hi friends

I am working in Ge turbine frame 5 mark 6 25 MW open cycle (10 combustion chambers, spark plug in 1st and 10th chamber, four flame scanners -2, 3, 7, 8). Last 20 days we are struggling to start GT.
On 20/3/2015 plant was blackout. After power resumed from grid we start our plant at normal procedure. But ignition failure happened more than 2 startup. We are using liquid fuel as our fuel. We people did following maintenance. we found the fuel pump damaged from inside and we install new fuel pump. then we started it 2 times, and it gave us 4 bar on compition, but with our start and it gave us fail to ignite.

friends what can i do for this problem
[email protected]

 

khairi,

Have you observed any liquid fuel coming out of the false start drains? That's the BEST way to see if liquid fuel is flowing into the machine during firing. I burnt liquid fuel will foe out of the false start drains.

Does the exhaust temperature increase at all during firing?

Do you observe any "white smoke" or a thin white vapor coming out of the exhaust stack after the failed starts? White smoke is an indication of combustion in one or more combustors--but not all of them. If there's any appreciable temperature rise and white smoke that means there is flame in one or more combustors--and perhaps there is some problem with the flame detectors (not likely, but not impossible). There would probably be some flow out of the false start drains, but not a lot.

If you observe a thin white vapor that's an indication that liquid fuel is flowing into the combustors and is being atomized--but is not igniting. This would be accompanied by flow out of the false start drains.

Have you checked the igniters?

Is the Booster Atomizing Air Compressor running? It's difficult to see a pressure rise in the AA system at such low pressure, so sometimes it's necessary to install a needle valve and low-range pressure gage to temporarily monitor AA pressure during firing. Just be sure to close the needle valve after flame is established or the gage will probably be damaged.

Have you removed and tested the flame detectors?

It's probably likely that there is still air in the piping between the fuel pump and the nozzles and it's just hasn't all been purged out. If so, then it may start the next attempt or the one after that--be patient. And be checking for flow out of the false start drains and watching the exhaust stack. Check the spark plugs to be sure they're dry and working.

If the machine is dual fuel (gas/distillate), check the tell-take leak-off to see if the liquid fuel purge check valve(a) are leaking.

If the machine was burning heavy fuel oil/crude when it tripped, have you properly purged the fuel lines?

Fuel, spark and air--including Atomizing Air--are all that's required. Make sure the spark plugs (igniters) are working. Make sure there's no air in the fuel lines downstream of the fuel pump. And make sure there is Atomizing Air pressure during firing.

While it's not likely, the liquid fuel check valves might be not working properly.

And,stable liquid fuel supply pressure at about 4-4.5 barg is also important.

Please write back to let us know how you fare!

 

khairi,

Another thing--when firing the fuel pressure to each combustor fuel nozzle must be greater than the cracking pressure of the liquid fuel check valves, usually about 100-125 psig (check the Device Summary for the cracking pressure of VCK1-10).

Use the manual selector valve handle below the gage at the liquid fuel flow divider to monitor the pressure(s). It must be in positions 1 through 10 for the respective combustor (1 through 10).

It will usually take 10-20 seconds at most after firing starts for the fuel pressure to build up to be at least greater than check valve cracking pressure, and at that point one person should be recording the fuel pressure for each nozzle while a second person should be rotating the manual selector valve handle through positions 1 through 10.

Oh, and since you found mechanical damage in the fuel pump it might have resulted in bits making their way into the liquid fuel flow divider and preventing rotation of the flow divider internals. If this happens, the pressures to each of the fuel nozzles will likely be high--but there won't be any flow to the nozzles, so there won't be much if any flow out of the false start drains.

So, there are several things to check. If it was me, I would first check the igniters (spark plugs), then station two people at the flow divider (one with a digital camera, or the old-fashioned pencil and paper) to monitor and record fuel pressures, one to rotate the manual selector valve handle through the detente (beginning at position 1), and one or two people monitoring false start drains and the exhaust stack, and then try starting again. It can take 10-20 seconds after firing stops to see flow out of the false start drains. If it doesn't fire and the fuel pressures were higher than cracking pressure and there was little or no flow out of the false start drains then I'd be looking at seeing if the flow divider is stuck.

You can monitor FQR (liquid fuel flow reference) and FQL (sometimes FQL1 or FQLM1), the liquid fuel flow feedback from the flow divider, on the operator interface during firing. 10-20 seconds after firing starts FQR and FQL (or FQL1 or FQLM1) should be approximately equal. If the flow feedback doesn't increase very much then it's likely the flow divider is stuck. If flow feedback is very unstable it's likely there is still air in the fuel lines--but it could also indicate the flow divider is not spinning freely. But, you should also have people stationed as suggested during any start attempt until the unit finally does start. There may be more than one problem.

Hope this helps!!!

Again, write back to let us know what you find! And if you need more help, please provide the answers to all the questions asked (even if they don't seem relevant to you).

 

Hi,

Please check these following things before start gas turbine:

1) Ignition system to be checked

2) Flame detectors #1 & 110 to be checked & exhaust temperature increase or not?

3) Liquid fuel pressure to be checked a nozzle or fuel pressure at pumps

4) Air intake filter self cleaning system to be checked

 

rvsaradhi,

Per the original post, the flame detectors are located on combustion cans 2, 3, 7 & 8.

How would a malfunctioning self-cleaning inlet air filter system prevent ignition?

 

dear friends

we found problem on fuel flow driver. we install one ok, but the speed pick ups are not ok. we adjust it to 0.05 and it gave startup fuel flow excessive. how could i adjust it friends?

waiting for feedback.

 

khairi,

Congratulations on finding the problem.

The gap for speed pick-ups should be 0.050 inches, +/- 0.010 inches, for Speedtronic turbine control panels from Fuel Regulator to Mark IV. Mark V, Mark VI and Mark VIe turbine control panel speed pick-up input are much more sensitive, and the gap for liquid fuel flow dividers should be approximately 0.040 inches, +/- 0.010 inches.

HOWEVER, that's NOT going to stop the excessive fuel flow issue. That's caused by air in the lines, and until all of the air is purged from the lines that alarm is probably going to keep occurring.

And, the simplest way to get the air out of the lines is to make sure all filter vessels are bled of air, and then just to keep trying to start the unit. Unless the sequencing/application code in the turbine control panel is non-standard, when the alarm "FAILURE TO IGNITE" is annunciated the turbine will NOT trip, but will just continue to crank. Just select CRANK, wait about 30 seconds, then re-select FIRE or AUTO and the unit will go through a purge and fire sequence. This prevents having to wait for the unit to coast down to zero speed, or near zero speed, to initiate another START. Again, this works for the majority of Speedtronic turbine control systems. A failure to ignite is not usually a trip or shutdown condition (though many people mistakenly think it is), and just selecting CRANK and then re-selecting FIRE or AUTO re-initiates the purge/start sequence. No other action should be required on a failure to ignite.

Write back to let us know how you fare!

 

Well said by CSA,

Few false start too useful to purge the air in the pipelines and it'll ensure fuel is flowing into the unit.

take care
g.rajesh

 

dear friends

thank you friend. we adjusted speed pick up on 0.043 and it was ok the turbine working smoothly.

thanx friend about helping
regards

 

khairi,

Thank you for the feedback!

 

khairi,

Please clear how to change it all and please explain step by step to change it.

thank you
Bronx9669

 

>The gap for speed pick-ups should be 0.050 inches, +/- 0.010
>inches, for Speedtronic turbine control panels from Fuel
>Regulator to Mark IV. Mark V, Mark VI and Mark VIe turbine
>control panel speed pick-up input are much more sensitive,
>and the gap for liquid fuel flow dividers should be
>approximately 0.040 inches, +/- 0.010 inches.

This is reverse. In case of mark-VIe due to high sensitivity air/liq gap need to be higher, in our case it is 0.5mm mark-VIe. previously when it was mark-IV the gap was 1mm.

>HOWEVER, that's NOT going to stop the excessive fuel flow
>issue. That's caused by air in the lines, and until all of
>the air is purged from the lines that alarm is probably
>going to keep occurring.
>
>And, the simplest way to get the air out of the lines is to
>make sure all filter vessels are bled of air, and then just
>to keep trying to start the unit. Unless the
>sequencing/application code in the turbine control panel is
>non-standard, when the alarm "FAILURE TO IGNITE" is
>annunciated the turbine will NOT trip, but will just
>continue to crank. Just select CRANK, wait about 30 seconds,
>then re-select FIRE or AUTO and the unit will go through a
>purge and fire sequence. This prevents having to wait for
>the unit to coast down to zero speed, or near zero speed, to
>initiate another START. Again, this works for the majority
>of Spectroscopic turbine control systems. A failure to ignite
>is not usually a trip or shut-down condition (though many
>people mistakenly think it is), and just selecting CRANK and
>then re-selecting FIRE or AUTO re-initiates the purge/start
>sequence. No other action should be required on a failure to
>ignite.

This part is well said. I want to add one thing, check diesel engine temperature during continuous repetitive start up as said here, normally DE gets overheated specially it's exhaust line gets red hot and causes permanent damage. So keep a close watch by field operator.

 

Khairi,

I think you must do Air Removal regurarly to remove some air on the line, especially after replacing flow divider. On the Mark V we used to force some points (L4_XTP, L20TV1x, L20FL1X, L20FD1/2, etc. and manually running 88FD & 88HQ (for opening 20FL).

Thanks

 

CSA,

at what percentage of TNH speed or pressure of compressor discharge, the false star drain valve will close?

 

Chiranjeevi,

You have not specified which GE-design heavy duty gas turbine, or which packager.

For decades, most false-start drain valves were operated solely based on compressor discharge pressure. That is, there was no solenoid-operated valve controlled by the turbine control system which controlled the flow of CPD to the actuator(s). The output of the axial compressor discharge was connected (sometimes through an adjustable pressure regulator (again not controlled by the turbine control system) to the actuator; so as CPD increased, the actuator closed the false-start drain valve(s).

Some packagers (GE and GE Belfort, most notably) started using solenoid-operated valves to control when the false-start drain valves closed and opened, but usually only on F-class machines (including FA and FB, and newer models--but sometimes on E-class machines).

So, there's no one answer to your question.

Most false-start drain valve actuators which were not controlled by the turbine control system (i.e., did not have solenoid-operated control valves) required approximately 20 psig of air pressure to close, and when the pressure dropped below that then the actuator would allow the valve's fail-open spring to open the valve.

Most false-start drain valves did not have limit switches to indicate position/status--that is until an F-class machines false-start drain valve didn't close and hot combustion gases (from one of the combustors) ignited fuel vapors in the gas turbine drains tank--which is when the valves started being controlled by solenoids and the turbine control system, and limit switches were added to indicate position).

Hope this helps!