We have Frame 6 Gas Turbine which operates in Droop mode with IGV temperature control ON in AUTO.


When the Load on GT is increased, the maximum load reaches at around 77 DGA of IGV opening. When the GT reaches base Load and IGV opens full (85 DGA), then the power load reduces by about 1MW.


This problem is chronic but as it doesn't cause any trouble, it has been neglected.


What could be the possible reason for this? Does this show some impending problem or it has to do with IGV calibration?


Please let me know if more data is needed.

 

Seeker1988,

1) What kind of combustion system does the turbine use?

2) What Process- and Diagnostic Alarms are active when the unit is running (even the Alarms it's believed are not relevant to this particular issue--just list ALL of them)?

3) Why do you suspect the IGV LVDT calibration?

4) What fuel is being burned when this problem is being noted?

 

Hello CSA,

1) What kind of combustion system does the turbine use?
DLN type combusters.

2) What Process- and Diagnostic Alarms are active when the unit is running?

Regulator not on primary FSR tempp ctrl
Bentley Nevada monitor fault.
Wheelspace temp high.
Load compt cooling fan trouble.
Acc. compt cooling fan trbl.
Naphtha scavanging trbl.
Turbine compt vent fan pressure low.
Wheelspace temp. differential high.
Turbine air inlet trbl.
Turbine compt vent fan overload.
Diagnostic alarms:
TCE1 TMR chck trouble, ETR2
TCE1 TMR chck trouble, ETR1
Voter mismatch, <R> L4ETR2
Voter mismatch, <R> L4ETR1
Voter mismatch, <R> TNH1

3) Why do you suspect the IGV LVDT calibration?

During last shutdown, IGV was calibrated. After start-up the base load reached at 88DGA opening of IGV while earlier it used to be 85 DGA open at base load.

4) What fuel is being burned when this problem is being noted?
Natural gas

 

Seeker1988,

1) What is the value of TTRXP (the primary exhaust temperature control reference)?

2) What is the value of TTRXS (the secondary, or back-up, exhaust temperature control reference)?

3) What is the value of TTRX (the low-selected value of TTRXP or TTRXS)?

4) What is the value of CPD?

5) What is the value of CPR?

6) What is the value of AFPAP?

7) What is the value of CSRGV?

8) What is the value of CSGV?

Please provide answers to all the questions for <b>BOTH</b> the 77 DGA IGV angle condition <b>AND</b> Base Load condition.

The first Process Alarm in the list you provided is saying, "The control system is not operating on the primary exhaust temperature control reference." Which means there is some issue with either the inputs affecting the calculation of the primary exhaust temperature control reference, or the back-up exhaust temperature control reference was not determined properly--or a combination of the two. There should naturally be a slight difference between TTRXP and TTRXS, with TTRXP being the lower of the two values. When TTRXS is the lower of the two values then the unit output may be reduced--as you are describing.

The concept of a back-up exhaust temperature control reference is that if for some reason the CPD input failed the condition would be alarmed but the unit could continue to run <i>with the condition being resolved as quickly as possible.</i> That was instituted before the days of DLN combustion systems--when CPD (CPR) measurement wasn't as critical as it is for DLN control. If the CPD inputs are lost on DLN units, the control system trips the unit, which it hasn't done in your case. And, if the back-up exhaust temperature control reference parameters weren't calculated properly then if TTRXS is less than TTRXP then output <b>may</b> be limited.

However, operating on back-up exhaust temperature control for long periods of time is <b>NOT</b> ever recommended, especially on DLN machines, and you say this output differential issue has been chronic, implying it's been ongoing for some time which also implies the control system has been operating on back-up exhaust temperature control for some time. How long? Since commissioning?

The list of alarms you provided is also telling you that something is amiss with <R>'s value of turbine shaft speed (TNH1), AND that <R>'s two emergency trip relays think the turbine should be tripped. If something drops out either of <S>'s or <T>'s ETR relays the unit will trip. This issue should also be resolved quickly.

9) Using the Pre-Vote Data Display, what are the values of TNH1 for <R>, <S> & <T>?

10) Using the Pre-Vote Data Display, what are the values of TNH_OS for <R>, <S> & <T>?

That's a pretty serious list of alarms indicating compartment cooling issues--which could lead to instrument or field device failures.

Finally, the IGV issue is very troubling; sounds like LVDT calibrations at your site are not being done properly. And, worse, if there is a large discrepancy between CSRGV and CSGV then it seems like in the absence of alarms about the discrepancy some "tweaking" of Control Constants may have been done. And/Or there is a servo current polarity issue.

11) Using the AutoCalibrate Display to view the IGV control parameters, what are the values of CAGV for <R>, <S> & <T>? (Using AutoCalibrate to view parameters when the unit is running WILL NOT affect unit operation! It will likely cause a couple of Diagnostic Alarms, but they can be reset a minute or two after the AutoCalibrate Display is closed.)

Please provide answers to all the questions for the conditions asked if you wish to continue this thread and troubleshoot the original issue. It does not seem--from the information provided--that the low output is related to IGV angle, but more information is needed to get to the root cause.

 

12) What is the value of CSKGVMAX?

 

I don't have the recorded values of the parameters you asked at 77 IGV opening and Base load. I will surely check those when power demand increases and we get to raise power generation. At present, plant load is low.

At the present PART LOAD condition, the affected GT has TTRXS < TTRXP (TTRXS=577 and TTRXP=583 and TTRX=577). The S core reads lower value of CPD by 0.5 bar as compared to R and T cores. AFPAP=771mmHg, CPD=8.1 bar and CPR=8.8 bar. The GT is at 23MW.

For IGV, there is no difference in the value of CSRGV and CSGV (57 and 57.1).

Both the vent fans of the compartments run. The other fan starts when these alarms appear. I will revisit the issue.

About diagnostic alarms, I checked and found the R core mismatch values and have informed the Instruments department to check the issue. Thank you very much for the feedback you gave here.

I have provided very few data you asked for but i will reach out again with full data when we have sufficient power demand.

For the time being, please clarify the significance of CPD over FSR for deciding TTRX. What if 96AP probes (AFPAP) are not in line or are faulty? Does it take some default value?

 

CSKGVMAX is 86 DGA.

 

Seeker1988,

To quote E. Edwards Deming, the founder of quality improvement, "Without data, you're just another person with an opinion." I try very HARD not to deal in opinions--but sometimes fall victim to the trap when there is no data.

DLN combustor-equipped machines use CPR to calculate the maximum allowable exhaust temperature limit for a given operating condition, TTRX. This is primary exhaust temperature control. CPR is a fancy CPD--one that uses the atmospheric pressure transducer(s) to calculate the compressor pressure ratio to use to calculate TTRX (non-DLN machines use simple CPD). So, the calibration and accuracy of the atmospheric pressure transmitters is VERY critical to proper operation of the unit--just as the calibration and accuracy of the CPD transmitters is ALSO very critical to proper operation of the unit.

The maximum allowable exhaust temperature control limit is also used as the Base Load control reference--meaning the control system will dump as much fuel into the machine as possible to make the actual exhaust temperature equal to the maximum allowable exhaust temperature control limit, also called the exhaust temperature control reference.

During Part Load operation (below Base Load) DLN machines use the IGVs to control air flow through the combustor and this results in elevating the exhaust temperature. But, the same exhaust temperature control limit/reference value is used to prevent the IGVs from closing too much and causing exhaust temperature to exceed the maximum allowable exhaust temperature. (Non-DLN machines, before gas turbines exhausted into HRSGs (boilers), just opened the IGVs pretty much as soon as possible--but that didn't have an adverse affect on combustion/flame stability like it does with extremely lean DLN combustion.)

Before DLN combustion system, in order to improve reliability GE added a back-up method for calculating TTRX: it used FSR (or the precursor to FSR, VCE) as a "parallel" calculation method. And, the two were sent to a Minimum (or Low) Select function, choosing the lower of the two as the maximum allowable exhaust temperature limit/reference.

DLN machines use generator output, DWATT, as the main input to the secondary (back-up) exhaust temperature control limit. So, the calibration and accuracy of the DWATT signal is ALSO very critical to the calculation of the correct secondary (back-up) exhaust temperature control limit/reference. And, this is why redundant MW transmitters are used on most DLN combustor-equipped machines.

To prevent oscillating or "fighting" between the two references (primary and back-up, or secondary) the parameters for the back-up reference are chosen such that the reference it produces is two or three degrees, sometimes higher, than the primary reference. So, if the parameters chosen for the back-up reference are incorrect (presuming the parameters for the primary reference are correct!) then it's possible, and it occurs from time to time, that the secondary (back-up) exhaust temperature control reference, TTRXS, is less than TTRXP which can reduce output slightly.

Conversely, if the inputs to either calculation are not correct then then the relationship between TTRXP and TTRXS can be incorrect--which can limit output. (It should be noted, that just because TTRXS is slightly higher than TTRXP under normal operating conditions--which would allow for a slightly higher output if some problem caused TTRXP to be less than TTRXS, that <b>DOES NOT</b> mean that the unit is not running at maximum possible power for any given condition, <b>OR</b> that it's acceptable to run on TTRXS for any long period of time. Again, the concept of a back-up temperature control reference is to operate for a brief period of time <i>until such time as repairs can be affected to return primary exhaust temperature control to service</i>.<pre> |__________ Isothermal Limit (same for TTRXP & TTRXS)
| \ *
Exhaust | \ *
Temperature | \ *
| \ *
| \ *
|
--------------------
CPR (\) or DWATT (*)</pre>
In the graph above, the negatively-sloped line with the "\" symbol represent TTRXP, calculated using CPR. The negatively-sloped line with the "*" symbol is TTRXS, calculated using DWATT. In the graph above, TTRXP and TTRXS have the proper relationship--they are basically parallel to each other and don't ever intersect. Again, the offset is intended to be very small under normal operating conditions with calibrated and accurate inputs and properly chosen operating parameters (Control Constants).<pre> |__________ Isothermal Limit (same for TTRXP & TTRXS)
| \ *
Exhaust | \*
Temperature | \
| *\
| * \
|
--------------------
CPR (\) or DWATT (*)</pre>
In the graph above, TTRXS intersects TTRXP, which means at some point that TTRXS will be less than TTRXP, which will have an effect on unit output. (It's difficult to "draw" in control.com posts, so the graph is somewhat exaggerated, but the concept should be clear.)<pre> |__________ Isothermal Limit (same for TTRXP & TTRXS)
| * \
Exhaust | * \
Temperature | * \
| * \
| * \
|
--------------------
CPR (\) or DWATT (*)</pre>
In the graph above, TTRXP is to the left of TTRXS, meaning that TTRXP is higher than TTRXS. The problem could be caused by incorrectly calculated secondary (back-up) exhaust temperature control parameters (Control Constants), or a problem with the MW transducers, or a problem with the inputs to the primary exhaust temperature control limit/reference calculation (the atmospheric pressure transmitters, and/or the CPD transducers).

Beginning with DLN machines, the control system was programmed to warn the operations department when the primary exhaust temperature control limit/reference was not the lesser of the two exhaust temperature control limits/references. And, the operations department should get the instrumentation department technicians to troubleshoot and resolve the problem.

In some cases, if the atmospheric pressure transducer input(s) is(are) outside of limits, the control system will choose a value that is chosen to be the average atmospheric pressure for the site (considering elevation, temperature range, humidity range). Most often, there is a Process Alarm to warn the operations department of this condition, to get the instrumentation department technician(s) involved with troubleshooting and resolving the problem.<pre>
-----------
| |
TTRXP----->| MIN |
| SELECT |----->TTRX - - - --->FSRT - - - --->FSR
TTRXS----->| |
| |
-----------</pre>
TTRXP and TTRXS feed into a MINimum SELECT block (function), the output of which becomes TTRX. TTRX is used to derive FSRT (Exhaust Temperature Control FSR) which also feeds into another MIN SELECT block, the output of which becomes FSR. (The other inputs to the FSR MIN SELECT block are FSRSU, (Start-up FSR), FSRACC (Acceleration Control FSR), FSRN (Speed Control FSR), FSRMAN (Manual Control FSR).)

Under normal operating conditions, between the time when the unit is first synchronized (0 MW) and Base Load, FSRN is in control (Droop speed control FSR). When Droop speed control FSR (FSRN) tries to make FSR greater than exhaust temperature control FSR (FSRT), then FSRT becomes the controlling FSR--and at that point, the IGVs should be at CSKGVMAX (the maximum IGV operating angle) and that is considered to be Base Load. (When BASE LOAD is selected and active, FSRN is actually driven to be slightly higher than FSRT, and to track FSRT by a constant amount--this to prevent the control system from oscillating between FSRN and FSRT preventing the two from "fighting.")

I hope this is what you were asking for. I also hope it helps to understand what might be causing the problem. Under typical operating conditions, even when there are issues, the Process Alarm "Not on Primary Exhaust Temperature Control" (or whatever it is on your particular unit) doesn't come in until the unit is at or near Base Load.

There are likely several things amiss, including something with the accuracy of the IGV LVDT calibration and/or a problem with the IGV servo current polarity (which may or may not be annunciated with a Diagnostic Alarm). But, if TTRXS is less than TTRXP at 77 DGA IGV angle, which should be a lower load than Base Load under most normal operating conditions, something is amiss. It's not clear if it's the atmospheric pressure transmitters (which seem to be abnormal), and/or the CPD value (sometimes there are three (3) CPD transmitters, one that is connected to <R>, one to <S> and one to <T>) which can cause the three processors to calculate different TTRX's and FSRs, also. You can use the Pre-Vote Data Display to monitor most of the signals for these functions for each of the three processors.

So, the mission here is to make certain ALL of the associated transmitters are properly calibrated and accurate. And, verify the accuracy of the IGV LVDT calibration, as well as the polarity of the servo currents being applied to the IGV servo-valve. Then, if the problem still persists, you likely need to contact the turbine control system provider and get them to review the Control Constants for both primary- and secondary exhaust temperature control to make sure they are correct and proper for the unit at your site.

We (the readers of these threads at control.com) would greatly appreciate you keeping us informed about the status of this machine and how the issue(s) is(are) ultimately resolved. This is a great learning opportunity--not only for the site personnel, but for all those who read and follow these threads on control.com. Please write back to let us know how you fare in resolving these issues (because there are several in this thread). "Feedback is the most important contribution!"(c) here at control.com. It's what sets this site apart from most other similar sites--knowing what worked and what didn't, and how problems are finally resolved.

And, data (what is called "actionable data") is extremely important to any troubleshooting. Without data, there are only opinions--and opinions rarely solve problems, unless the people expressing them have lots of experience and good judgement.

 

CORRECTION: The first sentence of the description of the third graph should have read:

"In the graph above, TTRX<b>S</b> is to the left of TTRX<b>P</b>, meaning that TTRXP is higher than TTRXS."

Apologies for any confusion.

 

CSA,

Thank you for the detailed explanation. Currently, the GT is under shut down the check the speed probes and other s/d jobs. We are trying to sort out all the issues.

At present (during s/d) the IGV opening is 36.5 DGA. We will also get it calibrated.

We have 3 transmitters of CPD for RST cores. As the S core was reading 0.5 unit lesser than the other 2 cores, do you think the improper opening of the valve in the impulse line could be the culprit? Can that bring about a difference of 0.5 unit or its the probe itself? We will check the valve though after machine cools down.

Thank you again.

 

And, it was W. Edwards Deming, who gave that perfect bit of wisdom. Sorry, Mr. Deming!

Unless that unit is "special" the IGV angle when it's shut down should be approximately 32-34 DGA, if the mechanical stops are set correctly AND the LGV LVDTs are calibrated correctly.

0.8 barg (7.15 psig) is a HUGE difference!!! And, while a slightly closed impulse line isolation valve isn't good, it's really just an orifice, and since there's no flow through the impulse line/tubing, the pressure across the slightly closed valve (orifice) will eventually equalize. It's just the rate of change of the pressure that is affected by the valve opening/closure (orifice size). The less open the valve is, the longer it will take for pressure to equalize in the impulse line--but it will eventually equalize and the device should read the same as the others.

So, it's not likely the impulse line valve position caused a lower-than-normal reading. It's either the calibration of the device, or the device is defective.

Please write back to let us know what you find. If AutoCalibrate is being used to calibrate the IGV LVDTs, make sure the ACALIB.DAT file is properly configured for the IGVs. In other words, the fully closed angle needs to be accurately measured using a machinist's protractor at several locations around the bellmouth, averaging the readings to get a reasonable value of the close angle. Then the fully open position needs to be measured at several locations, averaging the readings. Those two values need to be entered into the proper section of ACALIB.DAT (usually SVO5 (Servo-Valve Output 5) in the POSITION_POS_SAT (for the closed position) and POSITION_NEG_SAT (for the open position). These changes need to be saved when exiting ACALIB.DAT, and BEFORE opening AutoCalibrate. This is how AutoCalibrate knows what the fully closed and fully open IGV angles are when the servo current is in positive saturation (to drive the IGVs fully closed), and when the servo current is in negative saturation (to drive the IGVs fully open). The default values from the "factory" are usually 34.0 and 84.0--and that's NOT what the physical readings are in the field!

Hope this helps!