Power generation demand in my region has been sluggish in the recent years probably due to slow growth in big industrial ventures. Due to this, our power plant almost always be in partial load and often times asked to shutdown 1 unit during night time and started up again the next morning. Due to this, we have several problems with our system which I dont want to step into but one interesting problem is the MOOG servovalves.
As a general background, we have 2 units of Frame 6Bs, and have been running for 12 years or so. Control system is MarkVI TMR.
Recently during a routine startup after an overnight shutdown, the machine trips on GCV not following reference. Upon investigation, GCV could not be stroked past 92% and SRV could not be stroked past 89%. Looking at the servo currents, it seems like the VSVO could not drive the servos past these values. We replaced both servos and tried calibrating it again, same problem. We have encountered similar problem in the past 2 year where a GCV just straight up failed to operate and the only remedy is to replace the servovalve.
Funny thing is, during calibration, the servo can be driven up to 100% but immediately during manual stroke test, it fails to do so. We then move up to the controller debating the need to replace VSVO for T core which is the core that is not driving the servo(to my understanding, the servo valve can still be operated with 2, even 1 core) and we ended up restarting the whole controller.
Problem magically disappear and we returned to normal operation.
Any ideas what is happening here? Hiccups on MarkVI?
We are expecting a lot more problems to rise from these dispatch trends and just putting this out here for discussion.
Hmmm.... First, if replacing the servo has solved the issue or a similar issue in the past, then it kind of says to me that the problem is NOT the Mark VI.
Second, you did NOT say what Diagnostic Alarms are being annunciated by each of the three VSVO cards.
Third, you are correct. It is possible to open/close any servo-operated device with the current from a single processor/controller. And, that should be done EVERY time a servo is replaced--it's called a polarity check. (The rubbish in the Control Spec about "jerky" movement indicating incorrect polarity is just that--rubbish.)
To perform a polarity check, manually position the GCV (in this case) at something like mid-stroke with all three servo coils connected to the TSVO/VSVOs. Then disconnect one of the servo coil wires of the <S> processor and one of the servo coil wires of the <T> processor (at the TSVO). The device will lose a couple percent of stroke--but it should remain open. If it does NOT remain open, and closes, the servo current being applied from the <R> processor is incorrect; exchanging the two coils wires for <R> will cause the valve to open again.
Once <R>'s servo current polarity is verified to be correct, connect the loose wire of <S> processor, and disconnect one of <R>'s servo wires. The GCV should remain open; if not, then the polarity of current being applied to the servo by <S> is incorrect.
Once <S>'s servo current polarity is verified to be correct, connect the loose wire of <T> processor, and disconnect one of <S>'s servo wires. The GCV should remain open; if not, then the polarity of current being applied to the servo by <T> is incorrect.
Once all the polarity of all three processors has been verified, re-connect all the servo coil wires.
Fourth, you also did not provide ANY servo current values for ANY processor during manual stroking....
The servo (whether it's manufactured by Moog or Abex or anyone else) is just a device for converting an electrical signal (in this case, three electrical currents from three control processors) into a hydraulic flow to a hydraulic actuator driving the GCV (in this case). The servo doesn't move anything--the hydraulic actuator does.
There are three coils in the servo--and the algebraic sum of the torque produced by the currents (magnitude AND polarity) applied to the three coils determines how much hydraulic flow--and in which direction (to or from) flows through the servo to/from the hydraulic actuator.
The magnitude of the servo current determines the flow-rate--and more current (negative or positive) determines how fast the hydraulic actuator moves the GCV (in this case). The polarity of the servo current determines whether the flow-rate of air or fuel through the device the hydraulic actuator is positioning is increasing or decreasing. (I'm negating the effect of null bias current--which doesn't change.)
It's entirely possible that <T> was experiencing some kind of "issue"--BUT it would be HIGHLY UNLIKELY that NO Diagnostic Alarm(s) was(were) annunciated to alert someone to a(the) problem(s) when you were experiencing the issues you described. It's not impossible--but it would be very highly unlikely.
It's also possible that the shield drain wires of the twisted, shielded pair cables connecting the Mark VI (TSVO) to the servo coils are not grounded properly--maybe only one is incorrect (<T>???), but if any are incorrect, it can cause intermittent problems.
Also, have you verified that all the servo Configuration Constants/Setting are the same for the VSVOs are the same--and that all hardware (Berg) jumper positions on the TSVO are all correct?
Recently, I have seen one servo wire for one processor incorrectly terminated on the SIMPLEX terminal of the TSVO causing similar problems (though on a Mark VIe). So, check to make sure that all servo wires are terminated on the proper terminals of the TSVO card (while checking the polarities of the servo currents being applied--especially AFTER the polarity check!).
Have you checked the servo coil circuit resistances of all three servo circuits from the TSVO through each individual servo coil? Are they all the same?
Have you checked the wiring (using a meggar) to make sure there are no grounds or shorts? (Don't use the meggar on any coil when the other coils are still connected to the Mark VI. Disconnect all three coils (both wires of all three coils) from the TSVO before meggaring the leads, and only at 500 VDC. It's also recommended to disconnect the interconnecting wiring from the servo coils when meggaring, but as long as you don't go above 500 VDC and don't leave the meggar connected (at 500 VDC) for a long time (more than 10-15 seconds or so), it should not harm the servo coil(s).)
Last, how often is your oil condition tested/analyzed? Because the single biggest cause of servo failures and intermittent problems is dirty oil. Full stop. Period. Servo valves are made out to be these magical, mythical devices--and nothing could be further from the truth. They can't be calibrated using AutoCalibrate (contrary to EXTREMELY popular belief). They are delicate, industrial devices (in that they require clean hydraulic fluid. Make sure the polarity of the current being applied to each of the servo coils is correct, and make sure the oil is clean, and servos have been known to operate for decades without problems. Yes--decades. (Oil refiners have changed the formula for turbine lube oil--it's much better at lubricating bearings. BUT, the change has caused problems for turbines that use lube oil for hydraulic oil--something that worked flawlessly from the 1950's to the early 2000's, when the oil refiners started changing their turbine lube oil formulations. Servos have taken the brunt of the blame for something that is NOT their problem. At least one major oil refiner (BP Castrol) has produced a turbine lube oil specifically for turbines that use lube oil as hydraulic fluid--and it has been proven to alleviate so-called "servo problems/failures.")
Thinking logically about all the things you have said, they don't really make sense. And, without knowing what Diagnostic Alarms were being annunciated (specifically on the three VSVO cards), AND without actionable data (servo currents and conditions) it's even more difficult to say with any degree of certainty what might have been happening.
Hope this helps! Give us better information, and more of it, and we can probably be more help. (And, don't overlook Diagnostic Alarms.)
A couple of additional questions. First, when you are using the Manual feature of AutoCalibrate to stroke the GCV, what is the maximum reference you are entering? 100%? What happens if you try 128%? Or 127%?
And what is the current for EACH processor when you put in 127% when using Manual to stroke the Valve? And, what is the ACTUAL physical position of the GCV when you put 100% or 127% in the Manual Reference field (not the LVDT feedback position, the actual physical position)?