hello guys,

while starting on diesel fuel, a strange phenomena happens which is from the moment the L3FIRE is true,and the machine fires, during the 60 sec wrm-up, there is very high fluctuation in FAL from 36% to -7%, then followed by a flactuation in diesel flow A60FQI from 0.2 gpm to 11 gpm, this caused the flame to come on and go off several times, the FSR1 was fixed at 19.5%, and the FQROUT was 3.6%, and the FQL1 was flactuating between 0.6% to 7%, (the TCQA requlator current-gain is 1.55 & current-bias is 2.71).

these parameters were flactuating until the speed reached about 46%, at this speed, all became much smoother, and all flactuations normalized,and the machine proceeded to accelerate, until the speed reached 90%, at this speed, very huge increase in diesel fuel took place, which went up from 14 gpm to 37 gpm, followed by very sharp flactuation in FAL, FQROUT, and the FSR1, it went on until a change-over to gas command took place. a previous change from Gas to Diesel on the same machine was carried over before, but the machine tripped on Turbine Underspeed trip, which caused by very high diesel flow which might cuz the flame to be less tensive, so i really wish if i get any feedback analyzing the phenomena.

i tried to post a sample of the view2 here, but it didn't work, so if any of you guys interested, i can send you the text document of the view2 to better analyze this problem.

any feedback will be highly appreciated.
thank you in advance.

Omar
[email protected]

 

This is EXACTLY what happens when there is air in the distillate fuel piping upstream of the high-pressure liquid fuel pump.

Air not only gets through the piping and through the high-pressure liquid fuel pump and liquid fuel flow divider, it gets trapped in high points in the liquid fuel piping--even in the tops of liquid fuel filter canisters. When air is trapped in the piping, because it is compressible it compresses and expands and causes the liquid fuel flow to be erratic--that is what the Mk V is RESPONDING to.

You say the reference (FSR1) is stable but the servo-valve current (FAL) is varying--that's because the servo-valve output regulator is trying make the liquid fuel flow feedback equal to the reference.

Another thing that happens when air is trapped in the piping is that when the liquid fuel flow gets high enough the air can be "pushed along" and make it's way to the high-pressure liquid fuel pump and through the flow divider.

There is one other thing which could be causing your problem--and that is the liquid fuel forwarding skid discharge pressure regulator. If it's unstable, this is also the kind of thing that can happen. Of course, if there is air in the piping the stability of pressure regulator will be affected, also. So, it's kind of difficult to pinpoint.

MANY times the people designing and installing the liquid fuel forwarding piping system do NOT put vents (valves) in the highest points in the piping as it travels from the discharge of the forwarding skid to the inlet at the accessory base. These vents are absolutely necessary if the piping goes up into a pipe rack and then down into a pipe trench and up into the accessory base inlet--or up and down through any possible configurations to get to the accessory base. Go and determine if you have any high points in the piping, and if so, if there are any vents that can be manually opened to vent the air from the piping.

As previously mentioned, there ARE (or there SHOULD BE) vents in the tops of the liquid fuel filter canisters. Sometimes there are two sets of duplex liquid fuel filters: a low-pressure set before the high-pressure liquid fuel pump and a high-pressure set after the hig-pressure liquid fuel pump. Open the vents and remove the air from both sets, if so equipped. And open the transfer fill valve between the filter canisters and open the vent valve on the standby filter canister(s), also.

There is more than likely NO problem with the Mk V or the Frame 5. The throughput of the high-pressure liquid fuel pump is a function of the pressure at the pump suction (from the liquid fuel forwarding skid--controlled by the pressure regulator) and the ABSENCE of air in the liquid fuel piping system between the forwarding skid and the high-pressure liquid fuel pump. If the pressure is unstable either because of problems with liquid fuel forwarding skid pressure regulator or air in the line, the Mk V will behave EXACTLY as you described.

markvguy

 

Sounds like either mistuned or faulty regulator, other startup problems on GE units sometimes related to noise attenuation on the flow divider speed pickups which can be filtered by connecting in parallel to spare speed pickup inputs. I would like to know does the unit bounce in and out of Startup FSR and Acceleration FSR control if it does the controls are not performing correctly. A correctly tuned system after warup complete should NEVER leave Acceleration control until FSNL

 

thanx guys,

so far, the idea of the air tripped in the diesel system piping is the most possible, after we vented the system from air,it seemed much smoother, but we still have that kind of delay in the diesel flow, from the other side, when we do a changeover from gas to diesel,the speed drops to 99.1%, which cuz the drop in load when the machine is loaded, my question is:- do you think guys that the only problem is the air in the system? or there might be another reason for that?

thank you mkvguy and Eoin

 

thanx guys...

we found out that the presence of air in the diesel system piping was a serious problem, after we vented the system, we gave a try, and all the flactuations in the previous trials were smoother, but there is still a delay in the diesel flow in both start-up and changeover to diesel, and this causes a drop in speed when the machine is not loaded, and when the machine is loaded, the drop will be in Load, about 2.8 MW.

my question is: is it only the air that causes the this problem or there might be another problem?

thank you mkvguy and Eoin for your participation.

 

No; thank YOU for the feedback. It can't be said enough: Feedback is what makes this forum work best for everybody!

There may be some tuning on the system which can be done during a fuel transfer, things like slowing down the transfer to allow the forwarding pump/pressure regulator to keep up with demand. But, that's usually what most people don't want--to slow down the transfers; they want to speed them up.

Are there sufficient vents in all the high points in the piping--and are they properly placed? They should be in the highest point(s) of the piping, and closest to the downward turning elbow--as air is pushed along to the end of the horizontal run when there is flow. Sometimes it takes days to vent all the air which can be pushed along to the filter canisters--if it can be. Operating at high loads (fuel flow-rates) on liquid fuel and bleeding air from canisters and properly placed high point vents usually works best, and it usually takes a few tries to get all the air out--especially if the Liquid Fuel Forwarding Stop Valve allows fuel to drain back to the storage tanks during periods when the unit is not being operated on Liquid Fuel....

Fuel transfers can usually be made to be as "bumpless" as possible--but not at any load and not if the unit isn't operated frequently or at least periodically on liquid fuel. How often is your unit operated on liquid fuel? Are you sure that all the purge air check valves and liquid fuel check valves are working properly? (Check the liquid fuel tell-tale leak-offs (see the Liquid Fuel and/or Purge Air piping and instrumentation/piping schematic diagram) to see if the purge checks are all working properly; they might not be sealing immediately, so watch them as the transfer begins.)

markvguy

 

Why do not you compare the plot trend of dist flow with the plot trend of speed or dwatt when the unit is at full speed or online? If they are noticed the same, the air may be the only cause.

 

Okay, so it seems that the unit start-ups are better now but the transfers from gas to distillate are a problem--is this what you're saying? Are you saying that during a transfer the fuel flow doesn't "start-up" as fast as it "should?"

A 0.9% drop in speed during a fuel transfer at FSNL (Full Speed-No Load) isn't very bad at all, and a 2.8 MW swing in load during a fuel transfer isn't all that bad, either (though that's about a 10-15% change in load for a Frame 5...). Most field engineers will tell you that a 10% load swing during a fuel transfer isn't too bad (in fact, most are thrilled to get that small a bump during a transfer).

During a transfer from gas to distillate there are MANY things which happen--the Liq. Fuel Forwarding Pump is started; then the Liq. Fuel Forwarding Stop Valve is opened and the Speedtronic usually checks for sufficient Liq. Fuel pressure upstream of the Liq. Fuel Stop Valve before opening the Liq. Fuel Stop Valve and energizing the Liq. Fuel Pump Clutch; then the Liq. Fuel Bypass Valve is modulated to control Liq. Fuel Flow-rate through the Liq. Fuel Flow Divider. At some point during the process, the Liq. Fuel Purge air has to be shut off and, if so equipped, the Gas Fuel Purge air has to be enabled. There are MANY things which have to happen in the proper order and at the proper rate during fuel transfers, so there are many possibilities for problems.

If you are experiencing a momentary drop in speed/load during a transfer from gas to distillate you are going to have to determine where the delay is coming from. As Muggsy suggests, using the Real-Time Plot trending feature may be helpful.

You haven't said if the Liq. Fuel Forwarding Skid Pressure Regulator is properly maintaining the desired pressure during the fuel transfer--this is very important to a stable change-over. Many times, when the Liq. Fuel Stop Valve is opened the forwarding skid discharge pressure drops as flow increases--and this causes a problem. The pressure regulator must be capable of and adjusted to maintain desired skid discharge pressure from low-flow conditions to full-flow conditions, and this usually takes some "tweaking".

This author has also seen sticky Liq. Fuel Forwarding Stop Valves that took as much as 20-30 seconds to open that caused problems establishing proper flow-rates. There have also been problems with Liq. Fuel Pump Clutches slipping that have caused drops in liq. fuel flow-rate during transfers.

During a fuel transfer from gas to distillate, FSR1 (Liq. Fuel FSR-Fuel Stroke Reference) is ramped up at the same rate that FSR2 (Gas Fuel FSR) is ramped down. Depending on the vintage of your Mk V, the signal names might have changed but usually the servo-valve output regulator uses FQROUT as its reference and adjusts the servo-valve output to make FQL (sometimes FQLM or FQLM1) equal to FQROUT--check your CSP to get the proper signal names.

You also haven't said if there is any liquid fuel flowing in the tell-tale leak-off(s) during a liquid fuel transfer--as was said earlier, if the purge check valves don't seat properly there can be a problem.

The problem is most likely NOT in the Mk V, it's something else. You've already seen what the presence of air in the system can cause. The Mk V, as sophisticated as it is, requires that many components in external systems operate properly--and gets incorrectly blamed when they don't. Since the Mk V doesn't control the Liq. Fuel Forwarding Skid discharge pressure, it's at the mercy of the pressure regulator. If the Liq. Fuel Forwarding Stop Valve doesn't open fast enough (it should be open in less than 1 second), the Mk V can't compensate for that. If the Liq. Fuel Pump clutch slips or the Liq. Fuel Purge Check valves don't seat properly, the Mk V can't do anything but try to compensate for that.

So, make sure all the external components are working properly--and there's no air in the lines.

Lastly, if your unit hasn't operated very much on Liq. Fuel in the past, and just "recently" the need to operate on Liq. Fuel has arisen and the unit has been primarily operated on Gas Fuel with very few start-ups or transfers to Liq. Fuel, you're actually VERY, VERY lucky that these are the only problems you are experiencing. Usually, there are all kinds of problems with clutches and check valves and pressure regulators and air in the fuel lines and bacteria (YES--bacteria!) in the Liq. Fuel Filters with units which haven't been operated on Liq. Fuel very often and "suddenly" need to be run on Liq. Fuel. Everybody just thinks, "Oh, it's a dual-fuel unit! We'll just switch to Liq. Fuel!" and there won't be any problems--WRONG!!!

GE recommends that if you have a dual-fuel unit, that you regularly, at least once-per-week, perform fuel transfers to make sure that everything works as it should when it's vital to perform fuel transfers. Problem is, most fuel transfers during initial start-up and commissioning don't go very well, and if fuel transfers are only performed once or twice a year they don't usually go very well, either. So, most plants don't perform fuel transfers until the Speedtronic does it for them (on loss of pressure)--and those don't usually go very well, either!

But when one stops to consider all the devices which must work properly and in the proper sequence to perform a successful fuel transfer, it's really fairly surprising that they work as well as they do when they do work....

Now, if you're saying that when the unit is loaded and you perform a fuel transfer that when the transfer is complete the load is 2.8 MW lower than it was on gas fuel, then what needs to happen is a procedure that's called "fuel matching". And that's outside the limits of discussion in a public forum such as this--there's just too many variables which most people won't take the time to answer to be able to say how to properly perform fuel matching.

markvguy

 

All of you have made some very good points about liquid fuel system reliability and have touched on many issues which significantly impact operational capability of the liquid fuel system on dual fuel configurations.

In my experience, coking of liquid fuel check valves has been the most prevalent cause of fuel transfer issues. Specifically, while there is more than adequate liquid fuel pressure to force coked liquid fuel check valves to open, the return springs are not strong enough to force them to close completely. Leaking check valves combined with unequal combustion pressures in combustion cans will create a phenomenon called cross-talk. Cross-talk is defined as the migration of combustion gases from higher pressure combustor to lower pressure combustors. The combustion gas circulation path is past the leaking check valve seals, through the flow divider and out to the lower pressure combustion cans. This is what causes the complete evacuation of the liquid fuel from your system when you are operating on gas.

A company called JASC, http://www.jasc-controls.com, has solved the coking check valve problem. They developed a water cooled liquid fuel check valve which eliminates coking and allows the check valve to maintain a class 6 seal. This improved functionality results in liquid fuel systems which start and transfer reliably. Whether the turbine operates at baseload or is a peaker, the design can also be applied on any dual fuel turbine engine. To date, approximately 17 engines have been outfitted with this technology. Collectively, these units have availability and reliability which exceeds 98% for the liquid fuel system. Obviously, there are only a few instances where installation of this valve design alone can bring about such dramatic results. However, there are comprehensive solutions available from this company. They have designs for both the purge air and water injection systems as well.

Next, flow divider upgrades are also needed to round out the reliability package. This can be accomplished by having Roper Pump either refurbish your existing flow divider to the new DuraFlow design or purchasing a DuraFlow for a swap out during your next outage. This upgrade is required because the check valve upgrades will allow you to exercise your turbine at intervals of once every three or four months without sacrificing reliability. As the DuraFlow design is corrosion resistant, entrained water in fuel will not cause the flow divider to rust and corrode during these long periods of inactivity.

Finally, I have witnessed first hand the benefits of these new technologies and in some cases helped to develop them. I would be happy to share what I have learned with anyone who is interested. I can be reached at 864-350-4332 and my name is Schuyler (Skyler) McElrath. Please feel free to either call me or send an email. You can find presentations which I have given at recent Gas Turbine User conference on the JASC website.

Schuyler McElrath

SMTC, INC.
PO Box 17203
Greenville, SC 29606
Phone: 864-350-4332
Fax: 864-220-5913
Email: [email protected]

JASC
Jansen Aircraft System Controls
2303 West Alameda Drive
Tempe, AZ 85282
Website: www.jasc-controls.com
Phone: 602-438-4400
Fax: 602-438-4420
Email: [email protected]
Email: [email protected] Technical product information

 

What if you are having problems with nuisance trips during transfers due to failed purge check valves, is there anything on the market to prevent this from happening?

 

GE had a very big initiative several years ago to identify the root cause of liquid- and purge check valve problems. As a result, they tested and and tested and tested many different styles of check valves and now have some very good check valves available which are very effective at reducing the types of failures that were being previously experienced.

In most cases, the old P/Ns (Part Numbers) have been superseded by new P/Ns in GE's parts listings--but it's always best to ask about the details of any check valve you order.

The biggest thing operators can do to prevent this type of failure is to regularly exercise the Liquid Fuel System by performing scheduled fuel transfers (like on Saturday evenings at midnight--nothing like adding a little spice to that shift!). This keeps good fuel in the filters, helps to keep air out of the lines and filters, and helps to keep the high-pressure Liquid Fuel Pump and flow divider all in working order, not to mention exercising the the liquid- and purge check valves and the Atomizing Air system. Run for at least half an hour or so on liquid fuel, then transfer back. Watch the Tell-tale Leakoffs for evidence of failing purge check valves.

It's also a good idea to stop and start on liquid fuel once in a while. It's best to transfer to liquid fuel then shut down. Then, start on liquid fuel--that way there's no air in the lines when starting and everything's pressurized since the unit's already been running and was shut down while running on liquid fuel.

markvguy

 

I have a solution to the failure mode caused by water injection check valve on dual fual combustion turbines.

Lets take a GE 7EA dual fuel combustion turbine, there are 20 3/4" Parker water injection check valves that fail and allow air to back up to the isolation fail. Without knowing which check valve has failed all twenty must be inspected. Downtime can be very costly at certin times of the year. Below is the solution and vendor information.

If a DFT inline silent check valve is installed on the main water supply line near the turbine in the turbine room it will not allow the air from the failed waterjection check valves on the turbine to reach the isolation valve. Below is a common configuration.

3" 600# DFT check valve sized per flow with a 2 PSI cracking spring and viton soft seat. Sizing the valve is very important in not allowing fludder in the valve. The 2 PSI cracking spring is also important to keep a tight seal.

TVA in Tennessee has had great success with this valve stopping the failure of the parker water injection check valves causing failure modes on these dual fuel combustion turbines.

Steve Calvert
DFT - Check Valves
Regional Manager
2177 Cincinnati, OH 45244
513-307-6523
[email protected]