White thick smoke from gt during start up and shutdown


Thread Starter


Hii Everybody,

We are running a 9E with liquid fuel and Mark VIe controller.

Combustion inspection was carried out in July and the unit has been running smoothly till two months ago.

About two months ago, we started experiencing white thick smoke from the exhaust which starts at a 17% speed and the smoke becomes thicker and thicker till the GT speed gets to about 30% FSNL and stops.

The alarm that comes is FLOW DIVIDER FUEL FLOW HIGH DIFFERENCE which clears after a short while.

During a normal shutdown,a similar thing happens(white thick smoke) with the GT loosing flame between 59% and 47% speed with "CHAMBER FLAMED OUT DURING SHUTDOWN" alarm

The atomizing air system has been checked and it looks OK.
The exhaust temperatures were trended and the values fall between 340 degree F to 480 degree F

Liquid fuel check valves have been replaced but the problem still persist.

Fuel pressure going to the individual cans after firing was checked at the flow divider side and each read 10bars.

I would be very glad if anyone can share his experience with me.

Many thanks.

White smoke coming from the exhaust during start-up/shutdown when firing on liquid fuel is the result of incomplete combustion.

One or more combustors have lost flame and the unburnt liquid fuel is passing through the combustor and turbine section into the exhaust where it's "expanding" because of the heat of the gases from combustors which do have flame (kind of like water flashing to steam when exposed to heat), but it's not fully burning. This is where the white smoke comes from. The more combustors without flame, the thicker/heavier the smoke.

"Chamber Flamed-out During Shutdown" means that one or more of the combustors with flame detectors have lost flame before they should have lost flame. When that happens, the Speedtronic then closes the fuel stop valve.

There's either a problem with one or more fuel nozzles, or a blockage of atomizing air passages, or bad liquid fuel check valves. But, it's not a control-related issue.

If you look at the exhaust temperatures during and prior to the smoking you will see very high spreads (the unit does not alarm or trip on high spreads below 14HS, 95% speed). Also, if you look at flame detectors during start-up/shutdown just before the smoking you will likely see them flickering or have low intensities/fluctuating intensities.

The liq fuel flow divider feedback problem is a sensor or wiring or noise issue, and most likely has nothing to do with the smoking problem.
Dear CSA,

Thanks a lot for your reply.

We will check on the atomizing air passage and recheck the liquid fuel check valves even though we replaced with new ones, we didn't recheck with a pressure pump calibrator their breaking pressures.

I would be very glad if you can throw more light on the following for me please:

1. Can a plot of the exhaust temperature values reveal the suspect fuel nozzles? Or is there any other way of checking the fuel nozzles?

2. If the there is one or more problematic fuel nozzles, why does the smoke comes only between 17% and 30% speed but not throughout the running period of the Gt.

3. What could cause flame out at about 57% speed during normal shutdown to bring the "CHAMBER FLAME OUT DURING SHUTDOWN" alarm.

Many thanks.

>1. Can a plot of the exhaust
>temperature values reveal the suspect
>fuel nozzles? Or is there any other way
>of checking the fuel nozzles?

Only if you know the swirl angle, and that's difficult, if not impossible at low speed.

Many times the fuel nozzle passages become plugged, and at low flows they don't work well. I sincerely question that all 14 fuel nozzle pressures are the same (10 bar) just prior to the appearance of white smoke. It's very difficult for one, two or even three people to check and record fuel pressures quickly at low speed--and it usually takes 20-30 seconds after a combustion problem to see smoke, so you need to checking pressures quickly and shortly after firing/warm-up.

>2. If the there is one or more
>problematic fuel nozzles, why does the
>smoke comes only between 17% and 30%
>speed but not throughout the running
>period of the Gt.

Good question. When you discover the reason at your site you will know the answer.

Another thing that can cause problems is if, during warm-up the acceleration (TNHA) exceeds the acceleration ref (THNRA) when war-up is complete fuel will be reduced to FSRMIN which can cause combustion problems and smoking. When TNHA finally gets in line with TNHRA then fuel will increase and the flame will be re- established in all combustors and then the smoking will stop. So, if you can compare TNHA/TNHRA before and after the HGPI you may find the torque converter is increasing speed too much during warm- up AND/OR the value of warm-up FSR is too high.

>3. What could cause flame out at about
>57% speed during normal shutdown to
>SHUTDOWN" alarm.

Incomplete combustion or too little fuel during shutdown or too little atomizing air flow. It's not normal for Frame 9Es to go much below 60% TNH doting a fired shutdown without some smoking--if not heavy smoking.

Also, has the fuel supply changed--meaning is the liquid fuel you're now burning nearly the same as before the outage?

Who supplied the fuel nozzles/HGPI parts, and who performed the outage? The OEM or someone else? Were the fuel nozzles refurbished or replaced doting the outage?
I had almost the same issue but it was duel fuel and it was happen only when operating on liquid fuel.

Could you please tell us what is the outcome of your troubleshooting

what was the root cause?
Am a Frame 9E gas turbine control room operator. So i can categorically tell you that you can operate the 88QA from local/manual. I have done that before even just last week here.

Thank you.