Smart questions
Smart answers
Smart people

Member Login

Remember Me
Forgot Password?
Join Us!

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips now!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!

Join Eng-Tips
*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Donate Today!

Do you enjoy these
technical forums?
Donate Today! Click Here

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.
Jobs from Indeed

Link To This Forum!

Partner Button
Add Stickiness To Your Site By Linking To This Professionally Managed Technical Forum.
Just copy and paste the
code below into your site.

aolalde (Electrical)
12 Aug 05 14:46
We have four Synchronous Generators Diesel Engine driven and a Static Reactive Power Module 20 MVAR, all synchronized to the utility power line. The four generators are identical size and manufacturer; 15 MVA, 12 MW, 400 RPM, 13.2 kV. 60 HZ, field Voltage 267 VDC, and current 231 DCA. The generators are brushless excited, with a rotating 6 diode rectifier and AC 59 kW, 26.7 HZ, 3-phase armature. The main rotor field has 18 poles one circuit series, the rotor outside diameter is 10.5 Foot.

The four generators under normal conditions provide 10.5 MW and 6 MVAR each one.( Very conservative load).
After 7 years in service, generator #1 failed. We found 3 of the 17 rotor field series jumpers fused.
Three poles were re-wounded, the jumpers were replaced and the unit reconnected in line. It last 2 days and the very same type of failure developed in the field connections again.
This time all 18 poles were rewound and all field connections replaced. The unit ran five days and developed the very same type of failure. Four jumpers making the interpole connection were totally melted and six pole windings were grounded. This time one pole shoe had five amortisseur winding bars totally fractured, each pole shoe has 12 bars. The rotor paint shows signs of overheating.
The stator winding does not shows signs of any damage and it test OK,  insulation resistance, PI and comparison surge test.The rotor failure has been similar but involving different random poles.
The electric records during the days in operation show, that this generator #1 was generating an average of 10.5 MW but only around 3.9 MVAR and for the last 1.5 hours it worked as induction generator, taking reactive power from the line bus.
1- Has somebody experienced a similar problem?
2- Why the series connections are melted but not the pole windings?
3- In theory when one jumper in the series connection melts, the field circuit is opened, what feeds the current to melt three more jumpers?
Any comments will be appreciated.
Helpful Member!  edison123 (Electrical)
14 Aug 05 10:15

Is it wound-on-edge rotor winding ?
edison123 (Electrical)
14 Aug 05 10:40
You could also refer IEEE 492-1999 (Guide for Operating and Maintenenance Hydro-Generators) clause 7.65 (Unsual Operating conditions - Loss of field excitation)for damages to salient pole windings like your DG's.
Helpful Member!  electricpete (Electrical)
14 Aug 05 12:00
I have no experience and no good idea.

If I focus on your question #2 and #3 and the fact the machine was operating vars-in before failure....  I am remembering that for our generator one of our limits from underexcited operation arises from "stator core end iron heating due to fringing axial flux"

I am not sure exactly what that means (I'm hopeing you guys are more familiar with it than me) but it might suggest in this mode (underexcited) there is abnormal flux at the ends of the machines which could perhaps result in abnormal heating of the jumpers and not the windings... and a little further out if the jumpers are large the heating effect may be eddy type heating which doesn't require a current loop outside the jumper?

Just a s.w.a.g. and probably off-base (you guys feel free to shoot it down).

Looking outside the rotor for what cause might be in common to multiple failures - you mentioned you have checked the stator - have you also checked the exciter?

Eng-tips forums: The best place on the web for engineering discussions.

electricpete (Electrical)
14 Aug 05 22:39

The above link is a photo of page 194 of "Power System Stability and Control" by Kundur 1994 which describes end region heating of syncronous machines during underexcited conditions.

Once the first jumper opens or the rotor circuit is opened for any reason, the machine would become severely underexcited and perhaps end region heating causes the remaining rotor pole jumper damage?

Eng-tips forums: The best place on the web for engineering discussions.

electricpete (Electrical)
15 Aug 05 7:38
"The electric records during the days in operation show, that this generator #1 was generating an average of 10.5 MW but only around 3.9 MVAR and for the last 1.5 hours it worked as induction generator, taking reactive power from the line bus."

I assume this was because of loss of the field. Shouldn't there be some kind of trip upon loss of field?

Eng-tips forums: The best place on the web for engineering discussions.

aolalde (Electrical)
15 Aug 05 14:58
Thank you Guys for comments.

Edison123: I do not know if these could be considered wound on edge field coils, I guess it is. The whole pole is bolted to the rotor rim and the coil is removed from the pole body. The coils are wound on molds; it has a single layer with 86 turns of flat copper 1.771in (45 mm) x 0.078 in (2 mm). The turns are flat, with the flat face (45 mm) perpendicular to the pole body. After insulated and cured it is assembled and wedged and the pole bolted to the rim.

Electricpete: I have the same opinion, a loss of field relay should protect the generators, but actually we do not have any similar protection.
byrdj (Mechanical)
15 Aug 05 19:02
I'm mainly mechinical but I've seen where operating a generator with no excitation caused what was explained to me as surface currents.  since this rotor was not designed for surface current, there was severe arcing acrross the rotor slots that held the copper coils.  Thus I would be on the same side as electricpete's swag of reduced excitation(untill a genrator specilist explains different)
electricpete (Electrical)
15 Aug 05 19:54
That's different than what I was saying but some of it makes sense to me (more sense than what I said before).

The rotor is designed for syncronous operation with no relative motion between rotor and the field so not much driving force to create eddy currents.

When acting as an induction generator there is slip between rotor and the field which can create eddy currents.  Can it operate there in steady state without overheating?

But why would it tend to heat the copper and not the steel. If I remember right, steel melts around 2000F and copper doens't melt until 3000F.  So if this were true there would have to be some reason the copper has much more heating.  Copper has lower resistance which favors more heating. Large pieces close to airgap might be more subject to heating than small pieces further away but I have a feeling there are big chunks of steel on the rotor in addition to laminations. Why wouldn't they melt as well?

Eng-tips forums: The best place on the web for engineering discussions.

electricpete (Electrical)
15 Aug 05 19:57
One last thought - focusing on the question how could multiple copper jumpers melt if all in series it seems like there are only two possible scenario's: 1 - very high current for short period of time or 2 - some kind of eddy heating like we talked about above.

Eng-tips forums: The best place on the web for engineering discussions.

aolalde (Electrical)
15 Aug 05 23:17
Electricpete; I appreciate your interest in helping.

If you can read what Edison123 suggested”IEEE 492-1999  Operating conditions - Loss of field excitation” , you will find interesting concepts. BTW copper melts at 1083°C (1981.4 F) and steel at 1480°C (2696F)
petronila (Electrical)
16 Aug 05 14:49
hi aolalde,

I was talking with other engineer about your problem,If the rotating fields are connected straight series (no parallel paths), it could be the discharge resistor is open, then the rotor windings are being spiked with the very high generated voltages during starting. ¿ how that could damage the jumpers so though? Usually, the high voltage causes turn-to-turn shorts,or even ground failure in the rotating poles.
The exciter has been rewound too? If so, the connection may be incorrect, causing a change in the output frequency. equency is much higher than 60 Hz.


aolalde (Electrical)
16 Aug 05 16:02
Hi Petronila thanks for your comments.
This application is a generator, the diesel motor accelerates to full speed, the field is applied and when the voltage window is close to match the line phases it is synchronized. There should be a transient while synchronizing but the other 3 generators are working under the very same conditions.
The exciter has not been repaired; the six rectifying diodes test OK. I am recommending replacement of the exciter and diodes since from the beginning this generator produced less KVAR than the other 3 (3.9 MVAR as compared to 6 MVAR on those working OK).  
edison123 (Electrical)
16 Aug 05 18:29

This gen has brushless excitation. So no field discharge resistor.
petronila (Electrical)
18 Aug 05 16:41
Hello aolalde,

When One Jumper is Open then no current flow by the others,should not be damaged.The melted jumpers could be damaged by Induced voltages during Starting. To avoid this The Slip Rings would have a discharge resistor in order to bleed off the excess induced voltage during starting, In a Self- Excited unit there is normally potted assembly on the end of the rotor shaft or near the Diodes, This is called for manufacturers a "Synch-Pack" or similar name, for this will be good a carefully looking at the circuit ROTOR-DIODES for some evidence to indicate this.Another important think is the damaged amortisseur, that indicates the unit was motorized or put on line Out of Synch. If you can find the "Synch-Pack" and it is damaged this could be the failure.


Skogsgurra (Electrical)
18 Aug 05 17:22

There is usually a shorting thyristor parallel to the diode pack. It is controlled by the induced voltage in the rotor winding and as long as that voltage is above a set limit it will short out the diodes so that starting is essentially with a short-circuited rotor. A resistor would not be sufficient during start. And if you make it enough low resistance to really help, you will have a lot of losses in it during operation - if not disconnected by the thyristor. Most (all) generators have no resistors in the rotor circuit.

Gunnar Englund

ScottyUK (Electrical)
19 Aug 05 3:00
Hi skogs,

On big machines it is common to find a field discharge or dump resistor which is connected across the field sliprings as the main field contactor opens. it has to be a make-before-break action, otherwise the arcing makes for a hell of a firework show (and lots of repair work).


Does the rotor forging show severe burning around the areas where the ground faults occurred? I'm trying to picture the likely path of the current which melted the jumpers. I will continue to ponder this.


One day my ship will come in.
But with my luck, I'll be at the airport!

electricpete (Electrical)
19 Aug 05 9:53
Scotty - no slip rings in a brushless machine.

Eng-tips forums: The best place on the web for engineering discussions.

ScottyUK (Electrical)
19 Aug 05 12:51
Sometimes I only open my mouth to swap feet...


One day my ship will come in.
But with my luck, I'll be at the airport!

djs (Electrical)
19 Aug 05 12:53
The only time I have seen the field coils of a generator electrically damaged is when the unit was subjected to large amount of reactive VARS. Also when units are parrelled, I believe that the manner to share VARS between them is by adjusting the field excitation between the various units. KW is controlled by the throttle setting of the prime mover.
aolalde (Electrical)
19 Aug 05 13:10

One corner of the laminated pole face, close to the rotor rim shows signs of flashover. That flashover was between the last field turn and the pole side lamination very close to the rim and it  jumped over insulation 2.5 mm thick ( 0.098”). The jumpers did not leave a clear path of flashover but it seems to be against the rotor rim body.
ScottyUK (Electrical)
19 Aug 05 13:27
Thinking aloud here -

If the rotor is reasonably clean, then the flashover required some significant voltage to initiate it. That would require two conditions: an existing fault to ground on the rotor, then either a second ground fault at the location of the burn or opening of the the rotor circuit to initiate the arc.

It seems we're discussing flashover damage from a relatively brief arcing fault rather than more severe damage which would probably have occured if the rotor formed part of the path which carried current for long enough to melt the jumpers.

From the damage to the amortisseur winding, is it possible that this machine has pole-slipped at some point? It seems unlikely that a diesel prime mover would withstand that kind of shock without some signs of damage. Had the broken bars actually broken, or had the end brazes suffered thermal failure?

I'm a little out of my depth with slow-speed salient pole machines because I'm used to looking at high speed turbogenerators which are quite different animals. If I'm adding confusion I apologise.


One day my ship will come in.
But with my luck, I'll be at the airport!

aolalde (Electrical)
19 Aug 05 13:42
Two poles had fractured damping bars. The brazing joint seems in good condition. The fracture occurred right at the edge of the inner face of the short circuit ring ( for this case the ring is made of individual copper plates interconnected with bolted copper jumpers.
petronila (Electrical)
19 Aug 05 17:49
Hello aolalde,

Did you take some pictures? can you send ?


petronila (Electrical)
20 Aug 05 10:11
Hello aolalde,

Taking acount your last comments That seems to indicate arcing of the high voltages induced in the rotor
poles during starting. In units With self-excitation, there should be something to do the  the discharge.( a thyristor on the rotating rectifier).In a conventional rotating  6 diodes Rectifier(3+ and 3-) mounted on two separate plates (1 + and 1 -), there should be  a thyristor across the 2 plates, to act as the discharge protection for the rotating fields.Without such protection, at the moment of starting, the motor behaves as a transformer with the  stator winding as the primary, and the rotating poles as the secondary. With all poles connected in series, the turn ratio is too high. There could be several thousand volts built up in the rotor during starting. It discharges across to ground wherever the windings are close to ground potential.Perhaps there is something missing, or a failed component, on the rotating rectifier package.

Best Regards,

edison123 (Electrical)
20 Aug 05 11:03
Aren't we talking about diesel generators ?
ScottyUK (Electrical)
21 Aug 05 4:45

The rotating rectifiers on the brushless excitation system fitted to large turbo alternators have a standard six-pulse configuration, sometimes with two or more parallel diodes per limb in the bridge, but no thyristors and no clever tricks. I suspect that those fitted to a low-speed machine are not so far different.




Is it possible that the flashover to the core at the jumper was a result of the broken jumper moving due to rotational forces? I ask this because we may be trying to incorporate it in the root cause explanation when it is actually a residual effect of the problem which caused the jumper to fail. Sometimes the residual effects add a lot of confusion when assessing the underlying cause.

The ground faults on six rotor bars: where did they occur, and what was the failure mechanism of the insulation? Were the fault locations burned from carrying large current or were the faults so solid that they would have had significant current-carrying capacity? Instinct is telling me that these ground faults were the underlying cause and that most of the other damage is more-or-less traceable through progressive failure of the rotor.


One day my ship will come in.
But with my luck, I'll be at the airport!

edison123 (Electrical)
21 Aug 05 9:11

I posed the last question since petronila seems to have misread the equipment as motor (and I don't agree with his comments about starting as motor and high turns ratio affecting the field winding, since I have seen many synchronous motors started as induction motors without any problems).

As for discharge protection, I have seen them in small, LV generators and definitely none in large beasts connected to the grid.

Now coming to aolalde's poser, for me it seems to be a classic case of both asynchronous operation and loss of field operation along with the root cause of thin (2 mm thick) pole jumpers failing due to centrifugal forces. The symptoms he tells are exactly the same covered in the IEEE 492 Std (of which I have an official copy with its usual caveats about not reproducing its contents without their 'permission').
electricpete (Electrical)
21 Aug 05 11:26
I see where you mentioned IEEE492 (1999). I looked at paragraph 7.6.5 regarding loss of field.

I see where they talk about rotor overheating. The cause is not exactly clear to me.

Machine speed increases and acts like induction generator.  I guess I can see that results in some increase in centrifugal force but I have a hard time understanding the role of centrifugal force in this scenario.

Eng-tips forums: The best place on the web for engineering discussions.

aolalde (Electrical)
22 Aug 05 8:56
Thanks to everybody for interest and comments.
I am almost convinced that high voltage was induced in the pole field windings, but when? The ease to explain condition is when the generator worked as induction generator for more than one hour. I am convinced that most of the damage to the amortisseur winding and to the field poles was done during that time.
The speed has to be increased but we do not have a record of the over-speed, we know that 10.5 MW were generated under asynchronous condition but the slip could be 3%, 5%, 10%? I do not know.
Rotor centrifugal forces at 13,194 Ft/min normal synchronous 400 rpm are not small. When the generator worked over-speeding it could bend those jumpers.   
My problem is to find the trigger that initiated the failure.
motorspert (Electrical)
22 Aug 05 12:43
You mention in your original post that the generators are diesel driven. I assume they are directly coupled (no flexible coupling). A diesel imparts considerable pulsating torques on the shaft system and the inertia of the generator rotor is used to damp some of this pulsation out.

An effect of the pulsations is to shake the poles which could cause the connectors between the poles to suffer from stress fractures - copper "work hardens" very quickly. Are the connectors the same as the original ones? I would check the size, material and shape - some connectors are design to be very flexible to cope with these mechanical and thermal stresses. I assume the diesl has been checked over for misfiring etc?  

This phenomena is not unusual and would explain one failure, not sure how to explain the other two simultanopus failures. Did all of the non-failed connectors look ok?

The brushless exciter shouldnt have caused this - although the fact that this machine failed to deliver the full VARS suggest another problem with the excitation system.

Hope tjis helps

ScottyUK (Electrical)
22 Aug 05 13:59
As a possible source for the high voltage, what if a field conductor suffered a sudden open circuit failure? The stored magnetic energy in the rotor iron has to go somewhere and according to Lenz's Law it will show up as voltage.

I'm not sure if the damage attributed to HV flashover looks like a relatively short duration event, or if it has been a persistent fault. A persistent fault would rule out my theory above. Perhaps you can comment?


One day my ship will come in.
But with my luck, I'll be at the airport!

petronila (Electrical)
22 Aug 05 15:18
Hello electricpete,

I looked too the IEEE 492 Paragraph 7.6.5 and This  bit about operating as an induction generator is only true when the load is 'over-hauling' - meaning it can drive the rotor.For example:  A pump motor pumping against a high head, or a conveyor.


electricpete (Electrical)
22 Aug 05 15:25
petronila -  what load are you talking about. This is a generator.  Upon loss of excitation the sync generator will cease absorbing torque. If prime mover torque does not go to zero (or no-load friction/winding value), then the machine will accelerate.  Machine operating above sync speed will act like induction generator.

Eng-tips forums: The best place on the web for engineering discussions.

motorspert (Electrical)
22 Aug 05 17:26
yes, a sudden open circuit leads to a very high induced voltage in the field, the exact value depends on the turns ratio between the rotor and stator and the slip.(percentage speed the rotor is above/below synchronous speed)

But yes, induced voltages can cause flashover
edison123 (Electrical)
22 Aug 05 19:26

I agree with motorspert about torque pulsations in DG's. Usually DG's have a flywheel to smooth out torque pulsations. Does yours has one ?

I would look at some sort of anchoring the pole jumpers to the rotor rim to strengthen them.IMHO, the root cause was mechanical failure of the jumpers with collateral damages. I would also recommend loss of field protection to prevent such failures.

What is the present status of the machine ? Is it under repair / investigation ?

electricpete (Electrical)
22 Aug 05 21:00
The generator has always been generating low vars.  That is the first symtpom seen so presumably closest related to the root cause.

It seems to me the exciter system (either rotating or stationary portion) is suspect and should be very carefully checked.

Eng-tips forums: The best place on the web for engineering discussions.

electricpete (Electrical)
22 Aug 05 21:11
Is there any record of vibration spectra when the machine was running?

Eng-tips forums: The best place on the web for engineering discussions.

motorspert (Electrical)
23 Aug 05 6:35
I am not sure the low VARS is related as the generator ran for 7 years like this, I suspect the AVR has been set up that way, maybe the original commissioning records give a clue.

If all generators are identical then a look at the machine logs should give an indication, a table of exciter current, voltage, current and PF would enable the generators to be compared. There is a slight possibility that the exciter is running on a single phase.
aolalde (Electrical)
23 Aug 05 11:06
I agree with your comments Electricpete. I am considering to recommend replacement of the exciter and rectifier bridge to ensure the field is properly excited.

In addition,loss of field protection will be another recommendation.

Actually all 18 poles are in our shop for test and rewinding of defective ones.

We do not have yet a solid root cause of failure.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!

Close Box

Join Eng-Tips® Today!

Join your peers on the Internet's largest technical engineering professional community.
It's easy to join and it's free.

Here's Why Members Love Eng-Tips Forums:

Register now while it's still free!

Already a member? Close this window and log in.

Join Us             Close