Servo suicide Mark VI

T

Thread Starter

Tarun

Dear All

We are having GE Frame VI gas turbine control by Mark VI control system TMR (R,S and T) having dual firing facility Gas as well as liquid. Control system was retrofitted from Mark IV to Mark VI Jan 2007. After commissioning we have problem of Servo Suicide.

VSVO card is inserted in slot 5 is for SRV (Regulator1), GCV (Regulator2) and Liquid fuel bypass servo (Regulator3). At present we are using 100% gas fuel.

VSVO card is inserted in slot 7 is for IGV (Regulator1)

TSVO termination card is common for all regulator.

Following diagnostic alarm appeared frequently after commissioning.

1. T controller Slot 5 : Fault code 47; Servo current#1 disagrees w/ ref, Suicided (regulator1) (SRV)

2. T controller Slot 5 : Fault code 47; Servo current#2 disagrees w/ ref, Suicided (regulator2) (GCV)

Above both diagnostic alarms were appeared on same day with half hour time difference. and

3. R controller Slot 7 : Fault code 47; Servo current#1 disagrees w/ ref, Suicided (IGV)

We need to reboot the processor to reset the problem. We have carried out lots of action as suggested by OEM. Like resetting null bias current, setting regulator gain, checking terminal tightness, checking cable for grounding and of course calibration on shutdown once in a year.

The setting for servo current disagree suicide is 5%. I am still confused with the alarm and how control system will calculate current difference. In Mark IV there is no such alarm was given.

And again it is required to reboot controller to reset the alarm. After reset, in two to three month continuous running same alarm came again.
 
The controller calculates an amount of servo current that is to be applied to the coil and then compares that value to the actual current physically flowing in the circuit. If it differs by more than 5%, the alarm is annunciated.

It's interesting that the alarm "originates" on the same card in the same slot--the VSVO card in Slot 5.

You say you have carried out lots of actions. Have you changed the VSVO card?

Have you replaced the TSVO card?

What positions are the hardware (Berg) jumpers in on the TSVO card for the SRV and GCV outputs?

Are you certain the removable terminal boards on the TSVO cards are fully seated in the stationary bases? The barrier-style terminal boards used on TSVO cards can be removed and replaced, and they fit into a stationary base on the card. It's very important that the terminal boards are fully inserted in the bases, and it's also been found that sometimes the pins and sockets of the removable terminal boards/bases can be damaged and not make good contact. It takes a fair amount of force to fully seat the terminal boards (not so much that the TSVO board can be broken!) but be sure the removable bases are fully inserted. Use an inspection mirror and torch (flashlight) to check along the length of the two 24-pin removable terminal boards to see there is no gap visible between the terminal boards and the stationary bases on the TSVO card.

Have you changed the servo?

Have you verified the polarity of the current being applied to each individual coil?

Have you measured the resistance of each of the servo coil circuits <b>from the wires at the TSVO card</b>? (In other words, what is the entire circuit resistance, not just measured at the coil wires in JB25/25A?) What were the resistances of each coil circuit?

Have you meggared the servo coil circuit wires from the wires at the TSVO? If so, what were the results?

What is the exact value of null bias current that is being applied to each coil of each valve by the Mark VI (from the .m6b file)?

What is the gain being applied to each servo by the regulator (from the .m6b file)? If you've changed the regulator gain settings, what were they originally before you changed them?

Is the hydraulic pressure stable and at rated?

When you manually stroke the valves, what is the difference between the position reference and the LVDT feedback?

When you manually stroke the valves, what is the difference between the position reference and the actual physical position of the valve (not the LVDT feedback; the measured position of the valve opening)?

Most of the Mark IV-vintage Frame <b>6</b> gas turbines (not Frame VI) have the SRV and GCV servos mounted in an area which is effectively at the top of the L.O. tank, which means hot vapors can be present in the area where the servos are located. Temperature changes resistance, which can change current values.

Please be specific in your responses; don't just say we've checked that without providing the data (like the values of resistances of the servo coil circuits, and the null bias current- and regulator gain values). We can't help you without specific information. If you don't know the values or haven't obtained the information, just tell us. But, get the information because it's important to the understanding of the current configuration and the diagnosis of the problem.
 
Dear Tarun,
just to add a few things to CSA, since I have seen this same error a couple of times only, and in those cases it was due to wire problems. The MKVI basically is able to monitor the desired current to each servo, and if what it wants to see and what it actually sees differs by more than your disagree setpoint, then it will "suicide" that cores output.

This could be due to a problem with the VSVO card, TSVO board, wiring, servo, or configuration. I would compare the configuration of all servos, and look for any glaring differences in the settings. But since settings are for each regulator and not for each "core" it seems like a configuration problem should not be causing this issue.
First of all since yours is a retrofit, I would first be suspecting a wiring or configuration problem, but you say your retrofit was done in 2007. Has this problem really been occurring since 2007, or did it just start recently?

You say you have to reboot the processors to clear the alarm, this seems strange since a diagnostic/master reset should reset this alarm if the problem has cleared itself.

You say "TSVO termination card is common for all regulator." I question this since each TSVO board can only support 2 servo valves, so to control 4 servos (SRV,GCV,LFBV,and IGV) you need at least 2 TSVO terminal boards. Since you have 2 VSVO cards I would actually suspect you have more than 2 TSVO terminal boards since a VSVO card can support 2 TSVO boards.

Since the problem seems so intermittent I understand it will most likely be difficult to pinpoint the fault right away. Since you are getting errors on both VSVO cards it seems less likely it is a card failure, but anything is possible. It is even more interesting you are seeing errors on 3 of the 4 of the four servo's under control of the MKVI.

As CSA said. I would suggest checking all the wires from the TSVO card to the servo. Use the MKVI trender to monitor the individual servo currents from R,S, and T cores and see if you can catch the problem. Refer to GEI-100559 the MKVI system guide for good detailed descriptions of the VSVO card and TSVO boards. Hopefully you have good prints from the retrofit, I would study them and make sure all wiring is correct for each servo and the 3 coils from each.
Please write back and let us know what you are able to find. Hopefully by spending time with this you will be able to find the problem/or problems.
 
Sorry for delay in reply.

We have not replaced any card (TSVO / VSVO). Once we have replaced SRV servo because valve was not operating.

On TSVO card Jumper JP1 to JP6 are on position 10 ; i.e. +-10 mA Servo current Limit; and 1 KOhm Resistance of coil which is right.
There are three TSVO card. One is for GCV and SRV (R, S, T Slot 5), One is for Liquid fuel bypass servo (R, S, T slot 5) and one is for IGV (R, S, T slot 7)

Terminal boards on TSVO are fully inserted in the bases.
Polarity of servo current in R, S and T are same for both GCV and SRV.
We have measured servo coil resister from TSVO board by removing cable; it was around 1Kohm. Right now I have no value for same.
We have also meggared wire by removing coil wire and wire from TSVO TB. Found ok. No value is with me right now.

Gain for GCV - 2.2 and SRV - 2.0 %current/engineer unit. We have checked by increasing gain value 2.5 for both. But problem was not resolved.

Reg Null Bias GCV - 2.4 and SRV - 2.9 %current
Hydraulic oil pressure is constant around 82 barg
At present 1% difference in GCV position and reference (Ref = 56% and Position = 54.9%).

We have two LVDT for GCV, GRV and IGV. And Reg Type - 2_LVposMax is configured for all. Means maximum of Two is selected for feedback signal.

Further I want to add following.
At present Following diagnostic alarm appeared.
1. T controller Slot 5 : Fault code 47; Servo current#1 disagrees w/ ref, Suicided (regulator1)(SRV)

2. T controller Slot 5 : Fault code 47; Servo current#2 disagrees w/ ref, Suicided (regulator2)(GCV)

3. R controller Slot 7 : Fault code 47; Servo current#1 disagrees w/ ref, Suicided (IGV)

Now under this condition Servo Output current of T controller GCV, SRV and R controller IGV was displayed freeze in toolbox to last value. While measuring servo output current for same was zero actually. Same way position feedback of this T controller in GCV, SRV and R controller in IGV was displayed freeze to last value. Now when we press master reset and diagnostic reset button, sometime all above diagnostic alarm got reset. And some time one or two diagnostic alarm got reset. But here servo output current value was displayed freeze and same time actual value was also zero by measuring servo output voltage (zero). And same way position feedback was not updating and freeze as it is though diagnostic alarm reseted. Then after some time (especially during variation in valve position) again same all alarm appeared. We need to reboot the controller to reset this suicide.

Let me know for further data required.
 
Dear Turan,

thank you for the additional information. At this point without being onsite I think I would recommend trying a new VSVO card for "T" core. If you have truly verified all wiring for the servo valve and LVDT's then this might be the next logical step.

As I said before it seems somewhat unlikely to have two bad VSVO cards, but stranger things have happened. Hopefully by focusing on 1 problem at a time, in this case either choose slot #5 or slot #7. Now just to be perfectly clear, you have verified that all servos for each valve are 3 coil servo's? Now typically the TSVO cards are wired in this order, SRV,GCV,LFBV,and last IGV.

For the first TSVO board for slot #5 there are wires landed on terminal block screws #25-30 for the SRV? Same TSVO board, wires landed on screws #33-38 for the GCV? The wires for each servo coil have the same resistance?

Second TSVO board for slot#5 should have wires landed on screws #25-30 for the LFBV, all servo coils have the same resistance?
TSVO board for slot #7 should have wires landed on screw #25-30 for the IGV, all coil resistance's should be the same.
Your post says:

"1. T controller Slot 5 : Fault code 47; Servo current#1 disagrees w/ ref, Suicided (regulator1)(SRV)
2. T controller Slot 5 : Fault code 47; Servo current#2 disagrees w/ ref, Suicided (regulator2)(GCV)"

Servo current #2 for the GCV should be fault code #48, since I think fault #47 is just for servo #1.

Have you tried setting up a high speed trend (40ms)with the individual servo currents for a single regulator like the SRV when everything is operating normally to try and see what the individual servo currents look like?

Under the configuration of each servo output do the values look similar? What are the values for curr_suicide, Fdbk_suicide, and TMR_Difflimit?

Keep looking and report back when possible.
 
Hi,

Wondering if you could help with similar issue. Getting 'Servo Current #1 Open Circuit' 'Servo Current #2 Open Circuit' 'Servo Current #3 Open Circuit' with I view diagnostics on VSVO card. Voltage at terminals of DSVO card are 15VDC, 15mA and 1033 ohms. Only one coil on moog valve. Any suggestions. Thinking of changing card.
 
I

Instrument Engineer

Dear All

<b>Moderators Note: this is an update from the original post.</b>

We are having GE Frame VI gas turbine control by Mark VI control system TMR (R,S and T) having dual firing facility Gas as well as liquid. Control system was retrofitted from Mark IV to Mark VI Jan 2007. We have commission data historian in Jan- 2011

VSVO (version H1B) card is inserted in slot 5 is for SRV (Regulator1), GCV (Regulator2) and Liquid fuel bypass servo (Regulator3). At present we are using 100% gas fuel.

VSVO card is inserted in slot 7 is for IGV (Regulator1)

There are three TSVO (Version H1BCC) card. One is for GCV and SRV (R, S, T Slot 5), One is for Liquid fuel bypass servo (R, S, T slot 5) and one is for IGV (R, S, T slot 7)

Following diagnostic alarm appeared frequently. On TSVO card Jumper JP1 to JP6 are on position 10 ; i.e. +-10 mA Servo current Limit; and 1 KOhm Resistance of coil which is right.

We found problem with R & T controller diagnostic alarm Slot 5 VSVO("Servo current disagrees w/ ref, Suicided"). The phenomena is random. Alarm for servo slot 5 may come in R controller or T controller. It will require to reboot the controller to reset the fault. This problem we are facing since last two year.

We refer our case on 20/08/2011. We got R computer VSVO Slot -5 Diagnostic alarm ("Servo current#1 disagrees w/ ref, Suicided (SRV)") on 20/08/2011 at 9:10 AM.

While looking in historian trend, it is found that GCV, GRV & FAL Servo Current (FAG_NVR, FAGR_NVR, FAL_NVR) of R computer got freezed on 20/08/2011 at 5:20 AM before 4 hour of diagnostic alarm time. No alarm appeared at 5:20 AM. Position of GCV and GRV is also showing freezed in R controller. With freezed value displayed in m6b, R controller Servo output voltage for GCV, GRV and FAL was 0 VDC measuring with multimeter.

This same thing has happen many time R- controller as well as T- controller.

So with no diagnostic alarm how the value of GCV, GRV and FAL servo current got freezed? and we got suicide alarm due to this fixed value is compared with other controller value. With Master Reset and Diagnostic reset alarm not getting reset.

If problem in field side than with single servo problem will suicide that regulator only not all three regulator. But to clear all doubt, We have carried out following action/Checking

1. Terminal boards on TSVO are fully inserted in the bases

2. We have increase suicide current difference limit from 5% to 10%.

3. Polarity of servo current in R, S and T are correct for both GCV and SRV

4. We have measured servo coil resister from TSVO board by removing cable; it was around 1Kohm

5. We have also meggared all wire between field JB and control room (servo and LVDT) by removing coil wire and wire from TSVO TB. Found ok.

6. Gain for GCV - 2.2 and SRV - 2.0 %current/engineer unit. We have checked by increasing gain value 2.5 for both. But problem was not resolved.

7. Reg Null Bias GCV - 2.4 and SRV - 2.9 %current

8. Hydraulic oil pressure is constant around 82 barg

9. We have two LVDT for GCV, GRV and IGV. And Reg Type - 2_LVposMax is configured for all. Means maximum of Two is selected for feedback signal.

Please let me know for further data requirement. Please suggest to resolve long lasting issue.
 
I

Instrument Engineer

Dear CSA,

please help me out. The severity of problem is increasing.
Please let me know for any data requirement.
 
This machine is probably still under warranty.

What does the packager say about the problem?

I am entirely unclear about the current state of the problem(s). Use the Mark VI manuals and Toolbox help. MIKEVI has also provided some very useful information to which you have not fully responded.

Best of luck with your problem(s); please write back to let us know the successful resolution.
 
I have recently seen a similar problem on a site with multiple turbines that were all installed and commissioned at roughly the same time (within a couple of months of each other) that was ultimately traced to wiring problems during initial commissioning that caused the back-plane of the Mark VI processor rack(s) to weaken and fail. The failures were intermittent and increased with increased ambient temperatures in the Mark VI turbine control panel.

During extensive troubleshooting and discussions with personnel who were on site during the commissioning it was learned that servo wiring was improperly terminated in the Mark VI, and also in the field on this one particular unit, which lead to replacement of the VSVO cards and TSVO cards.

Unfortunately, the intermittent problems didn't start until several months after commissioning, and were not associated with control panel temperature until extensive troubleshooting started.

It's virtually impossible to tell from the information provided in this post exactly what the problem(s) at this particular site may be, or what might be causing them or what might have caused them. This information is being provided in attempt to introduce recently learned information from another site, which may--or may not--be applicable. (Replacing processor racks is expensive and time-consuming and may not solve the problem(s) being discussed in this thread.)
 
Dear Friends,

Recently we had a similar problem in IGV control, we are operating GE 7FA DLN 2.6 Gas turbines in co-generation mode,

Here is the extract from trip report

Unit-1 was running on baseload. At 04:52hrs GT tripped on exhaust temperature high. Subsequent alarms were Mk6 R/S/T controller Slot10 VSVO diagnostics. Rigorous fluctuations of IGV control servo current feedback R/S/T were noticed from DCS trends. IGV actual position & feedback difference started to increase. GT started to de-load from base to 135 MW automatically

During investigation it revealed that VSVO card sensed the IGV servo coil suicide in T and R controller simultaneously which caused the IGV malfunction. Due to malfunction in IGV control it closes from 83 deg to 70 deg which has reduced the amount of air flow though the gas turbine, eventually the turbine exhaust temperature went high and tripped the GT on high exhaust temperature as per GT protection logic.

My question is why GT did not trip on IGV command and feedback difference? as per the logic when difference between command and feedback increased to 5% it will give us alarm " IGV position trouble" alarm and if the difference increased to 7% GT will trip on " IGV position trouble" trip logic.

I believe there should have some time delay to operate protection logic on IGV position trouble, there should not be any time delay to operate the protection logic on "Exhaust over temperature" trip protection

What are the reasons behind the servo suicide? is there any oil quality issue (varnish buildup) or Servo v/v aging problem? please advice

Regards
KKK
 
The reasons for suiciding of servo-valve outputs are listed in the Toolbox Help files and the troubleshooting section of GEH-6421, the Mark VI System Guide. The descriptions, while not self-explanatory, do not include every contingency or auxiliary component. So they need to be reviewed in context with the system (LVDTs, servo-valves, oil condition/cleanliness, compartment temperatures, operating conditions, etc.) to work towards a likely cause.

We have no way of knowing the condition or cleanliness of the oil at your site, nor the operating conditions or the maintenance practices.

Have you checked the "last chance" filters upstream of the servo? What is the condition of the servo-valve filter?
 
Top