Hi,

We have GE frame-VI gas turbine. We have problem of Ignition in Gas tubine during start up. There are two retractable-electrode spark plug.

During first time we took two trial to start gas turbine. Turbine trips on ignition failure on both trial. Then we observed and did following thing.
1. We took candle test of flame detector. found ok
2. Operation of GCV and GRV found ok.
3. Interstage gas pressure found ok.
4. Exhast temp increase 65 degC to 85 degC.
5. We also checked intensity of spark Plut putting outside and forcing login. We observed low intensity of spark. so we replace core assembly and terminal extention of both spark plug.

Then after on next trial our Gas Turbine got ignition and our machine run successfully.

Then after few day we got shutdown of machine for 20 hours. and then after during start up on first trial our Gas Turbine failed ignite. We did nothing and take second trial. The tubine got ignite and machine run successfully.

What are the possibility we need to check during next shutdown to come out from this problem ?

Please let us know the suggestion for that.

regards,
Jatin
e-mail: [email protected]

 

Jatin,

Interesting problem. hope is not the solution…
We are missing some relevant information such as, number of number of flame detection (A,B and/or C,D) during firing, the control panel (MK5?), ignition time, running hours etc.

In general, for a GT running on FG the ignition should take place within one (1) minute (assuming that the ignition system is functioning properly). If this is not the case, the FG system is highly suspected.

The bad actors for this problem could be;
1) Ignition system

1.1) Spark Plug

1.2) Spark plug cable (the usual suspect)

1.3) Ignition transformer

1.4) MCC ignition relay

1.5) Mechanical failure, such as jamming of the igniter plug.

Since your problem appears to be intermittent, there is possibility that its a mechanical failure. Verify post trip whether the igniters are not jammed in retracted position. This test can be done in cold condition of the GT (in shutdown status). You should be able to pull the igniter rod and it will release it self automatically (spring loaded). The rod should also able to move freely, without snap action.

If we will assume that the loop of the ignition system is checked and found functioning, follow the next steps.

2) Fuel Gas system

2.1) Despite you are confirming (assuming) that the inter stage pressure was ok, you did not give us the FG parameters during firing?

2.2) Verify whether the 20VG-1 valve is not passing.

2.3) Verify any other vent valves upstream are not passing (if used).
Again, if the GT flame detectors does not detect flame (all 4 detectors A,B,C,D)within one (1) minute, the bad actor is most likely the 20VG-1.

Did you ever change this valve?

Note: Ignition timer parameter is K2F which 60 seconds (MK5).

Docendo Discumus

 

Hi,

Three things are required for ignition: fuel, spark, and air. For a unit firing on liquid fuel, part of the air must come from the booster atomizing air compressor, usually an electric motor-driven roots-type blower, which is sometimes isolated from the rest of the system by solenoid-operated valves and almost always has a gravity-operated check valve.

You haven't told us what fuel you are trying to start the unit on, and if it's liquid fuel have you confirmed proper operation of the booster atomizing air compressor and all of the components in the system?

Also, liquid fuel-fired units can have ignition problems caused by liquid fuel "cartridge" problems. The liquid fuel assemblies are made up of several components, and they can come loose creating excessive clearances and leading to incomplete or poor mechanical atomization of the liquid fuel. Units which operate on both gas- and liquid fuels can also experience "coking" of the liquid fuel nozzles which can cause problems for starting on liquid fuel.

Have you checked the individual fuel nozzle pressures using the selector valve and gauge at the liq fuel flow divider? Problems with nozzles and liquid fuel check valves can usually be pinpointed with pressue readings; they should all be relatively equal, and if any are much higher or much lower, then that's an indication of problems with the check valve or the nozzle.

I have seen this exact problem, failure to fire on first attempt but successful on second attempt, on liquid fuel machines which have air in the liquid fuel supply piping. It's a problem for many units, and it's not always evident. If you are watching the liq fuel flow divider feedback during firing, it will usually be characterized by spiking when the air passes through, high peaks and then nothing and then high peaks and nothing. Sometimes the fuel will ignite and then go out when a slug of air goes through the system, and you will see slight rises in exhaust temperature as you have reported.

Gas fuel systems are a little easier. But, if the GCV LVDTs were recently "calibrated" and now this problem has just started or worsened, then the calibration procedure is suspect.

But, you haven't really provided us with enough information about the conditions of starting, though you did provide a fair amount of information about what you've done to try to troubleshoot the problem. Spark plugs do require maintenance and a weak spark can be problematic, though it's not usual to see weak spark on both plugs at the same time unless there's a problem with the spark plug power supply (ignitor exciter or transformer); that's why there are two plugs, for redundancy and hopefully to ensure that if one plug isn't working or isn't working at optimum efficiency the other plug will be working properly.

If the GCV LVDT calibration has recently changed, and it wasn't done exactly as previously done, and the firing FSR was borderline to begin with, this could be the problem.

So, you have some things to check and some more information to report. (Remember: When you're posting for help, you need to provide as much information as possible in your original post to get the best answer possible in the shortest period of time.)

Regards

 

Thank you for reply. and sorry for incomplate information.

Plese see your required information

We have GE Frame VI Gas Turbine. Turbine contorl system is Mark IV. Out gas turbine is modified for dual fuel firing facility for Gas as well Liquid fire.

But presently we are using 100% natural Gas of NCV around 8700-8900 Kcal/Sm3.

There are 10 nos combution chamber. Spark plug installed at chamber no 1 and 10. and flame detector located at chamber 2,3,7 and 8.
During Ignition trouble problem there is no flame observed in all detector. Yes Ignition time is 60 Sec.

Spark Plug at chamber no 10 suspected Jamming during trip.During machine trip rod of spark plug at no 10 remains outside and when we free mounting bolt it moved inside.But we able to pull the igniter rod and it will release it self automatically. The rod of spark plug at Chamber no:1 move freely while at chamber no 10 move somewhat hard.

At firing following parameter we observed interstage gas pressure around 2.5 bar, FSR around 14 %.

regards
Jatin

 

Fuel.

Spark.

Air.

If you have one working spark plug, that should be all thats' required to provide the ignition source (spark). (The redundancy thing.)

If you're firing on gas fuel, the air should not be a problem presuming the IGVs are at the normal position during firing and are not open more than normal.

So, that leaves fuel.

I have seen a similar occurrence on units which have some kind of purging system on the gas fuel supply system (most do not, but some do). There is simply insufficient flow during the first firing attempt to clear the purge medium (nitrogen or some other non-combustible gas) and permit fuel to flow through the system. The length of the firing timer (60 seconds) on gas fuel seems a bit long for newer machines (which typically only use 30 seconds for gas fuel firing timer setpoints, but 60 seconds was used for gas fuel firing timer setpoint for decades before programmable electronic control systems were used). So, the length of the timer leads one to believe that there may be more to the fuel gas system than we know about (like a purging system).

The P2 (interstage) pressure you cite seems about normal, and the Firing FSR about right also, but that's just a guess.

Conspicuously you did not mention GCV LVDT calibration. If the Firing FSR value is a little low and the GCV LVDTs are calibrated such that the valve is at 12% physical position when the LVDTs (and the turbine control system) thinks it's at 14% position, then there may be insufficient fuel flow. That doesn't explain why it always starts on the second attempt, but it could be a part of the problem. I believe in the original post you stated that the operation of the GCV and SRV were checked and found to be okay. How did you check the operation of the valves? How recently were the GCV LVDTs calibrated?

It's important to remember: The turbine control system of most GE-design heavy duty gas turbines does not monitor actual gas fuel flow during operation for anything other than scheduling water- or steam injection flow-rates. So, the control system does not calculate air/fuel ratios and adjust the valve position to try to obtain the proper ratio. The presumption is that if the GCV LVDTs are calibrated properly and the P2 pressure is correct and the fuel characteristics haven't changed much from the original values that if the GCV is at the position specified for firing that fuel flow will be sufficient to establish flame. But there is no gas fuel flow-rate reference being compared to gas fuel flow feedback to obtain or maintain a specific air/fuel (or fuel/air) ratio at all, not during firing, not during acceleration, not during loaded operation.

So, it's very important that the GCV LVDT calibration be fairly close to accurate, especially during firing, when there is no reference or feedback other than valve position. (During loaded operation, GCV position is a function of turbine speed reference and turbine speed feedback during part load operation and when at Base Load it's a function of exhaust temperature reference and exhaust temperature feedback. But during firing, the only reference is valve position and the feedback is from the LVDTs.) And the presumption is that the fuel characteristices haven't deviation much since the unit was designed and programmed and that the GCV LVDTs are calibrated such that when the turbine control system thinks the unit is at 14% valve position the valve is phsyically at 14% of valve position.

You haven't told us how long it takes to establish flame on the second firing attempt. 20 seconds? 40 seconds? 55 seconds? It should only take 10-20 seconds or less, sometimes as much as 20-30 seconds, to establish flame when firing on gas fuel. This presumes there is no purge gas which must be "purged" from the gas line!

If you're confident that you are getting sufficient fuel into the combustors and you are confident the IGVs are not open more than normal during firing then the problem could be the spark plugs. With the retractable style of spark plugs, there should be a good sharp "pop" each time the spark plug fires. With the spark plugs installed in the unit, and with the compartment vent fans off and the unit at rest (at zero speed and/or on Cooldown), if you force the spark plug relay logic signal you should be able to hear the spark plugs "pop" each time they fire. When they are removed from the combustion cans and properly grounded (something a lot of people don't properly do during testing) you should see a fairly strong spark and hear an even stronger "pop" each time the plugs fire. One can usually hear the popping from several meters away from the plugs when they are removed for testing.

Again: The formula for establishing flame in a gas turbine involves having sufficient fuel, sufficient air, and an ignition source (usually provided by a spark plug). It's no more complicated than that.

I am <b>not</b> a proponent of increasing Firing FSR values without ensuring that <b>every</b> other possibility has considered and ruled out. <b><i>Then and only then</b></i> should one consider increasing Firing FSR setpoint values. This includes comparing an current fuel gas analysis with the one in the Control Specification originally supplied with the unit. If the fuel constituencies have changed considerably, it's worth having a qualified person or company review the data and suggest new operating parameters for firing, acceleration, and shutdown.

You might try increasing the Firing FSR a tenth of a percent on the first firing attempt, and if it doesn't fire leave the setpoint at its new value and initiate a second start. On a subsequent start at another time of the day or another day, before the first firing attempt increase the Firing FSR another tenth of a percent. If the unit fires on the first attempt, then you have a new Firing FSR value and you can leave the setpoint at the current value. If it doesn't fire on the first attempt, continue testing as above until you achieve a Firing FSR setpoint value that establishes flame in about 20 or 30 seconds into the firing timer sequence.

Lastly, there is a limit on how much one should change Firing FSR setpoint values, and to my mind that limit is approximately 1.0%, so if you have to go above 15% to establish flame in a reasonable period of time, then something else is wrong or the fuel constituencies have changed or the nozzle orifices are not what they were originally, something like that. (Nozzle orifices could also be a part of the problem. Many people buy nozzles from other than the OES (Original Equipment Supplier) and the quality control of third-party nozzles has been found time and time again to be less than desirable.)

Fuel. Spark. Air. There are a lot of things that affect these three simple requirements, and a systematic approach to examining and eliminating each possible problem is all that's required to solve the problem.

Let us know how the situation progresses.

 

Jatin,

No problem, now we know a bit more about your GT.

So you found spark plug of combustion chamber #10 with some trouble. That problem is probably caused due to the "pulling" force on the igniter rod caused by the loose hanging igniter cable which re-positions the igniter rod out of the igniter centerline. So ensure that this cable is properly routed.

Based on your information given for firing parameters, it appears that the parameters are taken just post ignition. The excepted firing parameters should be between 16 and 18.0% TNH and 1.0 to 1.8 barg inter stage pressure while the FSR should be about 18%. Most likely you have taken the readings just after initial ignition. Post initial firing the FSR value will be about 12.0 to 14.0% and increase accordingly the control parameters. Again, these parameters are also based on FG pressure upstream the SRV. The following inter stage pressure formula is based on SRV maximum stroke value of 1.3inch and GCV maximum stroke value of 1.5inc: P2= TNH*FPKGNG+FPKGNO. You may validate your P2 with this formula.

Resume, the GT should be able to ignite even with one igniter functioning. Assuming that the igniters are functioning properly, the next bad actor could the 20VG1 valve. Did you check also the 20VG1? Consider to change this 20VG1 the next opportunity. The affect of passing 20VG1 valve for proper ignition is beyond your imagination! You will hardly notice the pressure drop from the screen. Depending on the severity of the pressure drop, the calculated fuel to air mix ratio in the combustion chamber will be not matched which will lead to ignition problems.

Tip: To shorten the re-start time, you may consider switching back from the firing mode to crank mode prior expiration of the ignition timer (60s). With this switch over, the firing timers are reset back to zero and you can initiate additional firing attempts. Ensure that, the sequence in MKIV is programmed to do this (check whether the purge timer is also re-setted etc.). Be aware that manipulating the sequencing logic may lead to damage to your asset, environment and may lead to injuries to your self and other people!

Docendo Discumus

 

good day,

can you send me the GE reference manual that show the firing timer is set to 60 Sec?

 

Hi,

In line with the issue, I would like to state that on liquid fired units, the problem of ignition or more precisely 'failure to ignite' is more common than the gas units. Considering that the above mentioned tips by others, who have contributed to the article have been tried, I understand that it would be better to have your fuel lines completely filled prior to the firing sequence. In detail, what normally happens during a liquid purge (off base fuel skids) is that the L33PLO is indicating a high status. When the fuel starts flowing in to displace the Naptha fuel (primary HSD & secondary naphtha), upon expiry of the T2PUR, the purge sequence stops and the VS-1 shuts off. But if during the process of purge, if the purge valves are closed (no need to close on individual valves, a main header valve downstream can be first closed, the sequence would be terminated. The liquid lines from the fuel pump outlet to check-valves would be completely filled in and upon subsequent start, this problem could be eliminated. One more thing, mostly firing timers are of 30 secs duration since this is more than sufficient to have a flame ON indication (at least for the sequence to proceed, two scanner detections would be okay.

 

The old GE turbines with the hydraulic actuators that are under the SRV/GCV can have the seal on the piston in the the hydraulic cylinder (actuator) wear out. This allows the hyd oil to pass around the piston and the gas valve(s) will not open far enough. The way to spot this is the machine never shows fire in the flame detectors but the exhaust temperature starts rising. The fuel air mix is too lean and lights off down in the liner or TP instead of up next to the burner. If the signals are available compare the requested valve position vs, the actual position. If it's short replace/repair the cylinder. This problem bit me a few years back on a propane compressor and took over a solid day on site to track down.