Loss of Flame Trip


Thread Starter


Gas turbine was running on part load i.e., 15.8 MW (rated 20 MW) and suddenly naptha pressure before 3 way valve reduces to 0.17 (normally it is 5 kg/cm2) due to naptha controller malfunctioning. And then again without any appreciable delay, pressure increases to 2.59 kg/cm2. But the gas turbine tripped due to 'Loss of flame'. Then after fuel changeover to Distillate took place. Why did not fuel changeover take place before tripping?

The gas turbine is GE frame 5 machine running only on liquid fuel (naptha and HSD).
Most P&IDs of machines which burn naphtha that I have reviewed have three-way valves to switch between naphtha and liquid fuel upstream of the high-pressure liquid fuel pump. The three-way valve is usually motor-operated, and takes several seconds (10-30) to move from one fuel to the other. This is so that the turbine control system has time to adjust the liquid fuel control valve to maintain a stable load while transferring fuels.

A loss of fuel pressure from 5 kg/cm2 to 0.17 kg/cm2 is pretty severe, and that means there was not enough flow available to maintain flame in the combustors for a brief period of time, and the flame was extinguished.

It takes the three-way valve several seconds to transfer, and that's if the liquid fuel system is primed and ready to supply fuel (if there are liquid fuel forwarding pumps they have be started and running, to be able to supply liquid fuel at the required flow-rate).

There's simply no possible way for the control system to maintain flame when the fuel flow is cut back so severely, even for a brief period of time. The air flowing through the gas turbine is three- to five times that required for combustion--so the air will blow out the flame when there's no fuel, or when the fuel flow is very low, such as when the supply pressure drops very quickly. And, even if the transfer valve started switching from one fuel to the other--it can't do it fast enough to prevent low fuel pressure/flow from causing the flame to be extinguished.

Now, if the flow went down to only 3.5 kg/cm2 or so, the load would drop precipitously--but probably not enough to allow the flame to be extinguished. But a drop to 0.17 kg/cm2 from 5 kg/cm2 squared is sufficient to cause a loss of flame--probably before the fuel transfer can even get started and the fuel transfer valve starts to move.

Fuel transfers don't happen in a split second, or even a second or two. The flow-rate of the fuel that is being burned has be ramped down as the flow-rate of the fuel being switched to is increased. A three-way valve does this, but I've never seen one that moves in less than a second.

Hope this helps! There are some thing control systems can do, and there are some that control system can't. Transferring fuels in the blink of an eye isn't usually one of them--even if the fuel being transferred to was ready to flow at a moment's notice. If you look at the FSR when liquid fuel is being burned at some load versus the FSR when naphtha is being burned at the same load you'll find that the two FSRs are different. Hence, in order to keep the load constant during fuel transfers it's necessary to ramp one flow-rate while increasing the other--and in this way the control system will be able to adjust the control valve in order to maintain a stable load.