No Flame in Combustion Chamber #3

J

Thread Starter

JP

During the startup of the Gas turbine, It was observed that the only Combustion Chamber #3 did not see the flame after GT firing sequence. Rest 11 nos. of chambers get light properly.

* Non availability of flame in CC#3 leads to higher exhaust temp. spread and in turn some time leads to trip the machine in combustion trouble trip.

* However, It was also experienced that if GT was kept in crank mode for longer duration say 30 min or more & then start the firing sequence , some time CC#3 see the flame at HP rotor speed above 54% and immediately after that exhaust temp. spread reduced to min. value and then machine smoothly reaches to min. governor speed.

* Maintenance team has simulated and checked the Ignition system, flame scanner, IGV, SRV, GCV and found working satisfactory.

* Six month back HGPI was carried out & then after only mainly machine startup problem observed.

* The GT nozzles are duel fuel type but running only on Natural gas. Liquid fuel is isolated permanently . However, Atomizing airline is on and same is connected to the fuel nozzles.

-- Kindly help me to understand why CC#3 don't see flame during firing sequence and only after 54% of HP rotor speed see the flames.

-- what could be the root cause of this typical problem?
 
JP,

<b>Something has changed</b> with regard to:

--GCV LVDT calibration

--P2 pressure transmitter calibration is off

--Gas fuel nozzle orifice sizes

--Atomizing air flow during start-up/firing/acceleration

--Cross-fire tube issue between CC#3 and CC#4 or CC#5

--Axial compressor air flow

Failure to ignite one or more combustors during firing and acceleration is not uncommon, and it's also not uncommon for flame to be lost in one or more combustion chambers after fuel is cut back from firing FSR to warm-up FSR. And, flame is usually established in the combustion chambers at some point during acceleration. This can happen when there is too little fuel, or too much air (flowing through the axial compressor), and/or the nozzle orifices of the CC#3 (in your case) gas fuel nozzle are smaller than the other fuel nozzle orifices, and/or there is too much atomizing air flowing during firing/acceleration, and/or there is a cross-fire tube problem.

Fuel nozzle orifices for all fuel nozzles are NEVER exactly the same; usually, fuel nozzles with tested flow-rates within a certain percentage of each other (say, 10% or 5%--the lower the percentage, the better) are chosen to be installed in a machine as a set. If a particular nozzle flows much less than the others in the set, then that can be a cause.

It's not uncommon for atomizing air line orifices to not be re-installed or to be installed improperly after a maintenance outage.

If the SRV/GCV were refurbished it's possible the same internal components weren't used, or they weren't refurbished to OEM specifications. Also, if the LVDTs of the GCV weren't calibrated properly that could cause less fuel to flow into the combustion chambers than before the maintenance outage.

If the P2 pressure transmitter is not calibrated properly and is not allowing the proper P2 pressure during firing and acceleration (P2 pressure lower than specification) then this can cause the kind of problem you are describing.

If the IGV LVDTs are not calibrated properly, this can also cause too much air to be flowing during firing and acceleration.

Cross-fire tube alignment is also a possible cause, though not very likely. Usually, it would take both cross-fire tubes to be misaligned, or the combustion liner would have be improperly installed (very difficult to do) for a sufficient amount of fuel not to be ignited by flows through one or the other cross-fire tube. So, while this is not likely--it's also not impossible, especially in combination with one or more other possible causes.

It's also not uncommon for a combination of problems (excessive fuel nozzle flow-rate differences and poor P2 pressure transmitter calibration and bad GCV LVDT calibration, for example) to occur and cause a problem such as you are describing. So, there may be more than one problem.

The reason you are seeing CC#3 "lighting off" at higher speeds is because the fuel flow to that combustion chamber is finally high enough to establish the necessary concentration of fuel to be ignited by the hot gas flows through the cross-fire tubes.

As to the issue of tripping on high exhaust temperature spread during start-up, that's usually blocked from happening simply because it's not uncommon (though it is undesirable) for one combustion chamber to lose flame or fail to fire during starting and for it to later "light off" at some point during acceleration--just as you have described. So, rather than trip the unit on high spread during starting, the combustion monitor is blocked until the unit is at or near rated speed (usually).

If there is a higher than normal exhaust temperature spread when you reach rated speed/load, then it could be that CC#3 has a serious fuel nozzle orifice problem--including a blockage of one or more orifice openings. Small rocks, weld slag, scarf (drill twistings), etc., have been found in fuel nozzles after maintenance outages (sad, but true). So, it's probably a wise idea to pull the fuel nozzle of CC#3 and examine the gas fuel nozzle orifices for any blockage(s).

Please write back to tell us how you resolve the problem!
 
I don't have direct experience with Gas Turbines, but I can provide some suggestions.

If the gas nozzles are clear (in the generator), check the gas lines going to the generator and make sure they're clear and unobstructed. Check the signal to the gas control valve and make sure it's working as expected. Also, check to make sure you're getting enough air (oxygen) into that chamber. And lastly, check that the exhaust port/way/area is free of obstructions.

It sounds like that the generator has to create a big enough pressure difference between the rotor chamber and CC3 in order to get the chamber to fire, which suggests to me an obstruction somewhere, whether it's controller driven or a physical obstruction.

I hope this helps.

~Paul
 
Dear Sir,

Thank you very much for the reply.
We have checked the fuel nozzles orifice and found no blockages. Moreover we have tried to swap the fuel nozzle but result is same- No flame in CC#3.

The fuel gas nozzles and Liners were installed as per OEM suggested and supplied part No.

No maintenance activity was carried GCV/SRV in past. From the startup trend it is noticed that GCV/SRV ramping functioning is fine as per machine demand/sequence.

IGV has two LVDTs and system is not receiving alarm like tracking error or LVDT fault.
Also, possibility of too much air to be flowing during firing and acceleration will have impact on all fuel nozzles CC#1 to C#12,but in our case only CC#3 is getting effected.

I agree that Cross-fire tube and combustion liner alignment may be having issue but our survey crew confirm that they have installed Cross-fire tube and combustion liner as per OEM guideline and details.

We will check P2 pressure transmitter calibration in next available opportunity. However, All CC gets flame during startup except CC#3 so, i really doubt that this may be the cause for the flameout.if possible kindly explain the in detail.

GCV/SRV/IGV/NGV stroke checked was done in presence of GE representative and found satisfactory.

Hope to get your further advice and other technical checks to resolve and understand the specific CC#3 problem.

Thanks
 
JP,

Again, something has changed.

P2 pressure is the "push" that gets gas to flow through the GCV. If the actual P2 pressure is lower than the Speedtronic turbine control panel believes the P2 pressure is, then that could cause less gas to flow through the GCV.

So, you found no blockages in the fuel nozzle orifices. You swapped the fuel nozzle from CC#3 with another CC and the problem stayed in CC#3. That would suggest a problem with the flame detector in CC#3, a problem with the wiring between the flame detector in CC#3 and the Speedtronic turbine control panel, or a problem with the Speedtronic turbine control panel the flame detector is connected to.

OR, it would suggest a problem with another aspect of the gas fuel supply piping and/or the atomizing air piping supplying CC#3. I have seen rags and wads of duct tape stuck in flanges. I have seen Flexitallic gaskets that have unwound in flanges that have created a blockage upstream of fuel nozzles.

You HAVE NOT mentioned any high exhaust temperature spread during normal operation--higher than would otherwise be expected after a maintenance outage. If the spread at rated speeds is higher than would normally be expected after a maintenance outage that would indicate a problem with a particular combustion liner or transition piece or fuel supply to that combustion chamber.

I have very little faith that most people actually perform LVDT calibrations correctly. Mostly, they just sit in front of the HMI and tell the valve to go to a particular position and if the LVDT feedback is close to the reference they call it, "GE! Good Enough!" If the actual position is not measured and compared to the LVDT feedback and if the two values are not very close to each other, then the LVDT calibration is, in fact, not performed or verified correctly. Most OEM personnel don't perform the verification of actual, physical position versus LVDT feedback--so most site technicians don't think it's necessary either. Just goes to show how poorly trained the OEM technicians are--and how little they actually understand what an LVDT calibration is all about and how it works and what constitutes a successful LVDT calibration.

Most people think they are "calibrating" the servo-valve, or the fuel valve, or the IGVs--they are <b>NOT</b>--they are calibrating the LVDT feedback. (That's why we get so many questions about calibrating the LFBV (Liquid Fuel Bypass Valves) of many Frame 5s and Frame 6Bs which don't have LVDTS!!! It's not possible to calibrate a servo-valve with a Speedtronic panel, or a valve or IGVs; it's only possible to calibrate the LVDT position feedback!!!)

So, again: <b>Something has changed.</b>

And, again, it's not uncommon for flame to be lost during warm-up and acceleration and to be established during higher speeds. The Combustion Monitor is blocked during acceleration (and even shutdown) for this reason. It's not desirable, but it's not uncommon.

If the P2 pressure transmitter calibration is good, then it's possible that the Firing and/or Warm-up FSRs need to be increased slightly from their present values. It's possible that the current values are borderline to begin with, and worked well when the axial compressor wasn't so clean and the past IGV LVDT calibrations (measuring IGV LVDTs can be very time-consuming and frustrating if one doesn't know how to do it properly and doesn't have the proper instrumentation--so it's quite frequently NOT DONE; people just use the indicator on the compressor case--which the laborers use as they would use a step on a ladder...).

Please write back to let us know how you resolve the problem. If you write back for more assistance, provide the requested information about exhaust temperature spreads during acceleration and at rated speeds prior to the outage and now, after the outage.
 
Top