gas turbine that is only used at peak times was working ok but then failed to light, two burner cans removed and ignitors seen sparking. interconnecting burner tubes are ok. Tried adjusting fuel for leaner running, richer running no success. Atomising air pressure ok, nozzles ok. Any advice on what else to check appreciated...

 

davidmcc,

Failures to fire on liquid fuel can be very difficult to troubleshoot (but NOT impossible)--especially if the unit is primarily operated on natural gas.

Sparking ignitors are not necessarily generating enough spark to ignite liquid fuel, which must be properly atomized (pressure atomization, by the fuel nozzle(s), and by atomizing air) to be ignited. Ignitors do deteriorate with time and do require maintenance.

If the unit has DLN-I combustors and the ignitors do not retract during normal operation (i.e., they are stationary) the ignitor tip does not protrude very far into the combustion liner and the spark must be very strong.

Cross-fire tubes only work when at least one of the combustors with ignitors lights. There are two ignitors for redundancy; only one has to work to establish ignition, and then the higher pressure in the can with flame pushes hot combustion gases into adjoining combustors, until, when all the combustors are lit there are no pressure differentials and flow through the cross-fire tubes stops.

You did not tell us what alarms are being annunciated during firing--other than probably "Failure to Fire." If there is an alarm about excessive fuel flow, then it's likely there is air in the liquid fuel supply lines to the stop valve.

A GE-design heavy duty gas turbine which burns liquid fuel will have False Start Drains from the lower combustors, the combustion wrapper/compressor discharge casing, and the gas turbine exhaust compartment. When liquid fuel flows into the gas turbine combustors and does not ignite, it generally condenses (most of it, especially if the unit is cold when being started), and collects in the lower combustors, combustion wrapper/compressor discharge casing, and the gas turbine exhaust. Hence the False Start Drains, which are there to drain the collected liquid fuel to the gas turbine drains tank.

The False Start Drains are supposed to drain into "scuppers" such that flow out of the drain lines and into the scuppers can be visually observed by human eyes. This is how one can verify that, in fact, liquid fuel is making its way into the combustors--or not. So, after a failure to fire personnel should be quickly checking the false start drains to ensure there is flow--or not.

No flow, or very low flow, from the false start drain valves would be indicative of air in the line.

If the unit is dual fuel (gas/distillate), there will be a liquid fuel purge system, with check valves which are known to be problematic. There is a three-way valve (VA-19,I believe) which has a scupper drain called the Tell-Tale Leak-off. If any of the liquid fuel purge check valves are leaking (preventing liquid fuel from flowing into the fuel nozzle/combustor) there will be flow out of the Tell-tale Leak-off (and there should be NO flow through the Tell-tale Leak-off). Again, personnel have to observe the Tell-tale Leak-off.

Lastly, some newer Speedtronic control systems have had problems caused by mechanical vibrations being "picked up" by the speed pick-ups on the liquid fuel flow divider--causing the Speedtronic to think there is flow to the combustors when, in fact, there is not. (Again--this is where observing the false start drains can be helpful!) This prevents the liquid fuel bypass valve from closing. Unfortunately, I believe the unit you are working on is a Frame 6B and one can't see the valve stem of the liquid fuel bypass valve, and there are no LVDTs on the liquid fuel bypass valve.

Strong spark is important. If there is too much fuel, the spark plugs can get so wet they won't spark very well, if at all.

Fuel flow is important. If there is air in the line, that' a problem. If the liquid fuel supply pressure to the turbine is not stable, that can cause problems.

Lastly, there is a selector valve at the liquid fuel flow divider which can be used to monitor the fuel pressures to each of the combustors during firing. It has to be manually rotated, and so it takes two people to record the pressures during firing. The fuel nozzle pressures have to be greater than the liquid fuel check valve cracking pressure (usually 90-120 psig; check the Device Summary for the cracking pressure of your turbine's liquid fuel check valves). If the pressure is less than check valve cracking pressure, there will be NO flow to the nozzles (again, that would be evident by the absence of liquid fuel flowing out of the False Start Drain Valves).

Atomizing air is very important. It's very difficult to measure during firing without a separate gauge with a very low range. There should be a booster AA compressor which really just makes sure there is flow, but doesn't raise the pressure very much (a few psig at best, usually less than 10 psig).

Lastly, liquid fuel pump clutches have been known to slip. The best way to check this is to remove the coupling guard cover and monitor the clutch output shaft. It should be rotating smoothly during firing; if it's jerky then the clutch is slipping. If the output shaft is not rotating at all, then there's some problem with the clutch (electric solenoid-operated).

Please write back to let us know what you find. Refer to the Piping Schematics (P&IDs) for the Liquid Fuel system, Atomizing Air System, and if the unit is dual fuel the Liquid Fuel Purge System to see the various components mentioned above. It will be necessary to actually go out and identify these devices and their locations, but it will be time well spent.

Again, let us know how this works out!

Liquid fuel flow divider feedback is usually signal name FQL1 or FQLM1. Liquid fuel flow reference is usually FQR or FQROUT. You should be monitoring these during firing to see what the liquid fuel flow-rate reference is, and what the liquid fuel flow-rate feedback is.

Also, please tell us about any and all alarms (Process and Diagnostic) which are being annunciated prior to and during and after firing.

 

hi CSA,

Thanks for your comprehensive reply. Some of the things you mentioned we knew about but others we hadn't considered. We did check the drain and found it was blocked and fuel (distillate, we don't fire on gas) did come out. It still didn't fire after that but the following day it fired ok with our normal settings. So I think it was just a case of the fuel preventing the spark and even whe the fuel was drained out, it needed time to dry out maybe. I'll keep your advice on file should it ever happen again. There was no other alarms about fuel flow or atomising air or anything, just the fail to fire. Sorry it's a little bit inconclusive but it just looks like one of those things that might happen from time to time when the machine sits off on standby for a long time.

Best Regards.