Guide Vane Oscillations 2

[simboti]

we're a hydro-power plant with a rated output of 750Mw. one of our turbo-alternator's (Francis turbine)guide vanes (GV) have been oscillating arbitrarilly. initially we thought that this was due to the governor actuator loop and we suspected the main servo valve's position transducer and the feedback from the servomotors. however these were tested and re-adjusted as per the commissioning report. dispite these efforts the difference between the guide vane referrence and the actual guide vane opening was approximately 3%. this is unusually high. we therefore replaced the actuator (electrohydraulic convertor),the oscillations temporarily stopped, but the difference between GV ref and GV actual remained the same. we believe that there is no problem with the digital governor (PID controller). since we've checked the programe and the control parameters. how should we therefore proceed? what are the root causes of such oscillations?

 

[mc5w]

If what you have is an electronic servo computer the digital to analof converter could be shorting the +12 volt analog power into the 5 volt digital logic. This would send 0.7 amp pulses into the microprocessor. Because of the current limit in the power supply there would be a voltage dip every time the microprocessor updates the DAC. If this is happening you need to replace the DAC and the microprocessor. You should also build a daughter card for the DAC so that you can use buffer gates with "high voltage" open collector outputs to convert from 5 volt TTL to 12 volt resistor-transistor logic.

You could also have a broken wire somewhere and depending on iffy contact that could be your problem.

Also, optically isolated input modules sometime lose their noise suppression. When this happens ALL the input modules go bad. I do not see how static electricity can burn up just the noise suppression, but I have seen it happen.

 

[MikeHalloran]

OR, you could have a worn-out bushing in the linkage.

I'm guessing that the 'actual guide vane opening' signal is not derived from the physical position of an actual guide vane, but from the position of the servoactuator or some part of the ring that adjusts all the guide vanes. Any slack in the mechanism would allow the guide vanes' real position to be influenced by hydrodynamic forces on them.

The governor actuator loop would get involved because it knows the shaft speed and would move the vanes to try to correct it, but the system would end up hunting in a limit cycle because of the lost motion in the linkage.




Mike Halloran
NOT speaking for
DeAngelo Marine Exhaust Inc.
Ft. Lauderdale, FL, USA

 

[simboti]

thanks guys for the quick response. mc5w, the turbine governor that we employee is the HPC 640 GUIDE VANE CONTROLLER. It is digital and is implemented in an ABB Advant Controller 110 (AC110).
Mike,you guessed right the guide vane position is read by a mechanical link connected to one of the servomotors at one end and a position transducer on the other end. these however were checked and necessary adjustments were done.
my worry is that we've operated this machine for two years after refurbishment and we've never attended to the electrohydraulic convertor (actuator, TR10: ALSTOM), though we've opened it for the other machines. Now i'm faced with a situation were i have to check the hydraulic balance, mechanical tendency and the associated electrical compensation. i only have skeletal information and seriosly in search of detailed information.

 

[Skogsgurra]

mc5w,

This is very detailed information - and information that is not so easy to understand. Could you please expand your thinking a little? Especially how you arrive at the 0.7 A current.

 

[niallnz]

Simboti,
Have you ben in contact with ABB?, they did the upgrade I would think they should be of some assistance.

The electronic governor puts out a position demand signal to the actuator driver. The drive takes this and the position feed back, and controls the guide vanes. A 3% error could result if the actuator is simply operating on proportional gain only, although I think 3% is too much on a system that uses an intermediate distributing valve. The position feed back does also go back to the governor but it is usually used to clamp the output of the PID control and avoid integrator wind up.

Random movement of the guide vanes I would be checking the oil filters on the oil supply to the electronic proportional valve. As you said things improved for a while after you changed the actuator. These proportional valves (Bosch, Moog, Vickers etc) have only microns for clearances on the spool. If your control oil isn't clean they can get sticky. I expect that you've got a duplex filter with DP switches and indicators but it's worth a shot.