Frame 6B IGV Not at 34 DEG When Turbine Shudown

Our site has a Gas Turbine simple cycle GE Frame 6B MKV control system. During Turbine loaded at Base load every thing normal and IGV at 84 DEG. Turbine Trip Suddenly by IGV Control Trouble L4IGVT. investigate the signal result to trip. I get LK4IGVTX logic 1 when 14HS logic 1 result to trip Turbine. when check physically IGV after trip, we get the IGV not back to 34 DEG after turbine trip. IGV still fully open at 84 DEG without hydraulic oil pressure, and alarm L3IGVFLT, and Turbine not ready to Start. we turn IGV back by manual to 34DEG and release alarm by master rest. when start again and load to base load, IGV open normal till 84DEG. after one day happened again trip same last story. i was check LVDT and Servo, all feedback CSVG. CSRGV correct.

Appreciate if any guide for solution this problem.
 
SamySam,

What load was the unit running at when it tripped?

What Process Alarms were present and active just prior to the trip? EXACTLY what alarms were annunciated at the time of the trip--ALL ALARMS, and their times, please.

What Diagnostic Alarms were present and active just prior to the trip?

Was the load changing just prior to the trip?

Does the unit have conventional (diffusion flame) or DLN-I (Dry Low NOx I) combustors?

A simple cycle GE-design heavy duty gas turbine with conventional combustors will typically load as follows: At synchronization, the IGVs will be at approximately 57 DGA, and will remain at 57 DGA as the unit is loaded up until the exhaust temperature reaches 700 deg F or 900 deg F (I forget the exact value--but it's one of the two). At this point if the load is increased the exhaust temperature will increase above the (700 deg F or 900 deg F) setpoint. BUT, the Mark V will open the IGVs to maintain the (700 deg F or 900 deg F) setpoint. As the unit is loaded further and the exhaust temperature tends to increase as the fuel flow-rate is increased the Mark V will continue to open the IGVs to maintain the (700 deg F or 900 deg F) setpoint. At some point, the IGVs will reach the maximum operating angle (which seems to be 84 DGA at your site). As the unit continues to be loaded, the exhaust temperature will continue to increase--but the IGVs are already at maximum operating angle.

Once the actual exhaust temperature, signal TTXM, reaches the CPD- or CPR-biased Exhaust Temperature Control Reference (Limit), signal TTRX, the unit is deemed to be at Base Load, on exhaust temperature control.

Something similar happens during unloading and shutdown. The IGVs will remain open until the (700 deg F or 900 deg F) setpoint is reached, and then as fuel is decreased to unload the unit the exhaust temperature will start to decrease--but the Mark V will close the IGVs to maintain the (700 deg F or 900 deg F) setpoint, until the IGVs return to 57 DGA. At that point as fuel is reduced further, the IGVs are already at minimum operating angle and can't close any further (because the unit is at rated speed), and exhaust temperature will decrease as fuel is decrease as the unit is unloaded.

If the unit has DLN-I combustors, we can talk about what happens then; it's a great deal more complicated.

The IGVs are hydraulically operated, using an electro-hydraulic servo-valve which is controlled by signals from the Mark V to port high-pressure hydraulic fluid to one side or the other of a double-acting piston which moves a ring around the inlet casing to change the angle of the IGVs (it's a rack-and-pinion type of mechanism, with a large rack and multiple pinions on the shafts of the IGVs).

Without being able to see the CSP (Control Sequence Program) running in the Mark V at your site it's impossible to say for certain what is actually causing the turbine to trip, but it is possible to say, that if the IGVs were at 84 DGA when the turbine tripped it was at or near Base Load when it tripped. Which means the IGVs were at maximum operating angle. What we don't know is if the exhaust temperature was close to the setpoint (700 deg F or 900 deg F) where any reduction in load/fuel flow would have tried to make the IGVs close.

GE heavy duty turbine control systems have a nasty aspect: When a turbine trips, instead of blocking all "subsequent" trip conditions from alarming and only displaying the original cause of the trip, they show almost all trip conditions which might also occur after one condition actually results in the turbine trip. And, this causes MANY sites and people to totally misread and misunderstand the Alarm Display and the Alarm Log, because there might be two, or five, or six trip alarm messages--and they always think the LAST one in the list is the "original" cause of the trip, when, it's actually the FIRST one in the list that's the "original" cause of the trip.

Here's what could be happening at your site. The turbine is tripped (let's say it's because of a high exhaust temperature spread). And, as the unit is coasting down the IGVs fail to move back to their closed position--this would result in a alarm message to the effect that the IGVs had caused a turbine trip, but it was already tripped. Instead of blocking that alarm message that the IGVs failed to move after the turbine tripped, GE allows the message to be displayed (after the original trip) causing a great deal of confusion for many people.

The other condition that might be happening is that the IGVs are trying to close as the unit is being unloaded but they aren't moving or aren't moving fast enough to satisfy the IGV error-checking situation. Usually, there is sequencing to check to see that the IGVs are following the reference, and if not, there is usually an alarm to indicate they are not following the reference, and if the error between the actual IGV position and the IGV reference (setpoint) gets too great then the unit is tripped--on IGV trouble. But, we don't know what load the unit was at when it tripped, or what was happening when the unit tripped, or what alarms were annunciated and active prior to the trip.

You seem to have a couple of problems. First, the IGVs don't seem to want to move. You say you "manually" moved the IGVs back to the 34 DGA position. Why didn't you use the AutoCalibrate Manual positioning function to hydraulically move them back to the closed position?

You could have also used the Manual positioning function of AutoCalilbrate to "stroke" the IGVs to see if they move freely open and closed in response to position setpoint changes. This would help to understand what might be happening when the IGVs are NOT closing after a trip, if they don't freely open and close in response to signals from the Mark V.

There is a solenoid valve, 20TV-1, which, when energized (as it is during normal running operation) allows high-pressure hydraulic oil to flow to the double-acting IGV actuator piston to move the IGVs open and closed. When the unit trips or is shut down, the solenoid is de-energized, and it's supposed to allow high-pressure hydraulic oil to CLOSE the IGVs by porting it to the appropriate side of the double-acting piston of the IGV actuator. And, even if the solenoid wasn't working properly the position reference from the Mark V to the servo-valve should also be trying to port high-pressure hydraulic oil to the IGV actuator to close the IGVs. Neither of which seems to be happening, which is quite strange.

There are a couple of P&ID drawings which should help explain how the IGVs work: Trip Oil, and IGVs. Sometimes, depending on the age of the unit and the packager, the two are somewhat combined, but find those drawings and you can see how the flow of hydraulic oil is used to position the IGVs.

But, the fact remains: We don't have enough information to understand what was happening at the time of the trip, AND we need to have ALL of the alarm information from the time prior to the trip AND AFTER the trip, including times (in chronological order) to be of any further help.

In my personal estimation, either the unit is changing load and the exhaust temperature is near the setpoint at which the IGV angle should start changing and because the IGVs are NOT actually moving the unit is tripping on IGV NOT FOLLOWING REFERENCE. OR, the unit is tripping on some other condition and because the IGVs won't close after the trip the Mark V is annunciating an IGV trip, but the unit wasn't tripped for the IGV problem is was already tripped because of a different problem and the nasty propensity for the Mark* to annunciate any and all trips after the "original" trip is causing confusion--though in this case, the IGVs are reportedly actually NOT closing so that is at least one problem.

Something seems to be amiss in the hydraulic system controlling the IGV position. It could be mechanical binding, or it could be a hydraulic issue (broken tubing; problem with 20TV-1 or the dump valve it operates; etc.).

But, to be of any further help you need to provide a chronological list of alarms from before and after the trip, including Diagnostic Alarms!!! Without that, and a basic description of what was happening with the unit load prior to the trip, we can't do any more to help you.

Also, if the unit has DLN-I combustors, the IGVs are moving a LOT during loading and unloading. But, we need to know more about the unit and its combustion system.
 
CSA,

>What load was the unit running at when it tripped?
Unit at base load 31 MW exhaust temperature 565 C IGV under Temperature Control Opening at 84 DEG.

>What Process Alarms were present and active just prior to
>the trip?
no any alarm before unit trip

>EXACTLY what alarms were annunciated at the time
no alarms at the time

>the trip--ALL ALARMS, and their times, please.
08:32:38.156 GT1 1* Q 0141 T* IGV-IGV CONTROL FAULT
08:32:43.218 GT1 0* Q 0142 A IGV CONTRL FAULT
08:32:43:218 GT1 1* Q 0143 A*IGV-IGV POSITION FAULT

>What Diagnostic Alarms were present and active just prior to
>the trip?
Diagnostic Alarms
23:38:50.000 GT1 1 R 1282 TCQA THERMOCUPLE TC3 FAILD
10:26:16.000 GT1 1 T 1742 TCE1 POWER SUPPLY OUT OF LIMITS
11:55:38.000 GT1 1 R 1742 TCE1 POWER SUPPLY OUT OF LIMITS,
11:55:38.000 GT1 1 R 1736 TCE1 POWER SUPPLY OUT OF LIMITS,

>Was the load changing just prior to the trip?
load no change steady at 30.5 MW this base for this Turbine with inlet temperature and ambient temp 36 C CTD 358 C CPD 9 bar

>Does the unit have conventional (diffusion flame) or DLN-I
(Dry Low NOx I) combustors?
Unit have only Diffusion
unit was load at base load normal no any alarm previous the trip
the alarm sequence regarding time

1- first alarm 142 (L86GVA INLET GUID VAN CONTROL TROUBLE ALARM)
this alarm was logic change from 1 to 0 two times exist and clear it Self.

2- Second unit trip by Alarm 141 (L4IGVT INLET GUID VANE CONTROL TROUBLE).

3- Third IGV not back to 34 DEG with L20TV1X logic 1 and no hydraulic pressure found. This result to Alarm 143 (L3IGVFLT INLET GUID VANE POS SERVO TROUBLE) Turbine not ready to start. To move IGV should have hydraulic pressure. we need to start the Gas Turbine at Cranking Mode 1000 RPM by diesel engine. Means the Gas Turbine should be ready to start and IGV not at 34 DEG result to start check 0 (L31STCK0). Logic 0. then "Turbine not ready to start."
 
SamySam,

During Gas Turbine Trip the Trip Oil by SOL Valves L20TV1 and L20FG1 turn to Logic "1" by Mater protective. and drain all Hydraulic pressure. in this case should all valves drive by hydraulic should be stop. but in this case very strange. Turbine trip and IGV still at 84 DEG. the mechanical parts should be check and test to be confirmation for any stuck or fraction more to prevent IGV easy moving. of course this should be done after check the LVDT and Servo.
 
SamySam,

There are a LOT of problems with your replies.

I can't draw (very well) here on control.com, but I'm going to try. I also don't have access to the CSP running in the Mark V at your site, but I have a good deal of confidence that the CSP I'm referencing (which is for a GE-design Frame 6B heavy duty gas turbine) is very similar, if not exactly the same, as the CSP in the Mark V at your site. So, here goes....<pre>
--------------
| |
(CSGV -CSRGV)---|A |
| |A|>B--td-|---L86GVA
LK86GVA1-----|B | |
| | |
LK86GVA2-----|-------- | |
| |
--------------</pre>
What the above logic "says" is: IF THE ABSOLUTE VALUE OF THE DIFFERENCE BETWEEN THE IGV LVDT POSITION FEEDBACK (CSGV) AND THE IGV POSITION REFERENCE (CSRGV) IS GREATER THAN LK86GVA1 (5.0 DGA) FOR MORE THAN L886GVA2 (5.0 SECONDS) THEN SET LOGIC SIGNAL L86GVA TO A LOGIC "1". And when L86GVA is a logic "1" the Process Alarm "IGV CONTROL TROUBLE ALARM" is annunciated.

That means that if the reference is more than the LVDT feedback by more than 5.0 DGA for more than five seconds, <b>OR</b> the LVDT feedback is more than the reference by more than 5.0 DGA for more than five seconds, then the alarm "IGV CONTROL TROUBLE ALARM" will be annunciated. That's a LOT of difference (5.0 DGA)--and it's hard to understand how that much difference could exist when the unit was running at Base Load with the IGVs at maximum operating angle.

The maximum IGV operating angle of a GE-design heavy duty gas turbine should be one or two DGA less than the full open mechanical stop, Or said another way the fully open mechanical stop is usually approximately two DGA greater than the maximum operating angle. So, let's say the maximum mechanical stop was 86, or even 87 DGA, if the reference was 84 DGA and the actuator couldn't control the IGVs they should only go to 86 or 87 DGA (in our example)--which is LESS THAN a 5.0 DGA difference. (Typically, GE-design Frame 6B heavy duty gas turbines have IGVs designed such that when the axial compressor is turning the air flow through the IGVs will try to OPEN the IGVs, so the Mark V and actuator are really acting to keep them from opening to the full open mechanical stop. That means that if the actuator was unable to control the IGVs they would only go to 86 or 87 DGA (in our example)--which, if the unit was operating at Base Load with the IGVs at 84 DGA would mean the IGV CONTROL TROUBLE ALARM would not be active.

That means that for this alarm to be active the reference must have changed to something less than (84.0 DGA - 5.0 DAG = 79.0 DGA), which doesn't seem very likely either--while the unit is running and BASE LOAD was selected and active.

Now for the "IGV CONTROL TROUBLE TRIP" Process Alarm.<pre>
--------------
| |
(CSGV -CSRGV)---|A |
| A>B---td--|---L86GVT (Latched & Reset by
LK86GVT1-----|B | | L86MR1_CPB)
| | |
LK86GVT2-----|---------| |
| |
--------------</pre>
The above logic says: IF THE DIFFERENCE BETWEEN THE IGV LVDT POSITION FEEDBACK AND THE IGV POSITION REFERENCE EXCEEDS LK86GVT1 (7.5 DGA) FOR MORE THAN LK86VGT2 (5.0 SECONDS) THEN SET AND LATCH THE LOGIC SIGNAL L86GVT TO A LOGIC "1". (And, L86GVT can only be reset by a MASTER RESET if the difference between CSGV and CSRGV is less than 7.5 DGA.) That means that the IGV LVDT position feedback (CSGV) would have to be 5.0 DGA more than the IGV position reference (CSRGV) for more than 5.0 seconds. This means the IGVs were open more than 5.0 DGA above the IGV position reference. And, with the mechanical stop set at 2 DGA, or even 3 DGA, above the maximum operating angle, 84 DGA, it's difficult to understand how the IGVs could have been that far open.

L86GVT is <b>one</b> of two conditions which can cause the Process Alarm "IGV CONTROL TROUBLE TRIP". Here is the rung for the Process Alarm "IGV CONTROL TROUBLE TRIP":<pre>
L86GVT L14HS L4IGVT
----| |-------|/|------------------------------( )
|
L4IGVTX |
----| |--------------
|
L86MR1_ |
L4IGVT CPB |
----| |-------|/|----</pre>
The above rung says: IF L86GVT IS A LOGIC "1" AND L14HX IS <b>NOT</b> AT LOGIC "1" <b>OR</b> IF L4IGVTX IS A LOGIC "1" THEN SET LOGIC L4IGVT TO A LOGIC "1" AND LATCH IT TO A LOGIC "1" WHICH CAN ONLY BE RESET WHEN EITHER OF THE ABOVE CONDITIONS IS NOT TRUE AND MASTER RESET IS PRESSED. So, because the unit was (probably) at 14HS (which means the turbine-generator shaft speed was above 95% of rated--which we DON'T know for sure; the grid frequency may have been very unstable!) then even if L86GVT was a logic "1" (meaning that the IGVs were more than 7.5 DGA above the reference) that would NOT have resulted in a trip (because the normally closed L14HS contact would have been OPEN).

Let's now look at L4IGVTX:<pre>
--------------
L14HS | CMP |
---| |----------|--------- |
| |
LK4IGVTX------|A | |
| | |
| A>B--| |--|----L4IGVTX
| |
CSGV------|B |
| |
--------------</pre>
The above rung says: IF THE VALUE OF LK4IGVTX (37.0 DGA) IS GREATER THAN THE IGV LVDT POSITION FEEDBACK <b>AND</b> THE UNIT IS ABOVE 95% SPEED (L14HS IS A LOGIC "1") THEN SET LOGIC L4IGVTX EQUAL TO A LOGIC "1". So, first of all , L4IGVTX can ONLY be set to a logic "1" if the unit is above 95% speed. And the value of IGV position feedback (CSGV) has to be less than LK4IGVTX (37.0 DGA).

Let's go back to L4IGVT:<pre>
L86GVT L14HS L4IGVT
----| |-------|/|------------------------------( )
|
L4IGVTX |
----| |--------------
|
L86MR1_ |
L4IGVT CPB |
----| |-------|/|----</pre>
L4IGVT can only be a logic "1" when L86GVT is a logic "1" AND L14HS is a logic "0" <b>OR</b> L4IGVTX is a logic "1"--and L4IGVTX can only be a logic "1" when L14HS is a logic "1".

>08:32:38.156 GT1 1* Q 0141 T* IGV-IGV CONTROL FAULT
>08:32:43.218 GT1 0* Q 0142 A IGV CONTRL FAULT
>08:32:43:218 GT1 1* Q 0143 A*IGV-IGV POSITION FAULT

The above alarm list cannot be correct. It's possible that at 08:32:38.156 Alarm Drop 141 went to a logic "1", and that at 08:32:43.218 Alarm Drop 141 went to a logic "0". Alarm Drop 143 is usually "IGV POSITION SERVO TROUBLE"--which is driven by logic L3IGVFLT, which is a logic "1" when the IGV LVDT position feedback is less than 32.0 DGA OR when the IGVs LVDT feedback is greater than 37.0 DGA when the unit is below L14HA (which is usually approximately 50% speed for a GE-design Frame 6B heavy duty gas turbine).

Alarm Drop 143 is a start-check permissive, but not a trip--at least not in the CSP I'm looking at.

I still maintain after the above review and based on the information provided that something else tripped the turbine and only because the IGV actuator couldn't close the IGVs were the IGV-related alarms/trip and start-check permissives set.

Did the operators print the Trip History when the unit tripped? That would be VERY helpful if they did. If they did, let us know and we can tell you how to find out exactly what tripped the unit. (We also need to know what kind of operator interface is being used--an <I>, a GE Mark V HMI running TCI, or a GE Mark V HMI running ControlST and WorkstationST. If it's the latter (an HMI running ControlST/WorkstationST it's possible to use the WorkstationST Alarm Viewer to look at past alarms (in a range) to see what happened (from an alarm perspective). And, actually, if the unit has a GE Mark V HMI running TCI you can use Microsoft Internet Explorer to also search backwards through alarms (in a range).

Last, let's talk about L20FG1X and L20TV1X. Both of these must be a logic "1" to run the turbine, and if they go to a logic "1" then the turbine will be tripped. That's the way GE-design Mark* turbine heavy duty gas turbine control systems work: They energize the fuel stop solenoid (L20FG1X is the signal that energizes the Gas Fuel Stop/Ratio Valve solenoid to allow gas fuel to flow to the unit. L20FL1X is the signal that energizes the Liquid Fuel Stop Valve to allow liquid fuel to flow to the unit. If either signal is a logic "0"--the fuel stop valve for that fuel will be CLOSED and no fuel will flow to the unit.

Which brings up a good point.... The Mark V has a set of primary trip relays (called the PTRs) and a set of emergency trip relays (called the ETRs). BOTH the PTRs <b>AND</b> the ETRs must be energized for the fuel stop valve(s) to be open. If the PTRs go to a logic "0" then the fuel stop valve(s) will close. <b>OR</b> if the ETRs go to a logic "0" the fuel stop valves will close. The ETRs are driven by the TCEA cards.... AND there are Diagnostic Alarms associated with two of the TCEA cards! Which could mean that because of issues with the power supplies of two of the TCEA cards that the ETRs are dropping out (going to a logic "0") and causing the unit to trip, and THEN the IGV-related alarms are being annunciated (because of the nasty "habit" that Mark* control systems have). Now, if that happened (the ETRs dropped out and the fuel stop valves closed and flame was suddenly lost in the unit) what SHOULD happen is that the alarm "LOSS OF FLAME TRIP" should have been annunciated. But, according to the alarms provided that alarm didn't occur, so it's not likely the ETRs tripped.

I have read and re-read and re-re-read the above and hopefully I have corrected all the errors and typographical errors and mistyping so that it's correct and clear (as can be). Unless the turbine speed was fluctuating pretty wildly AND the underspeed protection of the Mark V and the generator protection relays has been seriously tampered with it's very difficult to understand how the turbine could have been tripped on IGV troubles when running stably at Base Load.

One thing which MANY people overlook when troubleshooting "unusual" problems like this is wiring problems. It's possible that in the IGV servo wiring (between the Mark V and the IGV servo valve) and/or in the IGV LVDT wiring (between the Mark V and IGV LVDTs) there is one or more loose wires or bad wiring crimps. This is one of the easiest things to check and often the most overlooked. The terminal screws DO NOT have be gorilla tight--they just have to be snug and firmly tightened (not over-tightened). It's also possible that crimp-on terminal lugs--especially on servo-valve and LVDT wiring (which is usually very small gauge wiring) can be loose. An light tug on the wires can often reveal loose crimps. So, open all the intermediate junction boxes for the circuits (IGV servos and IGV LVDTs) and check all the wiring for loose crimps and loose terminals, and also at the Mark V. (Overtightening terminals on the Mark V terminal board has caused small gauge wiring to be cut by the terminals.... And, it has also led to an inability to ever remove the wire from the terminal(s) again, forcing a replacement of the terminal board.)

I hope this helps. It's a LOT of writing and some difficult "drawing" but it's how a typical GE-design Frame 6N heavy duty gas turbine Mark V operates the IGVs and how the protection sequencing works. The very definition of Base Load is: The IGVs are at maximum operating angle, and the actual exhaust temperature is equal to the CPR- or CPD-biased exhaust temperature control reference. And, usually when at this condition the operation of the unit is fairly stable--UNLESS the grid frequency is NOT stable, and then all sorts of untoward things can and have happened. If the grid was unstable at the time of the trip(s), then that could definitely be part of the problem.

If the operators at your site don't already print the Trip History WHENEVER the turbine trips (even if it's certain the reason for the trip is known)--they should start doing so. The Trip History can be VERY helpful, once one learns how to read it. And, that's not difficult to understand or to learn. But, without it, and without a printout of the alarms prior to and after the trip (which are in the Trip History display--at least the most recent 128 of them!) it's virtually impossible to determine from afar precisely what happened.

It is possible to say--again--the GE Mark* turbine control systems WILL annunciate any and all trip conditions which occur AFTER the turbine was tripped for another condition. And that makes it unnecessarily difficult to determine exactly what tripped the turbine. A working Alarm Logger printer is essential to troubleshooting a turbine--or someone who is familiar with searching a GE Mark V HMI which has been properly configured to store alarms on the hard drive. But, since GE didn't always put the Alarm History files in the same locations on the hard drives, and because they also didn't properly enable the logging of alarms to the hard drives on every HMI, it's hit or miss when trying to troubleshoot without <b>printed, hard copies</b> of alarms, and the Trip History display. The Trip History isn't the greatest tool for troubleshooting, but in the absence of anything else it can be pretty darn good. And, at a minimum, it has a fairly good list of alarms prior to and for a short time after a unit was tripped. As well as a small list of operating data.

It's STILL VERY ODD that after the trip the IGVs did not go back to the closed position--because when L20TV1X goes to a logic "0" that should allow high-pressure hydraulic oil to get to the actuator to move the IGVs to the closed position. L20TV1X is a logic "1" when L4IGVT is a logic "0" AND L14HR is a logic "0"--so that means L20TV1X should be logic "1" any time the unit is above zero speed <b>AND</b> there is not "IGV CONTROL TROUBLE TRIP". Or, if the unit is running and there IS an IGV CONTROL TROUBLE TRIP alarm L20TV1X should go to a logic "0" which should allow high-pressure hydraulic oil to move the IGVs to the closed position. (I have personally NEVER seen a Frame 6B GE-design heavy duty gas turbine without a 20TV-1, driven by L20TV1X. I also haven't seen every Frame 6B every built. But, it's pretty important that the IGVs close during a trip, because it's part of the axial compressor protection.)

Thanks for reading this far! Please write back to let us know what you find.
 
SamySam,

It's worth noting that on most GE-design Frame 6B heavy duty gas turbines the IGV actuator, and associated LVDTs, are located in a very difficult to access location--under the bellmouth at the 6 o'clock position. For this reason, the actuator and LVDTS get very little love, or care, or maintenance.

It's also worth noting that MANY of these actuators have failed in various ways over time with little or no maintenance. One of the more common methods of failure is that the bolts that hold the actuator stationary on the crossbeam under the bellmouth loosen and actually wear. The bolts can become very thin, and the holes in the actuator plate can become enlarged, and both can contribute to a lot of slop in the IGV action.

There is also a very large Heim joint on the actuator, and, again, because of the location, it doesn't get the lubrication it should and wears out--also causing slop in the IGV action. It has also caused binding and sticking in some cases.

You said you manually moved the IGVs to the closed position--can you describe how this was done?

Has anyone had a look at the IGV actuator--and the LVDTs attached to the IGV actuator?

Has anyone compared the IGV LVDT feedback to the pointer on the side of the axial compressor casing? (The pointer is rarely, if ever, adjusted during or after maintenance outages, and it usually gets stepped on and used as a step during maintenance--so it may not be very accurate, but it's better than nothing.)

The LVDTs are attached to the actuator and measure linear movement of the actuator. The actuator is connected (via the Heim joint) to the ring gear which surrounds the axial compressor casing. So, the LVDTs don't actually measure IGV position, but since the linear stroke of the actuator is directly proportional to the IGV angle it's a very good indicator of IGV angle. When the actuator and/or Heim joint get worn the IGV LVDTs can be affected, and one of the things that happens is that the movable core of the LVDT can scratch the interior of the stationary armature of the LVDT, and when that happens the feedback from the LVDT can be erratic when the core is in the area of the damaged armature.

Since the Mark V looks at the feedback from two LVDTs (usually) and selects the higher of the two, when the feedback from one is erratic when the LVDT armature is damaged that can cause erratic operation. I've even seen the insides of both LVDTs damaged because of damaged actuators or misalignment of the LVDTs during installation or replacement.

It seems you don't have an AC motor-drive Auxiliary Hydraulic pump to use to stroke devices, such as fuel control valves or the IGVs. That's not unusual--and many sites have to crank the unit to develop hydraulic pressure in order to use the manual positioning feature of AutoCalibrate to stroke devices. In this case, I think it's very warranted and necessary--you need to know what is causing the IGVs not to close after a trip (whatever caused the trip...). I would suggest after starting the unit in CRANK mode to use the manual positioning feature of AutoCalibrate to open and close the IGVs two or three times while observing the operation of the IGVs (meaning someone physically watching the IGV ring and pointer on the side of the axial compressor casing). I would also suggest after the IGVs are opened to the fully open mechanical stop to start closing them in steps, starting with 84 DGA, then 75 DGA, the 65 DGA, then 57 DGA, then 45 DGA, then 34 DGA, and recording BOTH the indicated LVDT feedback from the AutoCalibrate display (for BOTH LVDTs!) and the indicated position from the pointer on the side of the axial compressor casing.

After that, I would open the IGVs to 84 DGA and then force L20TV1X (that's the typical name of the solenoid output that energizes/de-energizes the IGV solenoid, 20TV-1). At this point the IGVs should close to the fully closed mechanical stop. If they don't, I'd be looking VERY CLOSELY at the Trip Oil and IGV P&ID drawings, as well as considering pulling the IGV solenoid and inspecting it, as well as looking at the IGV dump valve (sorry; I don't recall the device number and I don't have any P&IDs at this writing). They could be gummed up with varnish and need cleaning.

If the hydraulic/trip oil systems on your unit are typical of most GE-design Frame 6B heavy duty gas turbines when 20TV-1 is energized it should be possible to open/close the IGVs using AutoCalibrate. And, if the IGVs are open and 20TV-1 is de-energized, the IGVs should move to the fully closed position. This should be verified, and if it doesn't operate this way then either the system on your unit is atypical (not impossible, but also not likely), OR there is something amiss with some component of the system.

And, don't rule out a worn actuator causing mechanical binding. Usually, there is an inspection "door" (plate) in the bottom of the inlet plenum ductwork which can be removed to look at the IGV actuator from above (without having to crawl under the unit to get to the actuator's location). It's not the best access, but if someone can also be watching the actuator when stroking the IGVs with AutoCalibrate, even using a smartphone camera to video (record) the movement, that would be a very helpful thing for troubleshooting.

So, here's a few things that can be done to further investigate what might be happening.

Hope this helps! Please write back to let us know how you progress in resolving the problem.

If the unit trips again, it would be EXTREMELY helpful if you could get the operators to print the Trip History Display BEFORE they issue a MASTER RESET. If you have an older operator interface, please provide the top 20 (twenty) alarms from the alarm list, and dates/times of each alarm. If you have a newer operator interface (a GE Mark V HMI running ControlST/WorkstationST), please provide the bottom 20 (twenty) alarms from the alarm list, and their dates and times.

Also, the data should be divided into sections called POST and PRE (or something like that). If you could provide the dates and times of the POST data rows (just the dates and times), that would also be helpful.
 
SamySam,

A couple of people have written me to say I was not very specific about L20TV1X and L20FG1X.

In the test I suggested to check the operation of the IGVs/IGV actuator/IGV LVDTs, when the unit is CRANKing you will find L20TV1X a logic "1", and to test the IGV actuator to see if it will close on a turbine trip you will need to force it to a logic "0" in order to de-energize 20TV-1. (When you have completed the test, you will need to unforce L20TV1X.)

L20FG1X is the signal from <R>, <S> and <T> that drives one set of PTRs (Primary Trip Relays). It controls the application of -125 VDC to the Gas Fuel Stop/Ratio Valve solenoid, 20FG-1. In the philosophy used by GE for control and protection of their heavy duty gas turbines the fuel stop valve solenoid must be energized to allow fuel to flow to the turbine combustors, and when the fuel stop valve solenoid is de-energized the fuel is very quickly shut off. The fuel stop valve solenoids are energized when BOTH the PTRs and ETRs (Emergency Trip Relays) are energized. If EITHER the PTRs or ETRs are de-energized, the fuel stop valve solenoid will be de-energized and the fuel flow to the turbine combustors will stop very quickly. And, in the Mark V when the logic signal L20FG1X is a logic "1" the PTRs should be energized and applying -125 VDC to one side of the fuel stop valve solenoid. (The ETRs apply +125 VDC to the other side of the fuel stop valve solenoid.)

Neither L20FG1X nor L20TV1X should be forced to a logic "1" to allow the turbine to run OR to keep the turbine running. That's EXTREMELY dangerous (along with forcing many other signals). The signals can be forced when the unit is NOT running, <b>BUT</b> before forcing L20FG1X (or L20FL1X) the fuel supply(s) to the turbine should be manually isolated to prevent any fuel from flowing to the turbine.

Hope this helps to clarify how 20FG-1 (and L20FG1X) and 20TV-1 (and L20TV1X) work. Again, please write back to let us know how you fare in resolving this issue. Many people read these threads--now and in the future--and all are interested in learning what was discovered and how the problem was resolved. If you need more assistance, you will need to provide the information requested.
 
CSA,

I would like to deep thanks for you and appreciate about the detail description for control flow against the problem. this turnip my direction to analyses the diagnostic alarms it wasn't care before from my site.

You have right the most operators when alarms came make Master Rest and release the alarms. our MKV <I> not have Event alarm Historian. unfortunate the Alarm printer not work in this time. the Trip Log Display register last event whatever trip or normal shutdown.

- (First Step has been done) the condition to start the turbine IGV should less than 35 DEG .this get the (L31STCK0)start check logic 1 and can start the turbine. our turbine starting means by Diesel Engine.so the hydraulic oil pressure build up by gear pump when starting diesel Engine at cranking speed 1000 RPM.The Gas Turbine should be Ready to start. We Turn the IGV manually by Push down the gear ring by two men and one column arm catch the gear ring and push down till reach the Pointer at 33 DEG and Master Rest to release alarm L3IGVFLT and can start the Turbine normal.

- (second step) Turbine start normal and at Cranking mode The IGV regulator Checkout .establish L20TV1X = 1 and L4_XTP=0 and from IGV page we start opening 50 DEG I get the feedback and matching with the pointer 50 DEG Exactly .rebate at 75 DEG ,86 DEG. and back to closed 75 DEG ,50DEG,34 DEG the IGV normal every thing normal. we confirmation IGV normal 100%. the trip happened from other thing not discover yet.

- (Third step)I check carefully the diagnostic alarms as you detail list. its seem the TCE1 has power supply out of limits,P5,P335 and DCOM for R,S and T alarms TCE1 (1736,1743,1736,1742).

but if the Turbine trip done by trip relays PTRs and set ETRs no any alarm indicate?. I am still check how can adjust or clear the TCE1 power supply. but the reasons for IGV not closed at Turbine trip still under analyses and check.
 
>but if the Turbine trip done by trip relays PTRs and set
>ETRs no any alarm indicate?

Under normal circumstances, if the PTRs (which are on the TCTG card in <P> with the TCEA cards) are the cause of the trip there will be an alarm (presuming the machine was configured and commissioned correctly)--something like high-high vibration, or low-low L.O. pressure, or exhaust overtemperature trip, etc. If the ETRs trip the turbine, there will be only a couple of alarms--and both will be speed-related. Either an Emergency Overspeed Trip, OR, a rate-of-change Emergency Overspeed Trip. So, either at least two of the three TCEA cards detect an overspeed, OR the rate of change of at least two of the three TCEAs detected a rate-of-change of their emergency overspeed trip inputs exceeded a limit. This usually happens when the signals are intermittent, changing very quickly from one value to another, such as from 5100 RPM to 0.

<b>However,</b> if the TCEA power supplies are not working correctly then the PTRs and ETRs (which are also located on the TCTG card) might be affected, and in that case there may be no alarms--because the power supply(s) to check the alarms might be too low.

I don't have a Diagnostic Alarm Help file to look at as of this writing, but it's also possible that the TCEA power supply alarms are related to the flame detectors. Yes; the flame detection is handled in the <P> core. If the flame detector power supply(s) are affected, the unit could be tripped on "LOSS OF FLAME TRIP" (it's a Process Alarm, that affects the PTRs). The alarm confuses many people--because what it means is that flame was lost BEFORE any other alarm condition was detected that would result in trip (low-low L.O. pressure; high-high vibration; exhaust overtemperature trip; etc.). If the fuel stop valve closes without being told to close--or fuel supply pressure/flow-rate falls below what is required to maintain flame--flame will be lost before any other trip condition is detected. If the fuel supply pressure/flow-rate drops low what USUALLY happens is there is an alarm, and load will drop as fuel supply pressure/flow-rate drops. The fuel control valve will usually open fully to try to maintain load. But as fuel supply pressure/flow-rate drops the load will drop and eventually flame will be lost. (Speed won't change if the grid is stable.)

>but the reasons for IGV not
>closed at Turbine trip still under analyses and check.

The "standard" is that every condition that results in a turbine trip will have its own Process Alarm. BUT, occasionally some times the factory programmer makes a mistake--which should be caught by the commissioning field service person. Or, a trip is added by commissioning field service person, or some other field service person, and they didn't add a Process Alarm.

Look, I know dot matrix printers seem like old technology, and they can require lots of maintenance (printer ribbons; tractor feed paper, that get's skewed; and a LOT of paper is used)--but, it's critical to operation and troubleshooting. Critical. CRITICAL. <b>CRITICAL.</b>

Some sites have been able to connect the Alarm Logger (Printer) output to another computer and which can save the alarm strings to a hard drive. That's an option you can (should) look into, if you don't want to or can't maintain the Alarm Logger dot matrix printer.

<b>OR,</b> you could leave the Alarm Logger turned off until there is a turbine trip. You would then clear the Alarm Logger cache/list by going to a command prompt and typing the command:<pre>
LOGGER FLUSH</pre>and then turning on the Alarm Logger and printing the Trip History Display.
 
CSA,

During Investigation the TCE1 in core <R> ,<S>, and <T>. and the cause of Diagnostic alarms;

//<R>,and <T> 1742 ;TCEA POWER SUPPLY OUT OF LIMITS,P335

this alarm Cause by flame detector (A) not connected. The loss one P335 supply should not result a problem. so we can ignore this alarm for effected the IGV or trip unit because we have other three Flame Detectors normal in Service. i am remove and re-connected the communication wiring between DCC card and other cards with cleaning well by contact cleaner.

for alarm:
// <R> 1736 ;TCEA POWER SUPPLY OUT OF LIMITS, P5.
This alarm main problem. it's cause the Processor to malfunction. this very harmful effect in MKV may be this result to trip the unit without indication alarm. my action going to order request the new TCEA card to replaced for core <R>, and i will see the result after that.

i think PTR and ETR if Energized should give alarm when turn to logic (1) but no any evidence for that.
 
samysam,

>i think PTR and ETR if Energized should give alarm when
>turn to logic (1) but no any evidence for that.

PTRs and ETRs are DE-ENERGIZED to trip (the logic signals driving them go to a logic "0" when the turbine is to be tripped). In other words, the PTRs and ETRs have to be "energized to run"--they have to be a logic "1" in order for the fuel stop valve solenoid(s) to be energized to allow fuel to flow to the turbine combustors.

If the any PTR (or ETR) is commanded to be a logic "1" or a logic "0" and it's NOT a logic "1" or a logic "0" there should be a Diagnostic Alarm associated with that condition. But, if they are a logic "1" or a logic "0" and they are supposed to be a logic "1" or a logic "0", respectively, there will be NO Diagnostic Alarm associated with that.

Further, if the power supply voltages on the TCEA cards are not what they should be then the ability of the firmware on the TCEA cards to check and report on Diagnostic Alarm conditions can sometimes be impaired. So, things which would otherwise be reported (when all the power supply voltages are proper and stable) may not be reported.

I'm only offering this as one possibility. I don't even really consider it to be a likely possibility--and without Alarm logs of some kind I don't think it's wise to point to any condition as a likely cause of the tripping. The most prudent thing to do is is to get the Alarm Logger repaired and working again, and then start the machine after you've investigated as many things as you can and fixed any problems you have identified, and prepare for another start.

One of the things you should do in preparing for a start is to get ready to collect some date using VIEW2--which is a high-speed data-gathering software "tool" on the Mark V <I>. You can create lists of data points you want to gather values for, and you can start the VIEW2 program and have it running to collect data. (Of course, you will need two working <I>s--one to operate and monitor the operation with, and one running VIEW2 to collect the data). That way, if the turbine trips again you can have some actionable data--actual running, high-speed data on many more points than available in the Trip History Display. And using that data MUCH more information about what happened will be available.

BUT, you NEED the Alarm printer to be working. OR, you need someone to supervise the operators to prevent them from immediately clicking on MASTER RESET without investigating, or writing down, with dates and times, EVERY alarm message and having someone double-check that the alarms as written down match what's on the Alarm Display before any actions are taken.

If a fuel stop solenoid slammed shut before it was supposed to for ANY reason there WILL BE a "LOSS OF FLAME TRIP" Process Alarm annunciated. And, that can be useful--especially if you have information obtained using VIEW2. And, if the IGVs don't close, then the data collected from VIEW2 may be helpful in understanding what signals were sent by the Mark V and when to close the IGVs.

Without actionable data and actual alarms (and dates and times)--it's really all just speculation. And speculation can lead to LOTS of wasted time (which is MONEY), and lost generation (which is MONEY). And replacing things which don't need replacing (which is costly in terms of parts and time (both of which cost MONEY). You NEED actionable data and a concise list of alarms.

Full stop.

Period.
 
CSA,

I am agree with you .new dot-matrix printer from old stock installation has been done. I was very excited to solve this problem.

The problem solved and all the past Diagnostic alarms clear by replacement new TCE1 card in core <R>. The Gas Turbine start and reach to Base load perfect.

We thank you for the valuable discussions that have benefited a lot.
 
Top