Hello dear friends
Thank you in advance for your countributing in troble shooting
Background:
After a 6-month gas operration , we commenced the startup procedure for a GE Frame 9E gas turbine using diesel fuel (gasoil). The following pre-startup checks were performed:

The gas and diesel Check Valves and atomizing air system were tested satisfactorily.

The cracking pressure of the check valves was measured at 8 bar, which is acceptable.

The diesel fuel line filters were inspected and found to be clean and normal.

The stop valve and the VC3 control valve were stroke checked; both operated according to their reference signals correctly.

The fuel line wasfilled up and removed air.

Initial Problem:
While in the shutdown state, we filled up the line The pump and pathes observed liquid fuel from fals start valves. Then the turbine was placed in crank mode to completely purge the fuel from combustion chamber system. After 15 minutes, we attempted a diesel startup.

Startup Failure with failure to ignite:
Despite having a supply pressure of 6 bar to the pump and a discharge pressure of 10 bar from the pump, the startup attempt was unsuccessful. The unit tripped on a "Failure to Ignite" alarm.

Investigation of Control Valve Bypass:
We investigated the operation of the control valve bypass. The trend data (provided below) indicated severe instability and fluctuation in its behavior. also the fuel pressure fluctuated.

Subsequent Gas Startup & Diesel change over :
Following a second purge, the unit was successfully started on gas fuel and brought to Full Speed No Load (FSNL). The fuel transfer from gas to diesel was commanded and was successful in initiating and completed . However, the diesel fuel flow (FQL1) exhibited significant instability, fluctuating between 0.6 and 3.5 kg/s (or corresponding units).

Further Control System Checks:
We then examined the control system signals:

The feedback signal from the diesel flow transmitter (FQL1) was oscillatory.

The output signal from the controller to the valve (FQROUT) showed the same normal pattern.

Crucially, there were no alarms indicating that the control valve was not following its reference signal.

During this period, the turbine speed (TNH) also exhibited minor fluctuations of approximately 10 RPM.
That is main question , control valve is blaming or TCQA cart or Flow divider mag pick up or ...
thanks for your effective comments.

 

@abed.a,

In recent years (or longer, maybe) GE has removed LVDTs from GE-design Frame 9E heavy duty gas turbine Liquid Fuel Bypass Valves (the valve used to control liquid fuel flow to a GE-design heavy duty gas turbine). While it's not really necessary for proper operation it does provide a good means of troubleshooting valve/flow instability issues using electronic means.

What the lack of LVDTs on the Liquid Fuel Bypass valve really hampers is the ability to manually move (stroke) the Liquid Fuel Bypass Valve valve when the machine is at rest (not running; at zero speed). So, again, while LVDTs aren't necessary for proper operation while burning liquid fuel the lack of them certainly makes troubleshooting liquid fuel flow issues more difficult--not impossible, but more difficult.

Over multiple decades I heard when arriving on site that, "We vented ALL the air from the liquid fuel supply piping and filter canisters," which was soon found not to be fully correct. MANY sites with GE-design heavy duty gas turbines that burn liquid fuel have liquid fuel supply systems that use centrifugal pumps to send liquid fuel from storage tanks to the ;
low-pressure liquid fuel filters and Liquid Fuel Stop Valve. Many of these system have diaphragm-type pressure regulating valves which can trap air in the diaphragm and cause serious problems with liquid fuel supply pressure stability.

A liquid fuel supply pressure of 6 barg up to the liquid fuel stop valve (which is essentially at the inlet to the high-pressure liquid fuel pump) is high, in my personal experience. It's not very much below the liquid fuel check valve cracking pressure, which also seems high in my experience (it's usually around 6-7 barg in my personal experience)--which would possibly make it difficult to control liquid fuel flow during START attempts and possibly even at FSNL. Pleas verify the design liquid fuel supply pressure and the design liquid fuel check valve cracking pressure.

Many liquid fuel supply systems have piping that runs in overhead pipe trays and a lot of those piping systems were built without adequate means of venting air from the high points in the piping system (including not installing pitched piping systems to help the air move to high points in the piping where they can be more easily vented.

You also mentioned checking the "... gas and Diesel check valve..." cracking pressure. I'm not aware of gas fuel check valves on any GE-design Frame 9E heavy duty gas turbine, and I wonder if you checked on or two or four or eight or all of the liquid fuel check valves cracking pressures. And the same for the liquid fuel purge check valves.

The data you provided is not good, especially because we don't know the rate at which the data was gathered and we can't see other signals which might be helpful. Yes; it does seem that something is causing unstable fuel flow--but it's difficult to understand what is causing the instability from the graphs. You mention a TCQA card, so I'm presuming the turbine control system is a Mark* V or some perversion of a Mark* V (possibly one of the flavors of Mark* V Life Extension, of which there are several). We also don't know how the data was gathered (from Mark* V Short Term Trending, or ToolboxST, or a GE-provided Historian, or some MODBUS-based system). What's really required is high-speed data (multiple times per second, not one every second or two seconds or three seconds). We also don't know what the resolution of the data on the graph is (vertical resolution, that is).

The data also doesn't include the values of servo-current outputs from the control processors while liquid fuel was flowing.

You wrote, and I don't understand, these two sentences:

"The feedback signal from the diesel flow transmitter (FQL1) was oscillatory.

The output signal from the controller to the valve (FQROUT) showed the same normal pattern." (The emphasis was mine.) Please clarify.

Without LVDTs the Liquid Fuel Bypass Valve cannot be stroked using the AutoCalibrate feature of a Mark* turbine control system. The AutoCalibrate manual stroking feature uses a position reference to the device with LVDTs providing position feedback. Without LVDTs there is no position feedback from the Liquid Fuel Bypass Valve which only receives liquid fuel flow-rate feedback from speed pick-ups mounted on the liquid fuel flow divider. So, one can't tell the liquid Fuel Bypass Valve to go to 25% position and monitor the LVDT position feedback to see if the device holds steady at 25% of stroke.

There are ways to use a frequency generator capable of square wave or sine wave output that crosses a 0V axis to simulate liquid fuel flow feedback and test the Liquid Fuel Bypass Valve hydraulic actuator--including the valve's hydraulic servo--but it's very complicated and time-consuming.

So, the only thing one can do without the proper equipment, set-up, configuration and monitoring (visual monitoring of the position of the Liquid Fuel Bypass Valve valve stem) is to:

1) Recheck that all air has been removed from the liquid fuel supply piping and filter canisters between the liquid fuel storage tanks and the gas turbine Liquid Fuel Stop Valve

2) Check that the liquid fuel supply pressure is steady at approximately 4 barg (approximately 60 psig, usually)

3) Check that the strainers of the liquid fuel supply pumps from the storage tank are clean and not choked

4) Replace the Liquid Fuel Bypass Valve servo valve and perform proper servo-valve polarity checks PRIOR to attempting to start on liquid fuel or transfer to liquid fuel. (Again, this is somewhat difficult to do but it is CRITICAL to ensuring the servo-valve is receiving the proper signals for proper operation from the three control processors of the Mark* turbine control system).

You would do well to also check the operation of the liquid fuel flow divider? Does it turn freely? Are the speed pick-up gaps within specification?

Unless there are Diagnostic Alarms on the TCQA card(s) related to liquid fuel devices I wouldn't suspect the TCQA cards--and I haven't asked about Diagnostic Alarms (you know them, the "nuisance alarms" nobody pays any attention to). What are any Diagnostic Alarms present when running on liquid fuel or STARTing on liquid fuel? Do you have any reason to suspect any issues with Mark* V hardware (printed circuit cards)?

You did well to START the machine on gas fuel and transfer to liquid fuel. Did you shut the machine down while running on liquid fuel and then attempt to START it on liquid fuel? If so, what happened? Did you see any exhaust temperature increase while firing on liquid fuel during the liquid fuel START attempt?

Was anyone stationed near the Liquid Fuel Bypass Valve to observe the valve's stem during the start attempt or during the fuel transfer? A video camera can be helpful when doing this, but without LVDTs (which were typically still provided when Mark* V turbine control systems were being supplied with new Frame 9E gas turbines--I'm presuming the machine doesn't have LVDTs on the Liquid Fuel Bypass Valve--or, rather, based on the description you provided and the data that the machine doesn't have LVDTs, which is really and assumption on my part (and I LOATHE assuming anything-sorry!) visual observation is the only way to see what the valve is doing.

Also, did anyone check the Telltale Leak-off drain during or after the initial START attempt on liquid fuel? Or during or after the liquid fuel transfer?

Was any white smoke observed coming from the gas turbine exhaust during the liquid fuel transfer or any subsequent START attempt?

Without LVDTs there's really no way to monitor the Liquid Fuel Bypass Valve to determine if it's following the reference or not.

If you were able to transfer from gas fuel to liquid fuel at FSNL and there isn't any liquid fuel exiting the Telltale Leak-off drain and there wasn't any white smoke coming from the gas turbine exhaust while the machine was running at FSNL on liquid fuel after the transfer, and TNH_RPM was only oscillating +/-10 RPM, I still have to suspect something is amiss with the liquid fuel supply system (air still trapped in the liquid fuel supply piping or filter canisters; plugged liquid fuel strainers; any liquid fuel supply system pressure regulator issue(s); etc.), the liquid fuel purge system (liquid fuel coming out of the Telltale Leak-off is a sure sign of Liquid Fuel Purge system issues), the check valves (liquid fuel and/or purge system), or the Liquid Fuel Bypass Valve servo valve.

 

thanks for your attention.

Air venting of the liquid fuel system was performed again.
By readjusting the 355 pressure regulator, the diesel fuel pressure downstream of the skid was set to 5 bar.
The discharge strainer of the transfer pump was visually inspected and found to be clean.
The flow divider was rotated manually and it turned freely.
The flow divider speed pickups were not checked.
No leakage was observed from the Telltale Leak-off drain.


A fuel transfer at 25 MW was carried out again.
During the first two minutes, the oscillations continued.
By adjusting FSR GAG, the load fluctuated up and down, and afterwards the oscillations were seen to be damped and became nearly normal.


However, with further load increase and upon reaching the Secondary Fuel Valve opening point, the turbine tripped on Flame Off.
Approximately three seconds after the Secondary Valve opened, the primary flame gradually extinguished; one secondary flame became intermittent and the other one fully extinguished, leading to a complete loss of flame and a turbine trip.

 

@abed.a,

Very important detail you didn't include in your original post--and I neglected to ask about: It appears the machine has a DLN-I combustion system. (I can't think of any reason for a Secondary (Liquid) Fuel Valve other than a DLN-I combustion system. If you can share the Liquid Fuel P&ID it would be very helpful (.pdfs and other files can be attached to Control.com posts.)

Usually with DLN-I liquid fuel systems fuel is introduced into the secondary combustion zone as load is increased--but usually fuel to the primary combustion zone IS NOT reduced. Without being able to see the Liquid Fuel- and Liquid Fuel Purge P&IDs it's really difficult to say what might have happened. I suspect there was air in the liquid fuel piping downstream of the Secondary (Liquid) Fuel Valve and that caused the issue you experienced--but that's just a guess without being able to view the P&IDs for the Liquid Fuel and Liquid Fuel Purge systems.

Every GE-design Frame 9E heavy duty gas turbine IS NOT like every other GE-design Frame 9E heavy duty gas turbine. Sure--the basic axial compressor and turbine sections and the number of combustors are all the same, and they can burn natural gas, LPG, butane/propane, and liquid fuels (distillate, naphtha, crude)--but they also do this with two types of combustions systems: conventional (including multi-nozzle quiet combustors) and DLN (Dry Low Nox) combustion systems.

I also presume that the machine uses Water Injection to reduce NOx emissions when running on liquid fuel, and we don't know if the Water Injection system was in use or came into use after the Secondary (Liquid) Fuel Valve was opened.

What you described seems odd to me, but--again, without the P&IDs I mentioned above it's really difficult to understand the system and what might be happening.

Please remember when posting to a World Wide Web forum like Control.com that the more information you can provide (especially the type of combustion system on the machine) is really helpful. Also remember that there very often slight but important differences in the auxiliaries provided with GE-design heavy duty gas turbines. Finally, if GE Belfort, France, was involved in the configuration and supply of the turbine and auxiliaries please, please also be aware that they are famous for making changes to established auxiliaries and operational sequences which can make troubleshooting difficult. Why do they do this? Because they think if something wasn't invented in Belfort, France, then it's broken and needs to be "fixed" and the engineering department at GE in Belfort is perfect to fix what (wasn't) broken to begin with. Sorry, but it's the truth.

Aways refer to the relevant P&IDs when trying to understand systems and system components. AND lastly--remember, MANY components, especially in Liquid Fuel and Liquid Fuel Purge systems ARE NOT under the direct control of the Mark* turbine control system (liquid fuel check valves; liquid fuel supply pressure regulators; liquid fuel purge check valves; water injection check valves; the liquid fuel flow divider)--and ALL of these devices MUST be working properly without any control system adjustment or even monitoring in most cases in order for the machine to work properly. Air in the liquid fuel system causing problems isn't just about air in the liquid fuel supply piping--it's also about air in the liquid fuel piping downstream of the liquid fuel flow divider. If the machine had DLN combustors there are also miniature liquid fuel flow dividers on the cap of each combustion can to split the liquid fuel flowing to each combustor into equal flows to each of the primary liquid fuel nozzles in the combustors--and these small liquid fuel flow dividers are also not controlled or monitored by the Mark* V. You said the machine hadn't run on liquid fuel for several months and that can mean that air can (and often does) get into the liquid fuel lines downstream of the liquid fuel flow divider (even though there are check valves which should prevent that! it does happen) so it's not outside the ream of possibility for air to have to be blown out of the secondary fuel lines on the initial operation of liquid fuel.

Finally, liquid fuel nozzle tips can be plugged with carbonized liquid fuel if not properly purged after liquid fuel operation. If the purge air flow isn't correct (also not monitored or controlled by the Mark*) it may not purge all of the liquid fuel out of the nozzle tip and the high temperatures in the combustor will cause the liquid fuel to burn (carbonize) and plug the small orifices of the liquid fuel tips. Liquid fuel and liquid fuel purge (including the Atomizing Air system (which is used for purging and cooling liquid fuel nozzles when not burning liquid fuel) are very intricate systems, especially with DLN-I combustion systems. And, proper operation and purging of the Water Injection system of DLN-I combustor-equipped machines that burn liquid fuel is also complicated and many of the components of this system are also not controlled or monitored by the Mark*.

GE recommends REGULAR operation of dual fuel machines that can burn liquid fuel--as in at least once weekly. This is necessary to keep the systems free of air and to identify problems before they prevent reliable operation. So, if this isn't done while running on gaseous fuels then problems are more likely to occur with liquid fuel systems when switched to. GE sells some very intricate systems to try to make liquid fuel operation more reliable--but they are also complicated in their own right and GE STILL recommends regular operation of the liquid fuel system to keep components reliable and to identify problems before they prevent or reduce operation on liquid fuel. Most of their publications about dual fuel machines all make this recommendation, and many installations fail to do so because when they first try to initiate such a program the machines trip and management doesn't want that to happen so they just abandon plans to follow the OEM's recommendation.

Best of luck! I suggest you become very familiar with the P&IDs and try another liquid fuel START and see what happens when load is increased and the Secondary (Liquid) Fuel Valve opens. There may be some more load swings as air is pushed out of the secondary liquid fuel lines but hopefully it won't result in a trip, but if it does then that means more investigation and resolution. And, look into regular operation of the Liquid Fuel capability as recommended by the manufacturer.

 

@abed.a
Tivemos uma situação semelhante após a unidade nº 2, Frame 9E, funcionar apenas com combustível gasoso por mais de um ano; ela apresentou o mesmo comportamento. A solução foi substituir os filtros de diesel de alta pressão. A unidade nº 4 apresentou comportamento similar ao da unidade nº 2, mas a solução foi diferente. A válvula de retenção (VCK101) do compressor de ar de atomização principal tinha uma mola quebrada, permitindo que o ar atomizado do booster retornasse (recirculasse para a saída e entrada do CA2) em vez de direcionar todo o ar para atomizar o combustível líquido na câmara de combustão. Nessas condições, a partida da turbina a gás era (e é) impossível, mas a troca de gás para diesel faz a unidade funcionar, pois ela assume que não está mais funcionando como booster, mas sim como o compressor de ar de atomização principal (CA1-1). Esta é provavelmente a causa do problema. Outra possibilidade seria a válvula reguladora de pressão a montante do filtro de baixa pressão. Como a turbina a gás opera com gás combustível por longos períodos, a válvula de controle de pressão do diesel permanece imóvel em uma única posição por muito tempo (no seu caso, mais de 6 meses), o que causa seu travamento. Por ser uma válvula mecânica sem feedback de posição, é difícil saber se está funcionando corretamente (o mau funcionamento da válvula afeta a pressão a montante da bomba de alta pressão). Recomendo que você verifique primeiro a válvula de retenção do ar atomizado e, em seguida, substitua o filtro de alta pressão. Se o problema persistir, verifique também a válvula reguladora do filtro de superfície. Não se esqueça de nos informar qual solução você encontrou para resolver esse problema.

 

**TRANSLATING @JULIÃO NGAIE'S POST #5, ABOVE, FROM PORTUGUESE TO ENGLISH**

@abed.a,

We encountered a similar situation after Unit No. 2, Frame 9E, operated exclusively on gaseous fuel for over a year; it exhibited the same behavior. The solution was to replace the high-pressure diesel filters. Unit No. 4 exhibited behavior similar to that of Unit No. 2, but the solution was different. The check valve (VCK101) on the main atomizing air compressor had a broken spring, allowing the atomized air from the booster to return (recirculate to the outlet and inlet of CA2) instead of directing all the air to atomize the liquid fuel in the combustion chamber. Under these conditions, starting the gas turbine was (and is) impossible, but switching from gas to diesel makes the unit run, as it assumes it is no longer functioning as a booster, but rather as the main atomizing air compressor (CA1-1). This is likely the cause of the problem. Another possibility would be the pressure regulating valve upstream of the low-pressure filter. Since the gas turbine operates on fuel gas for long periods, the diesel pressure control valve remains stationary in a single position for a long time (in your case, more than 6 months), which causes it to seize up. Since it is a mechanical valve without position feedback, it is difficult to know if it is functioning correctly (valve malfunction affects the pressure upstream of the high-pressure pump). I recommend that you first check the atomized air check valve and then replace the high-pressure filter. If the problem persists, also check the surface filter regulator valve. Be sure to let us know what solution you found to resolve this issue.

Tchau!