3.5 MW motor starting problems continue... 7

[ters]

A while ago, I started a thread

http://eng-tips.com/viewthread.cfm?qid=235885&page=1

which many of you contributed to, but that story has not ended yet...

So the issue is that a 3.5 MW motor driving a large fan has problems starting. The starting method is the soft starter. Rotational inertia is high.

We discussed several causes, such as inadequate motor, inadequate starter, week and inadequate power system, etc. But calculations show that although things are marginal this still should work, but it does not.

Today we managed for the first time to put two identical 13/20MVA transformer in parallel. Their impedance is 10.5% at 20MVA but each transformer also has a current limiting reactor 6%. However, 20+20MVA system still would no start a 4MVA motor...

System: two 13/20 MVA transformers in parallel, + about 5 MW of additional load shared between two transformers.

Motor: asynchronous, TECO-Westinghouse 3.5MW, In = 385Amp, 6kV, 50Hz, 1485 rmp, locked rotor current 2300Amp.

Fan: Large rotational inertia. The fan has no load at the present (no air, dampers closed). The Impeller weight is 15,000 lbs, and the Impeller moment of inertia (WR^2) is 66,000 lb-ft^2.)

Starter: constant current, voltage ramp ~10 sec, current limit 470%, starting time 50 sec, after which bypass contactor closes (when then speed is about 80%).

We had some luck starting and running the motor using one 13/20 MVA dedicated transformer but now with 2 transformers in parallel it would not work. We tried three times and none of the starts was successful. Each time the soft starter was manually shut down after about 20 sec since motor was creating too much vibrations and was not accelerating any more.

Problems still could be electrical, but for a change we started to suspect that there may be some mechanical problems, perhaps related to the motor rotor axial movement. If the rotor indeed does try to move, either it may not have enough space to move to the electrical center or the shaft is moving back and forth around its electrical center for some unknown reason and hitting the trust.

The attached charts provide more info. During the voltage rump time, which was about 10 sec, motor runs smooth and with two transformers in parallel the initial voltage drop is rather modest - falls from 6.4kV to 6kV. This is good news, with one transfomer we had much larger initial voltage drops. However, the bad news is that after the voltage is ramped up high enough that the current reached the preset limit (470% In), some vibrations start.

As you can see, the current and voltage outline forms on the charts start at some point having "teeth" like a saw. Once stabilized there is about 3 - 4 “teeth”/second. Just judging by the ear, the frequency of mechanical vibrations is in the same range, meaning current spikes likely correspond with frequency of mechanical vibrations.

Any idea what might be happening?

 

[edison123]

ters

Aren't the wye/wye transformers with unconnected neutral supposed to have unstable neutral potentials ? Would that cause the soft starter to misbehave ?

I agree with pete 16.67 Hz is not a triplen harmonic.

If the motor is stalling all the time at a specific speed, I would say the stator/rotor slot combination is more of a problem at this point.

 

[waross]

The motor used to start.
A second transformer was added and the motor would not start.
The series reactor was jumpered and the motor would not start.
This lends weight to the suggestion that the system is incompatible with the motor/soft starter combination.
Why not try a start as it used to be, one transformer and the series reactor in the circuit? How about adding capacitors to shift the resonant point?
If you can start with one transformer the strip chart recording may give some clues.

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

 

[ters]

Electricpete, I have the same dilemma. As you said, 16.67x3^2 is 150, that is thirds harmonic. But the modulation I see on the chart has a frequency of about 3 times in one second, not 150 times, so I don't understand it yet either.

Edison, we have an option to ground the 6kV neutral through a reactor - there is disconnect swith for that. The Client keeps it usually ungrounded as the capacitance in the system is not (or was not) high enough. So we can try with grounded natural and see what happens. However, the soft starter measures the line voltage with two PTs in an open delta configuration, and uses that information for firing circuitry so why it would care about floating neutral?

Waross, we will go back to one transformer scenario when we can - at the present it is no transformer scenario as the substation is being rebuilt. However, as it behaved then, that mode is not sustainable as a long term solution due to long starts and dip and unstable voltage drops. We were able to do it only having one transformer dedicated for the test. However, as you pointed out, maybe adding capacitance would help in either case (one transformer or two) but I think we have to know what the exact problem is in order to add something which will compensate for it.

 

[Marke]

Hello ters

Induction motors have a series of slots in the stator and in the rotor. These slots should not be equal in number because if they are, there is a good chance that the motor will not start at all due to a characteristic known as cogging. The slots will align like a stepper motor.
For this reason, there are an unequal number of slots in the rotor and in the stator, but there can still be situations where the slot frequencies coincide with harmonic frequencies and this can cause torque modulations. The slots are skewed to keep an overlap on all slots to reduce this problem.
Another characteristic of induction motors, is crawling. There are harmonic fluxes developed in the gap due to the magnetics of the motor. These harmonics create additional torque fields. A common problem is with the seventh harmonic where the seventh harmonic creates a forward rotating torque field at one seventh of the synchronous speed. There will be a maximum torque just below 1/7 Ns and if this is high enough, the net torque can be higher than the torque due to the line frequency where at 1/7 Ns, the slip is high. This can cause the motor to crawl at just below 1/7 synchronous speed.
There is another crawl speed at 1/13 Ns.
When a motor is started with a soft starter, the gap harmonics are increased by the harmonic currents produced by the phase controlled SCRs of the soft starter.
A motor that has a tendency (small) to crawl when operated with a clean supply, will have a much greater tendency to crawl when controlled by SCRs and a chopped waveform.

The subject of harmonic torques and harmonic fluxes can get very complex when you try to analyse the effects of all the electrical harmonics, plus the magnetic harmonics plus the slot noise.

The seventh harmonic is the best known and documented crawl speed, but others can exist due to the total interaction of all harmonic sources.

The two critical speeds are 1/13 Ns and 1/7 Ns. If the motor crawls just below either of these speed, there is a definite interaction between the motor and the harmonics produced by the soft starter. At other odd order non triplen harmonic speeds, there can also be more complex interactions. - I have seen reference to interactions at 1/11 Ns but I can not recall the mechanism.

The design of the motor, no of slots in rotor and no of slots in stator, and the skew of the slots are all selected to minimise these problems, but all design is a compromise.

A motor crawling at say 1/7 Ns will produce maximum torque at a low slip (relative to 1/7 Ns) in the same way that the maximum torque is produced at just below Ns. There will therefore be a "slip" frequency at the maximum torque. This will cause a modulation in the current flow in the motor and this in turn can modulate the commutation angle of the SCRs.
Where the SCRs are triggered relative to the voltage wave form, there will be a modulation of the conduction angle of the SCRs as observed. This will tend to accentuate the crawl torque by effectively amplitude modulating the line frequency applied to the motor. Where the SCRs are triggered relative to the current waveform, the phase modulation will not affect the conduction angle of the SCRs and the effect will not affect the crawl torque.

Where the phase modulation of the commutation angle causes an amplitude modulation of the motor current, and there is a current control loop, it is possible that the response time of the loop can further amplify the modulation.
The greater the level of modulation, the lower the probability that the motor will accelerate through the crawl speed.
In the same way that the torque just below crawl speed is a maximum, the torque just above crawl speed is a minimum. There must be sufficient synchronous torque at the minimum to accelerate the load to full speed.

The major influence of the supply on this scenario, is that an increased supply impedance will result in increased conduction angles and reduced harmonics which will reduce the crawl torque. A reduced supply impedance will cause a reduced conduction angle for the same current and this will increase the harmonic content and result in an increased crawl torque.

At this stage, we have no positive proof that you are having crawl torque issues, but the characteristics that you have described suggest to me that this could be the issue.
The best indication is to determine the actual shaft speed where the motor acceleration stops. If this is at just below 1/13 Ns (115RPM) or 1/7 Ns (214RPM) then I would suggest that this is a very likely scenario.

In my opinion, the low frequency modulation of the current waveform does not indicate a rotor bar problem. If the motor was operating close to synchronous speed, then I would certainly be concerned about a damaged bar when presented with a low frequency modulation of the current, but the slip frequency in this case is very high, so bar modulation would also be at a high frequency.

Best regards,

Mark Empson
L M Photonics Ltd

 

[waross]

I did not mean the one transformer stat as a long term solution. I thought that it may be valuable to compare a recording of a successful start with the recording of a failed start. Something may become apparent.

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

 

[ters]

Mark, I really admire your desire to help. It is almost hard to believe that one can dedicate so much time and efforts to provide such a detailed explanation.

I don’t have much new to report, we cannot run the machine until about mid next week. At that time we will be able to try a couple of new things as we have not tried yet, and also do some more measurements.

If the increased system impedance does help to dampen oscillations, unfortunately by increasing it we will be encountering another problem, the voltage drop too big and lasting too long. With one TR (=large impedance), the system capacity becomes marginal with only 90 MVA SC level at the process switchgear (next to the soft starter). So these two things are rather conflicting. During successful starts with one transformer, in addition to voltage drop, we also had voltage oscillations (on the line side – before the soft starter). The transformer we used is a 13/20MVA, with 10.5% @ 20MVA, and we have 2 such units. Its capacity is rather barely adeqate for the start as the motor draws something like 20MVA during 50sec (at 470% current limit).

We hope to have some more luck with the new third transformer we are just installing which is 20/27MVA with 14% at 27MVA, it will not have a reactor, but will have some longer cable before it connects to the same bus, so the overall impedance will actually be somewhere between one old TR + reactor and two old TRs in parallel (with reactors). But it might behave differently to worse or better – the total cable length with the new transformers before it reaches the motor with be substantial – some 700m.

Once we get out of the puzzle, we are supposed to have three such 3.5MW motors running... Just wondering is one is already running, would starting the second one on the same bus make its start more difficult? Or maybe for some reason the first one which is running might actually help the second one to start?

Electricpete, we have a six-lead motor connected in WYE where the neutral in former in the termination box for differential protection CTs. So it is WYE (transformer) – WYE (motor), ungrounded, with the soft starter in-between. If the transformer neutral is not grounded though a Petersen’s reactor (which has been the case for years), then the only sort of reference to ground are switchgear PTs, there are 3 of them, a WYE connection.

 

[ters]

I don’t think I posted this before. The attached file are two charts, for two failed starts, both showing period 9 – 14 sec. The only difference is that one start was with two transformers in parallel (and their reactors in the circuit), while the other start is with one transformer only, with its reactor removed from the circuit (shorted).

As you can see, the impact on the line voltage is much severe with one transformer only, oscillations/amplitude modulation/resonance/disturbance, whatever you prefer to call it, is much more noticeable. These oscillations are even noticeable on the 110kV side.

The equivalent impedance of two transformers in parallel with their reactors in the circuit (not shorted) is not much different than having only one transformer with reactor shorted (transformers are 10.5% @20MVA and reactors are 6% @ 1000A). However, one transformer with no reactor still has slightly more impedance than two transformers in parallel with reactors. So this case may not support the explanation that less impedance makes it worse, although I think that this difference is oscillations is more due to marginal MVA capacity of one transformer vs. adequate capacity of two in parallel, since using a 13/20MVA transformer for a start which draws about 20MVA seems to be rather marginal.

If we dedicate new 20/27 MVA (14% @27MVA + 300 of cable) for those three large 3.5MW motors only, which we could do, that might prove to be some sort of optimization between the system impedance and capacity, which we just happen to have by a chance, but we will know if it works only when we are in a position try it, a couple of weeks down the road.

 

[ters]

For those still interested in this ever lasting post... Here is some news. We were in a position to do a direct online start today for the first time. With two 13/20 MVA transformers in parallel it was an easy start which took only 26 seconds with the fan coupled.

The voltage drop was not that bad, from 6.4kV to 5.8kV and the voltage remained firm and steady during the entire start, without swings as we used to have with the soft starter. The LRA current of 2300Amp existed for the first 200ms or so, and then stabilized at 2000Amp and remained there very constant until about sec 25 when it dropped to something like 130Amp at full speed.

This DOL start was actually noticeably softer than with the soft starter...

So we seem to have a defective soft starter, especially having in mind our last attempt to use it when SCRs would conduct for say 0.4 sec and then make a pause for 0.2-0.4 sec and kept doing it for about 50 sec, at which time it was tripped. A sample of such intermittent soft starter operation is attached. From the top to the bottom, values are: line voltage B-C, motor voltage B-C, phase motor voltage A and line current A. This was a start with one transformer, which in the past was not intermittent, but now it is all time. A start with two transformers behaved differently, the soft starter did not work intermediately but the motor halted at about 150RPM as previously discussed. So we will be replacing the soft starter with the spare one, and even though I'm sure the new one will behave differently, my felling is that using a soft starter for this particular application is rather waste of time, or at least it is waste of time trying to use a particular make and model which we have,unless the manufacturer is willing to get involved and do some modifications on it.

 

[burnt2x]

ters,
Thanks for the closure post. Good to know you have singled-out the defective part!

 

[ters]

Thank you burnt2x. Looks like we did narrow it down to the single component, and we now seem to see light at the end of the tunnel, but the tunnel is still fairly long

 

[Marke]

Hello ters

The problem may be a faulty soft starter, or a compatibility between the starter and the motor, or the motor may have a tendency to crawl which is exaggerated enough by the soft starter harmonics to make the motor crawl.
It is easy to blame one component when it is the combination that does not work.

Best regards,
Mark.

Mark Empson
L M Photonics Ltd

 

[ters]

OK Mark, you are probably right. Perhaps we should blame ourselves for not realizing that application of a soft starter for loads with high rotational inertia may be marginally useful, even if it works.

However, if the soft starter is defective, then the new one may still offer some advantages vs. DOL start, if it works say at 400% In, in which case the start with only one transformer will be possible what would simplify our life eliminating time consuming procedures for reconfiguring the power system for each start (paralleling transformers and reverting back to normal).

The soft starter we have been struggling with for 3 moths needed at least 470% In, drawing so the current in the range of 1800Amp with unstable voltage and current conditions.

The DOL start, with the voltage drop to 5.8kV drew 2000Amp steady, so even if the soft starter was working fine all the time (which was not the case) and was not causing voltage and current swings and offered shoot and steady start like it did most of the time, still the difference 1800 Amp vs. 2000 Amp is obviously not very dramatic, only ~ 10%.

I'm attaching the info (chart) of the DOL start for reference for those that one day may struggle with similar issues and will report back what happens when we replace the soft starter.

 

[edison123]

If your trafo has an on-line tap changer, you could try using it as a soft starter.

 

[LionelHutz]

Good feedback. Your last chart recording = OUCH! It appears the soft-starter went completely unstable.

I'd like to comment on your post. A soft-starter will work just as well with a high inertia load as it would work with any other load. In your case, you should have been able to achieve a current reduction down to about 1350A, but more realistic about 1600A = 20% current reduction. If that is enough to allow a start with a single transformer then the soft-starter is worthwhile. If it's not enough of a reduction to allow a single transformer start, then there is little point in using a soft-starter.

I'd like to expand on something I posted before. Digital control loops, in this case the current control loop, are not nearly as simple as they seem. I'd bet, conservatively, that we spent over $50k on the current control loop alone for our newest soft-starter model. I know who was involved and how long it was worked on. And, it does work much better than any of our previous starter models. With our older model, it was possible to pick some unrealistic settings and create an instability similar to what you are recording.

 

[ters]

LionelHutz, thank you for your comments. Yes the old transformers we have are rated at 1800Amp, so if the soft starter could start the motor with one transformer which is already partly loaded, then it would definitely be worth having it.

DOL starts may be even possible with one transformer with plenty of impedance in the circuit to work as a soft starter, but the disadvantage of using DOL start with one trafo would be that no other loads could be on the same bus. With two transformers in parallel, and initial voltage of 6400V, the voltage drop is not severe, so other loads on the same bus may be OK to live through a 25-sec modest voltage reduction. However, paralleling and un-paralleling is a bit time consuming.

We also have a new transformer rated to give 2600A, not commissioned yet, so there seems to be some other options on the horizon, but temp unavailability of one of the transformers may be an obstacle in any case.

Your last comment on the quality of the current stability loop does likely explain our case if not to the full extent, then at least partly.

edison, we have online tap changers on all transformers, but they aren’t fast enough to replace the soft starter in varying the voltage during some 30 sec

 

[ters]

Here is another closure statement on this. Probably not the final one

Meanwhile, we added one 110/6kV transformer, 20/27 MVA, 10% @20MVA. And looks like that due to its higher impedance and 700m of cable in the 6kV circuit, it works as a "natural soft starter"...

So we now start directly online with this new trafo and encounter a very smooth start lasting only 25-30 sec, voltage drops from 6.3kV to 5.6kV but it is stable and does not swing, starting current is average 1800 Amp expect for the inrush, when it may be reaching about 2200Amp, but it drops down with the voltage drop quickly and stabilizers around 1800Amp in ~ 1 sec, before it finally drops to about 130Amp at full speed (dampers closed fan has no load except some air leaks)...

So it appears that the soft starter provided no advantage, since it also needed 1800 Amp as minimum current (470%), which is the same as for direct online starts... Plus, with the soft starter, the start was two times longer then with direct online start (55 sec vs. 25 sec). Also voltage swings were huge with the soft starter, but using direct online, the voltage remains rock stable after the initial drop of 10%...

So, in summary, the start without the soft starter is actually much softer then with the soft starter...

However, we are still planning to bring ABB in to try to determine why their soft starter is not performing any better (theoretically it should start the fan at 1400 Amp or lower), and if they find a fix and if it can start the motor with significantly less than 1800Amp, maybe we will still use it... But I'm not very optimistic. For this sort of rotational inertia, using a soft starter seems to be a rather suboptimal solution... The fan cannot be started without some 400% In as a minimum, while the direct online start is theoretically 600%, but due to system impedance it never gets higher that 500%. So 400% or 500%, the difference may not be significant and worth the efforts...

For a change, we now have major fan vibration problems. All sort of rework has been done on bearings, pedestals, foundations, alignments, etc, still the same and gradually getting worse... Only the fan has such problem, motor bearings vibrations are within limits...

 

[Marke]

Hello ters

Do not let this experience convince you that soft starters do not work with fans etc. Soft starters do usually work under these conditions and give a very worthwhile result. The current reduction is a function of the load torque/speed curve and the motor speed torque curve, but in most cases there are worthwhile improvements.
In this installation, there appears to be a problem of some form that is causing the system to go unstable. This is not necessarily the soft starter, rather the whole system.

I note with interest

we now have major fan vibration problems

. Perhaps the real gremlin is starting to show it's head??

Best regards,

Mark Empson
L M Photonics Ltd

 

[ters]

Thank you Mark. We always had higher fan bearing vibrations (higher than anticipated by the manufacturer). However, it does not seem it has got much to do with the (soft) starting problems, but could be related, who knows. Earlier, it would stall at something like 150 RPM and vibrate, but the vibrations I’m now referring to are vibrations at full speed (1480 RPM) and they initially were 2x "expected" ones, then 3x, and now 4-5x. We just redone sole plates and pedestals' arrangements, bearings, checked alignment, coupling, anchor bolts seem to hold fine (torqued properly), etc, but it absolutely made no difference. None... It seems that somehow impeller critical speed when some mechanical resonance happens is about the same as nominal speed...

The vibrations cease swiftly upon the motor is shut down, within about 2 sec they drop to half and continue progressively dropping, while the fan speed does not drop that quickly due to inertia.

Several skilled experts measured vibrations, but nobody came up with a positive conclusion what is going on. The motor shaft does not vibrate that much, in average 4x less than the fan.

 

[Marke]

Hello ters

It sounds like you have some resonance somewhere, and it is possible that the harmonics were being excited by the action of the soft starter. Once resonance occurs, the whole system loop can begin to chase the resonance and exaggerate it.

Best regards,

Mark Empson
L M Photonics Ltd

 

[ters]

Once resonance occurs, the whole system loop can begin to chase the resonance and exaggerate it.

Very encouraging words . So what can we possibly do to avoid resonance being chased and exaggerated?