We use 6.6 kV Squirrel Cage Induction Motor 4pole 6.6kV star connected in our blast hole drill application. The motor is a 500 hp motor, driving an air compressor at one end and set of hydraulic pump on the other side. The motor is started using Auto Transformer starter connected as Korndorffer Starter(closed transition). We find flash over taking place while switching from 80% tap connection to 100%. The flash occurs at Star contactor of the vacuum contactor or line to ground in the contactor. Can some one explain any such experience and possible solution.

K.Gurumurthy
Manager Splecial projects
Ingersoll-Rand India Limited
Peenya Industrial Area
Bangalore 560 058
INDIA
Ph: 91 80 8395791 extn 3103
Fax: 91 80 8394651
email: [email protected]
mobile: 98440 30417

 

Responding to K. Gurumurthy's Fri, Dec 7, 11:61 am, query:

The two most frequent reasons for flashover are:
a) transfer current is too high, which means that transition occurs before motor has sufficiently accelerated, thus, inrush current is too high.
b) voltage across AT starter contacts is too high, whivh means that the transition between the 80% and 100% taps is too long so that the motor's residual terminal voltage is well out of phase wth the supply.

Thus, adequate analysis will depend on system parameters. The following will help:

1) Is this a new or old(er) installation?

2) What is ratio of motor to load inertia?

3) What is motor torque characteristic, that is, NEMA B, C, or D? And, its starting kVA ratio (locked-rotor to rated current)?

4) Is transfer to 100% determined by motor speed? Or, is it a fixed time?

5) Does motor attain nominal speed with 80% tap?

6) What is ratio of closed-transition resistor ohms to motor impedance (or motor size)?

7) Are you able to determine flashover path, that is, between phases? Or phase-to-earth.

8) Is 6.6 kV source solidly earthed or impedance earthed (resistor, reactor, other)?

9) Is AT deta or wye cconnected? If Wye, is AT neutral earthed? Solidly? Or with Resistor?

10) What is ratio of source transformer kVA to motor Hp?

Regards,
Phil Corso, PE
(Boca Raton, FL)
[[email protected]]

 

Further to my earlier response to K. Gurumurthy's Fri, Dec 27, 11:16 am, query:

Question 6) refers to a closed-transition resistor used in (some) European practice. It is not normally supplied in USA designs.

Regards,
Phil Corso, PE
(Boca Raton, FL)
[[email protected]]

 

Further to my earlier responses to K. Gurumurthy's Fri, Dec 27, 11:16am query:

Assuming that more than one flashover has occurred, that location is random, and that vacuum bottle integrity has not been compromised, then, an approach to the solution for case a) follows:

Use the speed-increment method to predict "normal" total acceleration time, Tat, at 100% volts. Note: for the "usual" combination of a centrifugal compressor & NEMA 'B' motor, Tat is 3.0 to 5 seconds! Then, the total acceleration time, T'at, for the 80% tap connection, will be K x Ts, where K = reciprocal of %-tap squared, or K = 1 / (0.8^2) = 1.5 x Tat! If the actual time-to-transfer is substantially shorter, then case a) is the problem. Caution: for the reciprocating compressor & gear-box case inertia plays a very important part... total acceleration time could be substantially longer!!

If you need additional detail, I can forward the calculation method.

Regards,
Phil Corso, PE
(Boca Raton, FL)
[[email protected]]

 

My earlier responses on this subject may indicate that I'm obsessively fixated on motor starting time! Well, I AM. I have witnessed or investigated numerous catastrophes involving excessive starting time. Those of you who have read my past responses are aware of my definition of catastrophe!

In short, remember that upstream equipment like captive-transformers, auto-transformers, starters, contactors, breakers, and their like, are not as forgiving as motors when it comes to exposure to abnormal currents. Let me emphasis... my position is that starting currents are abnormal, but inevitable, events, and they should "exist" for as short a time as possible!

Case in point: A startup assignment took me to a large refinery to "assist" in commissioning of a 1,000 Hp loading pump. When asked for comments, I questioned the status of the pump's discharge and inlet valves. Their collective reply "of course they are closed!" But, their wide-eyed stares said "typical dumb Americans, thinks he knows everything!"

The 6.6 kV motor was connected to an 11 kV bus via what is called a "captive" transformer! Starting time should have been just a few seconds. After 6 seconds, I yelled "shut it down!" Again, the wide-eyed stare which implied "what the hell is he talking about."

At 10 seconds the transformer blew up. The resultant short-circuit should have been cleared by the 11 kV breaker. It wasn't! The entire refinery was consequently tripped, the flare header overloaded, several unit fires ensued. Oh yeah, there were the endless "shouda, coulda, meetings.

On a happy note, everytime I visited later, I was asked to select the "entree for lunch" for the entire refinery staff.

My point is that in the minds of most... engineers in general, electrical engineers in particular... motor starting is trivial, by rote, a no-brainer! Nothing is farther from the truth,,, and its not based on "humble" opinion!

BTW, if interested in other stories, let me know.

Regards,
Phil Corso, PE
(Boca Raton, FL)
[[email protected]]

 

Phil,
More than one flashover occured. Vacuum bottles are OK. I am expecting our field engineer to return for a discussion. I shall get all the detail and hopefully post them on the mail on 1 Jan 2003. Thanks for the response. The field where this problem is occuring is far away and hence I am delaying the reponse. Pardon me for that.
Regards
__________________
K Gurumurthy
Manager, Special Projects

Phone: 91 80 3722570
Fax: 91 80 8394651
Mail: [email protected]

 

Phil,

Thanks for the leads.

1) The same machine system is working using auto transformer starter, with 80% tapping, at 3.3kV without anyproblem which was also recently supplied to the same site. There are 4 machines with 3.3 kV input and all are OK. Only these 2 machines of 6.6kV is giving problem. We are able to start these 2 machines also using DOL starting with 4 to 5 seconds startup time.
2) The ratio of motor to load is 33/10.2 = 3.235
3) Motor torque characteristics is closer to NEMA B and Locked rotor current as per test certificate is 4.5 times full load current.
4) The transition is timer set at 6secs. When the switching takes place, without motor pickingup speed, high current rush takes place and flash is observed.
5) When connected at 80% tapping and started, upto 6 secs, the motor was not reaching the speed for transition.
6) Ratio of closed-transition resistor ohms to motor impedance is 0.6 ohms/ 1.531 ohms. I am not sure, whether I am giving this value correctly. We assume the closed transition resistance means, that portion of the auto-transformer, through which current flows during transition and its DC resistance.
7) Since the panel doors are closed, the observation was after occurence of flash and is reported as from line to ground in the star contactor. It appears the energy available for the flash is not much and hence, the burnoff is localised to a spot.
8) 6.6kV source is resistance earthed.
9) AT is wye connected and neutral is not earthed.
10) Source transformer is 3.5 MVA 33kV/6.6 kV. Motor is 450hp with 1.1 service factor.

We are now proceeding to resolve the problem as follows: a) Set the timer for a higher time in 80% mode and see whether motor accelerates. If it does not, do not switch to 100%.
b) Unload the compressor by closing the air intake and try to start.
c) Change the auto transformer and try. (Luckily we have one successfully working at 6.6kV, made by another vendor)

Thanks for your inputs and our field engineer will take a few days to resolve the problem. I shall keep you informed, when we finally succeed. Your inputs were valuable and we appreciate.
__________________
K Gurumurthy Manager, Special Projects
Phone: 91 80 3722570
Fax: 91 80 8394651
Mail: [email protected]

 

Responding to K. Gurumurthy's Thu, Jan 2, 7:20pm transmittal providing additional data:

Urgent
Do not increase time-to-transfer, unless you have confirmed withstand capability of AT!

Conclusion
The data you provided indicates, to me, that net accelerating-torque (motor-torque less load-torque) is insufficient to bring machine to full speed! Thus, motor is 'hung-up' at a speed at which the inrush current is still high! This is corroborated by fact that all is OK for 100% case. Remember that at 80% motor torque is lowered to 64%.

My conclusion is corroborated by the fact that all is OK for the 100% voltage case. You must examine the motor's torque-speed curve, as well as that of the driven-load. I suspect that net-torque beyond the maximum pullout point, drops off very quickly.

Then one must ask:

1) What is the driven-machine?
2) Are you sure it is unloaded?
3) Are 3-kV and 6-kV motor designs identical?
4) Are motor parameters available including, but not limited to, inertia, no-load and locked-rotor tests, torque-speed curve?
5) Are the driven-machine's parameters available, including inertia, and torque-speed curve?
6) Is 80% tap really 80%?
&) Can you record runup speed?

I will reply to your other questions, later.

Regards,
Phil Corso, PE
(Boca Raton, FL)
[[email protected]]

 

Responding to additional comments in K. Gurumurthy's Thu, Jan 2, 7:20pm transmittal providing additional data:

1) Installation history provided.
2) Because load-inertia is less than motor inertia, assuming no intervening gear-box, then inertia is not the problem.
3) For math analysis, the torque-speed curve, NEMA or equivalent, will have to be obtained from manufacturer! Compressor as well!
4 & 5) Your reply that motor accelerates to full-speed w/100% volts, but doesn't w/80% tap, is the principle clue!
6) Not applicable! Resistor I referred to is not dc resistance of partial AT winding.
7) Is flashover mark evident at both neutral vacuum bottles or just one?
8) Low magnitude arc damage is due to main transformer neutral-resistor. For my benefit, can you provide its size?
9) OK!
10) Just curious. With a main transformer kVA to motor Hp ratio of 7.8, why is an AT required?

My urgent comment was to insure that the thermal capacity of the AT is not exceeded. It is far less robust than a power transformer. BTW, on the 80% tap, the motor should take 1 / 0.8^2 or about 1.5 times as long as the 100% volt case. Its operational duty should be restricted just as
you would a large motor of the size

Regards,
Phil Corso, PE
(Boca Raton, FL)
[[email protected]]

 

Further to my Thu, Jan 2, 5:59pm response:

Because the problem only appears with the 6-kV systems there are two components in question: a) motors; and b) auto-transformers. Thus, one approach would be to compare characteristics.

a) Motors The 3 and 6-kV motors should be relatively easy to compare from no-load and short-circuit factory test data. They should be nearly identical if motors are electrically equivalent.

b) Auto-transformers The 3 and 6-kV transformers can also be compared. If actual tap-voltage measurement is impractical or considered unsafe, then apply a low voltage, to one phase at a time. Apply, say 50 Vac, between supply terminal and its star-point. The 80% tap should show 0.8 x 50 or = 40 Volts. Measure input volts, amps, and watts. Compare the phases. Motors don't require disconnection because the starting controller should already isolate each of the 3 termnals per phase.

BTW, did I miss the fact that the star-point is closed with a 2-pole contactor, not 2 vacuum bottles? Please advise.

Regards,
Phil Corso, PE
(Boca Raton, FL)
[[email protected]]

 

Phil, Thanks for your mails. The auto transformer manufacturer has confirmed that I can go even as much as 30secs without damaging anything. However, we want to try with app 10 secs and see whether motor picksup. If it does not, we shall trip the motor and not allow it to switch to 100 %.

Motor drives on the drive end a screw compressor ( app 200 hp) and on the NDE side through a three output gearbox set of pumps. All the pumps are unloaded. However, compressor needs about 40% of 200 hp , even when when we unload and start. We unload by closing the inlet. We know very well that this system will never start with less than 70% voltage applied. This is one reason we do not go for star delta starter.

All motor parameters are available and I need a day more to put them together.

We do not have a right instrument to record runup speed.

We use a 3 pole vacuum contactor as star contactor for the AT.

The only extra information I got today after discussing with AT manufacturer is as follows: They supplied 3.3kV and 6.6kV auto transformers.

R at 60 deg C for each Limb 0.9214 ohms for 3.3kV 3.9933 ohms for 6.6 kV X value for each limb 0.4 ohms for 3.3 kV 1.580 ohms for 6.6. kV Z value from above for each limb 1.00447895 ohms 3.3kV 4.294513347 ohms for 6.6kV

I am unable to understand, why the values are 4 times.

As you have pointed out, the torque available is definitely marginal at 80% tap. Hence we are taking further trials unloading the compressor, which we did not consistantly do earlier. As I informed earlier, There will be some time lapse from field to get data about what is happening.

Phil, Thanks for your input. Wish you a very happy and great new year.
Regards
__________________
K Gurumurthy Manager, Special Projects

Phone: 91 80 3722570
Fax: 91 80 8394651
Mail: [email protected]

 

Responding to K. Gurumurthy's Fri, Jan 3, 11:53am question about AT vendor's parameters:

The resistance follows the square law. If kVA is constant and voltage is doubled, then resistance/impedance is quadrupled!

Thanks for clarifying my understanding of star-point contactor.

Regards,
Phil Corso, PE
(Boca Raton, FL)
[[email protected]]