VFD on step-up transformer problem

[tom kanzler]

We're having some trouble with VFD powered temporary equipment running through a 300kVA step-up transformer, from 208 to 480V. Machine specs shows 354A peak, 257A rated. Constant motor start/stop, with limited run time in between (hoist). Control system faults out when starting the hoist. We changed the transformer out for a 400kVA unit, and no more problems.

What I can't get straight is if the transformer size is actually causing undue voltage drop, triggering the fault. I wouldn't think there should be a problem at 282kVA (at 354A), and the transformer should be able to absorb an overload anyway. What I don't yet have is historical data on actual current and voltage at the drive, but I'm working on that.

Can the non-linear effects cause this? We have to use a grossly oversized generator when running this equipment, but that's for other reasons. This application is on utility power.

I don't want to use a blanket recommendation for an 'oversized' transformer without some justification, other than one bit of success, as the customer has to supply and connect the mains and transformer. And I can't be sure the transformer was the actual cure.

I've searched the web and this site, and found nothing to suggest that VFD's, as a general rule, require special treatment in transformers. Can anyone shed any light on this?

 

[GoldDigger]

Anything that rectifies the AC to produce DC, like the input stage of the VFD, and does not use a line reactor, harmonic filter or other power factor control will generate lots of harmonics that may require a larger transformer.
The high current pulses cause a much larger effective voltage drop.
For information specific to starting current when using a VFD, read the next post, from Jraef.

 

[Jraef]

You provided no details on the VFD, motor connected to it, type of load, etc., only the overall machine current requirements. Can't do much with that little info.

In general the answer would be no, but again, the devil is in the details. If the motor and VFD is let's say 200HP and it's for a centrifugal pump that you are trying to accelerate in 20 seconds, you should not be having any problems. But if it's a 200HP rock crusher and you are trying to accelerate it in 5 seconds, then yes, you are going to have a problem.

 

[templdl]

We're having some trouble with VFD powered temporary equipment running through a 300kVA step-up transformer, from 208 to 480V. Machine specs shows 354A peak, 257A rated. Constant motor start/stop, with limited run time in between (hoist). Control system faults out when starting the hoist. We changed the transformer out for a 400kVA unit, and no more problems.

What I can't get straight is if the transformer size is actually causing undue voltage drop, triggering the fault. I wouldn't think there should be a problem at 282kVA (at 354A), and the transformer should be able to absorb an overload anyway. What I don't yet have is historical data on actual current and voltage at the drive, but I'm working on that.

Can the non-linear effects cause this? We have to use a grossly oversized generator when running this equipment, but that's for other reasons. This application is on utility power.

I don't want to use a blanket recommendation for an 'oversized' transformer without some justification, other than one bit of success, as the customer has to supply and connect the mains and transformer. And I can't be sure the transformer was the actual cure.

I've searched the web and this site, and found nothing to suggest that VFD's, as a general rule, require special treatment in transformers. Can anyone shed any light on this?

Did I miss the frequency range that you are working with? You refer to voltages and current but your application is with s VFD we both no frequency range provided. By knowing the frequency maybe then we can address what affect it has on the transformer.

 

[tom kanzler]

I realize the info is rather thin, but details are hard to come by, as the manufacturer doesn't share much. I'm sure they didn't design the system, either, with the motors and drives were probably designed and furnished by their supplier (Seimens). The application is a hoist, with multiple motors and drives on the system, so acceleration includes the lifted load, plus rotational inertia of the motor (high speed) and drums (low speed). I'm trying to get info on actual peak current, and voltage, though the best I'll get is RMS without any detail on what the rectifier is doing to the wave form. But it's a start.

My question I suppose is somewhat generic regarding special treatment when sizing a step-up transformer to work with VFDs, as opposed to across the line starting (or reduced voltage starting, but no rectifier involvement) and running with conventional squirrel cage motors. I find it hard to believe 300kVA to 400kVA made the difference, but the only documented change was the x-former upsizing, so I can't ignore it.

 

[tom kanzler]

Did I miss the frequency range that you are working with? You refer to voltages and current but your application is with s VFD we both no frequency range provided. By knowing the frequency maybe then we can address what affect it has on the transformer.

It's a vector drive, and can run down to zero rpm under load. I don't know what the max frequency is, though. The manufacturer doesn't share very much.

edit: Acceleration time is estimated at 6 seconds on the 140kW (rated output) function. Hoist speed and load is consistent with that rated output figure, so that's no including acceleration transients.

 

[Jraef]

As a gross general rule, a VFD can provide you with full rated torque from the motor without exceeding the FLA. But that assumes you might have all day to accelerate it at that level of torque. So the faster you want to accelerate, the more torque, and current, you are going to need. A standard squirrel cage induction motor will potentially be able to deliver short bursts of "Breakdown Torque" of up to 220% of FLT, and a good Vector Drive (as these will be) will allow that to happen at any moment. To get there however, the current and torque become synonymous, so you need 220% current as well. The drive can only likely deliver that for up to 3 seconds, unless the drive is purposely over sized (again, one of the details). If your power source is too small at that moment that you attempt to accelerate the motor with 220% current, the voltage will drop and when it does, the VFD will compensate as much as it can. But in doing so, it will try to artificially limit current by extending the ramp time, or maybe shut itself off to protect itself if the programmer did not wish to allow for a ramp time over ride.

So bottom line, yes, it is possible that because of the load profile and what the system was being asked to do, the source transformer size might make a difference in it working or not.

 

[templdl]

It's a vector drive, and can run down to zero rpm under load. I don't know what the max frequency is, though. The manufacturer doesn't share very much.

edit: Acceleration time is estimated at 6 seconds on the 140kW (rated output) function. Hoist speed and load is consistent with that rated output figure, so that's no including acceleration transients.

Thanks, what fequency is the transformer designed for, 60hz? Run the frequency down less than that and the core will saturate the core and it will cease to function as a transformer. Higher frequencies should generally not be a issue.

 

[ron]

At no load versus full load (282kVA vs 300kVA), there will be voltage drop on the transformer secondary almost directly related to the impedance of the transformer. The voltage drop is made worse with lower pf.

So the increase to 400kVA, may have made the secondary volt drop less and reduced the control faults you were seeing.

 

[Jraef]

Ah, another of those devilish details. Is the step-up transformer AHEAD of the drive or BEHIND the drive? HUGE difference!

 

[petersonra]

Ah, another of those devilish details. Is the step-up transformer AHEAD of the drive or BEHIND the drive? HUGE difference!

why would someone put a xfmr downstream of a VFD?

 

[templdl]

Ah, another of those devilish details. Is the step-up transformer AHEAD of the drive or BEHIND the drive? HUGE difference!

Yes, but isn't fun to guess though? Using the transformer tty O step up from the VFD was the only thing that I could think of that made sense. What intregued me was the unexplained cause of a voltage drop which is characteristic of a transformer's core going into saturation when the transformer staps being a transformer.

 

[GoldDigger]

Yes, but isn't fun to guess though? Using the transformer tty O step up from the VFD was the only thing that I could think of that made sense. What intregued me was the unexplained cause of a voltage drop which is characteristic of a transformer's core going into saturation when the transformer staps being a transformer.

In general there is no way that any load on s transformer secondary can cause core saturation. That can only result from applying too high a voltage to the primary.
Putting a transformer after a VFD is, on the other hand, likely to expose the primary to high voltage pulsed DC, which it may not handle well at all.

 

[templdl]

In general there is no way that any load on s transformer secondary can cause core saturation. That can only result from applying too high a voltage to the primary.
Putting a transformer after a VFD is, on the other hand, likely to expose the primary to high voltage pulsed DC, which it may not handle well at all.

You would be correct if you defined the pri as the HV and sec as LV and are using the transformer as a step down. Core saturation can a lso occur when you reduce the frequency below that by which the transformer was designed. I ry to explain it as tty he frequency is decreased it be ones more like DC because of the reduced inductive reactance and the transformer stop bring a transformer and the Sec voltage fall. The same if you use the transformer to step up. Transformers are more forgiving of a higher than the rated voltage applied to the input side to a point where higher frequencies are normally not an issue where lower frequencies are.

 

[templdl]

In general there is no way that any load on s transformer secondary can cause core saturation. That can only result from applying too high a voltage to the primary.
Putting a transformer after a VFD is, on the other hand, likely to expose the primary to high voltage pulsed DC, which it may not handle well at all.

You would be correct if you defined the pri as the HV and sec as LV and are using the transformer as a step down. Core saturation can also occur when you reduce the frequency below that by which the transformer was designed. I ry to explain it as tty he frequency is decreased it be ones more like DC because of the reduced inductive reactance and the transformer stop bring a transformer and the Sec voltage fall. The same if you use the transformer to step up. Transformers are more forgiving of a higher than the rated voltage applied to the input side to a point where higher frequencies are normally not an issue where lower frequencies are.

 

[LMAO]

why would someone put a xfmr downstream of a VFD?

maybe VFD is not rated for 480V...

If control system is on the same source as VFD and it faults out, it may be due to harmonics which can be solved by line reactors upstream to VFD to reduce/isolate harmonic voltage and reduce harmonic current.

Controls can also run off a separate source like a UPS...

 

[Jraef]

why would someone put a xfmr downstream of a VFD?

At low voltage, you're right, it would not make sense because the higher current at 230V would make it a lot more expensive than just using a 480V drive. But it's done all the time in MV applications when people don't want to buy an MV drive but they have an MV motor with which they need to vary the speed.

It takes a special transformer design to be "best practice", however people do it with standard distribution transformers all the time, because 90% of the time they are for pumps that are not going to be used at low speeds anyway. Some pump skid mfrs in the oil patch industry (Centrilift) have standard packages built like this.

 

[tom kanzler]

Some more clarification. The transformer is on the line side, at 60Hz. The system is designed for 400V, but can be configured for 460V or 480V for use in North America. Same model units are operating fine on 480V generator power, though the generators are grossly oversized for the copper and iron, with the engines running lightly loaded.

Still waiting for current and voltage readings, and it doesn't look like that will happen until next week.

 

[LMAO]

Some more clarification. The transformer is on the line side, at 60Hz. The system is designed for 400V, but can be configured for 460V or 480V for use in North America. Same model units are operating fine on 480V generator power, though the generators are grossly oversized for the copper and iron, with the engines running lightly loaded.

Still waiting for current and voltage readings, and it doesn't look like that will happen until next week.

Magnitude of harmonic voltage distortion due to non-linear load is directly proportional to transformer impedance. Up-sizing the transformer (in other words reducing its impedance) brings down the voltage distortion on line. Harmonic current actually goes up but your controls are not affected by current distortion on line, only VOLTAGE.

This is the explanation I could come up with.

 

[mike_kilroy]

You said siemens vfd.... is your supplier aware MOST siemens vfds require 3% MAX IMPEDANCE xfmr? Unless custom built, ur xfmr is likely 5-6% Z.... one can CHEAT and get same effective Z by - ready for this? - oversizing the xfmr!

this is good chance why ur 400kva solved the 300kva problem...

Details details.....