GE Frame 5 IGV Not Following Trip

I

Thread Starter

Ibra Brown

Dear All,

GE 5001 Mark VI Unit Tripped by Alarm "IGV not following Trip." We checked the LVDTs feedback and excitation and everything was ok. We check IGV servo valve resistance and it was ok. we tried to calibrate IGV actuation piston the maximum angle we could get was 56 we couldn't exceed it.

We're highly suspecting that the actuation piston is worn and oil is passing through the piston rings. What are the practice that could help us emphasizes the piston issue?

your help is appreciated.
 
With the unit on cooldown (ratchet), and no hydraulic pressure it should be possible to use a long bar or strong pieced of wood and pry the IGV ring fully open manually. I have seen some sites use a hydraulic jack under the joint of the IGV ring, with wood blocks underneath down to the turbine base. (The IGV ring usually has splits/joints on either side of the unit at roughly the same position as the horizontal joint of the unit. There is usually a bolt holding the two halves together, and there is often a place to pry against, or to place a hydraulic jack under.) This would prove that the IGV can be opened to the full open position.

You can monitor the servo current (signal name CAGV) while you are manually opening the IGVs. As you slowly change the manual position reference to open the IGVs and approach 52-56 DGA, the current should remain relatively stable. If, as you increase the manual reference above 56 DGA (you said the IGVs would not hydraulically open above 56 DGA) the servo current should be steadily increasing in the negative direction as the error between the actual position and the reference position increases if the IGVs don't open. This would suggest that the increasing the oil flow-rate through the servo to the actuator does not force the actuator to move. In this case, it would be possibly indicating the actuator cylinder and/or the actuator piston rings are very worn in a particular position and allowing hydraulic oil to leak past. However, were this the case, it would seem that trying to close the IGVs from this position would be either very sluggish or they wouldn't close very much at all. So, this wouldn't be a conclusive test of much, if anything.

BUT, has anyone physically observed the IGV actuator while trying to hydraulically position the IGVs from the turbine control system? Most Frame 5s have the IGV actuator in a very difficult place to observe--but not impossible. Some even have a small access door in the floor of the inlet plenum which can be opened by removing a few bolts to be able to see the IGV actuator mechanism, with the LVDT(s).

The actuator mechanism is usually bolted to the I-beam that runs between the outside turbine support I-beams. I have seen two or three of the hold-down bolts sheared off, or so worn that the actuator isn't held stationary and shifts back and forth as the actuator moves. I have also seen the heim joint that connects the actuator to the IGV control ring be extremely worn, causing a lot of slop in the mechanism and not allowing the full range of travel.

It should be possible to get someone into the space to observe and video the actuator as it is being operated from the turbine control system. Or to observe/video from above through the access door in the floor of the inlet plenum.

Hope this helps. Please write back to let us know what you find!
 
I
Dear Sir

Thank you so much for you reply and I really made a good use of the ideas you provided.

we manually changed the IGV angle to 85 Degrees while monitoring IGV servo current and angle trend. we noticed that Angle and current started proportionally, and at the 65 degrees the current started to go higher and higher leaving IGV angle at 65 degrees. This emphasis our first suspect which was the IGV actuation cylinder. Now we are on the process of installing a new actuation cylinder.

What are the things that we should be paying attention to prior to installing the new cylinder?

how to get 0.7 excitation voltage on 42 degrees?

Thank you so much on advance.
 
C
After installing a new cylinder, ensure that the feedback ac voltage is around 0.7VRMS, but not the excitation voltage. adjust the allen screw for small changes of Vrms, and adjust a nut on the travel scale for adjusting both LVDT f/back voltage at a time. at minimum mechanical position it should be adjusted until Vrms feedback is nearly 0.7Vrms.

Excitation voltage is always constant and it is around 6.5 to 7 Vrms.
 
To adjust the "zero stroke" voltage to 0.70 VAC RMS, +/-0.02 VAC RMS, you need to loosen the lock-nut on the threaded rod of the LVDT when the IGVs are at the minimum mechanical stop (which should be slightly less than 42 DGA, if that's the minimum operating angle). Using a TRUE AC RMS voltmeter measuring the LVDT ooutput adjust the threaded rod in or out to achieve 0.70 VAC RMS, +/-0.02 VAC RMS (0.68 VAC RMS, to 0.72 VAC RMS).

Then tighten the lock-nut while monitoring the LVDT output to make sure it doesn't change outside the allowable range.

Hope this helps!
 
And, Chiranjeevi is correct--one adjusts the output voltage of an LVDT, not the input voltage to the LVDT, which is constant from the turbine control system.

The output voltage of the LVDT is the input voltage to the LVDT feedback circuit of the turbine control system.

Please write back to let us know how you fare!
 
I
Dear All,

Thank you so much for help and ideas.

After dismantling the actuation cylinder we checked piston and piston rings and they were completely worn. The unit has reached 12000 hr is it normal.

we installed a new actuation cylinder. we manually set the IGV to 42 degrees using a hydraulic jack and verifying the on field IGV indicator. we manually adjusted the LVDT feedback voltage to 0.7Ac (Blue and yellow) wires.

When we started the calibration process hydraulic pressure did exceed 12 bar as its normally reaches 30 bar with ratchet pump on and the IGV test valve open.

what might be reason for this?

Thank you so much in Advance
 
ibra_brown,

I want to answer the last question first. It seems the unit does not have an Auxiliary Hydraulic Pump, driven by an AC motor, and that there is some kind of tubing arrangement that can be used to direct the output of the DC Hydraulic Ratchet Pump to the hydraulic system. This is not very common, but it does prevent the need to CRANK the machine to establish hydraulic pressure in order to hydraulically stroke the IGVs and/or calibrate the IGV LVDT feedback.

As for why the pressure didn't build up as high as it did before? Well, it's probably the result of some "test" valve not being in the proper position when using the DC Hydraulic Ratchet Pump to produce hydraulic pressure. Or, it could be that the servo was installed improperly allowing oil to go where it shouldn't. But, it's most likely not a control system problem.

Anyone using the DC Hydraulic Ratchet Pump for stroking hydraulically-actuated devices instead of CRANKing the machine should review their L.O. P&ID, because if I recall correctly the DC Hydraulic Ratchet Pump draws its suction from the L.O. Reservoir, which means it is unfiltered oil, subject only to the any "pre-filter" which may be present in the hydraulic supply line to a servo-valve and hydraulic actuator. And those "last-chance" filters are basically just "rock-catchers"--they don't have a very fine filter in them. When they are installed, they are better than nothing, but only slightly. So, using the DC Hydraulic Ratchet Pump to provide hydraulic oil to hydraulically-actuated devices with servo-valves can result in some unfiltered or very coarsely filtered oil getting to/through the servo to the actuator.

Next, it's very common for the IGV control ring to have some over-travel, past the normal closed- and open positions. For example, if the minimum operating angle, CSKGVMN, is 42.0 DGA, and the maximum operating angle, CSKGVMX, is 84.0 DGA, then the IGV actuator can usually drive the control ring to approximately 2 DGA less than the minimum angle (so, to approximately 40 DGA), and 2 DGA above the maximum operating angle (so, to approximately 86.0 DGA). This is done so that there is no possibility that the control ring mechanical stops can prevent the IGVs from reaching their normal minimum- and maximum operating angles if they are not properly adjusted.

Why is this important? Because, when using AutoCalibrate to calibrate LVDT feedback it puts out maximum positive servo current (to move the device to the minimum mechanical stop) and then puts out the maximum negative current (to move the device to the maximum mechanical stop) and records the LVDT feedback voltage at those two points. In the case of the Mark VI, it then looks at the user-entered minimum and maximum measurements (taken BEFORE the calibration procedure is initiated!) and then calculates the offset and gain required to scale the feedback to match the actual physical minimum- and maximum measurements.

So, one has to physically close the IGVs to the minimum mechanical position, measure and record the angle, then open the IGVs to the maximum mechanical position, measure and record the angle, then enter the two recorded positions in the calibration fields for the IGVs in Toolbox, THEN initiate the LVDT calibration. Once that's finished, one needs to then verify the calibration was done accurately, usually by moving the IGVs to two, three or four positions (such as 50 DGA, 57 DGA, 75 DGA and 84 DGA) and physically measuring the IGV angles to be sure the indicated (scaled) LVDT position on the display matches the actual angle of the IGVs.

Now, it's pretty difficult to get most machinist's protractors into the IGVs of a Frame 5 when they are fully closed against the mechanical stop. And, most sites don't have a machinist's protractor or someone trained to use one to measure IGV angles, so they rely on the pointer on the side of the axial compressor casing for their measurements. That's okay--not great, but okay. (LOTS of site NEVER measure actual IGV angles.... and the turbines run fine, though they might not be producing as much power as they could be if the IGVs were properly calibrated.)

It's typical to adjust the "zero-stroke" LVDT output voltage when the IGVs are at the minimum mechanical stop (fully closed). And, when the IGVs are properly calibrated and they close to the minimum mechanical stop (which they will do when the unit is shut down or tripped) the indicated IGV angle will be less than 42 DGA--but it should be approximately equal to the measured angle recorded before the IGV LVDTs were calibrated--and that's all as it should be.

Hope this helps! It's not really difficult, and it doesn't have to be 100% perfect, either. But, one should know the proper procedure and methods--and the drawbacks of using DC Hydraulic Ratchet Pump for providing the hydraulic pressure for stroking/calibrating LVDTs.
 
I
Dear Sir,

Thank you so much for you help and your advises worked well.

To get over the hydraulic pressure issue We had to replace the IGV servo valve to get the 30 bars in order to stroke the actuation cylinder. the servo valve was not functioning properly.

Steps we followed:

we manually regulated the IGV angle using a hydraulic jack and a piece of wood to 42 degree. Then, we installed the IGV actuator and regulated that the IGV LVDTs feedback voltage to about 0.75AC.

We gained access to the LVDTs after installing the actuator by going inside the inlet plenum and removing the cover that exposes the actuator.

Then we started the calibration process and luckily after installing the new actuator we could stroke it to reach the maximum angle which we couldn't do with the old replaced actuator.

Additionally, I observed the proportional change between Servo current and IGV Angle change till they reached stabilization.

The unit now is back on duty and the IGVs are working properly.

Lesson Learned.

I'm really grateful for you assistance.
 
Top