Hello mkvguy,

Our GT-1 (MK-II System) was started after HGPI... During the shutdown we had calibrated the cards of MK-II TCP (Turbine Control Panel) as always.

The machine was run for 48 hours for performance checks. There was no problem noticed and the OES declared the machine ok and left.

Immediately afterwards, it was seen that when the machine was in base load, the VCE (the MK-II equivalent of FSR in MK-V) began to swing violently. (On a scale of 0 - 10V the VCE meter needle would violently swing back and forth between 7.5V to 3.5 volts...!!!) The machine rumbled and grumbled and in panic it was shut down.

Instrumentation people (our dept.) were summoned and given this briefing. We noticed that there was a flame failure alarm that morning and we thought we would attend it later in the month as it was difficult to attend it in the running machine. Since the Honeywell flame amplifiers source 300 VDC and take 28VDC as inputs, we thought that the cable might be grounded somewhere causing the loss of flame for 28FD-7 and "might" have caused the VCE (a DC signal) to sway as it did...!! (It was a pure conjecture, trying to link DC with DC.) We changed the sensor, cable and gave the machine a start. The machine went up to FSNL. Loaded to 5MW... It ran for 15 hours at 5MW and was ok... We felt that IT indeed was the source of trouble.

Next morning it was loaded till 15MW. All was quiet. I observed the running for nearly 3 hours and there wasnt any deflection of the needle. VCE (FSR) was held constant.

It was reported during the night that VCE has suddenly began to fluctuate again, almost as violently as it did the earlier time, and the machine was shut down.

For two nights and three days we kept changing cards after cards (MK-II has nearly 150 cards to change)... And the machine would not behave normally. VCE would fluctuate in FSNL only after few hours of run.

During HGPI we had changed the moog valve filter... and during the GCV stroke check it was found that some air got trapped in the moog valve somehow and the GCV didnt lift after the first stroke. But since we could zero in on the cause we felt it was ok now....

My questions are:
1) What are the things we might look for to trace the problem? (Apart from the TCP ofcourse which we are currently doing)

2)Our local GE rep has professed having very little information about MK-II systems and might not provide any support. Who else can? Is there any source, worldwide, where we can seek support from, on an urgent basis, if needed?

3) Knowing full well that VCE is "likely" to swing so drastically, is it safe to run the machine, after we change a few cards, to see if the problem is resolved? (I mean would it be mechanically dangerous to run the machine on test if we knew that VCE might fluctuate again? We can change cards but will not know if the problem is solved unless we run the machine... and if we run the machine there is always a possibility that it might be subjected to violent fluctuations of VCE again...!!!)

4) We have the readings of the GCV stroke calibration that we performed just before start up. It's "now observed" that the servo drive volts is -6.78 volts for each coil (Our Moog valve has two coils) at 0% stroke & -0.44 Volts at 50% and 100% strokes. We are unable to understand why is the voltage at the servo coils just -0.44 volts for both 50% and 100% strokes. The valve opening, nonetheless, was perfectly ok during stroke check - 19mm and 38mm respectively. We checked the past five years of records of GCV calibrations for both GT-1 and GT-2 (MK-II controlled machines) and these readings seems to be same year after year. (These readings are right at the coil inputs/terminals.

Regards

au revoir
Rahul

 

Rahul,

It would not be advisable to run the unit like this. Wildly fluctuating fuel pressure can cause wildly fluctuating compressor discharge pressure and can even lead to collapsed combustion cans...

This author has little experience with Mk II controls, but from what you have written it doesn't really seem to be a Mk II issue.

Every problem is not controls-related. Are you sure the hydraulic pressure is stable and the hydraulic accumulator is properly charged (presuming the unit is equipped with an hydraulic accumulator--many were, most are now-a-days)? What about the hydraulic relief valve and air bleed valve? Are you certain that the hydraulic tubing was firmly reconnected to the servo-valve? Are you certain the servo-valve was securely bolted to the manifold/block after replacing the filter? Are you sure the filter was properly installed in the servo-valve?

There are many ex-GE people around the world who have a lot of experience with Mk II Speedtronic turbine control panels. This author only knows of people in the United States--Pond and Lucier is probably one of the best (www.pondlucier.com). There probably really aren't many people in GE who have much experience with Mk II.

You could use two of three coils of a three-coil servo-valve IF the servo-valves flow the same amount of oil (flow-rate) AND have the same port arrangement.

ALL GE-design heavy duty gas turbine electro-hydraulic servo-valves operate the same: negative current opens the valve and positive current current closes the valve, and at steady-state position--when the LVDT feedback equals the position reference--the ONLY current applied to the servo-valve is null BIAS current, so it's not surprising that at steady-state position the servo current is -0.44 mA (probably per coil, for a total of -0.88 mA--which is probably exactly in the allowable range of total current at steady-state condition). (Currents are being quoted because the coil resistance (of each coil) should be approximately 1000 ohms.)

Having said all of this, the problem could also be the SRV (Stop-Ratio Valve).... The GCV will swing if the SRV swings (of course, the SRV will also swing if the GCV swings--so sometimes it hard to tell which is causing the problem). Things which might make the SRV swing would be intermittent/erratic speed feedback, intermittent/erratic LVDT feedback, or intermittent/erratic P2 pressure feedback.

There are a lot of Mk II panels still in service around the world--they are very reliable and fairly robust. So, make sure all the inputs are good, the outputs are good, the hydraulic pressure is steady and reliable, and the tubing connections and bolts are all tight and secure. The problem started after some period of time and has worsened--in spite of replacing Mk II cards....

markvguy

 

Without knowing much about your turbine, it is hard to pinpoint the problem. However, I assume the following:
1. Your turbine is MS5001A
2. There are no control system ground faults
3. There are no servo fault alarms.

You may proceed in the following sequence:

1. Make sure the gain of the speed controller is set correct. This is a pot on the SSZB card. The setting given in the document "Control System Settings" normally works. However, there have been occasions when the gain pot has to be tweeked a little to stabilize the controller.

2. Check / replace the magnetic pickups for speed control. Of the four pickups, two are used for over speed and the other two for control. Even if they pass the checks when they are cold, could cause problems when the unit heats up. Recommend to change both. However, make sure that the gap is set correct (50 mils).

If the problem still exists, let me know.

In the last part of your message you had mentioned the voltages measured at the servo coils. I think the reading for the closed valve position is incorrect. Normally it will be a positive voltage. The volgage of -0.44 at 50% and 100% positions are the expected values. Remember the servo valve has a bias spring inside. This negative voltage is applied to overcome the mechanical bias of the spring, else the valve will drift to the closed position. Once the actuator is at the required position, the servo amplifier will drive the current to zero (of course, with out the mechanical bias).

Good luck

Radhakrishnan

 

Rahul,

I have worked on MK-2 Speedtronic Controls system about 10 years ago and have seen these fluctuations of the VCE. If the fluctuation happened during normal running, I usually adjust the stability from the card. There is a port to adjust the gain and stability. You can refer this to GE specification, but sometimes doesn’t work due to problem on the Hydraulic system.

If the unit is shutdown, check the excitation of the LVDT reference which is usually about 7.2VRMS. Check the oscillator by using the speedtronic calibration box.
Readings should be 2800Hz - 3200Hz.

Try to stroke the SRV & GCV. Check the voltage for servo valve in the MK-2 terminal Control Panel. If the supply output Voltage to the Servo valve good but doesn’t move when stroked, replace the Servo Moog. If the SRV or GCV is able to move, keep the position at 50% opening and check the null Bias voltage. If the results are still not correct, then adjustment of the servo valve should be carried out. If spare servo available, then replace with new one.

During calibration record the values of both LVDT feedback from the MK-2 Speedtronic panel or directly from the JB. Usually the reading should be 0.7 VRMS at close Position – 3.5 VRMS at Fully open measured at JB terminal Wiring blue and yellow.

Make sure the gain and stability adjustments confirm to GE specification.

As Markvguy has suggested to check any possibility of leaks from the lube oil, trip oil and hydraulic oil. There have been cases of leaks from the MOOG servo valve due to broken ‘O’ ring. Make sure when tightening the lock nuts that all four bolts are tightened evenly. Also check the VMPA cards and replace if necessary, Check the Gap of Magnetic speed pick up should be within the specification, check if there is any ground fault it should be flouting Power supply system, Check if any loosing terminal connection on LVDT and Servo.

 

Thanks for the inputs... The problem seems to have been solved... ATLEAST FOR NOW because even in past the turbine was run for sometime before the problem reappeared.... Now it's been in operation for last 48-50 hours without any fluctuations... (We are careful to the extent that we dont even want to breathe near the panel lest the needle may begin begin to sway again...!!!!!)... Smiles apart, it's been a great relief for all.... I atleast had one place - control.com - to turn to, to seek some probable answers to the problem.....

A brief description of what we did, for people who still use MK-II and might find this important....

Problem : Violent fluctuations in VCE...

Grey areas : We just had a Hot Gas Path Inspection... We had changed the servo valve's filter... The Servo Valve did misbehave during the final Stroke Check exercise before start up... Hence a new servo valve was fitted and Stroke Check was successful... Turbine ran for several hours before showing the violent fluctuations in VCE(FSR)... The problem was intermittant.... We were also diverted by the fact that there was a 28FD-7 fault few hours before the emergence of the problem... And after replacing the cable and sensor the problem vanished for nearly 20 hours again....

Approach : We saw the connection diagrams and found that three cards mainly contribute to the final Calculated VCE... Viz., Speed Control Card (SSZB); Temp Control Card (STKK) and Start up Control Card (SSKC)... The minimum VCE selection is made possible by the presence of one diode at the amplifier (OP-AMP) output pin on each of these cards, the cathode of these diodes were hard wired to each other and connected to P12V...

The Card "willing" to assume control will lower it's amplifer output proportionately and allow the diode to conduct... The other cards would make the Amplifier output high so that the diodes were reverse biased and would not contribute to control.... This "assumption" became the clincher....

The efforts were directed towards changing and calibrating these cards... Unfortunately, it was found that out of Seven NEW temp control cards purchased in 1997 Six of them couldnt be even calibrated.... The New cards were all found faulty.... One card could finally be calibrated.... Two new Speed Control Cards SSZB failed twice during operation... Finally we have put another SSZB card, calibrated it and taken a last trial... Ever since the machine is in operation.... And I have some time to spend on net thanking you for your inputs....

Hope it continues to work...

au revoir
Rahul

 

Rahul,

I understand your problems with GE no longer supporting Mark I, Mark II, Combustion Monitor and ITS systems and I now hear Mark IV. My company, John Brown Engineering Gas Turbines, Ltd. in the UK still supports the full range of GE Control Systems and can provide technical support, spares, etc. If you have further problems please contact us on our service E Mail.

Regards,

Duncan M Wilson

 

Interestingly, our both the turbines are commissioned by John Brown Engineering... We are eagerly looking forward to purchasing spares for our MK-II systems... Shall get back soon with the "wish-list"....

Earlier we were in touch with some other parties who also supply MK-II spares/cards but typically our management desists practices of buying from parties other than the OES/OEM.... Given this case, it should be also easier for us to push our case to the management.... If possible kindly send me a contact mail ID at rahulpsharmaATGMAILDOTcom

au revoir
Rahul

 

Hi Rahul P Sharma
Thanks for all

We have the same problem, please help me , tell me how did you find a solution.

Our GT-5 ( TP5) (before 1995 equiped with MK-II System there was no problem).

after 1995 retrofited with MK5 Simplex and the problem of fluctuations appears and always after 70% of load.

We did a HGPI and calibartion all instrumentation. and the problem persiste until today.

Please help me