Experiencing problems with a GE MK_V turbine control system used to control a 68.5 MW GE steam turbine. Recently whenever the load command reaches 63% while in governor control mode, the load will not respond either up or down at this point. The only action that can be taken at this point is to switch to load limit controls after which load can be adjusted again.
In the past heavy fluctuations or non-response of control valves were corrected by servicing or replacement of the woodward EG3P servo due to water or contaminated hydraulic oil. However, what is strange about the current behavior is that the load can be adjusted once on load limit but not on governor once a certain load command is reached. Not sure if a dirty oil circuit would affect the controls in governor control but not on load limit.
Would appreciate any feedback thoughts on this.
If this is a relatively new occurrence, not related to oil condition, it would seem that some input to the Mark V is out of calibration. One would need to look at the CSP (Control Sequence Program) to see what is preventing the commands from affecting the output to the servo.
Unfortunately, steam turbine control is not as standardized as gas turbine control. There can be single-, double or triple controlled extractions. And, admission. And, the single- and double auto extraction control blocks are not well documented.
But, again, if this particular issue has just recently started I would be looking at inputs being out of range/calibration. Something like that. What Diagnostic Alarms and Process Alarms are being annunciated before and during this problem? Specifically, what modes are you transferring out of and back into? What is the configuration of the turbine (controlled auto extractions, or uncontrolled extractions; etc.)? Unfortunately, there isn't a lot of steam turbine control philosophy expertise here on control.com, but perhaps someone will have some idea(s) if we can get more information.
And, if this just started after some outage (planned or unplanned) where some work was done on field devices/instruments (even just removal and replacement) it's possible something was "disturbed" that isn't obvious.
Please write back to let us know how you fare resolving this. Many gas turbine control functions are the result of high- or low-select functions, where several references are fed to a HIGH or LOW select block and the highest or lowest one is passed through to the output to the servo. I'm not saying all steam turbine control functions are like this, but if this is one that would be the best place to start looking in the CSP. Gas turbines use L70R and L70L as the raise- and lower speed reference (load, when synchronized to the grid) signals. So, if the speed reference isn't changing when it should be one can look at those logic signals to see what's blocking the signals. I think steam turbines use something similar for basic load control, but I'm not sure (maybe L70LOAD?).
Hope this helps--at least a little!
Thanks much for your feedback. We are definitely looking into that possibility of an input being out of range.
We are reviewing the CSP to see if we can find anything that is limiting the output to the valves and will post whatever is found or whatever the final fix is.
Thanks again for your feedback.
Update on MK_V control problems.
The unit was taken off line to effect boiler tube repairs, therefore the opportunity was taken to service the EG3P Woodward servo actuator. It was cleaned with kerosene and particles found in the lube oil indicating dirty/contaminated oil. The processors were rebooted by shutting down all power to the MK_V panel. The unit was returned to service this morning and it appears that the issue of non response above 63% Load_CMD has been resolved.
Am still baffled as to why this behaviour would manifest itself on governor control only and not on load limit if it's a hydraulic oil issue but it is what it is.
Thanks for the suggestions recognizing that every problem encountered or every advice/suggestion is an opportunity to learn something new.