Today is...
Friday, August 18, 2017
Welcome to Control.com, the global online
community of automation professionals.
Featured Video...
Featured Video
Watch an animation of a conveyor stacking operation demonstrating the use of a move on a gear command.
Our Advertisers
Help keep our servers running...
Patronize our advertisers!
Visit our Post Archive
Exhaust Temp Spread Trip after Selecting IGV Temp Match
The GT trips on high exhaust temperature spread after selecting IGV Temp Matching (transfer from SC into CC)

Greetings

Our Gas turbine (7 Fa Dln2.6 Mark VI) trips on high exhaust temperature spread after selecting IGV Temp Matching (transferring from SC into CC).

After selecting Matching ON, spreads #1, 2, and 3 start increasing fast. After selecting Matching OFF, spread return to normal levels.
There are also some other alarms at the same time like:

1. Combustion Trouble
2. Correcting TNR Drift on IGV Temp Match
3. TNR outside IGV Temp Match range trip TM

This problem happened after changing the IBH Control valve due to mechanical issue (Stuck). After the change, I&C calibrated the feedback. Also I&C tuned the CTIM sensor.

We reported the issue to GE, and first they recommended I&C to calibrate IGV, SRV, Pm1, pm2, pm3, pm4.

This calibration didn't solve the problem. After that GE recommended to change some constant values (sorry I am not what they were).

All the constant values were changed according to GE recommendation but still the problem has not been resolved. Any idea please?

1 out of 1 members thought this post was helpful...

Aptx4869,

Explain how the unit transitions from Simple Cycle to Combined Cycle modes. Does it have an exhaust bypass stack, with a damper or diverter that has to change position to re-direct the exhaust flow from the bypass stack to the HRSG? Is this happening before or after you are changing from Simple- to Combined Cycle?

What is the TNR--and load--when you are trying to put the unit into Temperature Matching?

What TNR--and load--was the unit at BEFORE the IBH control valve was changed and Temperature Matching was being enabled? In other words, is the operation of the unit being changed for some reason since the IBH control valve was changed--and, if so, why?

What is the temperature differential Temperature Matching is trying to achieve? Is it trying to raise the steam temperature, or lower the steam temperature?

Is the unit being operated on Pre-Selected Load Control while all of this is going on?

What kind of sensor is used for the axial compressor inlet temperature measurement? (The only sensors I've ever seen used are Type K thermocouples--which don't require any field calibration....)

Did you ask this question of a local GE field service person, or of the vaunted GE PAC (Power Answer Center)? Not that it matters--because, again, one can't calibrate any device or servo. The ONLY "thing" that gets calibrated with AutoCalibrate is LVDT feedback--which has ABSOLUTELY NOTHING TO DO WITH ANYTHING OTHER THAN LVDT FEEDBACK. LVDT calibration has nothing to do with valve gain, or stability (unless there is a problem with the LVDTs). LVDT feedback has absolutely nothing to do with exhaust temperature spreads. And, if you no one checked the LVDT calibration (by comparing the indicated position against the actual position (at the device) before AutoCalibrating the "devices" then how does anyone know if the LVDT calibration was correct or not? How does anyone know if the LVDT feedback needed to be re-calibrated? Where there Diagnostic Alarms indicating "device" position trouble or LVDT troubles (other than the IBH before it was replaced)?

It's just ludicrously insane to waste time "calibrating" IGVs, or the SRV, or PM1, or PM2 or PM3 or PM4--because AutoCalibrate does NOTHING to the IGVs or the SRV or PM1 or PM2 or PM3 or PM4--or the servos associated with those devices. The ONLY thing AutoCalibrate changes is the scaling of the LVDT feedback from those devices. And if nothing was done to change the physical stroke of those devices, or the LVDTs weren't touched, then performing AutoCalibration "of" those devices doesn't do anything at all--except waste valuable time. And, this advice came from the OEM--which says volumes about the knowledge and training of their staff.

The alarm messages related to Temperature Matching ("TM") are trying to tell you that the unit is outside the allowable range when temperature matching can be used. It can't be used at any load, or any TNR; Temperature Matching can only be used in a certain range of TNR (load).

And, if the exhaust flow is undergoing huge changes in direction and back-pressure at the time the unit is trying to change from Simple- to Combined Cycle--and the unit is trying to go into temperature matching all at the same time, well it's probably going to trip.

Without knowing what Control Constants GE told you to change and why they told you to change them and what they were changed from and to it's practically impossible to comment on the effect that might have had on the unit. But, if this is coming from the same GE source that told you to "calibrate" the IGVs, and the SRV, and PM1 and PM2 and PM3 and PM4--it's no wonder it's not working.

Again, we don't know what's happening when the unit is transitioning from Simple- to Combined Cycle. And all that happens in the Mark* is that the IGVs are being used to try to maximize exhaust temperature when the unit is at Part Load. And, Temperature Matching is trying to control exhaust temperature to a VERY different setpoint in an effort to try to limit steam temperature to protect the steam turbine. In some cases, it MIGHT be trying to raise the steam temperature, but that would be an extreme case, in my opinion, when the steam turbine and steam piping is already warm, very warm.

Hopefully you can see that the two (Simple/Combined Cycle and Temperature Matching) can be mutually exclusive. One might be trying to raise the exhaust temperature as high as possible while the other is trying to limit or reduce exhaust temperature.

Further, exhaust temperature spreads have been covered MANY times before on control.com. It's HIGHLY UNLIKELY--nigh on impossible--that any valve or the IGVs is going to cause any problem with a small, particular group of exhaust T/Cs. If the output of the SRV supplies the PM1/PM2/PM3/PM4 valves, and they in turn supply the PM1, PM2, PM3 and PM4 manifolds which in turn feed the combustion can fuel nozzles (and there are multiple nozzles in each of the combustors)--how can any gas fuel valve cause a problem with a small group of exhaust T/Cs??? They can't. Anyone with any basic understanding of how the fuel system is constructed and works would know that the "advice" to "calibrate" the fuel control valves to try to resolve and exhaust temperature spread problem is categorically wrong and a waste of time--and any money spent obtaining that advice. If one fuel control valve, or all the fuel control valves, has(have) a problem--it's going to affect ALL the exhaust thermocouples--not just a small few.

The same goes for "calibrating" the IGVs. Since the "outlet" of the IGVs feeds the rest of the axial compressor, which goes into the combustion wrapper and makes a 180-degree turn, enters the combustor flow sleeves and makes another 180-degree turn and then enters the combustion liners--how can the IGVs cause a problem with a small group of exhaust T/Cs??? They can't.

My best guesstimate is that the exhaust gas flow is being re-directed from a bypass stack to the HRSG through some kind of diverter or damper and the unit is being put into Combined Cycle mode at the same time it's being put into Temperature Matching--at some load which is outside the allowable range of Temperature Matching--and all of these things are combining to cause combustion mode changes and IGV angkle changes (and likely IBH position changes) and all of this is likely causing the flame in one or most combustors (without a flame detector--remember, only a limited number of cans have flame detectors!) to be extinguished, which is causing a REAL exhaust temperature spread problem which is tripping the turbine.

But, re-scaling LVDT feedback--when it probably didn't need re-scaling to begin with--IS NOT going to solve the problem, which is likely a real combustion trouble problem causing the flame in one or more combustors to be extinguished (or to flash back--which is REALLY BAD!!!) which is causing the trip. (And the results of the "calibrations" speak for themselves--they did nothing to solve the problem!)

Please write back with the requested information if you want more help. We need to understand how the unit operates (specifically if it has a bypass exhaust stack and some means of re-directing exhaust gas flow from the stack to the HRSG), and what TNR and load the unit was previously put into Temperature Matching at (meaning before the IBH change-out and these problems all started happening), and what TNR and load it's currently being put into Temperature Matching at, and if the unit was in Simple- or Combined Cycle mode when Temperature Matching was enabled previously. Also, we need to know if when Temperature Matching is being enable if the exhaust temperature needs to go up or down to achieve the desired steam temperature.

And we need to know what values were changed--from and to.

Lots of questions--but, the real problem sounds most like the unit is just "going crazy" between being transitioned from a bypass stack to an HRSG, and being switched from Simple- to Combined Cycle mode and being put into Temperature Matching all at or very near the same time, and it's changing combustion modes very rapidly and some flame in some combustors is being lost--or there is a flash back situation--which is causing a real combustion problem which is causing the unit to be tripped on excessive exhaust temperature spread. And if the operators are doing this, then that says a LOT about the operations staff training and supervision.

You need help--we need more information. Lots more information.

And, ignore that stuff about "calibrating" devices when it's given in relation to exhaust temperature spreads. It's horribly bad advice.

1 out of 1 members thought this post was helpful...

At the time I retired, temperature matching on FA & FB GE gas turbines with no dampers involved a 2-step process:

First, with a cold HRSG, limit the GT exhaust temperature to whatever the HRSG vendor specified for initial warmup.

Second, raise the exhaust temperature to 100 degrees F above the steam turbine 1st stage metal temperature or, if the steam turbine is cold, some low exhaust temperature maybe around 725 F (my memory is getting a little clouded).

Then, after steam turbine reached inlet pressure control mode, the steam turbine control ramped up the gas turbine temperature control setpoint to its normal value.

While I don't recall any F class units with dampers there may be some. However, the dampers are not a very effective means of temperature control of the exhaust gas into the HRSG. They mainly control the flow from the GT to the HRSG; they don't mix in ambient air into the stream, so the gas temperature doesn't drop much.

I also am curious about calibrating the CTIM sensor. At one time, I recall an RTD being used instead of thermocouples, but that would not need sensor calibration either, and it wasn't used for control purposes.

1 out of 1 members thought this post was helpful...

otised,

Thank you for joining the thread!

I have heard of a couple of F-class turbines with bypass stacks--but only to allow operation when the steam turbine and/or HRSG were unavailable. The fact that the unit has Simple Cycle and Combined Cycle modes would seem to lend credence to that possibility.

Thank you, also, for the information about Temperature Matching. My experiences with Temperature Matching were all bad ones, because the temperature signal came from the DCS and they were never biased very well, and assumed the Speedtronic was able to instantaneously change exhaust temperature and hold a temperature indefinitely regardless of load--all the while the operators and in some cases the DCS were trying to raise gas turbine load to get into emissions compliance. A huge disaster, usually.

I have also heard of a move to use 4-20 ma transmitters for some measurements, and ones used for CTIM could use either a T/C or RTD for the sensor. BUT, being a critical parameter for F-class machines one would think there would be multiple sensors/transmitters for CTIM, feeding a Parallel Transmitter Selector block to derive CTIM. Or maybe there is a transmitter with multiple inputs and a selector function for a single output. I have also heard The General is using Profibus for some signal circuits.

In any case, the original poster has been pretty good about responding with most of the requested information, so we'll have to wait and see.

Again, thank you for the benefit of your experience and knowledge! It's always much appreciated!!!

Otised, I am very happy that you have joined the thread. Wishing to learn from your experience.

CSA, As always you are willing to help others solve their issues. Sorry It took may a while to gather the information you have asked for, since I'm out of the plant these days. There is a lot going on related to this issue so please let me describe it in my best way.

First, Yes the turbine is equipped with a diverter damper to allow it to run on either SC or CC. Even when we start the unit from turning gear using CC(Combined Cycle Mode), when Temperature Matching is enabaled, there will be a high spread when burning GAS

To describe how The Diverter Damper Equipped Gas Turbine operates when CC mode is selected from the beginning, please read:

Gas turbine will go through purging then fires and accelerates up to FSNL. The operator will select "Preselected Load" and loads the turbine up to 20 MW. Still up to this point the diverter damper is closed to the HRSG and open to the Bypass Stack. After the turbine reaches 20 MW, Temperature Matching is enabled. Once Matching is ON, IGV will open and IBH will close to "drive" the exhaust temperature down to 370 Degrees Celsius. Once 370 degrees is achieved the diverter damper is allowed to open to 70 Degrees (There are some permissives for the diverter damper to open and one of these permissives is the exhaust temperature has to be 370 Degrees). So the diverter damper will be at 70 Degrees for 10 Minutes, then it will fully open.

The load at which TM is enabled after replacing the IBH Valve is 20 MW and TNR is 100.7 % and The Droop Speed Setting is 4% (Base Load is around 160 MW). Sorry still I couldn't get the TNR before IBH Replacment since I'm out of the plant these days.

The constants GE requested to change are Tuning Control for IGV and GCVs like FXKPM1S1, FXKPM3S2, etc.

Why they requested this ? Because the same unit trips on LBO (Lean Blowout) when transferring into 6Q mode at approximately 95 MW. The Cans that experience LBO are 10,11, and 12. And "these are the same Cans that the spread is coming from when selecting Temperature Matching On"

GE have done tuning and now the turbine transitions into 6Q Mode without LBO problem. But still now the spreads are coming when TM is enabled while burning gas.

ONE IMPORTANT NOTE: If TM on is enabled while burning LIQUID FUEL, there will be NO SPREADS.

Forgot to mention something.

CSA, you are right. The CTIM sensor is a K Type Thermocouple. There are 3 of them at each GT. There are not calibrated ones as you have mentioned. The one who gave me the information about CTIM Tuning was incorrect.

Thank you!

1 out of 1 members thought this post was helpful...

Aptx4869,

Thanks for the information, but there is still much we don't know.

When did this problem with the LBO (Lean Blow-Out) and high exhaust temperature spread during TM (Temperature Matching) begin? Has it been ongoing since commissioning (when was commissioning, by the way)? Did it start after a maintenance outage--was it the recent outage to replace the IBH control valve?

So, the unit does have a diverter damper, but the location/operation of the damper is not clear. GE-design heavy duty F-class gas turbines have what are termed 'axial flow exhaust'--meaning that they exhaust in a direction parallel with the axis of the unit. Where is the diverter damper in relation to the axis of the unit? Does it swing horizontally left and right to redirect the exhaust flow to the bypass stack or the HRSG?

Where, exactly, are these exhaust T/Cs which always indicate the "spread"--and are they always indicating higher temperatures than the average of the remaining exhaust T/Cs, or are they indicating lower exhaust temperatures than the average of the remaining exhaust T/Cs when the unit is going through Temperature Matching?

Has this high spread problem always been a problem for this unit, or did it just start, and if it just recently started, can you relate the beginning of the problem to some activity? A CI? A HGPI? A MI? Were the fuel nozzles replaced during the activity?

Does the unit have Combustion Dynamics Monitoring equipment?

Was some work done in the exhaust duct during an outage just around the time the exhaust temp spread problem began?

As for why GE makes changes to fuel splits--they have some pretty sophisticated software that can analyze emissions and combustion dynamics to recommend small changes which can help improve emissions and combustion dynamics. They have HUGE amounts of empirical data they have accumulated over time, which they can use to fine tune the DLN systems. GE likes to promote the idea that DLN is a mature technology--that it is established, stable, proven and has been for a long time. The reality is that they are constantly making small adjustments and changes to various hot gas path components--ALL the time. Some are the result of analyzing a LOT of data, and others are science projects, which don't always work out as planned. So, this idea of a mature technology isn't exactly 100% correct or true. Yes; it's been around for a long time on many, many machines. But, when they keep making changes to hardware, even small ones, some of which are not fully proven, the technology isn't stable and fixed.

Since your site has F-class machines, it's a safe bet it has a LTSA agreement with the OEM, and they are "responsible" for the hardware installed in the machines during any outage. And, they use this leverage to do their testing--sometimes with the knowledge of the site operations and management, other times not. And, even if someone in the plant O&M staff know about the new hardware, they don't communicate it down to the rest of the staff. And, sometimes, GE even has site personnel sign NDA (Non-Disclosure Agreements) when new hardware is installed and tested--to protect their IP (Intellectual Property).

Fuel nozzle sets are all tested in the factory (new, or refurbished) and have a range of flow-rates that are allowed. Further, sets of fuel nozzles are put together based on their flow-rate test results to try to limit any flow-rate differences to certain ranges (lower differences usually result in lower spreads immediately after installation, and long term when the fuel is clean). But, some nozzles are at the high end of the allowable flow range, and others are at the low end of the allowable flow range--and if great care is not used when deciding where to place nozzles in the machine low-flow nozzles can be placed next to high-flow nozzles and that usually results in higher spreads immediately after installation.

Without a LOT more information--and a visual analysis of the design and operation of the exhaust duct diverter damper--it's really difficult for me to say what's happening and why. At first, I suspected something amiss with the exhaust flow patterns in the exhaust duct related to the placement/operation of the diverter damper (and it was just a suspicion the unit had a bypass stack and diverter damper).

100.7% TNR on a unit with 4% Droop and a rated load of "around" 160 MW, translates to approximately 28 MW [(0.7/4)*160=28], which is close to 20 MW, but not too close. Operating the unit on Pre-Selected Load Control is, ..., not my preferred method of operation--ever. It's a poor "crutch" for operators--and their supervisors. And, without being able to examine the application code running in the Mark* panel it's not clear what happens to the Pre-Selected Load Control command when TM is enabled. When the IGV are opened during TM (and the IBH control valve closed), air flow through the unit increases--which leans out the premix air/fuel mixture even more, which is probably what's causing the LBO event(s). And, if the unit is trying to put control fuel during this time to maintain some load, that's going to add to the possible problems with combustion.

Complicate this with changes in exhaust duct back-pressure and flow direction/patterns, and there's a lot of possibilities for problems.

Yeah; the I&C Departments (Instrumentation & Control) are always telling they are "calibrating" this, or "calibrating" that--when, in actual fact, they are only really verifying the operation of some device. They use "calibration" interchangeably with "verification"--and, of course, it always gives the impression they are "doing something" to help solve the problem--even if they don't understand what they are doing (as in the case of "calibrating" the SRV, or PM1 or PM2 or PM3 or PM4!) or why. Many don't really have good training, and are given to believing everything they are told without really questioning or trying to understand what they are told. "Tribal knowledge" and myths and falsehoods are very common in the controls business; it's really an occupational hazard that's very difficult to avoid.

Finally, there is NO premix combustion happening during liquid fuel operation. And, the liquid fuel flows out of VERY different orifices than gas fuel, too. The flame is diffusion flame--and is much more stable than the premix "pilot" diffusion flames when gas fuel is being premixed and combusted. The stoichiometry of combustion for gas fuel in a DLN combustion is borderline stable to begin with, whereas liquid fuel--burning in a pure diffusion flame--is pretty stable, especially compared to the gas fuel combustion.

Hope this helps! We would love to follow this problem as you work through the issues, but to be of real help it would require a LOT more information than you can probably provide in a forum like this. And, often times, pictures are worth a thousand words--and we can't do pictures here very well. (Yet.) So, keep us informed, and when we can comment we will.

1 out of 1 members thought this post was helpful...

CSA,

I would expect the diverter damper to be located right before the inlet duct to the HRSG. With the damper pivoting from the top so the damper is swung upwards against the exhaust flow to open the gas path to the HRSG and close the path to the bypass stack which is on top. This probably means the exhaust gas back pressure is somewhat unstable during the transition - try and visualize what's going on inside the duct when the diverter blade is being raised against the gas flow. That gas flow is also corkscrewing in the duct. And this is with a DLN combustion system that is not exactly known for its stability during transients.

1 out of 1 members thought this post was helpful...

otised,

Great explanation; couldn't have been spoken better!

I'll wager that it (exhaust temperature spreads during diverter damper opening while Temperature Matching is going on whilst on Pre-Selected Load Control!) was dodgy before some maintenance and/or DLN tuning, or that some problem has arisen with a combustor or its fuel nozzles due to the failure of the IBH Control Valve. And now it's finally reached an operational reliability problem.

To my mind, there's an awful lot going on when the diverter damper is moving--or stationary at 70°--perhaps the diverted damper position indication has drifted and is not as it was before....

Again, I say, there's SO much we don't know about the timeline of this issue and what's been done. We don't know how recently DLN tuning was done--and why. We don't know how recently a CI or HGPI or MI was performed. AND, we don't know how much any if the LVDT "calibrations" changed (if at all) after following the OEM's instructions.

But we do know--the SRV and PM1 and PM2 and PM3 and PM4, and I believe, the IGVs were "calibrated"--whether they needed it or not, eh? When I hear "instructions" like that, it's clear whoever gave them doesn't have any bloody idea what's causing the problem--but they want to be perceived as doing "something", even if it's wrong, to try to solve the problem. OR, they are trying to divert attention from what they do know is causing the problem.

The more I think about this, the more I think there might be a problem with one or more gas fuel nozzles in a particular combustors. If there's no spreads when running on liquid fuel, and the spreads always appear in the same location when running on gas fuel, that's a pretty strong indication of problems in one combustor's fuel nozzles, made worse by all the changes occurring during transitioning from Simple- to Combined Cycle AND Temperature Matching and Pre-Selected Load Control--along with any combustion mode transitions which might be happening, too.

CSA,

The location of the diverter damper is exactly as Otised have mentioned. I don't believe the problem is related to the diverter damper beacause the spreads are happening before the damper opens. In other words, the spread happens before the exhaust temperature reaches 370 degrees Celsius (which is a permissive for the DD to open).

There has been no work done on the exhaust duct in the previous T&I nor a major work on GT. Only regular preventive maintenance. The unit was commissioned back in 2008. I can't "exactly" relate this new issue to a previous event. But there are two main things happened before the spreads came.

1- The unit has been running pretty well since the last T&I, until an increase in the wheelspace temperature happened. The increase in the wreelspace has not been resolved until an OFFLINE WATER WASH has been done. The compressor was REALLY dusty before the offline water wash.

2- After the offline water wash (mentioned above) has been performed. The unit started and there was a problem with the IBH not following and STUCK.

3- After IBH Replacemnt, the unit has been suffering from the two issues mentioned in the thread (1- LBO when transferring into 6Q mode , 2- High spread when selecting Matching ON). The LBO has been solved by Tuning.

The way our GT behaves when TM is enabled and PRESELECTED LOAD is active is that the Control system will look to the IBH Valve position, if it is lower than 70 % (say 65 %), the control system will deactivate the preselected load and decrease the MW automatically until the IBH position is 70% at least. (because as MW increases, IBH opening should decrease and as MW decreases, IBH opening should increase.)

Yes, the unit has a combustion dynamics monitoring system.
.
When selecting TM, there is a group of T/Cs lower than the average of the remaining T/Cs. The following links will show location and the values of the 27 T/Cs.

T/C locations
https://drive.google.com/open?id=0B9R-rponHkn5eEkzekd1UlYwSGc

T/C values when spread occurs
https://drive.google.com/open?id=0B9R-rponHkn5R0hheUVoRG9CQlk

Aptx489,

The plot thickens....

Thank you for the photos, but they really don't help with my understanding of the problem.

I'm going back to my previous thread and sticking with my assertion that there is something amiss with one or more fuel nozzles in one combustor and that when the fuel/air mixture gets too lean that flame is being lost on one or more of the fuel nozzles which are flowing fuel in that combustor at that time.

I have also seen some very bad damage caused to fuel nozzles when the water/detergent from an off-line water wash wasn't properly purged/drained from the the lines after a wash/rinse and was eventually blown into the nozzles and caused cracking due to the temperature differential.

You keep saying the IBH was "STUCK"--always in capital letters. I presume, since on F-class machines the IBH control valve is opened during starting and initial loading that it was stuck OPEN--but you haven't responded to the question of WHERE it was stuck, open or closed or some mid-stroke position.

And, I'm also going to remind you that fuel nozzles (of which there are at least 70, I believe, in a 7FA unit) can all have an allowable range of flow, and that sets are put together based on the maximum allowable range of flow, and that the nozzles are to be installed so that the highest-flowing and the lowest-flowing are NOT in the same combustor or even in adjacent combustors.

You have said that tuning was necessary because of LBO--which means the fuel flow-rate through some nozzles was too low to support the diffusion "pilot" flame and they "flamed out" (lost flame) because of the excessive air flow. Hence the term Lean Blow-Out (LBO)--the flame was very lean and air blew it out.

And, also, as much as the OEM wants to say DLN is a "mature" technology, it really isn't very old. And, it does have its quirks. "Flame" stability (really, combustion stability) in a DLN combustor is quite difficult to achieve and to maintain under all operating conditions during starting and loading and unloading and stopping (and DLN-2.6x systems are in constant premix mode during starting and loading, and unloading, and stopping). The OEM--and owners of the systems--are continually learning new quirks and idiosyncrasies. And, with the way the OEM changes compressor configurations and combustion hardware designs all the time it's a wonder there aren't more problems, actually. At least in DLN-2.6x systems there aren't multiple combustion zones and fuel doesn't have to be "staged" (switched) from one to another and mixed to achieve premix combustion during loading.

Finally, such a sharp drop in exhaust temperatures in one area of a heavy duty gas turbine IS NOT likely the result of a control system problem. As you have said, the problem occurs only when operating on gas fuel (which is premix combustion--very lean, with a small diffusion, pilot flame), and not on liquid fuel--which is pure diffusion flame and uses entirely different passages in the fuel nozzles than used by gas fuel. (Find the cutaway drawing of the fuel nozzles in the Operations & Service Manuals provided with the units.)

If you don't want to believe that the problem could be caused by exhaust flow disruptions when the diverter damper is moving (and the sequence/movements you have described seem specific to your site--I've never encountered them before), and the LBO problem which SEEMS to have been solved by tuning occurs in the same area of the exhaust during Temperature Matching, then it seems the problem is something is amiss with one or more of the fuel nozzles in a particular combustor. Maybe caused by improper washing/rinsing valve line-up.

Based on the information provided, that's all I can offer in the way of possible causes. NO ONE likes to hear that spreads are NOT caused by the turbine control system (that damned turbine control system, is, after all, the root of all evil on a turbine site!)--but they very, Very, VERY rarely are. I know of several sites that always had high spreads during starting in exactly the same location, and when the unit reached 95% speed (14HS) the spreads calmly dropped to single digits. ALL of those sites had problems in the exhaust duct--one had put new diffusers on the compressor bleed valve discharge which directed the flow from the open compressor bleed valves on to nearby exhaust T/Cs, which resulted in very cool readings. On both gas and liquid fuels. Not a control system problem--but a self-inflicted problem because the drawings for the new diffusers had not been followed. AND, the exhaust duct work had been done at nearly the same time the turbine control system had been upgraded to a Mark VIe. And, the refrain was--for almost two years--"It worked FINE before that Mark VIe was put in there!!!"

I only mention the above because it's so VERY difficult to convince anyone that exhaust temperature spreads are the result of combustion problems--it just has to be that $%^@ turbine control system!

But, it rarely is (that wonderful turbine control system).

It would be great if you would update us from time to time with the progress and resolution of this problem.

Aptx489,

By the way, it's possible to scale trended signals individually so that the trends are MUCH more understandable. I believe if you left-click on the signal name, a set of options will appear in the pane in the lower, left-hand corner of the Trend Recorder window, and you can select the min and max scale values for the signal. Trend Recorder has a LOT of nice features, and people should, literally, play with it (using a saved trend file) and get to know how to use it. It's really the very best attribute of the Mark VIe--for operations and for technicians and for troubleshooting as well as for understanding how things work. I salivate when I get a chance to work with Trend Recorder--it's very powerful and very configurable. It doesn't do well with historical data, but it can be set up to start (trigger) on a logic signal, and record data for some time before and after the trigger. It's all in the Help files for Trend Recorder (which are little more than what's in the Manual; both are pretty good, actually). Become friends with Trend Recorder; it's very, very helpful and quite useful.

Thank you CSA, I appreciate your help

GE now has recommended to check the fuel nozzles (especially Can#1) using a borescope although I don't know how they have specified this Particular Can.

Can you put it in another way how you have reached the conclusion of one or more than one of the combustors is amiss?

As I said, the compressor was really dusty to the point they had to do NINE extra rinses. The usual case would be only one extra rinse.

The IBH got stuck near the zero position (closed position). You can say around 120 MW. I believe when the unit was shutdown for Offline Water Wash, the IBH remained in its closed position ( where it should be in a fully open position).

Thanks and will update you on the borescope findings.

1 out of 1 members thought this post was helpful...

Aptx4869,

The subject of exhaust temperature spreads and how they develop has been covered MANY times on control.com in other threads. And, in this case, GE has used their swirl angle charts to decide which combustor is having the problem. (And, from anecdotal accounts, it doesn't seem GE is very willing to share their swirl angle charts with every owner/operator. Sometimes they do; sometimes they don't. Consistently inconsistent--that's one thing which can always be said about them. ALWAYS.)

Very quickly, when a combustor loses diffusion flame the temperature of the gases in that combustor drop below the other combustors which still have flame. What isn't obvious is that the cooler temperatures DON'T mix with the hotter temperatures as the gases pass through the turbine section (through the stationary nozzles and rotating buckets, and the next set of stationary nozzles and rotating buckets, and so on). Rather, the cool gases stay together as they pass through the turbine section into the exhaust.

HOWEVER, when the IGV angle is less than fully open the entire gas stream rotates slowly in the direction of rotation of the turbine. This phenomenon of rotating is called "swirl." Swirl is most pronounced (largest) when IGV angles are lower, and is less pronounced (smaller) when IGV angles are larger. Swirl angle also depends on IBH, and fuel flow (to a certain extent), as well as ambient air temperature, and compressor cleanliness (more on that below). So, what the OEM has done is they have developed swirl angle charts which can be used to determine--approximately (though with better accuracy as the data used to develop the swirl angle charts improves)--what combustor is having a problem.

So, it seems their swirl angle chart tells them the problem is located in or near combustor ("can") #1. And, if it's not #1, then the combustors on either side of #1 should be examined, and so forth spreading out from combustor #1 until the problem is identified.

My conclusion is based on how the combustors work--and how gas streams from the various combustors pass through the turbine and into the exhaust and what effect that has on the exhaust temperature readings.

Exhaust temperature spreads can virtually NEVER be caused by one of the gas fuel control valves, or the IGVs, or even the IBH control valve, can cause the diffusion flame in one combustor (or even two combustors, or three) to be lost--it just can't happen (when running on gas fuel). The fuel from any gas control valve flows into a manifold which surrounds the axial compressor casing and splits into fourteen combustors (from each control valve) into the respective fuel nozzles in each combustor. So, it's simply impossible for a single gas fuel control valve to cause one combustor's fuel nozzles to lose flame--unless there is a problem with the fuel nozzle(s). But, the gas fuel control valve ISN'T the cause of the problem--it's still something wrong with the fuel nozzle(s). Blockages in the internal fuel nozzle passages or orifices, or some mysterious enlargement of internal fuel nozzle passages or orifices, cause the fuel to be improperly atomized and combusted. And, if the fuel flow is restricted, then the fuel/air mixture is unable to sustain flame.

Now, on to axial compressor cleanliness--and off-line axial compressor water washing. The axial compressor can get VERY dirty--but that suggests a couple of causes. First, there must be some kind of moisture (humidity) that helps the airborne contaminants to adhere to the axial compressor components. If the turbine air inlet filters are located near a busy roadway with lots of diesel lorry traffic the water vapour and particulates in the exhaust will cause axial compressor fouling. Being located next to a process plant with lots of exhaust (either from a cooling water tower system, or from the plant process (hydrocarbon vapours, for example)), can also contribute to the ability of airborne contaminants to adhere to the compressor. And, if it's hydrocarbon vapours and diesel exhaust, they will stick to the axial compressor themselves. Finally, if the unit uses any kind of inlet air cooling such as evaporative coolers or foggers then that is a source of humidity that can cause problems, too. Oil leakage from the #1 bearing housing has also been found to sometimes make it's way into the inlet duct near the bellmouth.

The second thing a very dirty axial compressor suggests is problems with the inlet filters, or the welding of the seams of the turbine inlet filter structure and/or the inlet duct that connects the turbine inlet filter structure to the axial compressor inlet compartment. Something is allowing large amounts of dust and airborne contaminants into the axial compressor. Filters may be ruptured, or improperly installed, or the welding of the seams of the metal structure and/or inlet ductwork may have been inadequate. Some ductwork can "twist" and warp in hot environments and increase any gaps between sections, allowing dirt and contaminants into the inlet air stream to the axial compressor. I have even seen sites operate with the air inlet implosion door opens which allow unfiltered air to enter the inlet air stream. (There should be limit switches to alert (alarm) on opening of the implosion doors, but sometimes they stick, and sometimes operators and their supervisors don't understand what the alarm is trying to tell them, and if the turbine doesn't trip when the implosion doors open, well, then it must not be a severe problem. Right? (Wrong!) And, not all turbines are configured to trip on open air inlet implosion doors, though it would seem F-class turbines definitely should, they are such delicate machines.)

Filters are NOT designed to block all types of airborne contaminants (dust; dirt; etc.). They can't. They have ratings for the size of contaminants (particle size) they can block, and the smaller the particle size the more restrictive the filter is--meaning less air flow through the filter. And, if the particle size is small, then the filters usually get choked (plugged; dirty) sooner than filters with a higher particle size. So, choosing a filter should be based on the types of contaminants on site, and some sites have real problems with very fine sand, or even cement dust. Every site is different and has unique requirements. So, sometimes it's necessary to work with filter suppliers to find the right filters for a site.

It's VERY common for off-line washing procedures to be improperly performed. There are a LOT of valves which mush be manually placed in the proper positions for washing and then manually returned to the proper position for running. And those valves are usually NOT identified with tags/valve numbers by the erection/construction/commissioning crew. (I recommend all the handles of manually operated valves which must be operated during a water wash procedure be painted to make them easy to spot. The valves which should be in an OPEN position during running should be painted one color, and the valves which should be in the CLOSED position during running should be painted another color. When preparing for a water wash, one person goes to the valves which are open and closes them, and the valves which are closed and opens them. When the wash procedure is complete, the opposite is done--one goes to all the valves to ensure those painted with the OPEN color are, in fact open, and those painted with CLOSED color are, in fact closed. One person should go out and manually move the valves prior to the washing--with a second person passing behind, later, to ensure all the valves are in the washing position. Don't send two people out together to move and inspect the valves at the same time--something usually gets missed. And, then prior to re-starting the turbine one person goes out and returns all the valves to the running position, with a second person again going behind--later--to make sure all the valves have been returned to the running position. Very few chances for problems with this method.)

And, the amount of detergent is also a HUGE factor when performing an off-line compressor washing. REMEMBER, the detergent manufacturer is recommending the absolute MAXIMUM amount of detergent for each use--because that way they sell the MAXIMUM amount of detergent. And, LOTS of detergent causes LOTS of sudsing and bubbles, and greatly increases the amount of rinses required. And, if the compressor isn't properly rinsed of detergent, the residue can also cause increased fouling....

So, it's NOT recommended to use the manufacturer's recommended amount of detergent when washing, AND, it's necessary to completely, properly rinse after every washing--no matter HOW LONG nor HOW MUCH water it requires.(I always recommend starting with one-fourth of the manufacturer's recommendation, but not more than one-third.) And, if, after an inspection of the IGVs and initial stages of the compressor after rinsing is complete reveal insufficient cleaning, then perform another wash--but, again, NOT using the manufacturer's recommended solution. And, performing another proper rinse. And, the amount of detergent ALWAYS impacts the length of the rinse and the amount of rinse water.

AND, proper soaking after the detergent is applied is also very critical. Too little time for the detergent to do its magic and it doesn't work very well.

There are MANY factors to properly performing an off-line axial compressor water wash.

Lots of details, which don't always seem obvious at first, but which can become painfully obvious if people aren't observant and diligent. And, has been said--if water/detergent gets trapped in piping low-points and not properly drained before re-starting, then eventually when the flow through that section of piping gets high enough the water/detergent will eventually be moved to some combustor--either through the fuel nozzles or the atomizing air piping or the compressor discharge. And water is MUCH cooler than the metal in the combustor, and can cause damage. And when this happens, it doesn't usually get blasted into EVERY combustor, but a small few. And, it doesn't take much water (a few litres) to cause problems.

By the way, compressor cleanliness doesn't have much of an effect on swirl angle--but it does restrict air flow, similar to closing the IGVs. It's not generally considered a big factor in determining swirl angle, but neither should it be discounted if the compressor is REALLY dirty. It might only change the swirl angle by a few degrees, at most.

Hope this helps!!! And we are looking forward to hearing the results of the borescope inspection, too. Thank you for keeping us informed of the progress of resolving this issue.