Current unbalance with a VFD

[thewellguy]

I have a site that is at the end of a feeder line that has lots of power problems with voltage fluxuation and current imbalance. Its a closed delta 460 phase supply. Every 2-3 years they burn up a 50hp submersible well pump motor. They had good motor protect but shut it off because the protect shut the motor off to frequently.

My question is, if I put this motor on a vfd (an ABB because that the brand I'm most comfortable with) will it help with the power problems. I have been told it will but the source didn't sound to confident about it. I was also told to derate by 15%. I think because the AC is going to DC bus and back to AC that the current should come out balanced, but I'm not sure. Any thoughts? Thank you.

 

[jraef]

The concept is sound if your power problems are only voltage imbalance and fluctuation (the current imbalance, if it changes, is most likely due to voltage imbalances). If you have other problems such as frequent grid switching or capacitor switching spikes, pervasive under voltage / over voltage or phase loss, you may end up trading one set of problems with another.

You are correct, a VFD converts the AC to DC then back to pseudo-AC. So one way to look at it is that the VFD is like a new power source for the pump, which just happens to use the utility power as its "raw material" to make that power. That is why most VFDs can even take a single phase input and give a 3 phase output. But in doing that for instance, ALL of the power supply must come from 2 of the 3 phases so the rectifier components will see 1,732 times the input current and must be sized accordingly. Plus, a full phase loss will incur additional ripple on the DC bus, requiring more capacitance. For that reason, we always say to double the size of a VFD when using a 1 phase input. This is likely the idea behind someone's recommendation to up your size by 15%, just in case your phase-to-phase fluctuations were severe. But if you also wanted to keep running even with a lost phase, just double the VFD size to be sure.

If you also suspect spikes from grid switching or capacitor switching (common problems at the ends of transmission lines), then also use a line reactor at the very least in front of the drive, even if the drive rep tells you you don't need it. Sales reps often tell you that because they believe their drives have better harmonic attenuation that XYZ brand, blah, blah, blah, but even if it's true, they are overlooking the fact that a reactor also helps to protect the DRIVE!

Pervasive under voltage is a problem that the VFD cannot fix; they cannot create potential that is not there. You possibly can program the VFD to respond to an under voltage by reducing the speed, yet keep the drive doing something at least, but this often is not feasible in a well pump if the reduced speed means the water doesn't make it to the top of the well. So check that out carefully before implementing.

Pervasive over voltage, if it is small, can usually be addressed with the reactor because it will drop the voltage slightly, but if large can eventually damage the VFD.

 

[waross]

A general question, Jraef:
If under voltage or voltage fluctuation is a chronic problem, may it be corrected by using autotransformers to boost the voltage and a VFD rated for a higher voltage?
For example, if you boosted the 480V to 540V and then used a 600V VFD, would you be able to supply the proper voltage to the motor (460V) despite input fluctuations from about 400V to over 500V? Yes I know it would cost more but occasions arise when cost is not the most important factor.
Thanks Jraef
Respectfully
Bill

 

[Marke]

Hi waross

Yes, you could certainly do that, but supplies that have a habit of dropping, also have a habit of rising, so you would need to ensure that there is plenty of headroom for the overvoltage situation. AC line reactors are pretty important to prevent the sudden changes in voltage from causing sudden current transients on the input.
Good idea!!

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[DickDV]

This would be an application for ABB's ACS550 series drive. That drive comes standard with the equivalent of 3% line reactance so I would not add any additional input reactance.

Since that drive is CE rated, it contains a noise suppression network that depends on the input power being balanced with respect to ground. In a soft power system like you are describing, I would remove the F1/F2 or EM1/2 screws as shown in the instruction manual to disconnect this network. If it remains connected and the input power goes out of balance to ground, the network will overheat and destroy the drive.

Otherwise, be sure to put the actual motor nameplate data into Group 99 and also be sure that the Undervoltage Control is turned on in Group 20 Parameter 06. Set up that way, the motor will automatically slow down to the highest acceptable speed under low voltage conditions and then return to higher speeds as the voltage goes back up.

 

[gepman]

thewellguy
Since the power system is a closed delta, if they apply a VFD they should take a look at the manufacturer's recommendation for applying a VFD with this kind of system. Early VFD's did not like delta systems since if one leg grounded the other two legs went to 480V. I am not that familiar with the ABB drives but the A-B PowerFlex 70 & 700 drives have a jumper to remove what I believe is an MOV to ground from their protection circuit.

It is also possible that they are losing the submersible pump motor due to problems on the closed delta system (depending on how they track and fix ground faults). They could install a delta/wye isolation transformer ahead of the pump VFD.

 

[Drivesrock]

Raising the incoming ac voltage and the voltage rating of a standard diode rectifier (front end) drive but maintaining the motor nominal voltage lower than the input is sometimes a solution for supplies with voltage dips. You should be sure though that your motor windings can withstand the higher voltage as the PWM voltage element of the inverter output will be the dc link voltage or roughly 1.35 times your new input ac voltage. (It's the mark/space ratio of the PWM voltage that produces an average voltage according to the V/f characteristics - directly proportional or square law depending on what is programmed).
So apart from your motor possibly needing to be inverter duty, the insulation specification must consider the highest incoming ac voltages.
Alternatively, an active front end drive (with IGBT rectifier) could be considered that will maintain constant dc link volts under a wide variation of ac input volts - according to its rating.
An added benefit would be negligable supply harmonics (depending on who's drive it is) over a standard drive.


Drivesrock

 

[jraef]

Drivesrock took the words right out of my mouth on the peak voltage issue. Probably not a big problem with a high quality inverter duty motor with specifications you can actually have access to (some don't bother).

On the active front-end issue though, as I understand it most AFE drives can be used for regeneration OR harmonic mitigation, but not both at the same time. The control algorithms on the firing schemes are different. Correct me if I'm wrong, inquiring minds want to know.

 

[Marke]

Jraef
I think it depends on what you mean by harmonic mitigation. My understanding is that the active front end will regenerate in braking mode, and reduce the harmonics drawn by from the supply while in motoring mode. If you wish to use it to reduce other harmonics from other sources, then it becomes a different animal. - This is of course manufacturer/design dependent.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[Drivesrock]

I believe there are IGBT based regeneration units to add on to a drive that allow regen but not harmonic mitigation. Then there are true IGBT front end drives with large (>10% impedance) input reactors that do both. There are manufacturers out there who can offer active filters to reduce harmonics from other sources. The active filter could be the front end of a drive. Now we're talking bigger bucks!
For our colleague who raised the original question I think that a proper site survey is required to establish the characteristics of the supply - magnitude and duration of the dips. From that a requirement spec can be produced and different solutions proposed - interposing transformer or drive (and type) - and costs put on the table.

 

[thewellguy]

First, Thank you all for your input. You insight has been very educational.

Driverock, I agree that a sight survey would be the best route. However, my customer will not pay for something like that. The do understand electric so trying to sell them that would be like selling snow to eskimos.

To clarify further. I was thinking about using a ABB ACS550-U1 series drive with a KLC PWM output filter. This is because the distance between the drive and the motor is 250'.

I guess I am still confuses on one point. Do I need an input filter.

Thank you

 

[LionelHutz]

Waross has a good idea. Using a 600VAC rated VFD will have some extra built-in protection against over-voltages. If you don't use a higher input voltage right away, you at least have the option of feeding a higher voltage later if voltage dips are problematic.

Personally, I would not sell a drive to that site without installing a isolation transformer with a grounded Y secondary.

As already explained, the VFD will recreate the 3-phases giving a balanced and stable output. However, it will only do so if you can maintain a stable enough input that the buss voltage is maintained fairly stable.

 

[fornhamspark]

So can i put this question, if i understand this correctly.
If the supply is increased to say 500v to a 600v normal drive the motor will see a higher voltage due to the increase in the dc bus level.

But say i use an active front end drive instead at 500v could you lower the voltage to the motor to account for voltage drop ouver a long distance.