Steam Turbine Failure South Africa

Status
Not open for further replies.
Quote:
I just finished a failure analysis on a part that has one similarity to this case -- it too had four protection systems that failed.

One thing I've learned in failure analysis, is that catastrophic events often happen because of a combination of unlikely events that come together.

One of our managers came up with a "Swiss cheese" analogy...each protection scheme has a hole, like a slice of swiss cheese and generally there are multiple slices acting as protection...some days all the holes line up.
 
Originally Posted By: maersk
By way of logical consequence, the centripetal force doesn't exist either. They're both fantasies.
grin2.gif



The centripetal force DOES exist. It's the tension in the string in the tether tennis set that causes the acceleration of the ball away from it's linear motion.
 
Originally Posted By: ToyotaNSaturn
Originally Posted By: Johnny
"A witness said that absolute chaos ensued. One man, who had been showering at the time, ran outside naked."

This was probably the guy that should have been at the control station to activate the emergency shut down.


Questions going thru his mind a few minutes prior to the fateful incident: "Hmmm, I wonder if I can I fit a shower in before the turbine stress test or not?"


Not quite sure, but the 4500RPM sounds like there was much more steam available than a typical overspeed test.

Turbine speed protection is typically:
* Governor (mechanical or electrohydraulic) monitors speed and positions steam valves to control it.
* Electric overspeeds, 3 off, polled that any two out of the 3 will trip the hydraulic primary trip circuit, dumping control oil, and shutting steam valves...usually set at 5-7% overspeed.
* Overspeed "bolt" or "ring", which when it gets to rated speed (usually 10%), flings out, and trips the primary trip circuit. Industry standard is one bolt, but my state specifies two (reason later).
* Often a hydraulic emergency trip that comes in at 112%.
* Often an Acceleration sensing device that picks up rate of change and closes in the valves, pre-empting Governor response.

Various testing that takes place.
* Valve tests (daily or weekly). Each and every steam control valve is driven closed individually, and at load to ensure that they will shut off. (oxide growth on stems has caused many catastrophes).
* Overspeed bolt tests...the tripping action of the bolt is disabled, oil injected in behind it, and the fact that it stroke to the trip position is proven. The reason my state has two is that there's still an active bolt in the 2-3 minutes per week that one is locked out for testing.
* Actual overspeed - required annually by the insurance companies to test the system. Electrics are checked first, then wound up to 12% to get them out of the way. Turbine is then oversped until the trip bolts/rings activate, and their response is proven. Electricals are wound back to normal.

The article reads like it is the actual overspeed test that failed, and I find it difficult to believe, as the throttle valves are nearly closed during these tests, and rates of acceleration low.

Could have been a trip while doing the on-load testing maybe ?

600MW of steam turned into acceleration can move things quickly.
 
Originally Posted By: IndyIan
So what does the 500,000hp generator look like at the end of these things? Or I guess they link a bunch of smaller ones together? Do they use a generator as a motor to overspeed the turbine without steam?
My Dad does some work at GE large motors in Peterborough which seems to get into fairly big stuff, but not 500MW... If I recall they don't spin many large, but relatively small, motors that fast.


Great big bean can from the outside...on coal stations typically only one of them. Rotor is excited at a few hundred volts and thousands of amps. Stator is 22,000 Volts, many thousands of amps. Stator is cooled with water running through the conductor bars, rotor is cooled by 60psi of hydrogen.

They are "synchronous", meaning that when connected to the grid, every generator on the grid is running in lock-step at exactly the same speed.

They are only connected to the grid at exactly the point where their speed and phase matches the grid. Off phase synchronising is bad, and connection to the grid off frequency (particularly stationary) is really bad.
 
Originally Posted By: Shannow
Originally Posted By: maersk
By way of logical consequence, the centripetal force doesn't exist either. They're both fantasies.
grin2.gif



The centripetal force DOES exist. It's the tension in the string in the tether tennis set that causes the acceleration of the ball away from it's linear motion.


What makes the concept of "centrifugal force" so attractive is that it's the tension you feel in a string when you're hanging onto an object swinging around you. But there's actually no force acting on you - you're standing still. If a force was acting on you, you'd be accelerating. On the other hand, the swinging object is clearly moving and accelerating, so the only force that's actually measurable is the centripital force.
 
Originally Posted By: Shannow
Great big bean can from the outside...on coal stations typically only one of them. Rotor is excited at a few hundred volts and thousands of amps. Stator is 22,000 Volts, many thousands of amps. Stator is cooled with water running through the conductor bars, rotor is cooled by 60psi of hydrogen.

They are "synchronous", meaning that when connected to the grid, every generator on the grid is running in lock-step at exactly the same speed.

They are only connected to the grid at exactly the point where their speed and phase matches the grid. Off phase synchronising is bad, and connection to the grid off frequency (particularly stationary) is really bad.


Is it correct then that once the generator is connected, the connection enforces the synchronization? Any phase error will be corrected by current flow in short order. Changing the level of drive (steam valve or whatever) only changes the contribution from the generator to the grid, not it's speed, right? In fact, if you turn the steam off, it will act as a motor and consuming only enough power to overcome friction. Am I correct?
 
Yep, every machine in the grid is locked to every other in speed. Phase (think TDC) will shift a little as power factor varies around the place.

Every time a load (even a light bulb) comes onto the grid, the frequency drops slightly, and is recovered by turbine Governors. Lose an aluminium smelter, and the frequency rises reasonably sharply, and Governor's recover it.

In theory, you could remove steam from the turbine and "motor" the generator, but protection systems stop it from occuring for more than a few revs.

Motoring a stationary unit typically burns the rotor badly (has happend through switching incidents in this state)
 
Originally Posted By: jaj
Originally Posted By: Shannow
Originally Posted By: maersk
By way of logical consequence, the centripetal force doesn't exist either. They're both fantasies.
grin2.gif



The centripetal force DOES exist. It's the tension in the string in the tether tennis set that causes the acceleration of the ball away from it's linear motion.


What makes the concept of "centrifugal force" so attractive is that it's the tension you feel in a string when you're hanging onto an object swinging around you. But there's actually no force acting on you - you're standing still. If a force was acting on you, you'd be accelerating. On the other hand, the swinging object is clearly moving and accelerating, so the only force that's actually measurable is the centripital force.



of course the force is acting on you, you are not moving because your inertia and friction between you and the ground has a greater effect than the small force in the string, hence no movement.
 
Originally Posted By: Shannow
Originally Posted By: Kestas
I just finished a failure analysis on a part that has one similarity to this case -- it too had four protection systems that failed.

One thing I've learned in failure analysis, is that catastrophic events often happen because of a combination of unlikely events that come together.

One of our managers came up with a "Swiss cheese" analogy...each protection scheme has a hole, like a slice of swiss cheese and generally there are multiple slices acting as protection...some days all the holes line up.

Common analogy used in security and risk analysis...makes a good powerpoint slide...your manager quoted it...but doubtful he "came up" with it...
 
Originally Posted By: crinkles
...of course the force is acting on you, you are not moving because your inertia and friction between you and the ground has a greater effect than the small force in the string, hence no movement.


You can feel it, but it's not making you move because, as you point out, there's an equal and opposite force through your feet that's canceling out the force on the string and keeping you in place.

An outside observer can only tell if there's a force on you by measuring whether or not you move. Since you're standing still, they have to conclude there's no force, regardless of what you're feeling in the string. That's why there's no "centrifugal" force - it's not observable.
 
No, the centripetal force IS a force acting on you, it is observable and measurable.

The centripetal force is the force that's being applied to an object to accelerate it (make it change its behaviour/direction)

If you were riding inside the ball on the tether, and had no other frame of reference, you would feel like you were being dragged to the outside, and describe it as a force, centrifugal force, but it is the acceleration away from linear motion, and your reaction with the ball that you are feeling.
 
 
No, the torque is making the turbine spin. Not the centrifugal force. That is a result of the spin itself, not its cause.

As for the centrifugal force not existing, that's tantamount to saying impact forces, for example, don't exist either as they to are an expression of inertia as well.

I really don't get why people get so hung up over refusing to call them "forces" or acknowledge their existence like it's a matter principle.

Or are we just arguing semantics?
 
where does torque come into it ?

Draw a free body diagramme of a ball on a thether...and then explain where the "centrifugal force" lives.
 
Yes radial acceleration is caused by the centripetal force...it's the force that stops the ball following it's straight path, by accelerating it towards the centre.

Tangential acceleration by torque...that's what turns the turbines.
 
Originally Posted By: maersk
No, the torque is making the turbine spin. Not the centrifugal force. That is a result of the spin itself, not its cause.

As for the centrifugal force not existing, that's tantamount to saying impact forces, for example, don't exist either as they to are an expression of inertia as well.

I really don't get why people get so hung up over refusing to call them "forces" or acknowledge their existence like it's a matter principle.

Or are we just arguing semantics?


It's angular momentum that keeps it spinning and kinetic energy that does the damage.

We're not arguing semantics, if we're arguing at all...

Or maybe we are - a force pulling on you is not "acting on you" because it's not producing a measurable response. As I said - you can feel it, but nobody else can measure it because you're not moving.
 
No a force on you is a force acting on you...whether it causes you to move or not.

The force in the string is the centripetal force that you are applying to the object to cause it to deviate from linear motion. If you were tied by the feet to the rotating end of the string, you would feel the centripetal force trying to tear your feet off.

You would also "feel" a force trying to rip your head off which doesn't exist...it's the centripetal force that your neck is applying to your head, dragging it into the curve, while your head wants to continue charging in a straight line
 
Status
Not open for further replies.

Similar threads

dave1251
  • Locked
Exxon Mobil
Replies
19
Views
14K
MolaKule
Back
Top