Hello,

WE have one of the new gas turbines frame 5 mark VI controller Moog problem during making dry fire (turning off spark plug & increasing ignition time).

When we reach ignition phase after purge 2 min. the stop valve 20fl1x is open and the servo valve contiguous open and close many times during the flow test and no fuel arrived after check valves, the time between servo valve open and close is in seconds (3-5 sec.)

The parameters before starting is:

1- liquid fuel flow reference = -25%
2- liquid fuel by pass valve servo current= 70%
3- liquid fuel flow = 0

The parameters during starting is :

1- liquid fuel flow reference = 4.5%
2- liquid fuel by pass valve servo current = -15
3- liquid fuel flow read very big value more then 900&
4- also magnetic speed pickup of flow divider same reading more than 500 - 1000 %

so when the flow detected very high value bypass valve servo current go to high positive value more than 74% and after 3 to 4 seconds return to the negative value between -12 to -15 %

Is that the problem in servo valve or in high reading of flow divider magnetic speed pickup so the control valve (MooG) try to reduce the value by increase the bypass?

Is it software problem?
Is the servo current is ok?

Thank you all
Ayman

 

Ayman,

You didn't tell us if this is the first time these units have been run on liquid fuel, or if the unit has previously run on liquid fuel but hasn't run on liquid fuel for some time.

From the information provided, it would seem one or both of two possibilities might be causing the problem.

First, air is in the liquid fuel supply lines to the liquid fuel stop valve, or air is in the liquid fuel filter canisters. A liquid fuel flow divider feedback signal that jumps very high, then very low, then very high, is pretty indicative of air in the line.

Second, there is usually a pressure regulator in the liquid fuel supply system from the storage tank(s) to the turbine. It is pretty common for these regulators to not be set properly or to have air in the regulator mechanism and to cause fluctuations in the supply pressure, which can cause similar problems. As pressure goes up, the flow-rate goes up; as pressure goes down, flow-rate goes down.

Another, unlikely, problem is slipping of the electric fuel pump clutch. This is more difficult to observe and witness, but it has been known to happen. I've even seen loose couplings cause slipping. You need to remove the coupling cover and observe the shafts/couplings to see if there is any slippage.

Positive current applied to the servo-valve shuts off the flow of liquid fuel (by opening the Liquid Fuel Bypass Valve). Negative current applied to the servo-valve increases the flow of liquid fuel (by closing the Liquid Fuel Bypass Valve). When the liquid fuel flow feedback is equal to the liquid fuel flow reference, the servo current should be approximately equal to the null bias current value; for a TMR panel that would be approximately -2.67% for each of the three processors. A properly configured SIMPLEX panel should have approximately -8.0% per processor. (1% equals 0.1 mA; 10% equals 1.0 mA; 100% equals 10.0 mA)

Did you observe any fuel flowing out of the false start drains after the firing attempt? If not, then it's not very likely there was any appreciable amount liquid fuel flowing into the combustors. If you observed any liquid fuel flowing out of the false start drain valves, then it's likely there was liquid fuel flowing into the combustors.

Usually, when attempting to start a machine on liquid fuel during a false fire when the fuel flow is steady and equal to the reference there will be a slight whitish "haze" coming from the exhaust stack. This is unburnt, atomized liquid fuel that did not condense in the unit (and would have flown out of the false start drains had it condensed). This usually takes a few seconds to see coming from the stack after firing starts and fuel actually starts flowing--longer if there is an HRSG ("boiler") in the gas turbine exhaust. But, once the fuel flow-rate is stable and equal to reference for some time the whitish haze should be visible against a clear sky (few to no clouds) during daylight hours. This is normal and to be expected when false firing.

I'm a little concerned about extending the firing timer to 2 minutes.... Once liquid fuel does start flowing into the unit at the required flow-rate you could be putting a lot of liquid fuel into the unit and exhaust. Be certain to CRANK the machine for several minutes before attempting a START with the spark plugs enabled.

But, from the information provided, it would seem there is not a stable liquid fuel supply from the storage tank(s) to the turbine, or there is some problem with the liquid fuel pump clutch. But, it's way too early to tell if there is a problem with the servo-valve (not likely, unless the L.O./Hydraulic flush wasn't done well; or the unit has run for some period of time on gas fuel and has never run or rarely run on liquid fuel).

I think O&G are not providing hydraulic accumulators on Frame 5s these days.... So, were you monitoring hydraulic supply pressure during the firing period? Was it stable or jumping around when the liquid fuel bypass valve was unstable?

What is the servo gain value for the liquid fuel bypass valve?

What is the null bias current value for the liquid fuel bypass valve?

It's also usually helpful to have some stationed at the manual selector valve/gauge at the liquid fuel flow divider during firing. The handle can be scrolled through the 10 combustion lines to see if there is any liquid fuel check valve(s) that is(are) not working properly. Also, there is usually a position to monitor liquid fuel supply pressure upstream of the high pressure liquid fuel pump and one to monitor pressure downstream of the high pressure liquid fuel pump.

Once a stable and steady flow of liquid fuel is obtained to the suction of the high pressure liquid fuel pump, the Liquid Fuel Bypass Valve will then be able to stably control the liquid fuel flow-rate. The discharge pressure of the high pressure liquid fuel pump must be just slightly higher than the cracking pressure of the liquid fuel check valves in order for liquid fuel to flow into the fuel nozzles/combustors. If any combustor has a very low pressure, consistently during successive firing attempts, then it's likely that liquid fuel check valve is not working properly.

If the unit has been running on gas fuel for some time, and has never run or or rarely run on liquid fuel the liquid fuel check valves are known to develop leaks in the reverse direction. This causes air to build up in the high pressure liquid fuel filter canister, and in some cases to back-pressure the liquid fuel stop valve and open it very slightly, allowing air to get into the low pressure liquid fuel filters (every unit does not have both high- and low-pressure liquid fuel filters; some even have two low-pressure liquid filter assemblies in series, so you need to consult the P&IDs for the units at your site for site-specific details).

Lately, it has also been observed that the liquid fuel flow divider speed pick-up gap is being set to a very small value and that causes some problems with mechanical vibration of the toothed wheel to appear as liquid fuel flow to the Speedtronic. Do you know what the gap settings for the liquid fuel flow divider speed pick-ups is/are?

Please write back to let us know what you find. But, if this is the first time the units have been run (or false fired) on liquid fuel, this is not uncommon. Lastly, there have been many instances of poor installation practices causing dirt, rocks, cigarette butts, weld slag, welding rods, and other "unusual" objects to be found in the liquid fuel check valves. There have also been problems with plugged liquid fuel filters on initial firing attempts because the liquid fuel supply piping from the storage tanks to the turbine weren't flushed, or weren't flushed properly. Also, the strainers of liquid fuel forwarding pumps have been found to be plugged on initial operation because the tanks weren't properly cleaned after construction, or the liquid fuel was not very clean.

 

Hello,

The unit is new installed and its the first time to to do false fire. the hydraulic pressure is stable and equal to 82 bar.

> What is the servo gain value for the liquid fuel bypass valve?
2.5
> What is the null bias current value for the liquid fuel bypass valve?
2.67

No fuel arrive after check valve because the servo valve open and close in few seconds

But I understand from you the high flow rate sensed by flow divider magnetic speed pickup lead the servo current to be high positive value right?

If the gap of magnetic speed pickup is small is it lead to read very big rate of flow or no?

The fuel is clean and the pipes from storage tank is flushed.

is that ok value that servo current is 74% during unit is stopped (zero speed) or it should be less than 74%? in other GT turbine its 61% when the unit is stopped.
and when the servo current is it ok to be -15 or it should be another value?

I tried to make calibration to regulator 2 of servo valve but i have the message appear (calibration permissive has been lost) and I didn't know how to make perssives ok. I am really need your help in this matter.

thanks for your answers
Ayman

 

Ayman,

First, I'm not writing this just to/for you, Ayman. Many people read these threads, so I'm putting a lot of information here. And, I want people (you included) to think <b>SYSTEM</b>. The Speedtronic monitors and controls SYSTEMS. Everything in the system has to be working properly for the Speedtronic to work properly. First, and easiest, thing to do when troubleshooting is to make sure everything outside the Speedtronic is working properly--and, amazingly enough, the problems usually go away. Before having to delve into the mystical and magical Speedtronic. There are many parameters and variables the Speedtronic doesn't control, or even monitor. Especially in the liquid fuel system. Commissioning a liquid fuel system takes planning, understanding, and patience--whether it's the first time, or after a maintenance outage. Think SYSTEM always, when working with a Speedtronic. The mysterious and magical Moog servo-valve is the last thing you want to replace.

You have to get liquid fuel flowing through the entire system before you can determine that the Moog servo-valve, or the Speedtronic, is or is not working properly. Patience. Read on. And ask questions if you need clarification.

You have said you are working on a Frame 5 GT with a Liquid Fuel Bypass Valve. That means it most likely does NOT have LVDTs for position feedback. That means there is nothing to calibrate.

It is a <b>myth</b> that when someone is calibrating LVDT feedback they are "calibrating the valve". Only the LVDT feedback is being calibrated, and only LVDT feedback can be calibrated. If people would stop teaching and saying, "We are going to calibrate the LFBV (or the SRV, or the GCV, or the IGVs)," and start teaching and saying, "We are going to calibrate the LFBV LVDT feedback (or the SRV LVDT feedback, or the GCV LVDT feedback, or the IGV LVDT feedback)," everyone would start to finally understand that "calibration" is only done on LVDT feedback.

Calibration doesn't mean a servo-operated device is going to be more stable. It only means that the feedback from the LVDT(s) installed on the device the servo is controlling the hydraulic flow-rate to/from is going to be scaled "properly." And I say "properly" because most people "calibrating" LVDT feedback don't measure actual device position and compare it to LVDT feedback to determine if the calibration has been done correctly. They compare the position reference to the LVDT feedback and if the two are "GE" (Good Enough) then they deem the "calibration" to be accurate and complete. And that's just patently false.

The liquid fuel system on digital Speedtronic systems converts FSR (Fuel Stroke Reference) to a liquid fuel flow-rate reference (lately it's a flow-rate expressed in percent), and then the regulator in the VSVO compares the reference to the flow-rate feedback from the speed pick-ups on the liquid fuel flow divider, which is scaled into a percent value (so that flow-rate reference, in percent, is compared to flow-rate feedback, in percent).

When the unit is shut down, you have said the reference to the LFBV is -25%. When the unit is not running, the feedback from the speed pick-ups on the flow divider is 0.0%, so the regulator--trying to make the feedback equal to the reference--puts out a LOT of positive servo current to the servo (positive servo current because positive servo current shuts off the flow of fuel).

When the unit is trying to fire, what usually happens is that the LFBV is closed (sometimes slowly, sometimes very quickly) to increase the fuel flow to the fuel nozzles/combustors by applying a small negative servo current (because negative servo current increases the flow of fuel). The amount of current will usually be a small negative value (a few percent) until the LFBV reaches a point at which it is fully closed and then sometimes, if there is no fuel flowing, the Speedtronic will keep decreasing the amount of current (to a large negative value) to try to increase the fuel flow.

When there is liquid fuel flowing, and the flow-rate feedback equals the flow-rate reference the current is reduced to a slightly negative value approximately equal to the null bias value. If the reference changes, the current being applied will change (usually by fractions of a percent) to increase or decrease the fuel flow-rate as necessary until the flow-rate matches the reference, and then the servo current will go back to approximately null bias value.

If the liquid fuel flow-rate from the flow divider speed pick-ups is going as high as you have said, then I would have to believe one of two things is happening. First, there is a lot of air in the fuel supply line that hasn't been properly bled out of the lines--this is the most likely problem.

Second, the scaling of the liquid fuel flow divider feedback from frequency to percent isn't correct. Possible; worth checking; but not very likely. (The Control Specification usually has a LOT of information about the liquid fuel system and configuration in Sect. 05.01.nn. There is usually a section called "Expected Liquid Fuel System Characteristics" or something similar, where expected values of FSR and liquid fuel flows are listed in a table. The scaling values for the flow divider feedback should be there, too.).

The liquid fuel flow divider is a device for dividing the flow from the high-pressure liquid fuel pump into 10 equal amounts to each of the fuel nozzles/combustors. To accomplish this, the flow divider has ten "wheels", all of the somehow physically coupled to each other. The flow from the pump enters a chamber/manifold that feeds each of the 10 wheels. Because no single wheel can spin faster than any other wheel (they are physically coupled together, either on a shaft (the horizontal flow divider) or by gears (the vertical flow divider) that's how it's divided 10 equal flow-rates.

Connected to those 10 wheels is yet another wheel (sometimes two) with teeth on it, and the speed pick-ups monitor the speed of rotation of this toothed wheel to provide the feedback to the Speedtronic.

The speed pick-up gap doesn't affect the frequency of the speed pick-up output; only the point at which the frequency is sensed by the Speedtronic. If the gap is too small, then the voltage produced will be higher, and that can cause mechanical vibration of the toothed wheel to appear as if there is liquid fuel flowing when there isn't. I only mention it because it has been a problem on some sites, but not all. The Speedtronic only senses frequency, but the pick-up output (frequency) at very low speeds is a function of gap distance. At higher speeds (flow-rates) the voltage doesn't affect the frequency--and the Speedtronic is only interested in frequency from the speed pick-ups, not the voltage level.

What usually happens when there's air in the fuel supply line is that a "slug" of fuel hits the flow divider followed very closely by a "slug" of air, and that causes the flow divider wheels to spin. They spin very fast as the slug of liquid fuel is pushed through the flow divider by the air (there is another slug of liquid fuel behind the air).

The speed pick-ups see the very high feedback from the flow divider and USUALLY the Process Alarm "Start-up Fuel Flow-rate Excessive" is annunciated, which trips the turbine and shuts the liquid fuel stop valve and stops firing. <b>Is the "Start-up Fuel Flow-rate Excessive" Process Alarm being annunciated during firing?</b>

If the flow divider feedback doesn't jump too high when air/fuel is flowing through it, then the servo-valve will behave as you described, where it puts out a "small" negative current trying to increase the liquid fuel flow-rate, then when the flow-rate jumps up excessively it puts out a large positive value to open the valve, and so on, and air and fuel pass through the flow divider.

We have no way of knowing how the liquid fuel supply piping is routed from the storage tank(s)/forwarding pump(s) to the turbine. If the piping goes up into an over head pipe rack, and then down to the liquid fuel stop valve inlet, the high point in the piping is an excellent place for air to collect. And, unless there is a high-point vent in the overhead piping, it's going to be very difficult to get that air out. Sometimes, it takes a long time to get all of the air out, depending the length of the pipe, and how high it is, how many ups and downs in the piping run, etc.

The <b>best</b> way to get all the air out of the liquid fuel system is disconnect the fuel supply lines from the liquid fuel check valves at the fuel nozzle, connect lengths of rubber- or plastic hoses to each of the fuel supply lines and run them out to barrels (one on each side of the turbine compartment). Start the liquid fuel forwarding pump, open the liquid fuel forwarding stop valve (presuming one is installed in the system), pressurize the system up to the liquid fuel stop valve on the unit. Go to each low-pressure filter canister and bleed all the air out of the canister. If the filters are duplex filters, use the transfer fill valve to fill the other canister, bleeding the air out of it as well. Do this until a solid stream of liquid fuel flows out of the vent(s).

During this part of the process, the pressure from the forwarding pump(s) should be stable. If not, adjust the pressure regulator on the liquid fuel forwarding skid (if present). There is usually a needle valve in the sensing line to the top of the pressure regulator that can be used to stabilize the pressure. If there is a lot of air in the piping, it will very likely not be possible to completely stabilize the pressure. But, many times it's possible, be venting air out of the low-pressure liquid fuel canisters, to be able to able to adjust and stabilize the forwarding pump pressure.

Then, you will need to establish hydraulic system pressure on the unit to be able to force the liquid fuel stop valve open, and the force the liquid fuel stop valve open. The liquid fuel bypass valve is OPEN at this point, and it is in a "bypass" loop around the high-pressure liquid fuel pump, so liquid fuel will flow through the bypass valve around the high-pressure liquid fuel pump to the flow divider, through the flow divider, and to the hoses running out of the turbine compartment into the barrels.

If there is a high-pressure liquid fuel filter near the high-pressure liquid fuel pump, vent the air out of the high-pressure liquid fuel pump canister. If there are low-pressure liquid fuel filters down-stream of the liquid fuel stop valve, be sure to vent the air from them as well.

By doing this you are using forwarding pump pressure to push liquid fuel--which pushes most of the air out of the system, and all the way to the liquid fuel check valves at the fuel nozzles (effectively). You have essentially filled all of the liquid fuel system components, including the suction and discharge of the high-pressure liquid fuel pump, and all of the 10 lines supplying liquid fuel to the nozzles.

Be sure to have an extra barrel on each side of the turbine compartment, and to have someone stationed at the HMI to unforce the liquid fuel stop valve to close it if there is an unintentional leak or spill.

When you have a stable flow of low-pressure liquid fuel flowing out of each of the 10 hoses into the barrel (no bubbles!) for a minute or so that is probably time to close the liquid fuel stop valve, and stop the forwarding pump, and close the forwarding stop valve (if present).

Remove the hoses; re-connect the lines to the liquid fuel check valves. You have now removed MOST of the air from the system, all the way from the storage tanks to the fuel nozzles. If there any overhead loops of liquid fuel piping between the discharge of the forwarding pumps to the liquid fuel stop valve on the unit, those are potential areas which might not have been completely "bled" of air. If the A/E (Architect/Engineer) didn't design high-point vents in those overhead loops (and they should also have pitched the piping slightly and place the vents in the highest point in the pitched piping!), then there might be some problems during initial operation on liquid fuel until the flow through the piping gets high enough to push the liquid fuel out of the piping. Sometimes this doesn't happen until there is some load on the unit.... Which can result in a loss of flame trip or some momentary instability as the flow pushes the last of the air out of the piping.

A lot of Start-up "Managers" and Commissioning "Managers" and Construction "Managers" think that when a liquid fuel pump is started and run that a solid "wall" of liquid fuel moves through the piping pushing all of the air out of the piping as it moves along, so no bleeding of air out of the piping is required. That's just not true. It doesn't happen that way. It has never happened that way. They just point the finger at that darned Mark VI and say, "That's the problem!" A LOT of start-up TAs/field service people also believe similarly (mostly because they haven't been properly trained or even considered how a pipe gets filled with liquid).

What you are experiencing is NORMAL. If there are any overhead piping loops in the liquid fuel supply piping between the storage tank(s) and the liquid fuel stop valve, there's likely going to be some air in the line. Even if there's no overhead loops; even if the A/E designed all of the liquid fuel supply piping to be below grade in trenches until it enters the liquid fuel stop valve, there can still be air trapped in the piping if there are any elevation changes between the storage tank(s) and the liquid fuel stop valve.

Even if there is no air in the liquid fuel supply piping, it takes some times and flow through the liquid fuel system on the unit to fill the 10 lines all the way to the fuel nozzles. It doesn't just get filled in a few seconds on the first, or second, or sometimes even the third firing attempt.

The purpose of a false fire--especially the first false fire during initial commissioning of a unit, or after a maintenance outage--is to get the air out of the lines all the way to the fuel nozzles. The evidence of that is when the flow feedback from the flow divider is stable and equal to the flow reference. That's why you monitor flow reference and flow feedback during a false fire. That's why you monitor false start drains, to make sure that liquid fuel has indeed made it into the combustors. That's why you look for a whitish haze coming from the stack after the false fire, to be sure liquid fuel vapours have made it through the unit.

But, most importantly, the purpose of a false fire is to make sure the liquid fuel bypass valve can control the liquid fuel flow-rate to match the liquid fuel flow-rate reference. And, if there's air in the lines, then, that can't be proven or demonstrated.

Sure, it's going to take some coordination and some time to do the procedure above. But, it will be worth it! If you don't want to do the whole procedure--or, more likely you can't convince the Start-up/Commissioning "Manager" to do it--then at least start the forwarding pumps and pressure the system up to the liquid fuel stop valve, and make sure the low-pressure liquid fuel filters are bled and filled. As long as the forwarding pump pressure is less that the liquid fuel check valve cracking pressure--and it MUST be!!!--you can even open the liquid fuel stop valve (with hydraulic pressure and by logic forcing) and use forwarding pump pressure to fill the suction/discharge lines of the high pressure liquid fuel pump, the liquid fuel bypass valve, the high-pressure liquid fuel filter, and even get some liquid fuel into the flow divider. As long as the liquid fuel check valves are working correctly AND the forwarding pressure is less than check valve cracking pressure, you won't be putting any fuel into the unit!

Then, you can re-try your false fire to get the desired results-- which are:

1) Actual liquid fuel flow-rate (from the flow divider) matches reference for 10 or 15 seconds during firing;

2) Liquid fuel observed coming from the false start drains a few seconds after flow has been established and is stable;

3) Thin, white haze observed coming from the exhaust stack (indicating liquid fuel vapours have made it through the turbine and exhaust).

STOP blaming, or trying to blame, or believing others when they try to blame, the Speedtronic when the false fire doesn't work the first time, or the second, or even the third. DO try to think about the entire liquid fuel system, and all of the possible problems that can occur, all of the possible places air can exist, and how much flow is required to displace all of the air and fully fill the lines with liquid fuel, and how long that can take at firing flow-rates (which aren't that high!).

GE has designed a fine turbine control system, but they haven't programmed it or configured it to be able to sense all of the problems with the external systems it monitors. And, the Speedtronic doesn't monitor--or even control--all of the parameters that are necessary for proper operation of the turbine. It's a credit to the early turbine system designers and engineers that the system works as well as it does.

I realize I've written a lot, but, I hope it helps. Again, to everyone reading this: Think <b>SYSTEM</b>.

 

CSA /Aymen,

Thank you for the DETAILED reply.

Yes, we too experienced may problems during the liquid fuel change over test time. We have decided to place a automatic vent valves in the liquid fuel lines. Probably two will be placed, where to install automatic vent valve is being discussed.

As we mentioned in past thread, once we replaced the servo valve in the VC3 everything works fine. So we may believe that servo failed but it not so.

Take care
G.Rajesh

 

Dear all,

The problem is not solved yet. We connect the rubber hoses to each of fuel supply (before check valve) and we start false fire test again. the unit start cranking after that purge for 2 min. we select fire mode so ignition phase happened and stop valve energized so the fuel arrived to the barrels. and the flow look normal in each supply line. (of course we connect rubber hose after check valve, but no fuel arrived because the pressure is not enough to crack the check valve in previous test). but the servo valve still open and close and the flow rate sensed by magnetic speed pickup is very very high value, its arrived from 900% to 20,000% and of course its not real value because the flow looks not high in rubber hoses. and also there is no air in the pipes because its filled and the fuel is flowed to barrels normally.

also I checked the Gap setting of flow divider speed pickups. its ok according the GE specification and the Gap is same in another GT in the plant. and also the scale setting of pulse rate is same in another GT working properly.

is that may be electrical noise happened to the signal wires? If so is that electrical noise lead to very high reading arrived to 20,000 %? I am very confused...

After that we re tried again but without fuel flow. we forced the fuel metering skid valve to 0 so no fuel arrived. and we separate the pipes from bypass valve and the result is the servo valve open and close case is not happened. the servo current changed to -12% the flow reference is 4.5% and certainly there is no flow, because there is no fuel. so the servo current still negative value to try to increase the flow to turbine and that is normal state. and I think by this experiment that the bypass valve and Moog work is properly right?

so i don't know how to solve the problem of very high reading of flow. And especially it's fantasy value.

Let's assume that the bypass valve is closed (full flow to turbine). is it possible to read this imagine value of flow?

Thanks All
Ayman

 

Ayman,

If I understand correctly, you have managed to get fuel to the turbine compartment (and to some barrels) so the lines have basically been purged of air. Some air will make it's way into the lines in the turbine compartment in the course of re-connecting them, but it should easily be purged during a firing attempt.

You say you have removed the piping from the liquid fuel bypass valve and have tried to stroke it. Because there is no LVDT <b>AND</b> because the regulator is configured to be a flow control loop meaning the regulator requires flow feedback to work properly, when you try to move the bypass valve without feedback (from the flow divider) it will just go fully closed or fully open (depending on what kind of reference you provide it--negative servo current or positive servo current). Without any feedback for the summing junction the regulator will just "saturate" in either direction and the valve will go fully closed or fully open.

If you have a frequency generator you can disconnect the flow divider input wires at the Mark VI and connect the frequency generator in parallel to all the flow divider inputs (I think they are just using two now; they used to use three in the past). Set the frequency generator to output a sine wave, or a square wave, or a saw-tooth wave--it doesn't matter which, <b>BUT the output ***MUST*** cross the 0-volt axis!!!</b> Meaining it must NOT have any DC offset. Many inexpensive signal generators have outputs that do NOT cross the 0-volt axis and vary only from 0-5 V, typically called TTL outputs, which will not work with Speedtronic panels.

With hydraulic pressure, set the frequency generator to something like 1000 Hz, at approximately 6 VAC peak-to-peak. You should be able to read some value in Toolbox. That value should scaled, in percent, based on the configuration values in Toolbox and which were downloaded to the Mark VI. Then put a small positive reference which is just slightly higer than the frequency you are inputting to the Mark VI on the servo output using the Manual positioning feature of AutoCalibrate, and the valve will start moving. The valve plug should start closing, and if you then make the reference equal to the input the plug should stop moving. You may have to play with ramping both the reference and the feedback up slowly to get the valve plug to be stable.

The purpose of this is to prove that when the feedback matches the reference the Mark VI will stop the valve plug from moving--meaning that the Mark VI and the servo are working correctly.

Next, I would disconnect the frequency generator from the Mark VI, reconnect the speed pick-up input wires, and go out to the JB next to the flow divider and disconnect the flow divider wires from the terminal board, and connect the frequency generator first to one of the inputs, and start inputting a frequency and monitoring it on Toolbox to see that it matches the frequency generator output and is scaled in percent properly. If that's successful, move the frequency generator to the other input(s) one at a time and do the same. You are trying to see if there is some problem with noise being picked up by the interconnecting wiring with this test. You should also check the shield drain wires at every junction box between the flow divider and the Mark VI to ensure that the the drain wires are properly terminated (at ONE end ONLY).

I don't understand how the Mark VI can display 900% or 20,000%. I believe that those values far exceed the maximum signal scaling of the percent scale code/type. So, that definitely seems to be some kind of problem. If you can get good frequency and scaled feedback readings with the signal generator at the Mark VI and from the flow divider junction box, then it would seem something is wrong with the wiring allowing noise to get on the signal from somewhere.

Have you tried using frequency meters to measure the frequency of all the inputs from the flow divider when you are performing a false fire? Have you trended all of the individual inputs during the false fire? Hooking up a frequency meter to speed pick-up inputs to a Speedtronic panel will NOT harm the signals or the Speedtronic panel (contrary to popular belief, Rahul P. Sharma).

I believe the reason the servo current is fluctuating is because the feedback is fluctuating. If it's calling for flow and it sees a LOT of flow then the Mark VI will reverse the current to try to reduce the flow, and if the flow then goes to 0 the bypass valve will reverse the current again trying to establish flow and if the flow feedback goes too high then it will reverse the current again.

<b>Based on the information provided,</b> it seems as if there is something wrong with the flow feedback, either the scaling is not correct, or the wrong scale code/type was chosen, or something like that. I hope you are using the Trend Recorder function of Toolbox to capture data during your firing attempts, and not looking at values on CIMPLICITY displays--that would not be a good thing to tell us (that you are watching CIMPLICITY displays and not looking at Toolbox trends!). I believe <b>based on the information provided</b> the Mark VI and the servo-valve are working properly. The test above (of course you will have to remove the piping from the liquid fuel bypass valve to see the plug move!) will prove if the regulator can control the position of the bypass valve if the feedback and the reference match.

And by using a frequency generator to simulate inputs at both the Mark VI and from the flow divider you should be able to verify if noise is affecting the signals or not. You have said that the Mark VI is reporting several thousand percent feedback. The percent value is derived from the frequency input. If the frequency input is within reason (you can measure the frequency input at another machine--again, it WILL NOT affect the running machine!!!) and input that frequency to this panel to see what percentage is calculated and displayed. I think there is something wrong with the configuration of the frequency input to cause those kinds of percent numbers--again, presuming the numbers came from Toolbox and NOT from CIMPLICITY displays. Use a frequency meter or two to monitor the feedback frequency during firing attempts to make sure the two are reading similaryly (they should be identical!), and are not reading excessive numbers (again based on data from the running panel).

DO write back to let us know what you find.

 

Ayman,

You didn't mention if the 'Start-up Fuel Flow Excessive' alarm is being annunciated (or if the logic is being forced to keep it from being annunciated and tripping the turbine.?.?.?).

Unlikely, but there may be a problem with the flow divider.... Those things are usually pretty solid, and if you have a new one it's not likely there's a problem with it. If it's made by Roper, it's a good one, and they have excellent technical support, too. Unless there is some really unusual gearing inside that flow divider, or some shaft/wheels have come uncoupled it's just hard to imagine the kind of values you are reporting. That's why it's important to physically measure the frequency coming back from the flow divider when fuel is flowing through the flow divider--even if it's only the liquid fuel forwarding pump pressure through the liquid fuel bypass valve through the lines to the check valves. That information is really critical to know.

You could record the flow divider nameplate information and contact the manufacturer for clarification about the number of teeth, etc. GE should be able to help with that, also.

My money is still riding on some configuration problem. Unfortunately, what with the "dongle" thing these days, I don't have access to any recent Mark VI .m6b files via Toolbox (don't have a dongle for the newer versions of Toolbox). So, I can't cite specifics of configuration issues in Toolbox.

There are some configuration values "buried" in Toolbox for the speed pick-up inputs. I think you are probably using the VSVO speed pick-up inputs for the liquid fuel flow divider inputs, and if so there is something there about type of wheel/number of teeth, something like that. Compare ALL the configuration values for the liquid fuel flow divider inputs to the running machine; something sure seems amiss.

Also, if you are using the VSVO speed pick-up inputs, there might be some issue with the VSVO. I think there were some available speed pick-up input points on the TTUR which have been used in the past for liq fuel bypass valves, but I don't know if more recent versions of Toolbox can or can't use those input points for the regulator on the VSVO card (I don't think they can, but I'm not 100% sure about that; sorry!).

Most likely, it's either a configuration issue, or a flow divider issue <b>based on the information provided.</b>

If you really have time on your hands, you could try reconnecting the hoses to the barrels and using forwarding pump pressure simulate flow divider feedback to see if the bypass valve would regulate flow.... That would be a LOT of work, and potentially messy, too. Fuel spills are tolerated very well in many parts of the world, but, if you're out of ideas....

 

Ayman,

You also didn't mention if you had observed the coupling between the fuel pump clutch and the high pressure liquid fuel pump to make sure the pump is turning and the clutch or coupling is not slipping.

If there is no fuel getting past the check valves and (again, you <i>didn't</i> say if you have had someone monitoring the pressures to the fuel nozzles with the manual selector valve at the flow divider) the pressure is low even though the bypass valve is closing then it may be a problem with the high-pressure liquid fuel pump, or the coupling, or the clutch. Not very likely, but the high-pressure liquid fuel pump is the only way that the pressure is going to build up high enough to overcome the check valve cracking pressure. If the clutch is slipping, or the coupling is slipping, or, unlikely as it may be, there is some problem with the pump, then there won't be sufficient pressure built up when the bypass valve is closing. If there is some problem with the pump and pressure starts to build up and then it decays for some reason that could explain (some) of the flow divider feedback fluctuations--though not likely the absurdly high values you are reporting.

Remember: Think <b>SYSTEM</b>. Troubleshooting is a logic process of elimination. Experience helps sometimes by understanding where the likeliest cause of problems can be, but sometimes it's just a slog of eliminating possibilities one at a time until, finally, the culprit is found. Sometimes, one even finds multiple problems, though that doesn't happen too often on new units.

One last thing: Because there are no LVDTs on the liquid fuel bypass valve, and because it's easy to view the valve plug or the valve stem, it's also virtually impossible to verify the polarity of the servo current from each of the three processors being applied to the servo coils. So, if you're not certain about that, it's advisable to pull the piping from the valve and perform a servo current polarity check while being able to view the valve plug. You will need hydraulic pressure to do that. If the polarity is correct at a non-firing reference (-25%, +40% current, I think is what you have written), the valve plug will not move--presuming it's in the full open position when you're looking at it....

 

Dear all,

The problem is solved by configure the pulse rate scale according to another same GT unit and downloading configuration. after that the servo valve working properly also the flow rate now is ok and become near to the flow reference.

But as a result of this problem there is hydraulic oil leak from the servo valve.

Ayman

 

The speed of the flow divider cannot possibly be 20000% There would be mechanical failure surely.

Any chance this is all about EMI or something with the mag pick-up cabling? Or a problem with the mag pick-up? (Did you change pickups to see what happens?)

 

All,

This is good news--and still further proof that the ever-mystical, mythological, marvelous, mysterious and magical Moog servo-valve is <b>NOT</b> the root of all evils. And should not be the first suspect or the first thing to be replaced when it doesn't behave as expected.

Thanks for the feedback!

 

Ayman,

Thank you for the feedback. This is what CSA is expecting from all.

take care
G.Rajesh

 

G.Rajesh,

Isn't this what <b>you</b> expect, also?

Don't you find the feedback to be of the most help when trying determine if the problem was resolved or the question was satisfactorily answered?