Dear Sir,
In Our plant we have 3 gas turbines MS5001PA generator drive 25MW and 3 Waist recovery boilers.

All 3 gas turbine A, B and C are running and the load for each one is 8 MW, the problem is as follow:

The exhaust temperature (TTXM) for units A and C is 351°C but the unit B is 371°C.

Discharge compressor CPD for unit A is: 6.4Bar
Discharge compressor CPD for unit B is: 6.2Bar
Discharge compressor CPD for unit C is: 6.4Bar
The IGV feedback for unit A is 56.9Deg
The IGV feedback for unit B is 60.1Deg
The IGV feedback for unit A is 56.9Deg

Please can you tell me why there is this difference of the exhaust temperature between Unit B and units A and C. Which one it seems to be correct 371°C or 351°C.

Thanks a lot in advance for your help.

 

Hi,

Combustion system inputs are Fuel & Air and Outputs are MW (Torque) & Q (Heat or exhaust temperature). so if the MW is the same and TTXM is different so sure the air flow is different. so let us ask you a couple of questions:

[1] What are the diff pressure values for the three Inlet filters??

[2] what had happened ?? i mean did unit B did had any soon inspections of any type?? or even A or C??

[3] One possible reason is the fouling in Compressor. do you have any means of water wash packages??

[4] what is the control system?

[5] During this case was the units in simple/combined cycles??

[6] IGV Calibration, did you recently verified the LVDTs feedbacks for the three IGVs??

[7] Did you verified the control constants for IGV temp control (if available) between the three units??

[8] What is the combustion system used??

Thanks and best regards
Gemi

 

From your description it would appear that the units are using IGV Exhaust Temperature Control at Part Load. I would also presume that Pre-Selected Load Control is also being used to maintain a load setpoint during operation.

The interesting thing is that you didn't mention what the fuel flow-rates were for the three machines, because that is the best indicator of the efficiency or condition of the turbine and axial compressor and exhaust. If one unit is burning more fuel for the same load as another unit then that would indicate, to me, at least, that the efficiency of the unit burning more fuel is slightly lower than the machine burning less fuel for the same load.

But, having said that, comparing output and fuel flow of machines operating at Part Load *and* with IGV Exhaust Temperature Control active is kind of foolish. IGV Exhaust Temperature Control is, by design, decreasing the efficiency of the gas turbine in order to increase the exhaust temperature and therefore the steam production of the exhaust heat (waste) recovery system. Even though the efficiency of the gas turbine is somewhat decreased, the overall plant efficiency (as measured by including the steam production in the calculation) increases. "Choking" the air flow through the machine by keeping the IGVs closed to try to maximize exhaust temperature at part load is limiting the mass flow through the turbine which slightly decreases turbine efficiency, but it does increase <b>overall</b> cycle efficiency when factoring in the steam production.

You didn't tell us how long these machines have been in service; how long since an off-line compressor water wash was performed on each machine; what the inlet filter differential pressures were on each machine; how long since a maintenance outage was performed on each machine, and whether the outage was a Combustion Inspection, a Hot Gas Path inspection, or a Major Inspection, and what components were replaced in each machine and whether new, uprated components were used in any machine when replacing components during the inspection. We don't know a lot about how these machines are operated, if they are in continuous duty or cycled daily, weekly, or ....

And, lastly, you didn't say if the units have DLN combustors or not, which can also have a lot of impact on operations if they have Inlet Bleed Heat and it's active at part load.

Internal clearances (compressor blade tips to casing; turbine bucket tips to shroud blocks; etc.) have a lot to do with incremental performance differences, and that has a lot to do with how the unit are operated.

You mentioned IGV angles, and the slight differences between them on the units. If the units are being operated with IGV Exhaust Temperature Control enabled then the IGVs will be modulated to whatever position needed in order to make the exhaust temperature be equal to the exhaust temperature reference--regardless of position feedback. Position is not the primary control, exhaust temperature is, and so the IGVs will be moved to whatever position (and indicated position) that is required to make the actual exhaust temperature equal to the exhaust temperature reference.

In other words, if the IGVs of one unit are calibrated such that the indication is spot on (61.5 DGA indicated is exactly 61.5 DGA measured angle) and another unit's IGVs are calibrated such that at 61.5 DGA measured angle the indicated position from the LVDTs is 60.5 DGA, it doesn't make any difference when at Part Load on IGV Exhaust Temperature Control--because the Speedtronic is going to position the IGVs to the angle that results in the exhaust temperature that's required to be equal to the exhaust temperature reference. If that position occurs at a measured angle of 61.5 DGA on both machines, then one machine will indicate 61.5 DGA and the other will indicate 60.5 DGA, but the actual exhaust temperatures of both machines will be equal to the IGV Exhaust Temperature references. No matter how the LVDTs are calibrated, the IGVs will be moved to the position (angle) necessary to make the actual exhaust temperature match the IGV Exhaust Temperature reference.

Usually, the CPD-biased exhaust temperature reference is used as the upper limit of IGV Exhaust Temperature Control, so if you look at TTRXP you should see that TTRX and TTRXP are about equal. I don't have access to any sequencing files at this time, but there is a signal, TXGVERR, as I recall, that you can look at to try to understand how IGV Exhaust Temperature Control works.

But, I need to say this: The odds that this is a controls-related problem are extremely low. Sure, some "things" could be done in the controls to make everything be equal, but that would just be masking other problems, and while some Supervisors and Plant Managers are fine with that, it's not the proper thing to do.

If you're confident that the CPD transmitters are calibrated correctly and that the other instrumentation is working correctly (and that there are not high exhaust temperature spread differentials between the three machines) then the "problem" is most likely mechanical in nature (internal tolerances too far out of spec; high exhaust duct back-pressure; worn hot gas path parts; etc.) or mechanical cleanliness (dirty bellmouth and/or IGVs; dirty compressor; dirty inlet air filters; etc.).

One can have three identical automobiles built in the same factory with identical components and even driven by the same driver (at different times, of course) and the gas mileage and performance of the three will not be identical. And, again, you didn't mention anything about fuel consumption or efficiency which is what your question should really be about: determining if there is a performance difference between the machines and what the source of that performance difference is.

But, it's most likely that the problem is outside of the Speedtronic turbine control panel; it's only as good as the inputs. And, it can only do what the axial compressor and turbine conditions will let it do. Dirty compressors and high internal clearances will adversely affect CPD and exhaust temperature control, even at Part Load when IGV Exhaust Temperature Control is enabled.

The topic of performance and comparing two or more of the same units sitting side-by-side on the same site in ostensibly the same operating environment has been covered many times before on control.com. There is a very fast 'Search' feature in the far right corner of the Menu bar at the top of every control.com page. There is also a Search Help function, and it's suggested to review the Help first.

The second suggestion is to try different search terms and words, don't just try one term or one word and then quit. People can describe the same phenomenon in many different ways using many different words, and still be talking about the exact same thing. So, try to think of different descriptive search terms and words and as you review some relevant searches you may find even more terms and words that will yield better search results. And, searching is free--it's just an investment of your time.

As much as some would like to believe this is some kind of controls-related problem, it's almost never the case. Even if the three machines are new, were built sequentially, and started-up at the same time by the same individual. There are just inherent characteristics which can cause slight differences in performance and efficiency. And, any discussion of Part Load efficiency, especially when IGV Exhaust Temperature Control is being used, are not really valid unless the efficiency is significantly different.

 

Hi Gemi,

The combustion system is fuel gas. The exhaust dusct of turbine is connected to the Heat recovery boiler.

> 1) What are the diff. pressure values for the 3 inlet filters:

Unit A: 30 mmH2O
Unit B: 26 mmH2O
Unit C: 25 mmH20

> 2) What had happened?? I mean did unit B did had any soon inspections of any type?? Or even A or C?

We received from GE a list of constants to be changed and we have changed for the 3 units.

> 3) One possible reason is the fouling in compressor. Do you have any means of water wash package?

We never did the washing for the compressor

> 4) What is the control system?

SpeedTronic Mark V
5) During this cas was the units in simple / combined cycles??
Combined cycles

> 6) IGV calibration, did you recently verified the LVDTs feedback for the 3 IGV’s

Yes We have checked the IGV’s and the feedback for the 3 units, We have did this due to the problem of the difference of TTXM

> 7) Did you verified the control constants for IGV temp. control (if available) between the 3 units??

I didn’t verify, can you tell me the signals name of the constants to be checked.

> 8) What is the combustion system used?

Normal combustion system, without DLN

Thanks and best regards

 

Another possibility to consider is the waste heat recovery boilers. What is the exhaust back pressure at the GT Exhaust (or boiler inlet) of the 3 units? I don't know if you have pressure transmitters on the GT exhaust - if so, signal name may be P96EP or something similar. Higher back pressure on one unit may indicate fouling on the boiler tubes or other obstruction in the gas path.

Is there supplemental firing on the waste heat boilers? If so, was that on? (Supplemental firing can also raise the back pressure.)

 

And, the most important question was neglected:

<b>When did this problem start?</b>

It seems to have started after some Control Constant changes were made. And, we have to ask:

1) Why were the Control Constant changes deemed necessary?

2) What Control Constants were changed?

3) What were the original values of the Control Constants that were changed?

4) What were the new values of the Control Constants that were supplied to your site?

5) Were the Control Constants that were changed the same for all three units? In other words, were the same Control Constants changed to the same values on all three units?

6) Have you used the Control Constants Adjust Display to be certain that the Control Constant values in RAM in all three machines are the values that were sent by "GE"? In other words, are you certain that the Control Constants actually being used by the controls processors (the RAM values) are the values they are supposed to be? (Remember: Control Constants exist in two (2) places in Mark V control processors: EEPROM and RAM. And the Mark V only uses the values in RAM to control and protect the turbine. When values are downloaded to EEPROM they are only transferred to RAM when the control processor is re-booted. Control Constants are something of an exception, since when one changes Control Constants using the Control Constants Adjust Display one is actually changing the RAM values. But, if one does not do something to change the EEPROM values and the control processors are re-booted after changing Control Constants using the Control Constants Adjust Display the changes will be lost when values are copied from EEPROM during boot-up/initialization.)