we have Ge frame 5 5001 P machine, RPM 0-5100, when start up, initial firing also exhaust temperature went high. This is long time pending issue. It's a dual fuel firing machine. In both fuel also same thing happens. From firing to FSNL it takes only 2.3 Minutes. Acceleration rate is 25.5 RPM/sec beyond 2400 RPM. we need smooth start up without exhaust temperature high. Initial FSR rate is set as 38 for liquid and 30 for gas. Total 18 thermocouples are there. Almost 8 thermocouples went high during start up. There is no error in measurement part. No error in valve operation. Can anybody suggest how we can tune the machine for the perfect start up?
We have created thermocouple rejection screen in HMI to avoid start up trip due to exhaust temperature high. I have doubt in start up time, 2.3 minute to reach FSNL. Is that so fast? then what's the optimum acceleration rate of GE frame 5 machine?
2.3 mins. from Fire to FSNL is way too short. What is your AccelFSR set at? I don't have a Frame V Control Spec with me, maybe someone has one and can tell us. That would explain your high TTXM during S.U> and should help with a lower Accel rate.
Please write back and let us know.
Acceleration rate set at 0.5 %speed/sec at 40 % speed and above. At 2400 RPM exhaust temperature avg is 970 F. Firing FSR is 30 %.P2 pressure control is ok.
Um, ..., er, ..., Does this turbine have a GE Mark* turbine control system? It would seem to be very difficult to create an HMI display in CIMPLICITY to "reject" exhaust T/Cs--BUT I have seen such screens in WonderWare and FactoryTalk PLC-based turbine control system HMIs....
Most GE-design B/E-class heavy duty gas turbines I have worked on had factory acceleration rate setpoints of 0.10 %/sec or 0.11 %/sec above 50% speed. 0.5%/sec is extremely aggressive.
Also, unless there's something we don't know about the fuels used at the site, Firing FSRs of 30+ % are also extremely high--and would most certainly result in exhaust temperature spikes after flame is established. For a Frame 5- or Frame 6 GE-design heavy duty gas turbine the liquid fuel bypass valve (if the unit was produced in the 1980s or after) would NOT have LVDTs), but the GCV would--and if it was calibrated improperly it might require that kind of reference for firing, and/or if the fuel being burned was low BTU fuel or something like that.
This doesn't seem like a "typical" controls set-up, which makes it seem more like there's a lot we don't know about the situation/circumstances.
The good news is: The GE-design B/E class heavy duty gas turbines were built to be tough and rugged and had lots of engineering design margin and safety margin built into them. They are robust machines and can generally take a licking and keep on ticking. However, overfiring during starting and acceleration (for whatever purpose) is NOT good for hot gas path parts (combustion liners; transition pieces; first-stage turbine nozzles, and buckets), and can even cause pretty serious damage to exhaust diffusers.
This T/C rejection business is something that older, analog GE Mark* turbine control systems used to have, and when control system integrators retrofitted those older GE Mark* turbine control systems they often created HMI displays and logic in the PLCs they used to replace the Mark* with that duplicated this functionality--even though it's not necessary with just about any modern-day, digital programmable logic control system. Modern-day thermocouples are more robust and don't fail nearly as often, and don't require manual cold-junction compensation like the older GE Mark* turbine control systems did. GE certainly doesn't provide this functionality on their newest Mark* turbine control systems when upgrading the older Mark* systems. But, most control system integrators who don't have much, if any, Mark* turbine control system experience just duplicate what they are replacing and programmable logic control systems lend themselves very well to that (whereas CIMPLICITY and digital Mark* turbine control systems don't; I don't know if there's a way to "reject" an I/O signal in most Mark VI or Mark VIe control systems). Sure, one could use the formula and maths capability of CIMPLICITY to do all sorts of things, but I don't know if one could program the Mark VI or Mark VIe to reject signals from the TTXM or TTXSP algorithms.... Probably could be done, but should it? (I can think of one way right now--but it's not pretty and it totally defeats the combustion monitor purpose. But, if the turbine control system has a combustion monitor function it's not usually enabled until some time AFTER reaching rated speed. If it's a Mark* turbine control system.)
Anyway, it seems there's a lot more to this story than has met our eyes. I would say someone with knowledge of GE-design heavy duty gas turbine operation and philosophy should be brought to site to review the operating parameters (fuel control valve firing settings; acceleration settings; who knows what else) and make recommendations. On the other hand, if the unit has been running like this with no problems for some time (years?) and the maintenance outages haven't revealed any major concerns, then it might be okay. Again, older GE-design B/E class machines were rock solid, and that's how GE got such a good reputation (which has now been squandered some would say). Some owner/operators would kill for the kind of acceleration rate this unit has--but most newer GE-design heavy duty gas turbines (particularly F-class machines) just couldn't last very long (probably not even months) with severe duty starts and acceleration like this unit is experiencing.
Something is missing from this story. Which is one reason I didn't comment when it was first posted....Will be following this thread as it unfolds.
Thanks CSA for the reply.
You are right; this platform is Allenbradly control logix L6 processor, with RS view Supervisory edition. This unit is operating since 2007 after control system upgradation from mark 2 to control logix. Now we want this machine as black start machine for new GE 9 HA machine. No operator intervention required to start the machine. Now operator manually rejecting exhaust thermocouple in start up and normalizing after full speed. Exhaust thermocouple starting setpoint is 550 F and depends on FSR rate it reaches 970 F. Now I want to tune the control system, for avoiding high exhaust T/C reading. For that I observed that, starting rate is more. So I have a query that this starting rate is normal or not. Our unit operates separate grid, means its a Aluminium industry. Two points am considering one is decrease the acceleration rate and reduce the FSR starting contant. So am considering different opinions regarding this.
Without being able to see the programming of the A-B ControLogix it's pretty much impossible. All we can do here is talk about typical GE-design Frame 5 heavy duty gas turbine control and protection philosophy. And, I don't have access to any Mark II elementaries; I can only address what was done for Mark IV, Mark V, Mark VI and Mark VIe units. Perhaps glenmorangie can talk about Mark II systems.
If there are high exhaust temperature spreads (the difference between the highest and lowest exhaust temperature thermocouple values, and the highest and second lowest, and highest and third lowest) during acceleration, and those high spreads decrease when the unit gets to FSNL the most likely cause is that flame has been lost in one or more combustors during acceleration. That's the most common cause--BUT, given that the firing FSR is SO high for the machine at your site if the fuel is typical natural gas that doesn't seem very likely.
I would guess the unit lights off (establishes flame) VERY QUICKLY when firing starts, and may even be accompanied by a low "bang". Which is not good for the machine. First, the "bang" is bad for the machine and the exhaust (diffuser and duct/stack), and if it does light off quickly it's very likely the exhaust temperature spikes very high--which is also NOT good for the machine. That represents a thermal shock when starting the machine from a cold condition.
I was at one site a few years back that always had a cold spot in the exact same location in the exhaust (meaning the same exhaust T/C values were always colder than the others) during acceleration, and when the unit reached FSNL the cold spot went away. It was discovered that when the exhaust duct had been rebuilt a few years back (the unit was pretty old) the discharge of the compressor bleed valves was not pointed in the proper direction--which meant that the cool compressor discharge air was blowing on the same T/Cs during starting, causing the cold spot. (We proved it by opening the compressor bleed valves at FSNL, and by forcing them to stay open when reaching FSNL. In both cases, the same exhaust T/C values always went low. Rotating the compressor discharge "cages" in the exhaust duct to point away from the exhaust T/Cs fixed that problem.
The other telltale thing about that situation was that during starting and acceleration the cold spot in the exhaust never moved--it always stayed in exactly the same place, the same exhaust T/Cs were always colder than the others throughout the entire acceleration process. A real combustion problem would have caused the cold spot to move with speed as the unit accelerated.
How do the exhaust T/C readings behave when your machine is starting? Is it always the same T/Cs that are colder (or hotter) than the others? In other words, do the operators ALWAYS reject the same exhaust T/Cs during starting? If so, then there's probably not a combustion problem.
Now, on newer Frame 5 GTs (conventional combustor-equipped) the combustion monitor is NOT enabled until the unit reaches FSNL. High exhaust temperature spreads during starting are effectively "ignored." So, I don't know why the operators have to reject T/Cs during starting. Does the programming in the A-B trip the unit during starting because of high spreads?
OR, do the operators reject the highest exhaust T/C readings because if they don't the unit will stop accelerating? Which would probably mean there is a pretty serious cold spot in the exhaust caused by something (no flame in one or more combustors or something amiss with the exhaust configuration). If the operators are just routinely rejecting the highest exhaust T/C values to prevent the acceleration rate from slowing, that's not good. Do the operators even really know why they are rejecting the exhaust T/Cs? Is it just because that's what they've always done, and don't know what would happen if they didn't.?.?.?
As was said before, most GE-design B/E class units (including Frame 5 Model Ps and other Frame 5 models) have acceleration rate references of approximately 0.10-0.11 %/sec from approximately 40% or 50% all the way up to FSNL. One other reason the acceleration rate may be so high is that there is a serious vibration at some point during acceleration--so the unit accelerates very quickly to try to "ride through" that critical speed without too much vibration, or without tripping the machine....
Again, without being able to examine the programming in the A-B control system it's very difficult to say anything with any degree of certainty.
I would say the firing FSR is too high. And, the acceleration rate of 0.50 %/sec is also too high. And, one should very clearly understand why the operators are rejecting T/Cs during acceleration, and if they are always rejecting the same T/Cs. If it's because of a true combustion problem (lost flame in one or more combustors during acceleration) that should be possible to solve. Maybe not an easy solution, but solvable. Again, GE control and protection philosophy ignores spreads during starting, and once the unit reaches FSNL the allowable exhaust temperature spread value starts ramping down, usually over a 60 second period) to the temperature dictated by machine conditions.
If the operators can un-reject exhaust T/Cs when the unit reaches FSNL and the exhaust temperature spread is low (15-40 deg F) at FSNL then either there's NOT a real combustion problem during starting, or there is--and it should be solved, because that's not good for the machine either.
You need to tell what exhaust T/Cs the operators are rejecting--and why, if they know (I'll bet they don't--they've just been taught to do it that way and they've always done it that way).
Typical FSRs for firing are in the range of 28-25%. A couple of seconds after flame is establish, a Mark* digital turbine control then slightly reduces the FSR to what's called "warm-up" and this is done to try to minimize the thermal stress of establishing flame in the machine. "Warm-up" FSRs are usually 1-4% FSR less than firing FSR, and that is held for 60 seconds. If fuel is cut back too far during warm-up, then often flame is lost in one or more combustors--and it doesn't re-establish often until the unit gets close to FSNL. Sometimes it re-ignites during acceleration, but not always.
After the 60 second warm-up period, the Mark* then starts ramping up FSR to help accelerate the unit--at a rate MUCH LESS than 0.5%...!
That's about all I can offer at this point. Hope it helps! Please write back with more information and to let us know how you proceed.
Thanks CSA for the detailed reply.
This is dual fuel machine, Natural Gas and diesel. We want to start GT as a cold condition as distillate fuel. Without exhaust thermocouple rejection, Unit will trip. I himself tried with this with operator. First I tried without rejection then I tried with rejection.
Why this much fast, no body answered in plant. Am new to this plant. I tried these experiment with 2 GTS. One GT started with diesel with rejection (GT1). One GT tripped on flame off, no flame established (GT2). is any flow divider pressure chart available?
14 GPM liquid flow, some combustors pressure is 1600 IB/inch2 and some 1100 IB/inch2 (GT2). Exhaust spread reached upto 270 F(GT1). My flow reference and actual flow is matching (GT1 & GT2). My actual idea is reduce acceleration rate by 0.1 %(0.5 to 0.4) and reduce firing constant by 5. That's what why am proposing. everybodys question is why this much fast as by design.
One thing also I want to tell, there is one fast start mode in logic, that rate is 1 % speed/sec, even though we are not using it. My ultimate aim is to start GT from cold using distillate fuel without any thermocouple rejection. If using distillate fuel firing is not happening, but flow is there, flow divider having different pressure. Is that nozzle problem (we checked igniter, that is perfectly working).
I want to preface my reply with this: It is very difficult to ensure reliable starting on distillate fuel of a duel fuel machine. Full stop. Period. It's just not an easy thing to achieve. Can it be done? Yes. Is it common (reliable starting on distillate fuel)? No. So, you and others at your plant should temper your expectations. There are many reasons for this, but mostly because there is a liquid fuel purge system (maybe--the time when units were still being produced with Mark II and when they were being switched to Mark IV was about the time that liquid fuel purge systems were being introduced on GE-design heavy duty gas turbines--so we don't know if yours have them or not), and many check (non-return) valves none of which are directly controlled by the turbine control system.
I also want to point out that NOWHERE in your response did you list the Process Alarm that was annunciated when the unit tripped when exhaust T/Cs were NOT rejected. Exhaust overtemperature? High exhaust temperature spread? Loss of flame? WITHOUT KNOWING WHAT THE CAUSE OF THE TRIP WAS, IT IS VIRTUALLY IMPOSSIBLE FOR US TO MAKE ANY RECOMMENDATION(S).
You were also asked what the exhaust temperature spreads were during starting with no exhaust T/Cs rejected?
And, you were also asked if the operators always reject the same exhaust T/Cs when starting?
Realistically, how do you expect to get good feedback and recommendations when you don't respond to the questions you are asked?
I want to correct a statement I made: The typical range for firing FSR is approximately 15-25% for most fuels (gas or distillate). Some machines have individual firing FSRs for gas fuel and distillate fuel; some don't (and use the same value for both--which can complicate matters).
I really don't understand the conditions you are trying to explain. Liquid fuel systems are complicated--yet very simple--at the same time. You really need to have and understand all the P&IDs for the various systems involved. First, there is or will be a liquid fuel "forwarding" system--to get the liquid fuel from a storage tank to the liquid fuel stop valve. That system has to be working correctly--especially when trying to start the turbine--meaning that the liquid fuel supply pressure has to be stable, and that there must be no air trapped in the liquid fuel supply piping upstream of the GT liquid fuel stop valve. This is one of the hardest things to maintain on a liquid fuel system: keeping air out of the liquid fuel supply piping, including any low-pressure filter vessels in the liquid fuel supply piping. Many times, the liquid fuel supply piping rises up and into an overhead pipe rack, and than back down to a trench in the ground--sometimes more than once--before it rises up to the liquid fuel supply flange on the side of the Accessory Compartment. If that is the case, the piping on every overhead run should be "pitched" (slanted) so that there is a high point in the piping AND there should be a vent (a valve) that can be used to remove any air from the piping. Because without it, once air gets into the piping--and it does!--it's very difficult to get it out. Yes, the vents are manual hand-valves--meaning that people have to climb up to the vent to open the hand valve(s) when the system is under pressure from the forwarding pumps to get air out of the system. But, without this scheme it can be very difficult to get air out of the piping--and that can cause pressure fluctuations which are difficult if not impossible to control, AND if the air moves in the piping (and it does move at the most inopportune times) it can cause firing problems when the air flows into the combustors.
So, stable liquid fuel supply pressure--at the specified pressure!--is very important. If the supply pressure is fluctuating, the liquid fuel control valve is going to be trying to keep up with the pressure changes and that's going to be very difficult to get stable flow to the combustors, especially at the low fuel flow-rates when trying to start and accelerate the turbine on liquid fuel.
When air goes through the liquid fuel control valve it causes the liquid fuel flow divider to feedback to fluctuate--which causes the turbine control system to quickly change the liquid fuel control valve position to try to maintain a stable flow-rate. Also, when air gets to the combustor it can cause the flame to go out. Everyone thinks it will quickly re-establish through the cross-fire tubes, but it's not always as quick to re-establish as people think it should be. So, this is one way flame can be lost and can be intermittent in combustors during firing and acceleration.
There are liquid fuel check valves at each combustor fuel nozzle. The purpose of these check valves during starting is to get the liquid fuel pressure at the nozzle up to minimum (usually approximately 100 psig) so that it atomizes properly as it exits the liquid fuel tip of the fuel nozzle. (Yes, there is atomizing air flowing during starting, but it's usually minimal, and while it is helpful for establishing flame most of the atomization of the liquid fuel during starting is due to the "mechanical" atomization caused by the components of the liquid fuel portion of the fuel nozzle.) When you look at the liquid fuel check valve information in the Device Summary, you will see the "cracking pressure" specification for the check valves. That means the check valves will not allow flow in the forward direction (during starting) until the pressure is approximately 100 psig (typically)--and that is, again, to achieve the best atomization possible of the fuel when trying to establish flame. As the unit speed increases the speed of the high-pressure liquid fuel pump will also increase which will increase the pressure of the liquid fuel as the unit accelerates to FSNL. After FSNL, as load in increased, the flow through the high pressure liquid fuel pump is increased and that causes the pressure to increase.
While it usually takes two people, using and having the pressure gauge at the liquid fuel flow divider is extremely helpful when trying to troubleshoot liquid fuel problems. The purpose of the flow divider is to split the liquid fuel flowing through the liquid fuel control valve into ten equal flow-rates to each of the ten combustors. When the liquid fuel check valves are working correctly, the pressures at each fuel nozzle should be within 10% of each other. Any individual fuel pressure reading which is more or less than 10% of the average of majority of fuel pressures is usually considered to be suspect. So, if the average of the majority of fuel pressures for each combustor is 400 psig, but one pressure is 350 psig or 450 psig, then that combustor's fuel nozzle or liquid fuel check valve is experiencing a problem. That's a general rule--but it is also very a good rule and works very well.
The reason it usually takes two people to take readings from the pressure gauge at the liquid fuel flow divider is one person has to rotate the handle and shout out the pressure readings and the other person has to write down the pressure readings. This makes it difficult to get pressure readings during firing, but not impossible. And, all the pressures should--again--be roughly equal. Any pressure(s) which are not within approximately 10% of the average reading is suspect, meaning that the liquid fuel portion of the fuel nozzle and/or the liquid fuel check valve of that combustor has a problem.
There may be a liquid fuel purge system, which uses Atomizing Air to purge liquid fuel out of the liquid fuel portion of the liquid fuel nozzles (when it works correctly). That system uses check valves very close to the liquid fuel check valves at the fuel nozzles. If the liquid fuel purge check valves leak when the unit is firing on or running on liquid fuel, that means that some or most of the liquid fuel which should be flowing into the fuel nozzle to be burned in the combustor is NOT--and it will be seen flowing out of the Tell-Tale Leak-off (see the Liquid Fuel Purge P&ID for details).
The check valves GE and packagers of GE-design heavy duty gas turbines used for decades were the best available at the time, BUT they were prone to premature failure. Leaking liquid fuel check valves can allow hot combustion gases to get back into the liquid fuel system, all the way out of the liquid fuel stop valve!!! This has been proven to occur many times. When hot combustion gases get into the liquid fuel system, they can cause the liquid fuel to carbonize--which means it will harden and cause blockages. Sometimes the pieces will flow into the liquid fuel check valves, and make them stick open, or they will flow into the liquid fuel portion of the fuel nozzle and block passages in the fuel nozzle--which can cause insufficient fuel to get into the nozzle, or it can cause atomization problems. I have also found metal shavings, welding rod pieces, mold and cigarette buts in liquid fuel check valves causing them not to seal properly.
Finally, if the unit has a booster atomizing air compressor (AC motor-driven atomizing air compressor that only runs when starting on liquid fuel) all of the valves in the system have to work correctly. Usually, there is at least one, and sometimes, two, check valves in the booster AA piping. Those check valves might be gravity-operated, or sometimes they have small springs in them. Those check valves have to open fully when the AA booster compressor is running, and they have to close fully when the AA booster compress is not running. It runs during starting, to help atomize the liquid fuel to help it ignite easier and to help it burn more completely during starting and acceleration. So, that has to also be working correctly for reliable starting on liquid fuel (distillate).
The check valves are the weakest link in the system--and they are NOT controlled by the turbine control system (usually they aren't--newer turbines are now using check valves which are opened and closed using air pressure which is controlled by solenoid valve(s) which are controlled by the turbine control systems. This greatly increases the complexity of the piping and tubing (air pressure tubing has to be added to the turbine compartment!), BUT these valves have proven to be more reliable than the spring-loaded, poppet-style check valves.
Some of the larger GE-design heavy duty gas turbines (F- and H-class) now have some very complicated liquid fuel reliability systems to keep air out of the liquid fuel supply piping and improve the starting--and fuel transfer--reliability of the units. The systems maintain fuel pressure in the liquid fuel supply piping to prevent air from getting into the piping, and some even have recirculation systems to keep mold and bacteria from developing in the liquid fuel supply piping. (Yes; mold and bacteria can form in distillate fuel piping and tanks.)
So, stable liquid fuel supply pressure; lack of air in the liquid fuel supply piping; working check valves (liquid fuel check valves, and liquid fuel purge check valves if so equipped); and regular maintenance are all important aspects to reliable liquid fuel operation (starting and transfer). It's a never-ending process of working through issues, only to find new issues and problems. It's not easy, but it's not impossible, either.
I'm sure the site personnel are now very accustomed to (and happy with!!) the very fast acceleration rate you have described. Reducing it is going to be difficult, but it is prudent to do so. If they want the machines to last a reasonable long time between maintenance outages, and if they don't want to spend a lot of money on outages and spare parts, and if they don't want forced outages caused by failures of nozzles and buckets they should think seriously about significantly reducing the acceleration rate.
Lastly, think about it. How can we understand what is causing the unit not to start or to trip if you don't tell us precisely what alarms are annunciated when the unit trips? We don't ask for useless information (contrary to popular belief). We're not there alongside you and we don't know what you know and can't see what you see. Rejecting T/Cs during starting is a poor workaround to whatever the real problem is. I don't care how long it's been working "successfully." Or why it was started--it's not a proper solution to the problem.
Hope this helps! Help us to help you--by providing the information requested. If it helps, write down all the questions (because I know I don't list them numerically) and then check them off as you answer them. But we can't help very much if we don't have good information.
I want to preface the remarks below by saying the sequences described are for a non-DLN combustor-equipped unit (in other words it is for a unit with conventional, diffusion flame combustors with a single fuel nozzle per combustor), with two flame detectors (possibly four). It's NOT for DLN units with multiple flame detectors (four or eight).
None of the Mark* heavy duty gas turbine control systems I worked on tripped the turbine on a failure to ignite. Not a one. A failure to ignite occurs when there is NO flame in any combustor with a flame detector when the firing timer or the warm-up timer expires.
The typical firing sequence for liquid fuel goes like this: After the purge time is complete if the unit is above minimum firing speed the spark plugs (ignitors) will be energized. Then the liquid fuel stop valve will be opened. Then the liquid fuel control valve will move to increase the flow of liquid fuel to match the liquid fuel flow-rate reference (FSR1/FSR is converted to a liquid fuel flow-rate reference.) In order for flow to start and increase it is necessary for the liquid fuel control valve to increase the liquid fuel pressure, through the liquid fuel flow divider, enough to cause the liquid fuel check valves at each of the fuel nozzles to exceed their cracking pressure setpoint. (For GE-design Frame 5 heavy duty gas turbines, this has typically been approximately 100 psig; the cracking pressure is specified in the unit's Device Summary drawing.) Once the liquid fuel pressures exceed the liquid fuel check valve's cracking pressure liquid fuel will start flowing into the combustor through the fuel nozzle. The reason the liquid fuel pressure has to be above the check valve cracking pressure is that is the minimum pressure deemed necessary for proper pressure atomization of the liquid fuel as it exits the tip of the liquid fuel cartridge of the fuel nozzle.
Fuel is admitted to the combustor through the fuel nozzles for the duration of the firing period if no flame is detected; the firing timer is usually 30-60 seconds, but never longer than 60 seconds (fuel should NOT be admitted into a combustor during firing (starting) for more than 60 seconds if there is no flame--ever). The spark plugs are kept energized for the duration of the firing timer; just in case the flame is lost in one or more combustors the energized spark plugs will help to re-ignite the flame if necessary..
The firing FSR is usually a little higher than the amount of fuel required to sustain flame in the hopes that it helps to extablish flame quicker because the flow-rate is a little higher. BUT, the gas turbine exhaust temperature (and turbine internal temperatures) will spike very high if the firing FSR is too high and it is maintained after flame is established, so while it's set to be slightly higher than what would be required to maintain flame for the air flowing through the machine it's not set too high--to limit the exhaust (and internal turbine) temperature spike.
Very shortly after flame is detected in one or more of the combustors with flame detectors during firing the Mark* reduces the fuel flow-rate to what's called the "warm-up" value. The Mark* holds the fuel flow-rate at the warm-up value for a period of time--usually approximately 60 seconds.
At the end of the warm-up period, if flame is still present in one or more combustors the fuel flow-rate will start to be ramped up to make the speed increase at the desired rate.
However, if flame is NOT present at the end of the warm-up--or firing--period (which is also usually approximately 60 seconds, but not usually ever longer) then the Mark* will de-energize the spark plugs and close the liquid fuel stop valve and the liquid fuel control valve will move to shut off the flow of liquid fuel, and the Mark* will annunciate the Process Alarm "Failure to Ignite."
At this point, the unit continues to CRANK--that is, the starting means is still providing torque to the turbine-generator shaft and the unit continues to spin at CRANKing speed. But, no fuel is being admitted to the unit and the spark plugs (ignitors) are de-energized. This is done to allow the operator a chance to do some troubleshooting AND it also helps to purge out any unburnt liquid fuel or fuel vapors in the combustors, turbine section, and exhaust.
If the operator wants to stop the turbine, all that's necessary is to click on STOP and the starting means will be shut down and the unit will coast down to Cooldown.
If the operator wants to attempt another START, all that's necessary is to click on CRANK, wait a couple of seconds, and then click on AUTO. This will re-initialize the firing sequence once the purge timer times out, and the above process is repeated.
BUT, a "Failure to "Ignite" during a START attempt DOES NOT typically result in a turbine trip--which would mean the liquid fuel stop valve is closed, the spark plugs are de-energized and the starting means is shut down and the unit will coast down to Cooldown.
A "Failure to Ignite" occurs when there is NO flame in any combustor with a flame detector when the firing timer or the warm-up timer expires. It's NOT typically a trip--contrary to popular belief. Again, the unit typically continues to CRANK, which helps to purge out any unburnt liquid fuel or fuel vapors from the unit.
Now, let's say there was flame at the end of the firing/warm-up timer, BUT for some reason it was lost in all the combustors with flame detectors AFTER the firing/warm-up timer expired, during acceleration. Then the Mark* would annunciate a "Loss of Flame Trip" and the Mark* would immediately close the liquid fuel stop valve and shut down the starting means. No more fuel would be pumped into the combustors and the unit would coast down to Cooldown.
A "Loss of Flame Trip" is exactly what it says: Flame was lost for no apparent reason after firing/warm-up is complete. No low-low L.O. pressure condition was detected, or no high-high vibration condition was detected--no condition that would have resulted in the Mark* shutting off the flow of fuel to protect the machine was detected--but flame was lost in the combustors with flame detectors while the turbine was running after the firing/warm-up timer had expired.
As was mentioned previously, a LOT of components have to work correctly for the liquid fuel system to provide a stable flow-rate of liquid fuel to the turbine. The liquid fuel supply pressure up to the liquid fuel stop valve and liquid fuel control valve has to be at the proper pressure and stable. The turbine control system typically DOES NOT control the liquid fuel supply pressure, and it typically only has a single pressure switch to indicate a low liquid fuel pressure. Some liquid fuel supply ("forwarding") systems have an adjustable pressure regulator, and if air gets into the pressure regulating mechanism it will make controlling the liquid fuel pressure very difficult. If there is air in the liquid fuel supply piping upstream of the liquid fuel stop valve--including any low-pressure liquid fuel filter vessel(s)--that will also make it very difficult to maintain a stable liquid fuel supply pressure and flow-rate up to the liquid fuel stop valve and liquid fuel control valve. Which will make it difficult for the liquid fuel control valve to maintain a stable liquid fuel flow-rate through the liquid fuel flow divider to the fuel nozzles. And, again--the turbine control system doesn't usually control or even monitor the liquid fuel supply pressure; it's done with local control components and the turbine control system only alarms on low pressure (or sometimes it tries to switch to gas fuel, depending on the programming of the turbine control system--but it will only initiate a fuel transfer if the unit speed is less than minimum firing speed or above synchronous speed, NOT during firing or acceleration).
The liquid fuel flow divider has a manual selector valve and gauge on it. The first ten detents of the selector valve correspond to the ten combustors/fuel nozzles. This makes it pretty easy to narrow a problem with liquid fuel starting or combustion problems by checking the fuel pressures to each of the nozzles/combustors. As was mentioned before, the general rule is: Any pressure which is 10% more or less than the average of the majority of the pressures is considered to be suspect. And, one checks the components of that indicated fuel nozzle or liquid fuel supply line. And, it's best to have two people recording the pressures--one to rotate the handle through the detents and call out the pressures, and one to write the pressures down. When trying to record the pressures during firing, it's a good idea to also station someone near the liquid fuel stop valve and have that person shout out when the stop valve opens. Shortly after the stop valve opens the liquid fuel control valve will start trying to increase the liquid fuel pressure above the liquid fuel check valve cracking pressure. So, it's best to wait to start rotating the manual selector valve handle until the pressure on the #1 fuel nozzle/combustor increases to above cracking pressure (presuming the #1 liquid fuel check valve is working correctly...!). Then the handle needs to be rotated quickly from one detent to the next while recording the pressure at each detent. There is only about 60 seconds during firing to get this information.
Any reading which is significantly lower than the liquid fuel check valve cracking pressure or is significantly higher than the liquid fuel check valve cracking pressure during firing indicates a likely problem with the liquid fuel check valve or the liquid fuel cartridge of the fuel nozzle. If the pressure is much lower then either the liquid fuel check valve spring is broken or something stuck in the check valve prevented it from closing or something in the liquid fuel cartridge is loose or has come apart and liquid fuel is leaking into the atomizing air passage. If the pressure is much higher then something is blocking the flow of liquid fuel--either something is stuck in the liquid fuel check valve, OR liquid fuel has carbonized (hardened) in the fuel nozzle and is preventing liquid fuel from getting through the nozzle.
If the unit has a liquid fuel purge system (you haven't told if it does or not), then if the fuel pressure is lower than the liquid fuel cracking pressure then the liquid fuel purge check valve is leaking fuel. That can be verified by having someone at the Tell-Tale Leakoff during firing to visually watch for fuel coming out of the Tell-Tale Leakoff (and going into the Gas Turbine Drains Tank or Oily Waste Tank).
Here is an attempt at a "drawing" of the arrangement of the liquid fuel check valve and a liquid fuel purge check valve at the entrance to a liquid fuel nozzle cartridge:
Liq. Fuel->------|/|-----> Fuel Nozzle Liquid Fuel Cartridge
Liq. Fuel Purge /
Check Valve ---
The liquid fuel purge check valve will have air entering the check valve when the unit is operating on natural gas fuel, or has completed a transfer from liquid fuel to natural gas. No air will be flowing into the liquid fuel purge check valve when the unit is starting or operating on liquid fuel. The purpose of the check valves is block the flow of liquid fuel into the valve that feeds air to the liquid fuel purge check valves when the unit is running on liquid fuel. The purpose of purge air is make sure, one, there is no liquid fuel in the liquid fuel cartridge of the fuel nozzle (because it will carbonize (harden) in the cartridge and cause problems when trying to start or run on liquid fuel, and, two, to prevent the flow of hot combustion gases from entering the liquid fuel cartridge and flowing backwards through the fuel nozzle which can cause damage to the fuel nozzle. (The purge air flowing through the fuel nozzle also cools the liquid fuel cartridge of the fuel nozzle. But, the cooling is a minor consideration.)
Again, the primary purpose of the liquid fuel check valve is raise the liquid fuel pressure when firing to above the valves' cracking pressure to assist with pressure atomization of the fuel as it leaves the fuel nozzle. The secondary purpose of the liquid fuel check valve is to prevent hot combustion gases from flowing backwards through the fuel nozzle and into the liquid fuel system. As was mentioned previously, the poppet-style check valves used for decades on GE-design heavy duty gas turbines have been known to leak in the reverse direction, which allows hot combustion gases (or purge air, if the unit is so equippped) to enter the liquid fuel system. The air/gases will pressurize the liquid fuel lines it is entering when the unit is running on gas fuel, and that pressure, working through the flow divider has been known to open the liquid fuel stop valve, allowing air/gases to flow backwards into the liquid fuel supply piping and low pressure liquid fuel filter, if present. Also, hot combustion gases have been known to flow backwards through a failed liquid fuel check valve, into and through the flow divider, and through other liquid fuel check valves into other combustors. When this happens, though, the liquid fuel tubing becomes VERY hot, AND the liquid fuel in the lines and the flow divider will usuall carbonize (harden). It's very rare, but it does happen. And, it's a source of hot air for carbonized liquid fuel found in check valves and liquid fuel lines and the liquid fuel flow divider.
The manual selector valve and gauge on the liquid fuel flow divider make troubleshooting liquid fuel starting or combustion problems pretty easy--because one can see the fuel pressures, which should all be roughly equal (that's the function of the liquid fuel flow divider). If any pressure is much lower or much higher that indicates that liquid fuel supply system to the combustor has a problem.
Note: The turbine control system DOES NOT control or monitor the check valves in the liquid fuel system, or the liquid fuel purge system if the unit has one. BUT, the check valves have to work properly--or there can be problems.
Next, the combustion monitor function of the turbine control system of GE-design heavy duty gas turbines is usually not enabled during starting and acceleration--for the simple reason that it is not uncommon to have large exhaust temperature spreads during starting and acceleration. It's not a good thing, but it's also not uncommon and happens a LOT.
If the unit is tripping on high-high exhaust temperature, that suggests the turbine control system, via the liquid fuel control valve, IS NOT able to control the liquid fuel flow-rate for some reason. If the liquid fuel supply pressure is unstable and fluctuating, especially if it's fluctuating quickly over a wide range, it's often not possible for the liquid fuel control valve to respond to these pressure fluctuations and excess liquid fuel can flow into the combustors causing an exhaust overtemperature trip.
BUT, rejecting high exhaust T/C readings doesn't protect the turbine from exhaust overtemperatures--it just fools the turbine control system into thinking the high exhaust temperatures don't exist. Which can be very bad for the hot gas path parts of the turbine.
Most exhaust temperature spreads of GE-design heavy duty gas turbines with conventional combustors (non-DLN combustors) are caused by low fuel flows to one or more combustors, not high fuel flows to one or more combustors. So, they are the result of cold spots--not hot spots--in the gas turbine exhaust. Something is causing either low fuel flow-rate to one or more combustors, either poor atomization of the fuel or some blockage in the liquid fuel check valve or the fuel nozzle, or leaking liquid fuel purge check valve(s) is allowing liquid fuel to flow away from the fuel nozzle not into the fuel nozzle.
If the unit has a booster atomizing air compressor (again, we haven't been told if it does or not), that has to operate properly. Yes, the motor driving the booster AA compressor is controlled by the turbine control system, but usually there is one or more check valves in the booster AA compressor piping (inlet and/or outlet) that are NOT controlled by the turbine control system, and frequently stick (usually because of rust). It's IMPOSSIBLE to measure the discharge pressure of the booster AA compressor using the AA pressure gauge on the Accessory Gauge Cabinet--it's just too low for the range of the gauge. And, it's really about air flow--not pressure. The booster AA compressor is, or should be, a postive displacement compressor meaning that if it's running its moving air. But, the amount of air might be restricted by a check valve which is out of position.
You now have troubleshooting tips and procedures for trying to understand why the unit isn't accelerating on liquid fuel. Use the manual selector valve and gauge on the liquid fuel flow divider. If the unit has a liquid fuel purge system, monitor the Tell-Tale Leakoff (it's usually on the right side of the turbine compartment, under the walkway).
And write back to let us know what you find!
The good news is it is possible to get the system working. The bad news is--especially if the units usually run on liquid fuel--its not easy to keep them working. There are just too many possibilities for problems, and components that wear out (especially the poppet-style check valves!).
Don't forget to check the strainers of the liquid fuel supply (forwarding) pumps. It's very common for them to get plugged and restrict the flow of liquid fuel from the storage tanks to the turbine(s).
You asked about some kind of chart for the liquid fuel flow divider--I've never seen such a chart. The purpose of the liquid fuel flow divider is to take the flow from the liquid fuel control valve and divide it equally into 10 individual flows to each of the fuel nozzles/combustors. The liquid fuel check valve cracking pressure is set to ensure the pressure to each fuel nozzle/combustor is equal to the check valve cracking pressure which should be high enough to ensure proper pressure atomization as the liquid fuel exits the the fuel nozzle into the combustor, making it easier to ignite and combust more fully and efficiently.
The key for liquid fuel starting is primarily that the fuel pressures for all fuel nozzles/combustors should be above the liquid fuel check valve cracking pressure when firing, and they should all be stable and relatively equal. Any pressure less than the liquid fuel check valve cracking pressure means the fuel flowing into the fuel nozzle is NOT being atomized properly as it leaves the fuel nozzle tip. If the unit has liquid fuel purge, a low pressure can also mean NO fuel is flowing into the liquid fuel nozzle, because it might be flowing backwards through a failed liquid fuel purge check valve and not into the fuel nozzle/combustor. Even if the fuel pressure is above the liquid fuel check valve cracking pressure if one or liquid fuel purge check valves are leaking that means little or no fuel is actually flowing into the fuel nozzle/combustor. And, if the fuel pressure is much higher than the liquid fuel check valve cracking pressure that means that something is restricting the flow of liquid fuel through the liqiud fuel check valve OR the liquid fuel cartridge of the fuel nozzle, and the fuel might not be burning in that combustor or it might be buring incompletely.
If you have someone watching the exhaust stack of the gas turbine during firing you should see one of two things. If the unit does not establish any flame at all (in other words, the average exhaust temperature does not increase at all during firing) it should be possible to see a light, white vapour (not smoke--but distillate fuel vapours) exiting the stack. If there is sufficient fuel flowing but it did not ignite at all, the atomized fuel (vapour) will flwo through the turbine and out of the exhaust. Some of it will condense in the turbine section, the exhaust diffuser and the walls of the exhaust stack--but it should still be visible (on a clear day).
If flame is established in one or more combustors--but not ALL combustors--then white clouds of smoke should be visible coming from the exhaust stack. That is unburnt fuel which is partially combusting in the exhaust. The fuel which is not burning in the combustors will mix with the hotter gases from the combustors where the fuel is burning and will cause the white smoke. If the unit fails to ignite (as described above) there may be large plumes of white smoke caused by flame being established in one or more combustors and then being lost before the firing timer expires.
Billowy, white smoke coming from the exhaust stack of a GE-design heavy duty gas turbine is an indication of incomplete combustion in one or more combustors. It is EXACTLY like when one sees white smoke coming from the exhaust stack of a lorrie (truck)--there is something wrong with one or more cylinders of that truck that is causing the diesel fuel NOT to burn in the cylinder(s) and it burns partially when it exits the cylinder and flows through the exhaust manifold and exhaust system--creating the white smoke. (It is EXTREMELY rare to see black smoke coming from the exhaust of a heavy duty gas turbine--but if one does it means the turbine is SERIOUSLY overloaded, or there's a fire in the exhaust.) "Thin" white vapours exiting the exhaust stack on a failed liquid fuel start attempt means no flame was established during firing. Either the liquid fuel flow-rate wasn't high enough, or, the spark plugs (both of them) were not working properly (they need maintenance or they got wet from excessive liquid fuel flow-rate, or the liquid fuel flow wasn't properly atomized--either by the fuel nozzle or from the atomizing air coming from the booster atomizing air compressor, if the unit is so equipped). So, a low liquid fuel flow-rate can cause problems igniting the liquid fuel, AND an excessive liquid fuel flow-rate can cause problems igniting the liquid fuel.
It should be noted that if the unit failes to ignite on liquid fuel, it is ALWAYS a good idea to let the unit CRANK for several EXTRA minutes before attemting another start. Or, if flame is lost during acceleration when starting on liquid fuel, you will likely see a big puff of white smoke, and it is also a good idea to purge for a few minutes extra before the spark plugs are re-energized.
The spark plugs can also be a problem when trying to start on liquid fuel. There are two spark plugs, though only one is required to be working; the second one is in case the other doesn't work. Contrary to popular believe, spark plugs require periodic maintenance, and parts need replacing over time. And while spark plugs may seem to be working fine when starting on gas fuel because even a weak spark is usually enough to ignite natural gas, the spark must be strong when starting on liquid fuel. Also, if the flow of liquid fuel is too high--or is not atomized properly--during starting the spark plugs can become "wet" with liquid fuel and that makes the spark very weak.
This about completely covers all the things that can happen when trying to start and run on liquid fuel. The weakest components in the liquid fuel (and liquid fuel purge system if present) are the poppet-style check valves. Especially the liquid fuel purge check valves if the unit normally runs on gas fuel; the poppets become very worn and do not seal, and the springs also wear and break.
It should be clear--there are many liquid fuel and liquid fuel purge components which are NOT controlled or even monitored by the turbine control system--the turbine control system ASS-U-ME-s the components are all working correctly, and they should. But, for units which run primarily on gas fuel, the liquid fuel systems get overlooked and forgotten--until the unit needs to start or run on liquid fuel. And, then the turbine control system gets blamed (because it always gets blamed for everything!) when the unit doesn't start or run on liquid fuel. And, it's usually NOT the turbine control system's fault--it's the fault of the owners of the unit, and the mechanical department who maintain the systems and components. If the owners/operators will not allow or schedule or pay for proper maintenance of the liquid fuel system, it's not going to be reliable. GE publishes several documents about liquid fuel reliability--and they recommend REGULAR operation of any dual fuel unit on liquid fuel to ensure reliability and to identify problems before they become REAL problems when the unit MUST start or run on liquid fuel. How often? At least once per week the unit should be transferred to and run on liquid fuel for a couple of hours. More often if possible. But, because most sites have LOTS of problems with liquid fuel operation and starting when they do try it, they will not follow the recommendations--when that is entirely counter-productive. The liquid fuel system of a dual fuel unit needs to be exercised, or it's going to be unreliable. Full stop. Period. End of discussion.
Hope this helps!
Thanks CSA for the detailed Reply.
Even GE manual also can't explain like this. When am stuck in problem, I know that control.com can help me. Got lots of information regarding liquid firing. Informed mechanical department to check nozzle check valves. Liquid purge system available in our unit. No atomization compressor available. Clutch driven fuel forwarding pump is using for the liquid fuel pressure. Unit tripped on high exhaust temperature trip alarm. Limit is 1050 F.Operator thermocouple rejection database not available. Informed operator to maintain database for the thermocouple rejection. After the mechanical clearance, I will update this post with more data and my tuning experience with GE frame 5 gas turbine. This posts boosted my confidence regarding liquid fuel firing. My reading regarding liquid fuel pressure in flow divider is out of 10% band.
I did see something about your liquid fuel flow divider pressure readings, and I dismissed them because they were too high for firing pressures. I now see they are probably from FSNL (Full Speed-No Load). And, YES--the pressures are way out of line.
The high pressure fuel pump is usually driven through an electric clutch by an output shaft of the Accessory Gear. That's not the same as the forwarding pump. A forwarding pump is used to get the fuel from the storage tank up to the unit--to the inlet side of the liquid fuel stop valve. Again, especially during firing and acceleration it is VERY important that the liquid fuel supply pressure from the storage tank to the inlet of the high-pressure liquid fuel pump be stable and not fluctuating by more than plus-or-minus 5 psig. If the supply pressure is fluctuating the liquid fuel control valve will be fluctuating trying to maintain a stable liquid fuel flow-rate (based on the feedback from the liquid fuel flow divider). It's really important that the pressure from the liquid fuel forwarding pump be less than the liquid fuel check valve cracking pressure, AND that it be relatively stable especially when starting on liquid fuel.
The gears and internal components of most liquid fuel flow dividers also need inspection and replacement over time. If they are worn, that can cause problems with the flow-rates to the individual fuel nozzles/combustors.
I haven't asked, and you haven't told, but does the unit have a variable displacement high-pressure axial piston liquid fuel pump with a servo-valve controlled swash plate to control the liquid fuel flow-rate, or does it have a screw-type high-pressure liquid fuel pump with a liquid fuel bypass valve to control the liquid fuel flow-rate? Not that it makes too much difference, but controlling liquid fuel flow-rate with a variable displacement axial piston pump during firing (low fuel flow-rates) can be difficult, and if the supply pressure is not stable then the flow-rates can be high and might cause an exhaust overtemperature trip (high exhaust temperature trip).
I'm trying to understand what might be causing the high exhaust temperature trip--and I keep forgetting to remember that to achieve the EXTREMELY fast acceleration rate you are talking about the turbine control system has to put a LOT of fuel into the unit! I would say that slowing down the acceleration rate from 0.5%/sec to something like even 0.25%/sec (which is STILL high in my personal opinion) might help with that.
BUT, certainly, if the units are having trouble maintaining flame during firing and acceleration that is a problem with the liquid fuel system (check valves; unstable supply pressure; etc.). AND, if the turbine control system is putting a lot of fuel into the turbine to achieve the questionable acceleration rate AND there are high exhaust temperature spreads then that's probably going to cause problems with high exhaust temperatures on those combustors which do have flame. And, rejecting those high temperatures isn't solving the problem causing the high temperatures--NOR is it doing anything to protect the combustors and turbine nozzles and buckets and exhaust diffusers from the high temperatures!
My guess is there are problems with check valves--both liquid fuel check valves and liquid fuel purge check valves.
I have alluded to the liquid fuel "cartridges" of the fuel nozzles of dual fuel units. Most of the liquid fuel cartridges I have seen of unit of that vintage were assemblies--there were several pieces which had to be assembled together, and then that assembly (cartridge) was screwed into the fuel nozzle base, with the atomizing air and gas fuel tips screwed over the liquid fuel assembly (cartridge). There was a threaded "nut" (for lack of a better term) that held the internal components of the cartridge together, and the center of the nut was a hole--a hole that a hex key ("allen wrench") was inserted into and turned with a torque wrench to hold the internal pieces in place. It was common to find this "nut" had loosened during operation, meaning that the internal components were not in their proper places, and that causes improper pressure atomization of the liquid fuel which is a real problem when starting on liquid fuel, but not such a problem once the fuel flow-rates increase at rated speed and when loaded. It was common for the mechanics to use at least the mid-strength thread locker fluid on the threads of the "nut" to try to keep it from loosening. The high-strength thread locker fluids sometimes couldn't be loosened without damaging the hex key (allen wrench) and the cartridge was then ruined. I've also seen some mechanics try to "stake" the nuts in place, but it's such a small internal passageway that staking was very difficult or impossible.
A copper crush gasket was then inserted in the fuel nozzle base (body) and the liquid fuel cartridge assembly was then screwed down into the fuel nozzle body base. There was a torque spec for the assembly, which was supposed to keep from over-crushing the copper crush gasket, but didn't seem to keep the cartridge assembly from loosening. When the cartridge loosens from the crush gasket liquid fuel leaks into the atomizing air passage of the liquid fuel nozzle, and doesn't pass through the liquid fuel tip. It was possible to stake the liquid fuel cartridges in place after torquing them down, to keep the cartridge from loosening. Also, I've seen some mechanics use thread locker fluid(s) to keep the cartridge assemblies from coming loose from the fuel nozzle body base.
If the liquid fuel nozzle cartridge assembly is coming loose from the fuel nozzle body and leaking liquid fuel into the atomizing air passage of the fuel nozzle that will usually be evident by carbonized fuel found in the inside of the fuel nozzle atomizing air tip.
The liquid fuel system, again, has MANY components which all have to work properly and together--especially during starting! Many of these components are NOT controlled or monitored by the turbine control system! And, yet the turbine control system usually gets blamed for starting problems.
Based on the information provided in this thread about this problem, there are probably several problems with mechanical components in the liquid fuel system. AND, the extremely aggressive acceleration rates the unit have are probably contributing to the high exhaust temperatures. Fixing the mechanical problems just requires a logical troubleshooting process and an understanding of the system and the components. Slowing down the acceleration rate will also probably help with the high exhaust temperature tripping. But, these steps should help to restore--at least for a short period--some liquid fuel starting reliability. As was said, liquid fuel starting reliability on dual fuel GE-design heavy duty gas turbines is difficult to achieve. Especially if the units run primarily on gas fuel, and only occasionally on liquid fuel.
If you find problems with the check valves (as suspected), you can work with various other suppliers to find other check valves that may be more robust. Some sites use pneumatically-operated check valves controlled by a solenoid or solenoids--but that greatly increases the tubing in the turbine compartment, though it does seem to greatly increase the reliability of the check valves and therefore greatly improves the reliability of starting and running on liquid fuel.
Please, do write back to let us know what you find. That person watching the liquid fuel stop valve during starting can also be watching the liquid fuel supply pressure gauge (it's usually mounted pretty close to the liquid fuel stop valve). And, again, it's best if that pressure is fairly stable (doesn't fluctuate by more than plus-or-minus 5 psig) during starting.
Best of luck! Also, thanks for the feedback!
I neglected to write that the liquid fuel supply pump pressure specification is usually about 55-65 psig; see the Device Summary, or if GE or the turbine packager supplied the Liquid Fuel Forwarding System there will be a P&ID for the system, and the pressure regulator (if included) setting will be listed in the Device Summary. Sometimes, the Liquid Fuel Piping Arrangement Drawing for the Accessory Compartment will also list the liquid fuel supply pressure for the flange on the side of the Accessory Base where the liquid fuel supply line (from the storage tanks/pumps) connects.
Again, the pressure MUST be less than the liquid fuel check valve cracking pressure, and it should be relatively stable with very little fluctuation, especially during firing. Most fluctuations are caused by air in the piping, air in the pressure regulator (if present), air in the low-pressure fuel filter vessel(s), and sometimes by clogged forwarding pump suction strainers. But if the units gets to rated speed and can be loaded clogged suction strainers are not usually the problem--but it doesn't hurt to inspect them when looking at the liquid fuel system for any problems; while the check valves are being removed, inspected and tested (and replaced!) it doesn't take long to look at the strainers.
Stationing people at the liquid fuel flow divider (two people, one person to rotate the manual selector valve handle and call out the pressures once fuel begins to flow (when the pressure at the first position gets above the liquid fuel check valve cracking pressure, and the second person to write down the pressures), the Tell-Tale Leakoff (one person with a video camera), and at the liquid fuel stop valve (one person watching and recording pressure swings and to tell the two people at the liquid fuel flow divider when the stop valve opens) during the next start liquid fuel start attempt will provide a LOT of valuable data. I would also suggest having one or two people watching the exhaust stack (hopefully it will be a clear day during the next liquid fuel start attempt), looking for thin white vapours or white smoke. This is important also.
If you haven't already done so, it would be good to reduce the acceleration rate, because that will likely have a large and good impact on the exhaust temperatures. I would also reduce the firing FSR on liquid fuel, by as much as 5-10% (so from 38% to 33%, or 28%).
Also note that if all ten liquid fuel check valves are removed for inspection, testing and replacement the ten liquid fuel lines from the liquid fuel flow divider to the fuel nozzles will mostly be drained of liquid fuel. So, when you try to start the next time, the unit will most likely have trouble firing and may not fire.
My solution for this problem is to pull the fuses or open the circuit breaker that powers the ignitors (spark plugs) before the first start attempt. Then, just like normal, select AUTO, and initiate a START. The unit will purge and "fire" (admit fuel and try to energize the spark plugs--but they won't be energized)--and this will have the effect of purging the air out of the lines. If you have someone watching the exhaust they will mostly likely see the thin (light) white vapours because the fuel did not ignite. The turbine control system should annunciate a 'Failure to Ignite' and keep on CRANKing. Those watching the liquid fuel flow divider pressure gauge will probably see a lot of fluctuating pressures, as will the person watching the liquid fuel supply pressure at the liquid fuel stop valve. The person watching the Tell-Tale Leakoff should see NOTHING (if the liquid fuel check valves are working correctly!).
Let the unit continue to spin for five minutes or so, then select CRANK for a few seconds, then re-select AUTO. The unit will then try to fire (when the purge timer expires) and the unit should start with few problems. The pressures at the flow divider should be stable and nearly equal, the supply pressure should be stable, and there should be NO flow coming out of the Tell-Tale Leakoff. There should be some white vapours for a short period of time, then possibly a little bit of white smoke, but if all the combustors light off there won't even be much white smoke. If there's a lot of white smoke, then the exhaust temperature spreads will also be high, and something is still not exactly correct.... But you will have data (pressure readings and video of the Tell-Tale Leakoff)--and if you have someone filming the exhaust (which I just now thought of ... DUHH!!!) you will have video of that as well. You could even have someone videoing (filming) the pressure gauge at the liquid fuel flow divider as one person is rotating the manual selector valve handle! Even more video evidence! And, someone could be videoing the liquid fuel supply, pressure as well. The only problem with these video recordings is trying to link the timing of them together--but if the phones all have the same time stamp it shouldn't be too much of a problem.
So, there you go--another use for cell-phone cameras!!! Won't do us at control.com any good, unless you post them to a free web-hosting site (like www.tinypic.com) and then post the link to them on a reply to this thread (which would be GREAT!!!).
Hope this helps!!!
While the unit is CRANKing after the expected "Failure to Ignite" when the ignitors (spark plugs) were intentionally disabled by removing the power to the ignitor circuit, it will be necessary to restore the power to the igniter (spark plug) circuit before selecting CRANK and then AUTO for a real firing attempt.
I call the first firing attempt with the ignitors (spark plugs) disabled a "false fire." The purpose is only to get air purged out the fuel lines between the liquid fuel flow divider and the nozzles, and because it's known that air is present in the lines between the flow divider and the fuel nozzles because the liquid fuel check valves and/or the liquid fuel purge check valves and/or the fuel nozzles were removed the chances of actually igniting the fuel and maintaining flame during warm-up and acceleration on the first START attempt after reassembly are very low. And because the only way to get fuel past the liquid fuel check valves is to START the unit and attempt to fire it, by disabling the ignitors the turbine control system will go through an entire firing sequence trying to get fuel into the combustors and there won't be issues with high exhaust temperature spreads and a possible failed START attempt, because the ignitors were intentionally disabled and a failed firing attempt is expected. Then, by restoring power to the ignitor circuit and selecting CRANK then re-selecting AUTO while the unit is still CRANKing and purging it avoids having to initiate a STOP and wait for the unit to coast down to Cooldown before selecting START, saving a lot of time.
Sorry for any confusion caused by not including the part about restoring power to the ignitors before trying to actually fire the unit.