Disastrous trafo failure in Russia 15

[edison123]

BBC reports

At least 54 people are missing after an explosion at Russia's largest hydro-electric power station killed eight workers, investigators say.

The accident at the Sayano-Shushenskaya power station in Siberia happened when an oil-filled transformer exploded in a turbine hall, they added.

This damaged the wall and ceiling of the turbine hall which then flooded.

http://news.bbc.co.uk/2/low/europe/8204860.stm

Aren't oil filled trafos banned in the generator/turbine halls to eliminate exactly this type of disaster ?

 

[rasevskii]

Having read Waross post on the Penetron work, and looked at their site, It would seem that this method only prevents further deterioration of the existing rebar.

We are not civil engineers but the drilling of holes in the concrete to inject the preserving material would in fact weaken the construction initially. Was this a factor?

The two machines that were wrecked were according to the photos not adjacent to each other. There was a unit in between that seemed to remain intact. Was it not running or unloaded only? In this case the load rejections that wrecked the two units involved two transformer banks not one. As wolf39 points out two adjacent? units are on one bank.

If the water flooded out units are in fact recommissioned it may be only at reduced permitted maximum output. However, these being Francis units the minimum permitted continuous load has to respect the cavitation area which may possibly be at, shall we say, less than 70 percent load.
Unless modifications are made. But that is the business of the turbine supplier.

Only some thoughts.

rasevskii

 

[rasevskii]

Further to this item, note that there are a series of excellent detailed photographs of the damage posted under theoildrum.com, comments, username ++syndroma++ dated on yesterdays Drumbeat, 21 August. Follow the link there to the Russian site which has the photos.

It would appear that the stators of the wrecked units have just disappeared/exploded into small bits/as if electrical forces were involved...The only stator section seen intact is in fact from the auxiliary generator which supplied the thyristor excitation system. Seen in one photo.

Forum members are invited to add their expertise here...

regards, Rasevskii

 

[wolf39]

It would take too much time to comment on all the drumbeat pictures. It is interesting, however, that the status quo of every single unit is documented.

The entire unit 2 is completely wrecked and this unit seems to be the one where everything started. The turbine runner is visible in the generator pit, amoungst all the other rubble. Generator units 7 and 9 are also complete losses, both stators having disappeared entirely. These units seem to have sped up uncontrolled but with the turbines left hydraulically intact. This can be seen from the video of 1:08 duration, where the water column is visible in the unit 2 region only. The roof is gone at this time already but in the region of unit 7 there is some heavy flashing. These flares shine through the front window and may have their origin in generator 7 which propably was still running in an excited mode, its stator winding then failing. All this despite the fact that the machine floor was flooded already. There is room for some speculation here. A failure of the main transformer bank belonging to unit 7 propably can be ruled out as this bank is placed behind the massive back wall of the powerhouse.

Unit 6 was under repair and not running. It therefore may have been flooded only and is very likely the first one to be operational again. However, the root cause of this accident has to be fully understood before re-commissioning can be initiated.

A video of 0:37 duration is showing unit 9 in the backgroud with a cloud of water spray visible inside the generator pit. It may well be that this unit was still in motion this moment, whirling the water around which was trapped in the generator pit. It must have taken quite a time to have all 10 stoplogs installed.

Is there anybody around who can translate us the Russian comments audible in all these videos?

Regards

Wolf

http://WWW.HYDROPOWER-CONSULT.COM

 

[itsmoked]

Next time guys - POST A LINK!

http://forums.drom.ru/hakasiya/t1151239745-p169.html

(scroll down)

Keith Cress
kcress -

http://www.flaminsystems.com

 

[edison123]

Thanks Keith for that link. May be some one can answer this.

1. Powerhouse seems to be curved. Is it because of the dam wall ? Wouldn't it make it tougher for a crane to travel in an arc ?

or is it an optical illusion ?

2. I don't see any crane rail on the tailrace side. ???

 

[waross]

The track on the tailrace side is probably along the walkway. I believe that there are one or two rails on the back edge of the walkway. The crane bridge will have gantry legs on the tailrace side. With the proper drive ratios side to side, it is probably no harder to run a curved track than it is to keep a crane straight on a straight track.
By the way, edison123, thank you for making the original post.
Thanks also to Keith for the link.

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

 

[wolf39]

Videos are placed in You Tube. Search for "Sayano Shushenskaya".

Regards

Wolf

http://WWW.HYDROPOWER-CONSULT.COM

 

[wolf39]

Sorry folks:

Go to

http://youtube.com

Regards

Wolf

http://WWW.HYDROPOWER-CONSULT.COM

 

[itsmoked]

http://www.youtube.com/watch?v=3cHMS_7oqvI

You can hear a big bang around 22 seconds.

http://www.youtube.com/watch?v=vPSwegB5oHM&feature=PlayList&p=BBC9DDB8D1BCF99F&index=0&playnext=1

This will step thru about 6 vids.

http://www.youtube.com/watch?v=PAR5ORBG8pM

Four years to repair.

Price of electricity has risen 5x.
Price to repair, "over a billion dollars".

Keith Cress
kcress -

http://www.flaminsystems.com

 

[wolf39]

The "Blog post" issued by Syndroma on August 21, 2009 at 1:03 pm (see for rasevskii's post dated August 22, 2009, 07:33) show some penstock pictures. These penstocks consist of large diameter steel pipes, encased by concrete. The surface quality of this concrete looks poor and one can only hope that the steel pipes under the concrete look better.

The power plant staff has undoubtedly done an heroic job by sealing the 10 penstocks inmidst of all the chaos. God bless them.

Regards

Wolf

http://WWW.HYDROPOWER-CONSULT.COM

 

[lz5pl]

There is nothing interesting in audio of these videos. In that one where they show the HPP during the disaster from the roadside and somebody is running towards the station they say on the back "Look, look, look ! ... May be it is a high time to run away from here... That is the end!" Normal comments of people who are under stress and probabbly still cannot realize the extents of the catastrophe.
Much more interesting are comments in the forum, but they are too long to translate


------------------------
It may be like this in theory and practice, but in real life it is completely different.
The favourite sentence of my army sergeant

 

[rasevskii]

Unfortunately Youtube is blocked at my location here.

A possible scenario/this is speculation/would be as follows-

1. A fault on the 500KV results in a load rejection of all running units, likely nine.

2. Unit 2 is destroyed due to failure of the spiral casing.

3. The resulting flood wipes out all the AC and DC supplies,
that is, vital cables are ripped up. It seems unit 2 is located near the center of the station, oddly.

4. The turbine governors on the other units fail to close the guide vanes or transfer to manual control, as all electrical supply is lost. Assumption is that the units have electronic speed pickups /several/ but no power supply is available. Or possibly were om manual before. See item on load limit settings.

5. The remaining units go to runaway speed and stay there until something fails.

6.The operators manage to close the intake gates/stoplogs/ manually inside the dam after some time eventually stopping the runaway units. The ones not destroyed that is.

7. No electrical protection trips operate/no DC supplies.

8. The 500 KV switchyard /some distance away/ would have its own DC system and protections, these may have operated
partially but some may have been interlocked with circuits in the station whose cables were destroyed.

In the good old days the turbine governor was independent of all outside supply having its own PMG on the turbine shaft. Possibly not the case here. There was likely also no mechanical-hydraulic overspeed tripping device on the turbine shaft, independent of all electrics. A usual feature of old stations before anybody trusted electricity..

regards, rasevskii

 

[wayne440]

Babelfish works effectivly on the forum pages "Itsmoked" pointed us to. The results are not perfect by any means, but the basic statements are clear enough. There are a couple of plant employees posting, and dicussion of a governor retrofit, the first starting of a recently repaired unit on that day, etc. Early on, the dialog indicated local news media as saying "hydroshock" was the cause, perhaps by abrupt opening of flow controls somehow.

 

[rasevskii]

Some new information can be found on page 170 of the Russian blog, item 3389. Translate this by babelfish.

I would not post this here.

It seems a new governor was being tested and it all went wrong on unit 2. Runaway, then an out of sync closure onto the grid. The rest is history...

Members are invited to read this for themselves.

Scroll down for a model view picture of the units also.

regards, rasevskii

 

[ciy]

I have some questions as below,

1. Is there anyone has the single line dia.(w/relay) of this plant?
Please upload for reference.

2. How to prevent the disaster like this from the viewpoint of protection and control?

 

[slavag]

Hi.
Please see attached, it's Russian, but photo isn't need comments. History of station and views.

http://4044415.livejournal.com/51285.html

For my pinion we need wait for the some real information.
On this moment, talk about problems in the unit 2 with control. Operation personal start with stop of turbine, but regular stop, not emergency stop and this step in some not stability zone initilizaied crash of system.

 

[slavag]

Please see attached SLD ( sorry for the qulity, it's from some old book).

Now, we have new version, problem with tubine constraction and long time w/o maintanace.

Best Regards.
Slava

 

[wolf39]

I'm at present studying the internet discussion going on in Russia. This is a difficult task as the babelfish translation into English is far from perfect. Still, it is obvious that quite a few of the comments contradict each other and that the posts are all highly speculative. Fact seems to be, however, that unit 2 was decommissioned mid January 2009 for welding repair on the turbine runner and that the turbine governor was replaced. Also, there was an increase of vibration prior to the accident.

Until an official accident report is published we all have to stick to the videos and photographs available. The video of 1:08 duration has been taken shortly after the accident occured. It shows a black cloud of smoke in the vicinity of the collapsed roof behind the machine hall (this is better visible in another video), this propably being an indication of a transformer fault. Then there is some flashing going on behind the machine hall window in the vicinity of unit 7 or 9, lasting for about 6 seconds. About 10 seconds later an air blast breaker bang is audible, suggesting that some power may still have been available in the powerhouse at this time.

At first I was most surprised that the generator stators of units 2, 7 and 9 were gone completely. They seem to have been pulverized. To me there is an easy explanation for this. The generator was running in a pit filled with water. The generator rotor with a circumferrential speed of 80 m/s or more was acting like the impeller of a centrifugal pump. The hydraulic pressure entered the stator core cooling ducts and the lifting force ripped off the nuts of the core pressure bolts. The stator core must have delaminated in seconds and the stator frame then disintegrated and was washed into the turbine pit below or was disposed off somewhere else. In some of the pictures taken early, before the cleaning-up activities started, some stator core laminations were lying on the machine hall floor but most of these may also have been washed into the pit.

Some generator units look quite intact with the top covers still in place. Maybe these covers were tight enough to prevent pit flooding and/or unit speed was low enough already when flooding occured.

Regards

Wolf

http://WWW.HYDROPOWER-CONSULT.COM

 

[rasevskii]

Agree with Wolf 39.

I am also babelfishing the Russian blog, and on page 262 are interesting diagrams of the governor system as it was newly upgraded.

The turbine guide vanes/aka wicket gates/are positioned by individual oil servomotors which are evidently controlled by a digital system. A discussion about Omron follows. It would appear that each servomotor gets a position command signal and follows that by some kind of comparator mechanical linkage, with electrical feedback.

The turbine pit is not the place for electronics.

In the classic water turbine all the guide vanes are positioned by a regulating ring which is rotated over a small angle by two large servomotors/cylinders if you like/each guide vane is connected by link to this ring. For emergency shutdown the main distributing valve which controls the main servomotors, is forced to the -close- postion by a separate device independent of the governor. This can be completely mechanical-hydraulic, and can operate even if all the electrics are disabled.

We are talking of an oil supply available, usually 40 bar, for this the large pressure tanks which are seen on the photos, at generator level, are provided.

In the case of the Russian disaster, possibly the governor system on unit 2 went awry, and the guide vane positions became unstable, resulting in loss of control and the oft reported vibrations.

We can say that at the time of the flood, all the control, governor, and protection cubicles were carried away which meant that all the other operating units had no governors anymore and went to runaway.

In other words the guide vanes did not close, but remained in an uncertain partly open position.

The assumption is that all the other units were upgraded with the new controls.

The flood reached far out into the yard next to the control building, it must have been meters deep at generator level. One can see many overturned and swept away cubicles.

That is a possible speculative scenario.

regards, rasevskii

 

[edison123]

yup. The next shiny thing in the 'puter controlling these big behemoths is "The turbine guide vanes/aka wicket gates/are positioned by individual oil servomotors which are evidently controlled by a digital system". This was to "improve" the efficiency of the turbine or something like that. And I said nuts to that when I heard it first time around. I asked exactly the same question, what if the damn servo motor fails ?

I am finding more & more younger crop sitting in front of ther laptops controlling these machines without even understanding how they work.

/rant over.