Callide Power Station 6

[hacksaw]

What is the likely cause?

CS Energy, which runs Callide Power Station Queensland AU), said there was a fire in one of the plant's turbines.

"That tripped the other three units, that then was a serious reduction to generating capacity in Queensland," CS Energy chairman Jim Soorley said.

The station was evacuated and a 550-metre exclusion zone put in place as fire crews worked to extinguish the blaze.

 

[byrdj]

I haven't been able to sort through NEMlog, BUT found this in another forum post
If I read it correct, it appears unit C3 dropped load prior to C4 (might be plotting squew due to sampling rate)
the C4 reduction (and c3 increase) had only occurred an hour or so prior Looks like B was able to ride excursion for a few minuts

 

[MDEAus]

I think you are right about the sampling rate squewing the plot, although it would still be interesting to see why Unit C3 still tripped first even if only fractions of a second(but due to polling rate of data it's logged as 5 minutes earlier).

Either fire fighting/electrical safety would have been the reason for B2 to trip. If GCB remained stuck closed after the turbine disintergrated then fire fighting would have been hammpered due to safety. Yard breakers may have been the only way to safely de-energize everything?

 

[MDEAus]

Actually judging by the frequency plot it appears that C3 did trip off before 13:45 sending the frequency through the floor (as lossing 400MW would do). C4 didn't have a much effect on the frequency after droppping out. Still 4 to 5 minutes between C3 tripping and C4 ceasing to exist.

 

[byrdj]

...I have been informed that this devastation is the result of a hydrogen explosion inside the generator which has led to the destruction of the generator and the turbine shaft in C4...

explosion of H2 in a running generator are rare. Unless the operators let the purity go to hell. A funny noise could have been a loose bar or shorted turn that eventaily cause a ground fault. then the generator "braking" far exceeded its load and rung the coupling off

maybe posible that the generator shaft seals were blowing by, bearing cavity filled with H2/air mixture, then a spark from the grounding bush/braid strap. wipeing the adjacent bears, locking up the inboard gen bearing

very interesting that is seems the AU politicians are more envolved with the operations of power plants

 

[MDEAus]

AU politicians are more envolved with the operations of power plants

Because in Queensland they are state owned. Other states have done 99year leases but Queensland for various reasons didn't, even though we started down the path of privitisation(starve for funds, and split into smaller companies).

 

[MDEAus]

AEMO Prelim report

I don't think anyone had posted it yet.

13:34 Unit 4 stops generating power while remianing connected to the power system(first big drop in frequency on the above graph)Generator starts drawing approx 50MW and 300MVAr.
13:40 CS energy report turbine hall fire.
13:44 C3 Trips (second big frequency dip).
14:06 Calvale Yard breakers are triggered. Dropping B2 of the grid triggering the cascade of other units, Total generation capacity lost 3045MW( I don't think all units counted here were at capacity so this is total potential capacity)

Anyone know what 32 minutes of a generator motoring will do? Especially with fire for approx 26 minutes.

Lot's of talk coming out of places with the usual "I didn't tell you nuthin". Workers being pressured not to "speculate" on cause of incident.

 

[RRaghunath]

MDEAus,
"Given that the generators casing is largely intact " - I was told that the generator casing is constructed to withstand H2 explosion and I think that's what we are seeing (End shield failure but the casing intact).
32 minutes of motoring - Even a few minutes of motoring could cause end stage LP turbine blade expansion that leads the blades rip through the LP turbine casing. Motoring protection is set generally with a couple of seconds delay.
Seems the machines didn't have Breaker fail protection as a back up to Generator Circuit Breaker (GCB) failing to open. There may be no instruction (in the operating procedures) for the control room operator to take action in the event of GCB failing to open on protection.

 

[itsmoked]

Why would the blades expand when unloaded?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[FacEngrPE]

Tip clearances are small at running temperatures. LP blades get rather long. If the turbine steam stop shuts, flow through the turbine will fall to a low value, so the blades are likely aerodynamically stalled. Considerable windage friction will result. If the motoring load was 50MW, a significant amount of that is heat input to the turbine.

Loss of steam flow likely results in a loss of vacuum, which aggravates the windage friction.

Turbine heats up, blades expand and start rubbing the casing.

 

[byrdj]

I'm not well versed in Generator electrical failure modes.

would it be that the generator had a bar to ground failure, thus sucking in the 50MWs. the "explosion" in the generator was initated by the BIG arcing and then release of H2 (by blowing out the seal casings, leaving end shields intack) then the blast of H2/air.

I didn't notice the seperation of the turbine from generator (by sheared coupling) in the sequence. maybe for that "motoring" period it was just current to ground (with generator field LOCKED)

thinking about it in this context, I can see that the design of the H2 seal oil casings could be a "relief" blowout, thus endshields and casing not being damaged

as implied by other, motoring of generator is not all that bad. Spinning a turbine (by motoring) with out "cooling" steam flow is real bad

notice the hood relief diaphragms look like they blow with a lot of force. I think they are design to blow if condesor goes 1 bar positive pressure

My first assignment (4 decades back) was the return to service of a generator that had a bar explode (loss of water coolant) but I did not remember the symptoms that occured during failure

 

[byrdj]

..., told CQ Today he had been informed by people on the ground that a hydrogen bearing seal failure was the cause of the breakdown.

“The reasons behind the generator and turbine failure at Callide C 4 are reasonably well-known in closed circles,” he said.

“Just very briefly, there has been a hydraulic bearing seal failure, which has led to hydrogen leaking from the generator, which has led to an explosion, which has led to generator failure and turbine shaft failure.

Looking at the generator wrapper, I am assuming that Toshiba might have the GE type dual ring design with once through treated seal oil

 

[MDEAus]

 reasonably well-known in closed circles,

What I am hearing from these closed circles is slightly different to this but I've been asked not to share as they have been told not to "speculate", not to say that this couldn't have been the reason for the explosion but it might not have been the root cause. The question is why did the seals fail?

That whole post is pretty much just playing wedge politics. Colin Boyce is the federal LNP memeber for that region. The QLD state government is Labor, much of the post will be politicking and slim on details. The LNP have been trying to get new coal generators built for a few years, Labor are resistant to that but also trying to appear sympathetic to coal miners. Note both take huge donations from coal mining companies. Just a bit more background info on the politics involved.

 

[byrdj]

(MDEAus)
from what little is available, it seems the other units have not returned to service.
Has there any public info on why such a long delay. best case, I would have though a couple days cleaning coal handling from the abrupt stop.

However, Damage from insuficeint bearing cooling oil (battery time on DC oil pumps)

I

 

[FacEngrPE]

Insufficient bearing cooling oil is a well known troubleshot.
During the 1965 NY blackout
Northeast blackout of 1965 The newly placed into service unit 30 (990MW) was damaged (no lube oil supply to the bearings resulted in serious wiping of the bearings), and out of service for many months. The lube oil system was powered only from utility power, and had no reserve power for black spin down.

Perhaps the station designers made the same mistake, and assumed complete loss of station power was not credible, or did not include enough battery for the entire coast down? Spinning down a large generator can take a many hours?

There are other ways an unsupervised trip can cause damage to large units. With the plants being evacuated, all sorts of things are possible Energy and Power - Important Guidelines To Startup and Shutdown a Large Generator

 

[byrdj]

"Big Allis" at Ravenswood.

the folklore was that operator instructions was to place the hot unit on barring gear asap. However when DC oil pumps ran out and with no heat removing oil flow over the stopped journals, the bearing babbit went plastic. So when AC power came back the attempts to put units on gear wiped the bearings. However the iceal of running oil pumps alone for several hours following such an event before TG was not pushed.

Also excessive force assisting barring damaged the close tolerance between bucket tips making contact due to rotor bowing again waiting more hours (day) for thermal equilization

 

[MDEAus]

The investigation is to be wide ranging, including human error. It is my understanding that there is uninterruptable power supplies to the oil pumps, if these failed or something in this system failed will be part of the investigation.

If the hydraulic oil pumps for both C units ran off the same UPS, then I could see why C3 tripped of first and why it would now take months to return to service if the UPS failed and subsequent loss of oil flow caused bearing damage.

 

[byrdj]

Just guessing and rambling here . . .

The prior designs similar to that turbine would have had a shaft driven oil pump. even though centrifugal, it would provide oil pressure / flow during coast down to about 1000 rpms. And as the shaft pump decade, a DC emergency bearing pump would sense and start to provide for the rest of the coast down. AND with the lost of vacuum, the rotors would coast to stop in about 45 minutes (or less).

So, turbine / generator could have been designed to coast to stop with no AC. However, the batteries for DC are only recommended to provide for coast down. And stopped with HOT rotors, the bearings will get hot shortly with no oil flow

Even farther, the plant could be design to survive a loss of switch yard and the turbine generator stay spinning providing “House” power. The latent heat in the tripped boiler could provide enough steam for operators to do boiler purge and get flames back in (30 / 45 minutes) . been there, seen that!

If 3 and 4 shared a control room, I expected to operators were concern for their life and not the grid

 

[TugboatEng]

In marine, oil pressure comes from a head tank that provides a minimum of 5 minutes of pressurized oil flow. We do also have the luxury of an astern turbine capable of shaft bursting which lets us bring things to a stop quickly.

Our feed pumps are also steam powered so s long as the boiler is hot enough to damage itself it's hot enough to keep itself cool.

Secure fires and open the superheater vent covers 90% of the fault responses.

 

[MDEAus]

Yes I believe these turbines have shaft driven oil pump and a back pump. I am not sure what it would take for both of those to be offline.

 

[TugboatEng]

Even at a few rpm, the turbine still requires pumped oil. A turbine spinning at a few rpm is neither capable of producing electricity to run the electric pumps nor provide sufficient speed to maintain oil pressure with the shaft driven pumps.