
1) line regeneration modules connected to the VFD, which will dump the regenerated energy back to the grid
2) additional DC link energy storage, so that the regenerated energy increases the DC bus voltage in a tolerable fashion

-Jon
 

Besoeker3

Senior Member
Location
UK
Occupation
Retired Electrical Engineer
When you say 'reciprocating pump' do you mean something like a 'pumpjack' where the beam is rising and falling pretty slowly?

My understanding is that unless these are very well balanced they commonly regenerate over part of the pumping cycle. When connected across the line the motor draws power for part of the cycle (say lifting the beam) and then regenerates to the grid on the other half.

If this is the case, then on a VFD you are regularly going to be regenerating a significant portion of your power consumption.

You can always use a larger braking resistor; resistors are separately rated in ohms and watts, and you simply need a resistor of the proper resistance (based on drive to give proper braking performance) and proper wattage (to handle the continuous regeneration). However once you consider the energy lost to this resistor it is almost certainly not the cheapest approach.

In addition to the options you are considering, you can also look at the following:
1) line regeneration modules connected to the VFD, which will dump the regenerated energy back to the grid
2) additional DC link energy storage, so that the regenerated energy increases the DC bus voltage in a tolerable fashion

-Jon
A nodding donkey?
I hadn't thought of that. Yes, regen could play a part in that.
I had taken reciprocating to mean a piston type pump as in positive displacement.
Maybe the OP could clarify.
 

petersonra

Senior Member
Location
Northern illinois
Occupation
engineer
A nodding donkey?
I hadn't thought of that. Yes, regen could play a part in that.
I had taken reciprocating to mean a piston type pump as in positive displacement.
Maybe the OP could clarify.

I had also assumed reciprocating meant PD.

By the way, you can "improve" the duty cycle of brake resistors by a lot if you put use a fan to cool them. Does not take a lot of air to cool them down. Best bet is to have an output from the VFD that comes on when the brake resistor is in use that feeds an off delay timer. The off delay timer than turns on the fan and lets it run for a short time after the braking transistor turns off.
 

Besoeker3

Senior Member
Location
UK
Occupation
Retired Electrical Engineer
I had also assumed reciprocating meant PD.

By the way, you can "improve" the duty cycle of brake resistors by a lot if you put use a fan to cool them. Does not take a lot of air to cool them down. Best bet is to have an output from the VFD that comes on when the brake resistor is in use that feeds an off delay timer. The off delay timer than turns on the fan and lets it run for a short time after the braking transistor turns off.
No disagreement with any of that. We have done similar.
But reducing decel rate seems the simplest option - if that's how the regen occurs. We need more input from the OP.
 

RumRunner

Senior Member
Location
SCV Ca, USA
Occupation
Retired EE
I am currently working a project that uses a 40HP 460VAC motor to drive a reciprocating pump. I have done several of these in the past without issue, however my boss decided to through me a curve ball and add a VFD to the mix. The motor SF is 1.15 so I selected a 50HP VFD (GS4-4050). The issue I am having is once the motor gets moving (over 50hz) I get a DC Bus Overvolt alarm. I believe I have traced this to the fact it is a reciprocating pump and the motor regeneration is causing the over volt. Now from what I have read is that there is 3 ways to solve this 1. Add mass (not sure I could sell this one) 2. Using a breaking resistor 3. Buying a drive that monitors shaft load and can back of on excitation to avoid over generation. Option 2 being the most economical I would love to be able to go that route but the Brakes for the drive I currently have are only rated to 10% duty cycle and I am not sure how to determine if my current load would exceed that. Anyone know how to calculate this? Alternatively anyone know of a good VFD that monitors shaft load?

Thanks for the help,
Paul
How long does it take before the over volt alarm 4 sec, 10 sec or 20 sec.?


A VFD over voltage fault can occur during power up. If this is sporadic anomaly, check the incoming voltage. If the voltage is within specs. . . look for a jumper that connect the common capacitors to ground. You may have to look at the circuit board where a standoff screw terminal where this jumper is located. This is usually removable.


Check your manual.


Your environment could be rife in ground noise. This ground noise would come through these capacitors that could cause the voltage rise in the DC bus.

Yes, there are three ways to solve a deceleration bus over voltage problem, and any of those you've mentioned could work. But you only cited your perceived deceleration over voltage which [rightfully] elicited assumptions and speculations in the hope that a suggested solution could serve your purpose.


There are four causes of VFD Over voltage Fault.



1. Acceleration during ramp up which is caused as I mentioned unstable power source and ground noise.


2. VFD over voltage fault at deceleration.


This fault is common if the braking torque requirement exceeds drive braking capacity. This situation is noticeable when you hit stop during ramp down. . . while the load is still spinning faster than the set frequency. When this happens the regenerated power is pumped back and stored in the the DC bus. . . another cause of over voltage. In this instance the motor is acting like a generator.


One solution is to increase deceleration time.


3. Over voltage fault during normal operating mode.


This usually occur if the driven machine has a clutch disconnect. The sudden drop in load in the event the drive shaft is de-coupled would cause the motor to over speed and regenerate power to cause rise in the DC bus. This is when resistors come to the rescue.


4. Over voltage when VFD is not active [sitting idle]



In an industrial environment where all kinds of loads are present and there are large inductive loads could cause fluctuation/spikes in the power distribution. This needs to be investigated. . . improper transformer taps could also cause over voltage.


Using a generic pocket sized meter (those cheap ones) LOL cannot capture this fast occurring event.


An oscilloscope could easily capture the history of anomalous voltage rise.


All the best
 

kwired

Electron manager
Location
NE Nebraska
When you say 'reciprocating pump' do you mean something like a 'pumpjack' where the beam is rising and falling pretty slowly?

My understanding is that unless these are very well balanced they commonly regenerate over part of the pumping cycle. When connected across the line the motor draws power for part of the cycle (say lifting the beam) and then regenerates to the grid on the other half.

If this is the case, then on a VFD you are regularly going to be regenerating a significant portion of your power consumption.

You can always use a larger braking resistor; resistors are separately rated in ohms and watts, and you simply need a resistor of the proper resistance (based on drive to give proper braking performance) and proper wattage (to handle the continuous regeneration). However once you consider the energy lost to this resistor it is almost certainly not the cheapest approach.

In addition to the options you are considering, you can also look at the following:
1) line regeneration modules connected to the VFD, which will dump the regenerated energy back to the grid
2) additional DC link energy storage, so that the regenerated energy increases the DC bus voltage in a tolerable fashion

-Jon

A nodding donkey?
I hadn't thought of that. Yes, regen could play a part in that.
I had taken reciprocating to mean a piston type pump as in positive displacement.
Maybe the OP could clarify.
I was going to ask about nature of the driven load myself. A single piston type of load could have more regen characteristics than a multi stage unit - presuming they try to balance the stages across a full revolution of the driving shaft.
 

Jraef

Moderator, OTD
Staff member
Location
San Francisco Bay Area, CA, USA
Occupation
Electrical Engineer
Way back in the dark ages when I got my feet wet with VFDs, my mentor had made an application manual for me and page one, in giant letters, said only “No Contactors, Clutches, or Crankshafts!” The reason for the No Crankshaft rule was precisely this issue, it’s very common that this would cause VFD over voltage stresses. Since then VFD technology has improved to allow for reciprocal load adaptation for sure, but the “No” is only modified to “Beware of” because not all drives are created equal.

In my experiences with this issue and having to use resistors, I have used the resistor sizing criteria that assumes the “braking duty” is 25%, meaning that in 75% of each revolution the motor is either motoring, neutral or only slightly regenerating. That has worked for probably 90% of these applications I have come across (prior to the advent of the reciprocal load algorithms that come in better drives now). The other 10% were drives in which the braking TRANSISTOR was under sized (i.e. 10% duty cycle max, common in cheaper drives) or the application was so severe that I recommended using line regenerative drives rather than resistors. Now that Regen drives are becoming more readily available and less expensive (than they were), I haven’t used braking resistors in a few years. Even when sized right they are still a high failure point in a system.
 

Besoeker3

Senior Member
Location
UK
Occupation
Retired Electrical Engineer
Way back in the dark ages when I got my feet wet with VFDs, my mentor had made an application manual for me and page one, in giant letters, said only “No Contactors, Clutches, or Crankshafts!” The reason for the No Crankshaft rule was precisely this issue, it’s very common that this would cause VFD over voltage stresses. Since then VFD technology has improved to allow for reciprocal load adaptation for sure, but the “No” is only modified to “Beware of” because not all drives are created equal.

In my experiences with this issue and having to use resistors, I have used the resistor sizing criteria that assumes the “braking duty” is 25%, meaning that in 75% of each revolution the motor is either motoring, neutral or only slightly regenerating. That has worked for probably 90% of these applications I have come across (prior to the advent of the reciprocal load algorithms that come in better drives now). The other 10% were drives in which the braking TRANSISTOR was under sized (i.e. 10% duty cycle max, common in cheaper drives) or the application was so severe that I recommended using line regenerative drives rather than resistors. Now that Regen drives are becoming more readily available and less expensive (than they were), I haven’t used braking resistors in a few years. Even when sized right they are still a high failure point in a system.

We've put regen VFDs on paper machines where there can be an overhauling section or two but mostly we fed them from a common DC buss and the net power requirement was positive or out if you like. We did have dynamic braking for use in the event of an emergency stop.

We need more information from the OP before we get into too much speculation here...............
 
Status
Not open for further replies.
Top