15MW Steam Turbine Generator coupling Breakdown 3

[gokulkrish2]

Hello Everyone,

Hope everybody is doing great in this forum.

I am right now working on a commissiong of a 15MW steam turbine generator and have some questions for you guys to answer

I represent the customer who bought the generator unit.

While we were commissioning this generator two weeks before, we had an incident where the coupling broke due to wrong synchronization(or atleast we think-investigation still going on). But the project is proceeding and we have shipped the generator to the factory to check its integrity

I am schedule to go to the factory to inspect and monitor the testing process and report back to the customer.

Do you guys have some checklist where i should be focusing and looking for damages in the generator. This is the first time i am going to the generator factory.

BACKGROUND.

The generator was almost commissioned and was synchronized to the utility.

The night before the incident the generator was set to output 10MW and it ran for the whole night

The next morning we were going to check an EStop circuit and Auto synchronization function

So, we had to trip the unit and so we gradually reduced the output of the generator

We brought it to 500KW and ready to trip

But the Hyndai representative wanted to do some final point to point checks on the wiring diagrams

We they were doing this we heard sudden noise and the next thing we know, the coupling at the Turbine-Gear box end was broke.

The emergency lube oil circuit was broke and oil was splattered all over the place

But the Turbine and the generator rotate atleast for 3 minutes after the coupling broke. The vibration monitors did go to the alarm stage but never to a trip level state.

When we questioned the Turbine and Generator representative, they would not give us any answer as they claim the investigation is going on and they wont be able to issue a statement until their underwriters are given one

We questioned the Electrical Technician who helped the Hyundai Generator Engineer and he claims that the Hyndai guy had him bypass the sync check relay.

Now i have two questions:

What do you guys think would have happened here.

What would be the worst case damage to expect. When i got the inpsect the factory testing what should i focus on? What could have gone wrong as far as the generator is concerned- Shaft? End winding? rotor? Stator?

I would appreciate any insights and advice i can get from experts like you people.

Thanks

gokul

 

[ScottyUK]

Well, the fact that you were exporting 500kW indicates that the generator main breaker was closed. The fact that the coupling broke indicates that some event took place which generated a massive transient torque. I'd suggest that the generator breaker may have opened and then reclosed fairly quickly afterwards. If the check-sync relay was bypassed as you suggest and a close demand from the generator control system was present then the breaker may have reclosed almost immediately. Even if the check-sync relay was bypassed the automatic synchroniser should have blocked the reclosure unless it too was bypassed.

What event history do you have from the turbine controller and the protection? A well-configured Historian on the turbine and generator control system is the best window you can have into what took place - I hope the OEM took the time to configure it, otherwise the investigation could well descend into speculation and finger-pointing.

Worst case damage? Broken coupling or bent shaft, possibly damage to the long LP turbine blading, perhaps damage to the stationary vanes if they have been hit by the blades. Perhaps bearing damage - a wipe is certainly possible as the rotor was thrown about. The vibration data is a positive sign that you may have been fairly lucky, in as far as broken coupling is 'lucky'.

Look for endwinding movement and loose stator blocks, on the rotor look for copper movement under the endbell and signs of a rub anywhere on the bearings or fans. It would be useful if you could perform an RSO on the rotor to identify any anomalies.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[gokulkrish2]

Thanks Scotty for your timely reply..

Forgive my ignorance but i do not know what is RSO means. If this is a test done on rotors then, i should inform you that this generator in discussion is a Permanent magnet Brushless generator. So can a RSO be done on this type of rotor too?

I too came to the exact conclusion on what may have happened. But regarding the auto synchronizer the synch selector switch was in the manual mode.

I have the event history from the SEL 300G relay which i downloaded and have attached. But i do not have anything from the Turbine PLC. They barricaded the whole area and the whole panel is switched off now.

I am also counting on being lucky.. but for record, the turbine rotor has been sent to the factory for inspection and repair.

I am schedule to go to the generator factory next week monday when they start tearing apart the generator. Thanks for you suggestion and locations where i need to focus. Is there any test that i should require those guys to perform on our generator that would prove the integrity of my generator?

Thanks again.

gokul

 

[rasevskii]

I second Scotty UK. The brkr must have tripped and reclosed out of sync.

Sorry to say this, but it is absolutely irresponsible to allow anyone to be doing any wiring or "point to point checks" when the unit is running online or even if it running at all. By point to point checks what is meant? Was someone using a megger or bell tester or putting grounds or voltage here or there in the control wiring? Unbelievable. All panels and related equipment must be locked up or secured so that no one, NO ONE, goes in and does anything of this sort. Wiremen and cable terminators have to be ordered out of the operating parts of the plant.

It was bad management of the commissioning that was at fault. Not a uncommon problem on sites where many different contractors and suppliers are present.

Introduce proper lockout/tagout procedures. If there is no cooperation or respect for the procedures, the job has to be stopped.

Otherwise someone may get hurt next time.

rasevskii

 

[rasevskii]

BTW, you should open the gearbox and inpect the gearing (if not already done so)several teeth may be broken or damaged. I heard of a GT where the gearbox actually exploded/disintegrated due a bad sync.

rasevskii

 

[ScottyUK]

RSO - recurrent surge oscillograph. yes, it can be done on a brushless machine if you disconnect the rectifier.

rasevskii's comments about the poor control of the area are very fair. If the OEM's engineer requested that the check-sync relay be bypassed then he will have some explaining to do. I expect he is denying that he said such a thing, but it would be interesting to know what he was trying to do.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[rhatcher]

I think that you had a major fault that caused the breaker to trip and that the subsequent breaker reclose without sync broke the coupling.

I am not familiar with the format of the data file that you attached to your 2 Jun 11 15:33 post but, I am good with numbers and this is what I see:

"EVENT","TARGETS"
"TRIP","TRIP 27/59"

27/59 under/overvoltage relays are general purpose single phase relays designed to be used wherever protection is required against abnormal voltage conditions.

http://www.abb.com/product/db0003db004281/c12573e700330419c1256eac00501b42.aspx

"VAB(kV)","VBC(kV)","VCA(kV)","VN(kV
-8.1,-5.7,13.8,0.1

The voltage is shown to be in kV. VAB and VBC show no voltage. This would indicate a fault in the 'B' phase.

"IA","IB","IC","IN","IG"
2,-4,-5,4,-7

I am thinking that the current readings are secondary current from a CT such that 5 amps equals the CT rating. I do not know the convention that is being used. I assume that a positive current is power out of the generator and a negative current is power into the generator.

Regardless of the current/power flow convention, I see that there is a significant ground current that is 140% of CT rating. There is also power (current) going both ways (in and out of the generator) in the phases and significant currrent flow in the neutral.

This is a total guess because I only have the information that is provided in the thread but, I think that a generator stator ground fault is indicated. Other possibilities include a ground fault failure in the leads between the stator and the circuit breaker.

the Hyndai guy had him bypass the sync check relay.

This is the cause of the broken coupling and any subsequent damage that resulted from the transient torque caused by reclosing after trip without synchronization.

the investigation is going on and they wont be able to issue a statement until their underwriters are given one

The good news is that you have this data file that recorded the event. A person who is familiar with this format should be able to tell EXACTLY what happened. I would recommend that your company hire an Expert and a Lawyer to begin an investigation on your own.

This is because I am assuming that there are major costs associated with this event and, in the end, somebody has to pay...

 

[gokulkrish2]

rasevskii,

While i agree that we should have not allowed nobody to do any check while the machine was running.... we had absolutely no idea the sync check relay was bypassed. I am not trying to defend myself here but just trying to provide more information. He was supposed to be the expert and in many cases we are compelled to go by what he requires.

Yes, as you suggested, the gearbox is already on its way to the manufacturer for inspection and repair.

We are also getting statement from whoever was there on sight and getting our lawyers involved in this matter.

Thanks for your comments. It never offended me but proved how safe you are. We were taking all kinds for safety measure when we first synchronized this machine... Thanks again

Scotty,

Thanks for the information. I will request the generator guys to perform the RSO on the rotor. Thanks again for the valuable information.

 

[davidbeach]

Analogs in the event report are instantaneous primary quantities, amps and kV. No significant ground currents anywhere in the event.

At the beginning of the event, the breaker is open and there is no current.

At relay time 10:46:12.250 in the event everything changes with the breaker closing. Way out of phase closing.

At relay time 10:46:12.246 VAB has a positive maximum. At relay time 10:46:12.251 VAB is at -6.5kV, about 108 degrees later. The next data point, relay time 10:46:12.252 (16 samples per each 60Hz cycle) VAB is 1kV rather than the expected (from prior cycle) -10.8kV. The first VAB negative maximum happens at relay time 10:46:12.257 (0.6875 cycles after the positive maximum). That extra 0.1875 cycles equates to 67.5 degrees.

The voltage collapses sufficiently that the instantaneous undervoltage element (supervised by 52a) trips the breaker as soon as the 52a contacts indicate closed.

The first positive peak of C phase current reaches 10082. This is a filtered 60Hz fundamental value and is the highest current seen. The actual peak would require a raw event have been downloaded from the relay.

Ugly, ugly, ugly. Shocked the he!! out of the machine.

 

[rhatcher]

There you go! Someone has interpreted the data file. You have to like this forum. LPS for davidbeach. I would think that he has shown that there is a good case to prove that the the vendor is at fault.

However, I would like some help in understanding why there seems to be a ground fault current recorded in the data that I quoted above. I am not questioning davidbeach by any means since he obviously knows what he is looking at.

Am I reading the data file wrong? I would think that the ground fault current indicates a stator failure.

DavidBeach, what am I missing??? Can you give a simple explanation so that I can understand what you are seeing?

 

[davidbeach]

The currents in the event file are primary amps. Anything small can be considered zero. The IG value is a computed residual current. The phase currents are too small to make a meaningful angle determination, so a vector sum is worthless. Each line in the event report is 1/16th of a cycle. Nothing of interest until well into cycle three. The trigger point for the event is four cycles into the event, things start getting interesting during cycle 3. Best is to view it using SEL's 5601 software. But you can plot the analog values in excel.

 

[gokulkrish2]

David,

Excellent Explanation. Thanks.

I would also appreciate if you have any insights or pointers to look for and catch on the repair on the generator now. As you can see Scotty have given me good information but anything you may want to add might help as well.

Thanks again.

 

[dpc]

Regarding repair of the generator - you might post a specific question on the Motors and Generators forum. You might get some additional responses, although Scotty's advice is on the money.

You had a bad day, that's for sure. Good luck.

David Castor

http://www.cvoes.com

 

[dpc]

Duh! I thought this was the general EE forum - Ignore my last message. In the words of Emily Littela -- "Never mind". I'm glad it's Friday.

Dave

David Castor

http://www.cvoes.com

 

[davidbeach]

I'm not really a machine person, so I will defer to others about what to look for in the machine. Personally, I'd be pushing for an all new machine and payment for the delay.

 

[ScottyUK]

I agree with David, although it will depend on how well written the contracts are. The client would appear to be a good position to re-negotiate terms under which he will accept the machine. As a starting point I'd be looking for significant reduction of capital cost plus a substantial warranty extension against latent defects resulting from this event.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[gokulkrish2]

yes,

It was a very bad day..And as our legal guys are saying its going to get ugly as well.

Our folks are working on points to sue for compensation for downtime. Warranty wise it is for me to make sure that my client get the machine in a good condition he paid for.

I will post some pictures once i complete my factory inspection.

This forum has always been an excellent source of information.

Thanks for all those who contributed.

 

[macmckim]

gokulkrish2

I am curious - so need to ask if the generator was a Hyundai Korea unit or Hyundai-Idal unit. I have a local client that has a bid from Hyundai.

 

[gokulkrish2]

macmckim,

It is a Hyundai Ideal unit. In all fairness to Hyundai, They are ready to work with us. They were really fast acting and mobilized their crews once the incident happent. If they take care of this issue as it is supposed to be, then i would not have no problems in buying an another Hyundai ideal unit.

Before this incident happen, the machine was so perfect we could hardly get any vibration and noise was next to nothing.

my 2 cents before you take any decisions.

 

[rmw]

ScottyUK gave you a pretty comprehensive laundry list of things to check. On the generator rotor, make sure they remove the retaining rings (or end bells as he refers or end caps as others do) and check for rotor tooth cracking in the ring fit area. Check the fit area on the rings as well. Don't let them just do a borescope to check the copper winding blocking - make them remove the rings.

With any luck, the coupling acted as a fuse and saved you from major turbine distress, but don't take that for granted. Check for cracking in anything that you wouldn't want to see accelerated suddenly - Scotty mentioned the LP blades - good place to start. Check the dovetail fits at the blade roots as well.

Rotor run outs all up and down its length.

Check the gear box frame. If it was cast you may be OK. If it was fabricated, it might be tweaked.

rmw