Unexplained torque output from large VFDs 1

[tgmcg]

I'm working on a project where we are planning to replace an older, non-supported VFD with a new VFD of similar 12-pulse LCI-type archtecture. The motor and driven equipment (centrifugal pump...>10,000 HP) remain unchanged. Reading through the original test report for the motor and VFD issued by the motor vendor, I have discovered mention of huge unexplained torque spikes exceeding 100% FLT, and other excitation of the 1st TNF. As a result, I am concerned about the likelihood of encoutering similar unexplained torque spikes after installing the new VFD supplied by another manufacturer. Ater all, how can a torsional analysis account for huge unexplained torque spikes? (rhetorical question)

I'd be grateful for any insights as to causes and possible mitigations to prevent the occurrence of such potentially damaging torque excitation after installing the new VFD. Are we planning to use the best possible VFD design to prevent such anomalies?

I have read recent technical papers on subsynchronous torsional interaction (SSTI) when using PWM-type VFDs, but nothing in those papers metions unexplained torque spikes as large as 100% FLT. I'm doubtful any of the VFD manufacturers will share this kind of information with us, so am hoping to elicit some comments from folks who have dealt with these issues first hand.

 

[waross]

If you operate the motor at a "forbidden speed", then reflected pressure waves from the first elbow downstream of the pump will create a build up of pressure pulses. I have seen these pressure pulses blow a section out of the pump casing about the size of a hand. (On a 60 HP pump).
I expect that there will also be torque transients.
If the time for a pressure impulse to travel in your fluid to the first point of reflection and back to the pump is equal to or a small multiple of the period between the passing of one impeller blade to the next impeller blade, you are in a forbidden zone. The effect is related to the speed of the pump and most VFDs have provision to program in forbidden speeds. The drive will not stay at a forbidden speed but will accelerate past it to a safe speed. There is a similar action as the drive frequency drops.
This may or may not be the issue.

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

 

[Skogsgurra]

I think that Bill's thinking is right. The problem with standing waves in liquid filled pipes is real. No doubt about it. I have even met guys that are afraid of that phenomenon in gas filled pipes. But never seen that it happens IRL.

You do not say if the torque transients are periodic (if TNF doesn't mean Torsional Natural Frequency - which I start to realize that it may) or if they are starting transients.

If they are starting transients, there may be an explanation in the way voltage is applied to the motor. When the three phase voltage is applied to the motor, there is a transient that "sucks" the rotor into the resulting zero speed excitation field. This can usually very easily be seen and can result in short torque peaks that exceed rated torque, but not starting torque.

Many PWM inverters have a parameter that can be set so that voltage is ramped up in one or two seconds and, therefore, avoid that starting transient. The same thing should be possible to do in a CSI inverter by ramping up the DC current. But I don't think that I have ever seen that used in the few CSI inverters that I have met in my life.



Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[mikekilroy]

Far less likely compared to the fluid wave idea is that the combination of some motors & drives LRC values can cause motor to self oscillate at some lower than base speed frequencies just by running them at that resonant speed.

We had one 60hp GE spindle motor that would shake on way thru 900rpm on accel; if left to run at 900rpm it would osc at near FLA. Nothing wrong with the vfd or motor, just combination of the total LRC values in the system.

 

[Skogsgurra]

We discussed some of this in-depth a few years ago: thread237-248895 (Thanks for that "More of this" thing, Doug!)

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[tgmcg]

The observed unexpected torqe spikes and fluctuations are definitely not fluid-related. They were observed during a back-to-back VFD/motor test at the motor vendors factory back in the early 1990's.

Numerous failures have been reported in the literature attributable to SSTI or white noise from the VFD. Sometimes it's attributed to the controller of the VFD.

I've recently done a torsional analysis on a fairly complicated VFD motor driven machinery train. The VFD drive vendor will provide a torsional output signature describing the motor air-gap torque which is used as the torsional excitation to the torsional analysis. The torsional output signature desribes the frequency and amplitude of the various harmonic components of the pulsating torque. However, these harmonics rarely exceed a few percent in magnitude...which even at this level can be enough to cause fatigue failure when operating at or near a TNF.

My main concern are these very large unexplained pulsations..20%, 66%, and over 100% of FLT.

The situation we have now is that we're planning to replace only the VFD and not the motor. The VFD supplier can proivide the voltage and current output characteristic, but this needs to be converted into torque...and somebody needs to provide a guarantee on noise to enusre we do not have anything like the unexplained torques mentioned above.

 

[jraef]

It may be related to the technology used in the older LCI drive. If old enough, the output may have been 6-step using the old GTO thyristors, not PWM, and 6-step inverter drives did not provide the same kind of finite control we now have with PWM. GTO thyristor drives were also prone to firing issues at low phase angles and were notorious for having poor low speed torque control. I don't think anyone uses either technology any more.

"Will work for (the memory of) salami"

 

[tgmcg]

jraef,

As mentioned in the OP, the existing VFD is a 12-pulse LCI drive.

As a mechanical engineer, I lean towards using a VFD with the lowest achievable torque ripple and have done a torsional analysis for a large PWM driven compressor in the past. However, others on the team lean towards using a 12-step LCI drive.

I expect the 12-step LCI drive to be noisier than an equivalent PWM drive...possibly a lot noisier. But do not know how the two drives compare in terms of cost and resistance to obsolescence.

 

[Skogsgurra]

CSI (Current Source Inverters) are still being used and I think that Siemens still delivers some. Or not.

Anyhow, they produce torque ripple and I would not use one of them today. The torque problems that Jeff mentions are probably not an issue for a pump application but the power factor is bad (speed dependent) and the torque riple, as said.

They have a good reputation for being reliable. But most VFDs are that today.
I have seen a few Robicon MV drives lately and they don't seem to have any problems at all. Bearing currents, for one thing, were expected to be a problem. They are not. I measured close to zero.

Or just go mainstream and run PWM. Then you can use your old motor.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[tgmcg]

Gunnar,

It's possible we'll end up with a PWM design...I'm for the design producing the lowest amount of torque ripple.

But even with PWM, there will still be a concern with SSTI and white noise excitation.

 

[mikekilroy]

I give. I have been around vfd's from SCR days in say about 1970 thru GTO days say around 1975 to PWM CSI (current source inverter) & VSI (voltage source inverters), but in this 43+ years, I cannot recall what LCI might stand for. Sorry.

We had SCR 6 pulse systems, we invented 12 pulse systems, but never do I recall some LCI thing.... Perhaps it is time I am put out to pasture as too old to keep up, but please.....

can you make LSI into 3 words please?

 

[mikekilroy]

And since this is a pump, why in the world are you trying to minimize torque ripple? We have supplied some of the world's most pure sine wave output drives to minimize motor heating as proved with thermistors applied all around the motor for testing and proof, but why in the world do you want "lowest torque ripple, especially for a water pump?"

 

[sibeen]

LC = Load Commutated, so I'd hazard a guess to say that the OP is asking about a 12 pulse thyristor (csi) inverter.

 

[tgmcg]

Sibeen...you are correct. LCI = load commutated inverter.

mike kilroy...because it's not actually a water pump. It's a turbomachine with low torsional damping. Operating on a need-to-know basis.

 

[jraef]

LCI = Load Commutated Inverter, it is a variant of a Current Source Inverter, but uses the motor (load) in the switching scheme of the output thyristors. They are typical for high HP low speed machines, exclusively Synchronous. Instead of using GTOs or Transistors to switch the output waveform, an LCI uses the rotating motor itself to force regular thyristors (SCRs) to an off state. You rarely see them used below 5000HP or so.

"Will work for (the memory of) salami"

 

[davidbeach]

Common as starters for combustion turbines.

 

[Skogsgurra]

I think that I/we need to understand what we are saying. I (mostly European experience) call all Current Source Inverters CSI. And I can't remember them being anything but load commutated (in the meaning that the load's counter-EMF is needed to commutate the thyristors). Are there CSIs that are commutated any other way? I can't imagine a GTO or SCR with quenching circuit shutting off the highly inductive current in an CSI.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[sibeen]

You rarely see them used below 5000HP or so.

There is an Australian manufacturer of UPS systems called Thycon, which, until very recently (and still may) made 12 pulse output CSI UPS systems. All thyristor, and what they called "Line Commutated". They built them down to about 5 kVA

Admittedly, the output section of the UPS contained a shedload of capacitance, and a 'variable' inductor, so that an output voltage could be set up, and therefore perform the commutation function.

 

[tgmcg]

Hi Guys,

Do you mind if we steer the discssion back to my question regarding noise, SSTI and unexpected torque fluctuations?

Does anyone have 1st hand experience with this, especially if it resulted in mechanical failures?

 

[Skogsgurra]

Obviously not. If someone had, it should have shown by now.

I think that it is important that we know what we are telling each other. Definitions and acronyms that are not fully understood by the participants need to be defined.

So, to start over. Pleasy explain in some depth what your acronyms stand for, please.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.