High Spread While Burning Liquid Fuel

Greetings to all.

I will try to provide as much details here, so we can learn from each other.

The unit I speak about is 7 FA turbine, dual fuel with DLN2. The unit has 14 Cans and 27 thermocouples to monitor the exhaust temperature. The igniters are on Cans 2 and 3. The flame detectors are on 11, 12, 13, and 14. The liquid fuel system is equipped with 3 way purge valves and re-circulation system.

Usually we keep the re circulation ON during summer when the unit runs on gas. We run the unit on liquid for testing in the winter. the nitrogen purge system is not ready yet. so when a problem happens in the re circulation system and get isolated, the liquid remaining in the 3-way purge valve cokes since it can't be purged with nitrogen (nitrogen purge system is not ready yet).

Recently, we run a turbine that had the re circulation system off during gas operation for long period of time. so high exhaust spread is normal and expected. The main purpose of this thread is that how can we determine the problematic Can from looking into the exhaust temperatures?

Before running the turbine, we made sure that the liquid lines are free from air. The unit started from turning gear [Liquid Mode]. The spreads are shown in the following link. After taking 10.25 Seconds for purging, the firing started. I know that there is a way to check which valve has the problem by using the selector valves, but two problem arise here.

1- The flow divider has 14 ports for 14 Cans. In some units, the selector valve is not working for some ports. I mean for example I can read the pressure for 12 ports while the remaining 2 ports has their selector valve damaged.

2- I don't think It is accurate to measure the pressures at each selector while the unit is accelerating, because the back pressure is increasing. and when the unit reaches 95% of rated speed, it would trip if there is high exhaust temperature. So, there is no enough time to get an accurate measurement.

3- One of CSA replies says that if the pressure at one of the 3-way valves is less or more than 10% of the average pressures, this valve would be suspected. But consider a case where 5 or 6 of the valves have problems and the other 9 or 8 valves are intact. wouldn't that affect this method.

Finally, Checking the selector valve during acceleration didn't come to my mind because I was told by a control engineer from GE that you can determine which 3-way valve mostly have the problem by using the exhaust temperatures from the 27 thermocouples. During our last start-up in liquid we brought the turbine up to approximately 3200 RPM then stopped the turbine to avoid trip. The exhaust temperatures are shown in the link, and according to the GE Control Engineer, he advised to inspect the 3-way valves at Cans 1 and 14. He also added that It is better to inspect Can 13 also since it established the flame late. Frankly, I don't know how he has reached this analysis.

Just to clear one more thing. The arrangement of the annular Cans are as such. The top one is Can 14. The lower one is Can 7. And counting is in anti-clockwise direction. Please don't direct me to previous threads, As I have already read many.



Best Regards
To update you on this case. It seems there is a GE internal tool to help you determine which Can. In the tool you put the MW and the cold thermocouple and guide you which Can should be inspected. As you change the MW value, the positions of the thermocouples change.

Furthermore, the I&C team actuated all the 3-way valve using Instrument Air and found the valve at Can 14 doesn't move. Will update you then.

I believe the tool you are referring to is called a "swirl chart." It's called a swirl chart because the hot gases traveling through the turbine section after leaving the individual combustors do not mix and homogenize, rather they "bend" as it were. So, if combustor #14 was not lit or was having poor atomization then the cold spot resulting from combustor #14 might appear at T/Cs 22 & -23, for example (because the swirl is with rotation, anti-clock.

Further, the angle of movement (the swirl angle) changes with load and IGV angle. Usually, when the unit is at Base Load, the swirl angle is near zero degrees--meaning that the hot/cold spot is nearly directly downstream of the problem combustor(s).

However, while I know for a fact that GE has a swirl angle chart/tool for loaded operation, I don't know that they have a swirl angle chart for acceleration. So, while this tool may be helpful in <i>trying to determine</i> which can(s) are causing the spread (hot or cold), it's only a tool. It's better for some machines than others, and it's only and estimate. I've seen more often than not that the combustor(s) indicated as problematic per a swirl chart were not the actual combustors with the problem, but the problem combustors were located one or two positions on either side of the combustor indicated on the swirl angle chart.

I will also add that GE has done a LOT more work on refining the swirl angle chart for F-class machines than for the other turbines in the fleet, and I have seen results of using a swirl angle chart be a little better for F-class machines in recent years.

The recirculation systems I have seen have nothing to do with purging the liquid fuel out of the liquid fuel nozzle passages while running on gas fuel. The main purpose of the recirculation system is to keep the liquid fuel system purged of AIR and keep it full of liquid fuel. I presume that in some cases, it is possible for liquid fuel UPSTREAM of the three-way valves to get carbonized and cause problems but my understanding of recirculation systems is to keep the liquid fuel systems purged of air and full of fuel so that if the unit needs to transfer to liquid fuel while running that it can do so quickly without load swings or trips due to air in the lines.

The purposes of the liquid fuel purge system are, 1) to purge liquid fuel out of the nozzles (including the distributor on the combustion can covers (the miniature fuel flow divider to evenly divide the incoming fuel flow to all of the liquid fuel nozzles)), and, 2) to provide cooling to the liquid fuel nozzles while running on gas fuel, and, 3) to prevent the backflow of hot combustion gases into the liquid fuel nozzles and distributor while running on gas fuel. But the recirculation system doesn't purge the liquid fuel nozzles and fuel distributor (the components downstream of the three-way valve).

A non-operating three-way valve could certainly cause problems, as could carbonized liquid fuel which made it's way into the flow distributors and fuel nozzles. This would be very difficult to pinpoint even with a swirl chart. To my way of thinking, knowing that the recirculation system wasn't working on this particular machine the thing to do is to pull all of the equipment downstream of the three-way valves--including the three-way valves--and check them for cleanliness and operation. This would also provide information about what may have happened which would definitely be useful in trying to mitigate further problems. (For example, if you found high amounts of carbonization in particular combustors, and you reviewed the layout/placement of the tubing/piping in relation to other combustors you might consider making and installing insulation to try to help minimize heat transfer. (I've always thought the flow distributors on the combustion can covers were ripe for problems, mostly because of the heat they are subjected to, and if insulation wasn't a possible solution.)

And, you are right in that trying to check flow divider pressures during acceleration isn't going to be very helpful because of the increasing pressures during acceleration. I don't know if F-class machines have FIRE mode in liquid fuel (I seem to recall the last units I worked on did). If the units at your site do, you can select FIRE (instead of AUTO) prior to STARTing the unit, and after flame is detected and warm-up is complete fuel flow will be held constant and you can then use the selector valve/gauge at the flow divider. And, repair the non-working selector valves. Perhaps buy a new one, and swap it for one non-working one, refurbish the non-working one and swap it for another non-working one until you have corrected the problem on all of the units. Really, for liquid fuel using the flow divider pressures obtained via the selector valve/gauge is the BEST way to pinpoint problem combustors. There is no such thing available for gas fuel--and the selector valve/gauge is really important for troubleshooting liquid fuel spread issues. Much better than the swirl angle chart.

Some sites actually have installed pressure transmitters on all of the liquid fuel lines (14) and connected them to the Mark* so that they can be trended/monitored at all times. This would be the ideal way to monitor pressures during acceleration/deceleration, as well as during loaded operation and firing--at all times, really. Fast trends can be very helpful during troubleshooting, and given the problems associated with liquid fuel if it is critical for liquid fuel to work when needed then this information available on trends is ideal and really necessary.

In my personal experience, running a dual fuel (gas/distillate) GE-design heavy duty gas turbine almost exclusively on gas fuel and expecting it to transfer to liquid fuel at a moment's notice without hiccup or to start and run on liquid fuel once or thrice a year is just not realistic. GE has designed several versions of recirculation- and purge systems and redesigned the check valves, replacing it with the three-way valves (which it seems are also subject to their own issues, though maybe fewer issues than the check valves were)--and still, owners/operators say they have reliability issues. The only owner/operators I know that don't have issues have a regular and periodic schedule for liquid fuel transfers--as GE recommends, which is really once per week. Is it risky (meaning, can the unit trip?)? Yes. But, over time and with experience the operator, technicians and maintenance personnel learn how to keep the systems (of which there are many) and the components working. And, over time, the reliability improves--usually greatly.

Know, also, that there will continue to be developments in the systems and components. If you own the machines, and they are not under some kind of LTSA or CSA agreement, you are free to talk to vendors who may have new or different components which you can install and test. The OEM is NOT the only game in town when it comes to finding good help and information.
Do you experience check valves stuck when not running on liquid fuel for some time and how do you deal with it?

Do you also experience MOOG valves become ungovernable for the same reason?

>Do you experience check valves stuck when not running on
>liquid fuel for some time and how do you deal with it?

I'm not the original poster, but from one of the screenshots in the OP's original thread it doesn't look like this unit uses liquid fuel check valves, or liquid fuel purge check valves. Instead, there is a three-way valve which switches between purging and liquid fuel as necessary. Kind of a neat way around the check valve issue. I believe the three-way valves are air-operated (pneumatic), and have solenoids (possibly more than one) to control the air to the actuators for switching. But, these valves have been known to stick, but seemingly don't have the issues that check valves have been known to have. Kind of a nice design if the reliability issue can be solved--and I'll wager a strong bet that the lack of recirculation may be a contributing factor to the valve that is reported not to have changed position during off-line testing. But, in my opinion, the three-way valve is a good solution to the check valve issue.

Some sites have gone to water-cooled check valves (see jasc-controls.com) and have good success with them--but there is more tubing for the water-cooling....

>Do you also experience MOOG valves become ungovernable for
>the same reason?

Servo valve problems are usually the result of bad oil quality, and for liquid fuel bypass valves (LFBVs) if the oil quality is good then the most likely causes of LFBV instability are: 1) air in the liquid fuel supply line, usually trapped in high spots in the supply piping, or in filter canisters; 2) unstable liquid fuel supply pressure (pressure regulating valve problems; or, 3) flow divider speed pick-up wiring issues.

The reference for the LFBV is liquid fuel flow-rate, and the feedback is the speed feedback, converted to liquid fuel flow-rate. So, if there is noise on the wiring (usually caused by incorrectly terminated shield drain wires), or loose speed pick-ups, or worn toothed wheels, or even sometimes wrong speed pick-up gaps, this can cause instability problems.

Many LFBVs use LVDT feedback, for stability purposes, and many newer units (Frame 9E, and F-class units in particular) don't have LVDTs any more. But, problems with LFBV LVDT feedback can also cause issues--but it's not common. LVDTs, properly maintained, are VERY stable devices and don't experience much drift (that's why the SAME LVDTs are also used on personal, commercial and military aircraft and rockets--they are stable and don't have many drift issues, like some transmitters can experience).

Hope this helps!

Thank you for giving a background about the swirl angle (my experience is only one year, and I am really happy to learns from experts like you), and yes as you said. The swirl angle is not really accurate. The GE engineer told me that when he suspects a particular Can, he adds one Can at each side so they become three instead of one.

About the Firing mode on Liquid Fuel, yes it can run on Fire Mode with liquid. But I believe the pressure of the fuel nozzles will be too much low at warm-up condition (maybe 15 Psig ?) do you think this would be enough ? .And add to the fact that the gauge of the selector valve is analog not digital. If the pressure of the nozzle is around 15 psig and the range of the gauge is 1500 psig, this would be difficult

About the recirculation system, I have decent information about it. The purpose of the recirculation system:

1- To keep the lines free of air using the vent standpipe (high point vent), this vent pipe has low level switch and high level switch. As air is being vented from the standpipe and liquid level increases, the liquid will activate the high level switch and the vent valve will close, once the liquid level drops due to accumulation of air, the high level switch will be deactivated and the vent valve opens for venting the air.

2- You know the fact that keeping the valves cool decrease the chance of coke formation. How can recirculation system achieve this ?
By transferring the liquid in the 3-way valves ( that would be heated inside the turbine compartment if stays in static condition waiting for changeover) so recirculation transfers that liquid into areas outside the turbine compartment and new liquid will come, and this happens all the time while recirculation is ON. Because we have a constant flow of liquid through the 3-way valve. By this way we keep liquid below the threshold of coke formation which is around 250 F

* The three way valve has a small opening that if the liquid can't go into the combustor during purge air mode, it will travel through that hole and then into small pipes and go outside the turbine compartment. Remember: Flow always there, liquid leaves the turbine compartment and new liquid enters the 3 way valve. (Keeps going like that if the recirculation is ON)

3- The three way purge valve is used for purging the fuel nozzles while operating on gas. The valve has two modes: purge air mode and liquid fuel mode.

The valve is biased into purge air mode by the help of two springs. Once the machine starts on liquid, the pressure of the liquid fuel will increase (because the liquid pump will start on changeover) this liquid pressure will shuttle the spool into liquid fuel mode. Once the pressure of liquid becomes low, the valve will return into its biased mode (Purge Mode).

4- One of the purpose of running the unit on liquid is to exercise the flow divider to prevent gears from binding. This is achieved by the recirculation system since you have a constant flow of liquid passes through the flow divider.

GE recommends running the unit on liquid quarterly, if you keep recirculation always ON during gas operation,
If you don't keep it ON, then GE recommends weekly transfer.

I'll try to get these selectors repaired and buy new spare parts anyway.
Dear Jolek

>Do you experience check valves stuck when not running on
>liquid fuel for some time and how do you deal with it?
>Do you also experience MOOG valves become ungovernable for
>the same reason?

Although some plants with 7FA Turbine have the conventional liquid fuel system (liquid fuel check valve + purge air check valve), but ours is supplied with 3-way purge valve.

The 3-way purge valve replaces the (liquid check valve + purge air valve) into one single valve. The benefit of the three way valve is clearly noticed when you keep the recirculation ON during gas operation. Once your the recirculation system isolated for long time of period, the 3-way valve would fail just like the conventional check valve. please refer to my reply to CSA for more information
Oh my bad! I forgot that the cracking pressure of the 3-Way Valve is around 120 Psig. So the pressure at the selector could be (120 + 15) Psig. So I think checking the selectors at fire mode worth the trial.

>About the Firing mode on Liquid Fuel, yes it can run on Fire
>Mode with liquid. But I believe the pressure of the fuel
>nozzles will be too much low at warm-up condition (maybe 15
>Psig ?) do you think this would be enough ? .And add to the
>fact that the gauge of the selector valve is analog not
>digital. If the pressure of the nozzle is around 15 psig and
>the range of the gauge is 1500 psig, this would be