Overspeed with a VFD 4

[MKMason]

Hopefully this is an easy question. Which motor is capable of going overspeed, induction or synchronous? How? What would be the maximum increase allowable (%)? Does service factor figure into the equation, ie, does a motor (power) SF of 1.15 limit the speed also to 1.15 x synchronous speed, or some other factor limit?

 

[rhatcher]

There was a similar, recent, thread titled '87 hertz motor' in which the idea was suggested that a motor could be operated at a constant volts/hertz ratio above the design frequency to produce a higher than designed power output.

As an example, it was suggested that a dual voltage 230/460V motor could be operated on the 230V connection with an input power of 460V/120hz to produce double the rated power output. Since the insulation rating would be suitable for 460V power and the volt/hz ratio was kept contant, the idea was that this would result in a 'free' increase in power output that was double the original design output.

While in principle this seems possible, the fact is that in practice that this will not work. The reason is that while the load current, and therefore the motor I2R losses, will remain constant with a constant v/hz frequency change, the core losses will change in proportional to the change in frequency. At a constant volt/hz ratio, a lower frequency results is a proportional lower core loss and a higher frequency results in a proportional higher core loss. In most cases, if not all, the increase in core losses as the constant volt/hz frequency increases above the design value will result in an increase in temperature rise that will exceed the designed temperature rise. The result will be overheating of the motor and a reduction in service life.

 

[mikekilroy]

While in principle this seems possible, the fact is that in practice that this will not work. The reason is...the core losses will change in proportional to the change in frequency.

...In most cases, if not all, the increase in core losses as the constant volt/hz frequency increases above the design value will result in an increase in temperature rise that will exceed the designed temperature rise.

The result will be overheating of the motor and a reduction in service life.

We should apply real world values to these statements.

So you are saying to run to 120hz in constant v/hz mode will cause the motor to overheat, even no load?

http://www.KilroyWasHere<dot>com

 

[ozmosis]

MKMason
I think you need to get back to the basic requirements, all the way up the chain to the core requirements of the application.
Forget a VFD for the moment, forget what a motor can and cannot do, forget your voltage supply.
You are pumping slurry. You need to change what you have and this demands a view on a few basic questions such as flow rates and required head. This will determine, in part, your system curve on your pump. I'm no pump expert( this is the stage guys like biginch and littleinch jump in and tell us electricals what a load of cr@p we have been discussing...) but I do know that if your pump selection is not correct, then you will be in a whole lot of trouble down the line with everything else. Once you establish your pump curve, then you can design the electrical equipment to support your mechanical requirements.

 

[MKMason]

Ozmosis:
I know my pumps pretty well, and understand the issues around pump curves. It's the Process Engineers that do not understand what can and cannot be done and there is my battle. Sometimes I must re-rate pumps with VFD's on the motors, sometimes buy new pumps, and other times the process demands more flexibility. Slurry is a constant torque application. I do know this: a 1.15 SF on the motor HP will only allow for an almost 5% increase in flowrate due to pump affinity laws. It won't be enough for me, I needed 25%, and I need to consider a higher speed system if I intend to get the flowrate up. My electrical engineer here, and I, are reluctant to run an existing pump into overspeed.
This has all been a very interesting discussion, I'll agree. The Reality is, for a refinery type application, long life and reliable service are the governing factors, not "what tricks can I pull?"
Once again, thanks to all of you for your input: you've outlined the project limitations remarkably well. Hats off!

 

[mikekilroy]

We who apply vfd drives for a living say 'throw away your SF 1.15 rating - it no longer has any meaning when a vfd is used:' assume a SF=1.0 - no extra overloading allowed over nameplate no matter what your particular motor says. It is a good rule of thumb you should consider using since you have referenced SF more than once....

http://www.KilroyWasHere<dot>com

 

[KllrWolf]

I second mikekilroy. As a motor manufacturer, a VFD automatically drops it to a 1.0 service factor on the motor.

 

[MKMason]

OK, I'll accept that. This means I really can run the induction motor to 150%(-ish) speed? And not freak about overheat if I'm staying at constant torque?

 

[mikekilroy]

OK, I'll accept that.

great!

This means I really can run the induction motor to 150%(-ish) speed? And not freak about overheat if I'm staying at constant torque?

Sorry, not sure how you concluded this from SF=1? Two totally separate issues.

Others already went thru the 150% speed and torque available there.... I will summarize all those answers and you can go back and reread the great posts above for details why.

I really can run the induction motor to 150%(-ish) speed?

YES

not freak about overheat if I'm staying at constant torque?

YES if you keep rated v/hz curve all the way.
NO if not

http://www.KilroyWasHere<dot>com

 

[ozmosis]

MKMason
I didn't mean my comment to be disrespectful in any way. It was just a comment to highlight the need to keep focused the basics in the application, and that always starts at the prime mover.
This is also coming from someone who is in the VFD business for a living and needs to ensure the application fits any recommendation for drives, if it is to be used correctly.

 

[jraef]

I think what we can all agree on is that old axiom, "There is no free lunch." You cannot get something for nothing, including getting more from a motor than it is designed to give.

But here is where the trick mentipned earlier can make it LOOK as if you are. So let's start with this to maybe help you understand, and taking your word for it that this is truly a constant torque application.

For the most part, torque and current follow each other.

Torque in an AC motor is also directly related to the ratio of voltage and frequency applied to it, the V/Hz ratio.

If I want to run a motor at 150% speed then, I need to accept two principals:
1) If I apply the motor rated nameplate voltage at 150% frequency, I am running it in Contant HP mode, meaning I begin to LOSE torque once I get over base designed speed. This is because my V/Hz ratio is dropping as Hz increases with voltage staying the same.
2) If I want to maintain the same torque at 150% speed, I must also then increase to 150% Voltage, so that my V/Hz ratio stays the same.

So under #2, since I am maintaining my V/Hz ratio, I am maintaining my torque, which means I am maintaining my current as well (in theory). Motor heating is mostly the result of current based losses, so I am not really over loading that motor in that sense, because my current is still within design limts. "HP" is just a shorthand notation saying xxx torque at yyy speed, so with the SAME torque and a higher SPEED, I am actually getiing more HP from that motor. But in reality, I am really still getting the same torque, which in the case of a constant torque application, is what I need.

Now the theory part. A little over half of the losses in the motor are associated with current, but not all. Friction and windage losses will increase with speed, and iron losses, which are based on applied voltage, will increase as well. So this robs your motor of capacity in that sense, because running at the higher speed will mean the non-current base losses become a greater percentage of the total. A good Vector drive however will be capable of optimizing the motor operation in terms of the some of the voltage related losses, so that will help. Still, a 1.15SF may not be enough if your motor torque requirement for that pump is right at the limit of what that motor can provide. If on the other hand you already had a 20% cushion in the design, plus a 1.15SF, you may be able to pull it off.

But still, it comes down to you being able to supply 150% voltage at that 150% speed.

"Will work for (the memory of) salami"

 

[waross]

In regards to higher voltage: This is in relation to motor rated voltage.
Another option is to keep the applied voltage the same and to reduce the motor rated voltage.
Star delta; If a star connected motor is reconnected in delta, the effective rated voltage drops to about 58%.
If a 460 volt star motor is reconnected in delta, the effective rated voltage is now 265 Volts. With a 480 volt supply, (460 Volt utilization voltage) you have enough voltage to go 173% over speed and over voltage.
Dual voltage; If a 460 Volt rated motor is reconnected for 230 Volts, a 480 volt supply will allow the V/Hz ratio to be maintained up to 200% speed and 200% voltage.
The point is that it may be easier and cheaper to reconnect the motor than to adjust the supply voltage.

I would determine the maximum HP needed at the maximum flow and head. Then I would consider trimming the impeller to give the best pump match for the best available motor speed.
Consider going oversize on the motor.

A question in regards to the losses jraef; When the frequency is increased, do the iron losses increase for the whole motor or just for the stator? I am thinking that the slip frequency and the rotor frequency may remain the same. Comments??


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[mikekilroy]

Also remember that although the losses DO increase SOME (it is NOT a major amount) at higher speeds, these losses are EASILY taken care of with the shaft mounted cooling fan going faster and blowing a lot more cooling air.

Someone a few posts back suggested the losses are so great this is not even possible to do; I would not want to have to be the one going back to my customers where we have run 230v motors upto 460v@120hz (and then on up even higher) and tell them it isn't possible! We probably have 400 various size motors (5-100hp) in the field over the last 30 years this way!

I have never run into a motor run this way, even with a separately excited fan, that had any heating issues maintaining rated torque above base speed upto 2x voltage.

http://www.KilroyWasHere<dot>com

 

[waross]

Thanks for the information Mike. I suspected that was so. It is good to hear it from someone with first hand experience.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[waross]

4160 volts was mentioned. At the HP we are looking at a 2300 Volt motor on a 4160 Volt VFD may be a good option.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[jraef]

MKMason, what is the voltage level, ie Medium Voltage or Low Voltage, that you are intending to use for the motor and drive? You mention having 4160V available, but is that a viable option for the user? A MV drive will be a lot more expensive for the drive, but eliminates the transformer and losses in the transformer in terms of operating costs. But in some cases, the user may not want that because they have nobody on staff qualified to even open the door of MV equipment, let alone service it because in some jurisdictions, working on anything above 600V (the definition of MV here) requires special qualifications or certification. So if we are to offer suggestions, it would be nice for you to narrow the field a bit if possible.

"Will work for (the memory of) salami"