We recently upgraded our GE Frame 6 Gas Turbine from a MarkIV control system to a MarkVIe. Since then we have constantly had false readings on our liquid fuel Flow divider speed pickups.

This is generating the diagnostic Alarm:
"Flow Divider Pickup Unhealthy".

We have checked our cable shielding, increased the gap setting for the probes, replaced the pick-up probes and the flow divider itself.

Has anybody else had this issue?
Is there anything we can do?
Currently it is filling our entire alarm history.

 

Congratulations on your choice of a Speedtronic. Is this a Mark IV-to-Mark VIe Migration or a complete replacement of the Mark IV with a Mark VIe? Is it a SIMPLEX, DUAL or TMR Mark VIe?

Have you locked out the alarm using 'Alarm Lock'?

What does GE say? If this problem started during commissioning, is it a Punch List item and they are going to get back to you with solution(s), or are they aware of the problem?

Or did it begin after they left site?

When does this occur? When the unit is at rest (on cooldown)? When it's running on gas fuel? When it's running on liquid fuel? During starting/acceleration?

Is it on all three speed pick-ups, or just two, or just one? (I presume they are using three speed pick-ups; please confirm or clarify)

What have you increased the gap setting from and to? I believe the Mark IV Control Spec value was usually 0.010 inches. And I don't believe the Mark VIe Control Spec lists a gap, or if it does, it probably lists the same value.

What happens if you disconnect the speed pick-up wiring at the Mark VIe (temporarily for a test)?

What happens when you disconnect the speed pick-up wiring at the junction box closest to the flow divider (temporarily for a test)?

Have you trended the raw signal of the flow divider speed pick-ups when this is happening? What does the trend look like (for each of the pick-ups)?

What inputs are the flow divider speed pick-ups connected to? TTUR or TCAT or TSVO(s)or ???

If the speed pick-ups are connected to inputs that can be either passive or active inputs, are the wires connected properly? How are the inputs configured in ToolboxST?

Please answer all of the questions; without being able to be there to look at the problem ourselves we have to ask these questions and work through these troubleshooting steps, and we could ask these questions over the course of several exchanges, or we can do it in one fell swoop. (Call me crazy, but I prefer one fell swoop.)

 

Thank you for replying to my Post.
I have tried to answer your questions below.


>Congratulations on your choice of a Speedtronic. Is this a Mark IV-to-Mark VIe Migration or a complete replacement of the Mark IV with a Mark VIe? Is it a SIMPLEX, DUAL or TMR Mark VIe?<

It is a MarkIV to VIe migration, so all the field cable I/O termination boards remained the same. It is TMR.

>Have you locked out the alarm using 'Alarm Lock'?<

Yes, we can use this as a temporary solution until will clear the alarms.

>What does GE say? If this problem started during commissioning, is it a Punch List item and they are going to get back to you with solution(s), or are they aware of the problem? Or did it begin after they left site?<

It began during the commissioning before they left site and as of yet they have not been able to offer us a solution.
A field technician returned to site yesterday, but he has not brought any solution or plans with him.

>When does this occur? When the unit is at rest (on cooldown)? When it's running on gas fuel? When it's running on liquid fuel? During starting/acceleration?<

We are getting the false readings while cranking, starting/running the unit on gas fuel. A thought is it may be generated from vibration?

>Is it on all three speed pick-ups, or just two, or just one? (I presume they are using three speed pick-ups; please confirm or clarify)<

All three pickups show the same problem.
The migration to the MarkVIe now only uses 2 instead of 3 the pickups?

>What have you increased the gap setting from and to? I believe the Mark IV Control Spec value was usually 0.010 inches. And I don't believe the Mark VIe Control Spec lists a gap, or if it does, it probably lists the same value.<

You are correct our original setting was 0.01 inch. The gap has been increased to 0.025 inches. (recommended by GE tech.) this lowered the level of the false signal however it still occurs, plus we are concerned about the probe becoming too far away from original turbine specifications.

>What happens if you disconnect the speed pick-up wiring at the Mark VIe (temporarily for a test)?<

When the probes are disconnected at the MarkVIe terminals. The false signals stop.

>What happens when you disconnect the speed pick-up wiring at the junction box closest to the flow divider (temporarily for a test)?<

When the probes are disconnected at the flow divider. The false signals stop.

<Have you trended the raw signal of the flow divider speed pick-ups when this is happening? What does the trend look like (for each of the pick-ups)?<

The trend is irregular and the signal bounces around. (This is while running on gas fuel and no liquid fuel flow should be detected).
We have looked at the signal with a scope and it appeared to be less than 100Hz and less than 20mVpk. (irregular).

Manual Specifications states for correct reading signal needs to be:
@2Hz 33mVpk and @12kHz 827mVpk.

>What inputs are the flow divider speed pick-ups connected to? TTUR or TCAT or TSVO(s)or ???<

The terminations come through the original MarkIV terminal board - into the TCAT / PCAA boards.

>If the speed pick-ups are connected to inputs that can be either passive or active inputs, are the wires connected properly? How are the inputs configured in ToolboxST?<

I do not know if there is a passive / active selector (e.g. a jumper switch). The hardware manual for the PCAA board shows only 2 terminals for each of the 2 Flow inputs.

>Please answer all of the questions; without being able to be there to look at the problem ourselves we have to ask these questions and work through these troubleshooting steps, and we could ask these questions over the course of several exchanges, or we can do it in one fell swoop. (Call me crazy, but I prefer one fell swoop.)<

So far we have also temporarily installed new cable - fully shielded (drain connected at the MarkVIe) to prove it is not a cable shielding issue.

We never had this issue with the MarkIV. It seems the sensitivity of the MarkIVe is a lot higher?

The probes also seem to work OK while running on liquid fuel. (e.g. changes over from gas to liquid fuel ok).
We do have issues on start-up with liquid fuel. A lot of white smoke is generated. (excessive fuel? currently GE are looking into this as well, but there is conjecture as to wether the two may be related).

Thank you for your assistance. If you would like to contact me directly please email:

[email protected]

 

Thank you for replying to all of the questions; that was very helpful.

First of all, locking out an alarm <b>that you are aware of</b> that is filling up the alarm queue is good alarm management. You don't want to fill up the alarm queue with nuisance and erroneous alarms. As long as you are actively working to resolve the problem so that the lock can be removed, that's the key. But, allowing a known, nuisance and erroneous alarm to fill an alarm queue when it can be locked out is poor alarm management.

Second, this is not a problem that is going to impact running reliability or availability. At this point, it's just a nuisance problem that can likely be resolved (read below). So, locking out the alarm is not the long-term solution; that was not the recommendation. The ability to lock and unlock alarms is there just for this type of condition, to prevent nuisance alarms from filling the alarm queue(s) and from distracting operators from responding to other alarms. Again, as long as you are actively working to resolve the problem so that the lock can be removed, that's the key.

Third, the sensitivity of Speedtronic speed pick-up input circuits is much higher these days than for previous versions of Speedtronic turbine control systems. Mechanical vibrations which cause the flow divider wheel(s) to appear to move when no liquid fuel is flowing can appear as liquid fuel flow because of the increased sensitivity.

This is not an issue which is peculiar to the Mark VIe. It also has been noted on some Mark VI retrofit installations for years. At some sites, the speed pick-up gap has been increased to as much as 0.048 inches to alleviate the problem.

Those same pick-ups are used for many other applications on GE equipment (and that's both a good and a bad thing in some instances), and the typical gap settings for those other applications can vary from 0.030- to 0.050 inches. The gap in the Mark IV Control Specification was for the Mark IV liquid fuel flow divider speed pick-up input circuit sensitivity, not for the flow divider. The Control Specification is not peculiar to the turbine, it's specific to the control system being used on the turbine, so the 0.010 inch setting in the Mark IV Control Spec is for the Mark IV. The presumption was that it would work for the Mark VIe, and it does in many cases, but it doesn't in all cases.

Now that a Mark VIe has been installed, it may be necessary to increase the speed pick-up gap to reflect the increased sensitivity of the Mark VIe's speed pick-up inputs and the presence of vibrations on the liquid fuel flow divider from the accessory gearbox and shaft vibrations at your site.

When there is no liquid fuel flow, there is nothing holding the flow divider shaft and toothed wheel(s) stationary. The flow divider shaft is simply immersed in liquid fuel and free to "shudder" with mechanical vibrations. Because of the increased sensitivity of the Speedtronic speed pick-up input circuits, this can be interpreted as fuel flow.

Unfortunately, this is not a problem on all applications. It hasn't been reported, to my knowledge, to be a problem for new unit installations with Mark VIe, likely because the flow dividers are new and generally of a different design with different toothed wheels and don't seem to exhibit the same behaviour in response to vibration stimuli when running on gas fuel. It may just be the placement of the liquid fuel flow dividers from one machine to another. Different packagers (John Brown, Alsthom, GE, etc.) located components differently, including different anchoring methods and locations which may cause "magnification" of external vibration stimuli. Some sites have even been able to induce false speed signals on flow divider pick-ups by tapping on the tubing used as conduit between the speed pick-up and the local junction box when the unit was not running and no auxiliaries (Aux. L.O. and/or Aux. Hydraulic Pumps).

So, the problem isn't the speed pick-up, or the flow divider, or even the vibrations. It's the proximity of the speed pick-up face to the toothed wheel(s) in conjunction with the increased sensitivity of the Mark VIe's speed pick-up inputs. It may also be related to the width of the toothed wheel in relation to the face of the pick-up. But, it should be possible to reduce or eliminate this problem.

Because this isn't a problem for all machines, it seems GE hasn't issued a "standard" for a flow divider speed pick-up gap setting. Work with the field engineer at your site to keep increasing the gap to try to reduce and eliminate the spiking. You can do this while the unit is running on gas fuel, or during cranking, by backing out a single speed pick-up slowly, noting the number of turns/partial turns until the spiking stops on that speed pick-up's input. You can then calculate the change in gap from the thread pitch and the number of turns/partial turns the pick-up was backed out. Then increase the gap for the other pick-ups to the same setting while monitoring. It may be necessary to do this a couple of times to get all of the speed pick-ups to go to a very low, or zero, reading. But, you should have this information to get any additional support from GE engineering if the problem can't be completely eliminated or if the gaps have to be significantly increased to eliminate the problem.

This is "field engineering", and is normal for such problems which are not problems for every application and every site. GE devours, and requires, data, and this can take time. But, if done in a logical fashion and with a plan, it can be gathered relatively quickly.

I would begin to be concerned if the gap had to be increased much above 0.055 inches, and you should ask GE to get some kind of comment if you have to go much past 0.050 inches. I would also suspect that if the gap had to be much more than that, then it's something very peculiar to that particular installation. The speed pick-up conduit tubing is very strained, for example. Or the flow divider mounting is such that it is magnifying external vibration stimuli.

In this case, the turbine control system is not just the Speedtronic turbine control panel. It's also the devices and instruments in the field. In the particular case of the flow divider speed pick-up "noise", it may be necessary to add additional bracing to the flow divider mount, or to replace the tubing/conduit between the speed pick-up and the nearest junction box. All problems like this simply can't be anticipated on every retrofit application, and, again, the control system includes devices and components which are not part of the scope of supply. The increased sensitivity of the Mark VIe is likely just pointing to a problem which just wasn't being "picked up" by the Mark IV and has been there for a very long time, possibly since the original commissioning but because of the low input circuit sensitivity and lack of Diagnostic Alarm for circuit problems on the Mark IV just wasn't detected or annunciated.

As for the smoking during start-up, it's likely not related. It's been said many times before on control.com: <b>White smoke during starting on liquid fuel is the result of incomplete combustion.</b> Incomplete combustion means improper atomization, lack of sufficient fuel flow, etc.

If you are having white smoke, I would highly suspect that there are high exhaust temperature spreads, indicating that one or more combustors does not have flame. Presuming the fuel nozzles are properly assembled, with the proper components with the proper tolerances, and that the liquid fuel check valves and purge air check valves are working properly, and that there is sufficient atomizing air flowing during firing and warm-up, this is either caused by insufficient fuel during firing to ensure flame is propagated all around the unit, or because fuel is cut back too much during warm-up (after firing) and that flame is being lost in one or more combustors when the fuel is reduced from firing to warm-up.

When this happens (smoking caused by loss of flame or failure to completely establish and maintain flame in all combustors during firing and warm-up and acceleration), usually as the machine is accelerated flame will eventually be propagated through the cross-fire tubes to all combustors and the smoking will stop and the exhaust temperature spreads will decrease significantly.

Sometimes, during liquid fuel false fire testing, white vapors can be seen coming from the exhaust stack. They are just unburnt liquid fuel vapors (remember the liquid fuel is atomized at the fuel nozzle). If flame is established in some or all of the combustors during firing, when fuel is cut back during warm-up and flame is lost in one or more combustors then white smoke will be visible as the unburnt fuel from the combustors that do not have flame gets heated as it passes through the turbine section and mixes with the hotter gases from the combustors that do have flame. This results in even "thicker", more "billowy" white smoke than a complete failure to establish flame.

So, the fix if the unit reaches FSNL with no smoke at FSNL and low exhaust temperature spreads is to increase firing FSR on liquid fuel slightly, and/or to increase the warm-up FSR slightly to eliminate the exhaust temperature spreads and thereby eliminate the smoke. This is again "field engineering". This will usually solve the problem, but in some very extreme cases it might be necessary to also increase the fuel during acceleration, but that's not usually required and if it were I would suspect other problems. If flame can be established in all combustors without a high exhaust temperature spike, and maintained in all combustors during warm-up without a high exhaust temperature spike, then when fuel is increased to assist with acceleration flame will usually be maintained in all combustors during acceleration.

White smoke due to incomplete combustion can also be caused by problems with purge check valves (check the tell-tale leak-offs; see the Piping Schematic drawings (P&IDs)), and problems with atomizing air during firing. Also, problems with liquid fuel check valves can cause issues. Most liq fuel check valves are set to open when the upstream pressure is about 100 psi, which is the minimum pressure required for proper atomization of the fuel through the fuel nozzle (so-called mechanical- or pressure atomization). If the liquid fuel check valves are opening early, then the fuel won't be properly atomized and smoking will definitely occur. If fuel is leaking through the purge check valves, then the pressure at the nozzle won't be high enough for proper atomization and smoking will definitely occur.

Again, the key is to monitor and try to reduce the exhaust temperature spikes and -spreads during firing and warm-up and acceleration. And, this takes tuning.

Now, the next comment you are going to make is, "The firing FSR is not the same on the Mark VIe as it was on the Mark IV." That may be true--for a lot of reasons. If the liq fuel flow divider feedback scaling constants in the Mark VIe are the same as they were in the Mark IV, it will definitely be true. But, if you're able to fire and accelerate to FSNL, and operate with low exhaust temp spreads while at rated speed, then the scaling constant difference is likely negligible and can be accounted for by slight changes in firing- and warm-up FSR values. Again, "field engineering."

Lastly, some older dual fuel machines were just never that easy to fire on liquid fuel without too much smoking. And on some machines, it was near impossible to shut them down without smoking on liquid fuel(after breaker opening during fired shutdown). It's all related to the engineering compromises made to accommodate the dual fuel capability, and to different combustor and fuel nozzle designs.

You're going to say something like, "But we didn't have that kind of smoking on the Mark IV." And that may be true for a variety of reasons. The Mark IV start-up algorithm/control was not nearly as accurate as the Mark VI or Mark VIe. There was no acceleration control, and fuel was just increased at a programmed ramp rate and it was hoped (literally) that the shaft speed would accelerate at an "acceptable" rate with low exhaust temperature spreads and no exhaust temperature spiking.

The tools to analyze exhaust temperature spreads during starting of the Mark IV just didn't exist, and as such, it is quite common to find that excessive amounts of fuel were used during firing, warm-up, and acceleration with Mark IVs just to eliminate smoking. That was done when the Mark IV was installed. The values for firing and warm-up FSR in the Control Specification were never meant to be hard and fast numbers, and were subject to adjustment during commissioning as necessary to achieve smooth steady acceleration with low exhaust temp spreads and minimal exhaust temperature spikes.

So, through field engineering of the retrofit installation, it just might be that higher fuel flows are what's required for the unit(s) at your site to reduce or eliminate smoking. Higher, that is, than might be expected, but as long as the exhaust temperature doesn't spike too high during firing and doesn't reach the isothermal limit during acceleration, if the exhaust temp spreads are good and the smoking is reduced to acceptable levels and the acceleration is not too fast, then that's what you're shooting for.

You can, and should, use the Mark IV operating characteristics as a guideline. But, as with any control system retrofit, it's not likely that every parameter will be exactly the same in the new control system as it was in the old control system. The Mark IV executed the sequencing at an 8 Hz rate in each of the three control processors independently of each other (they were not synchronized at all; the execution of each control processor started and stopped independently of any other control processor and was the source of many nuisance Voting Mismatch Diagnostic Alarms especially for water- or steam injection as well as for TNR increment/decrement logic signals!).

The Mark VIe executes the sequencing (typically) at a 25 Hz rate in each of the three control processors simultaneously and while very tightly synchronized (the execution starts and stops for each control processor at the same time for all three processors). So, there are fundamental differences, and improvements, in the way the two systems perform the control and protection. Which means they won't be exactly identical even on the same machine when nothing else has changed.

No two machines are ever exactly alike, even if they were built and installed and commissioned at the same time. And older machines which have been operated for years develop their own idiosyncrasies, just as most mechanical equipment does. A new control system isn't going to change that, and it's not likely going to be configured exactly as the old one was, either, just for those reasons. The key here is smooth, stable firing, with flame in all combustors without a spike in exhaust temperature, a warm-up with minimal exhaust temperature spreads, smooth, stable acceleration with minimal exhaust temperature spreads and minimal exhaust temperature spiking, and therefore minimal smoking.

Again, as has been said before on control.com, the liquid fuel systems of GE-design heavy duty gas turbines have a lot of components that are not controlled or even monitored by the Speedtronic. Fuel supply pressure regulators; liquid fuel check valves; starting atomizing air pressure; individual fuel pressures at each combustor/fuel nozzle; liquid fuel check valve operation; purge air check valve operation; fuel nozzle integrity; all of these are presumed to be working properly in order for the Speedtronic to be able to control fuel to provide the smoothest starting and acceleration with minimal exhaust temperature spreads and spiking and minimal smoking during starting and acceleration. Sometimes that's easily achievable, sometimes it's not. But it is achievable, and it's sometimes more difficult to achieve on older machines.

It's likely that the field engineer at the site was never able to witness a start-up of the unit on liquid fuel with the Mark IV. So, the amount of smoking is subjective. And, it's even more likely that the amount of smoke varied with each start with the Mark IV, as it likely does with the Mark VIe--until tuning can be done to achieve the optimal results <b>for that particular machine</b>.

What usually happens is that people tire of the testing and tuning, and so it stops. But, the complaints about the smoking never stop. And the refrain, "It worked just fine with the Mark IV," also never stops. And that is a subjective observation, usually made with respect to a single condition without data to support a complete comparison and analysis.

 

Firstly, Thanks again. Your input has been invaluable.
We sorted the issues with the white smoke thanks for your help.
The following variables were decreased and increased by the field engineer in an attempt to 'tune' the GT startup.

FSKSU_FI - Firing FSR%
FSKSU_WU - Warm Up FSR%

There is still a question as to whether there is some threshold/limits that should be observed while this is being performed? FSR at warm up was changed between the range of 12% to 19% by the field engineer.

My main question is, what are the short- and long-term effects of having too much or too little fuel for Firing and warm up and how does the field engineer actually decide where to leave these values?
The machine ignites at 12% FSR and it will also ignites at 16% FSR etc. How do they decide?

Also how will these be affected by ambient temperature conditions (we have swings from +40degC to 0degC)

A second field engineer came to our site and identified the true cause for the white smoke was that the Acceleration FSR was coming in early. The warm-up period only lasting about 20seconds. This was cutting the fuel back to FSR minimum and causing flame flickering.
When the logic was corrected (and the warm up period lasted the full 60 seconds) the smoke decreased drastically.

The only problem still existing is the speed pick-ups. The gaps were increased while the machine was on cranking and we thought we had solved all our issues with the liquid fuel.

Then we ran the machine on gas again and the Vibration increased and the pickups spike with false readings. We agree it is a nuisance alarm only and have tried to lock the Alarm out to prevent it filling the history overnight.

The problem is as soon as anyone (operators) perform a diagnostic Alarm reset the alarms unlock again and become active. The gaps for the pick-ups are close to 0.050inches.

The belief is the Flow divider vibration is not excessive and our Fitters have already replaced it once and do not wish to do this again without good cause.

However we are still currently investigating to try and reduce the vibration. It has just never been seen to be an issue with the lower sensitivity of the previous system.

Is it possible to wire the pick-ups differently or install a newer type of speed pickup? Could we put a band filter (RC) on the hardware input?

Is there anyway to disable the diagnostic alarm while the unit is not actively running on liquid fuel? (without physically wiring in a relay to open the speed pick-up circuits while gas fuel is selected)

Your thoughts are greatly appreciated.

 

Good news on the smoking issue.

As for adjusting the constants, it's part and parcel of field engineering. Identical cars with carburetors will not always have the same jetting adjustments to run smoothly, and without an exhaust gas analyzer it's impossible to adjust any two identically or even ideally; smooth is subjective. So, this is not unusual. Especially for liquid fuel. Proper atomization of liquid fuel during starting is very important, and lots of factors affect that.

Too little fuel will result in white smoke and even flame-outs. Too much fuel will result in high exhaust temperatures and is not good for the hot gas path parts, especially during starting when the air flows through the machine are low. So those are the concerns: Limit smoking and exhaust temperature spreads and flickering flame indications and flame-outs (too little fuel) and prevent excessively high exhaust temperatures during firing (too much fuel). This is kind of subjective, I realize, but so is adjusting a carburetor without an exhaust gas analyzer.

With regards to ambient temperature changes through the year, the Speedtronic panel uses temperature-corrected speed (TNHCOR) when calculating fuel flows for start-up and shutdown (or at least it did for many years) to try to protect against over- and under firing and the problems associated with that (namely insufficient fuel in colder ambients that might result in flickering flame indications, smoking, exhaust spreads, and loss of flame trips during starting, warm-up and acceleration).

As for the flow divider vibration, I differ on your assessment. The vibrations being "sensed" (and I would even say magnified) are excessive, because it has been eliminated on other jobs by increasing the speed pick-up gaps without adding additional components (RC filters). Now, the other jobs may not have had the exact same mounting method/location, but you have proved by using external cables and by disconnecting the wires at the JB closest to the flow divider that the noisy feedack signals go away that it's mechanical vibrations. Increasing the gap has decreased the noise, so that's another indication that the vibration is excessive.

We don't know what the width of the toothed wheel is, nor the number of teeth nor the shape of the teeth. I have no sympathy for Pipefighters, so let's not even go there. (After operators, they are the next biggest source of human frustration on a turbine site.) It's unfortunate that the unit at your site is not "responding" to the normal fixes which have been used at other sites.

It's my personal opinion that the speed pick-ups being used are too "big" for the toothed wheels in use on most flow dividers. So, changing to a pick-up that is only 1/2-inch in diameter or even 7/16- or 3/8-inch would probably help with the problem. All that would be required would be to install a bushing (most Pipefighters can do that without too much direction or complaining) in the existing location and then thread the new, smaller diameter pick-up into the bushing. I'm sure the speed pick-up vendor in your area could provide you with a comparable pick-up to what's currently installed with a smaller diameter.

Using a relay to switch the flow divider pick-ups in and out is risky, very risky, in my opinion. If you configured a relay that was energized to connect the pick-ups to the Mark VIe when running on liquid fuel, then if something happened to the relay when the unit was operating on liquid fuel and the feedback was lost because the circuit opened then the Speedtronic would just close the liq fuel bypass valve to try to increase the flow feedback--which would not be a good thing. Not a good thing at all. (Can you say "exhaust overtemperature trip"?)

If you used a relay that connected the speed pick-ups to the panel when it was de-energized and the unit was running on liquid fuel, and energized when the unit was running on gas fuel to disconnect the pick-ups from the panel, that would be safer, but if the relays failed during gas fuel operation then the Diagnostic Alarms would come back. This would be the better choice, in my personal opinion, if you wanted to go this route.

But, I would be looking at getting the flow divider more solidly mounted, at least with some temporary bracing of some kind (remember: too much stress on the speed pick-up conduits is not good either) at least at as proof the problem is vibration or it's not. Use some web straps and clamp the flow divider down (don't put too much tension in the middle of the flow divider!). And possibly changing to a smaller diameter speed pick-up. Before going through the risky operation of adding a relay to isolate the speed pick-ups during gas fuel operation.

You could also ask the GE field service person if the limits of the Diagnostic Alarm are adjustable in ToolboxST. GE seems to be putting some adjustability in the limits of detection for some Diagnostic Alarms; there just doesn't seem to be much logic behind which ones are being changed to allow modification or the order in which it's being done.

Another question you could ask of GE is if the Enhanced Speed Algorithm is enabled, and if so, is it necessary for the liquid fuel flow divider speed feedback application. That might help with the noise problem, that's at least partially caused by the increased speed pick-up sensitivity as well as the unusual amount of vibrations being transmitted to the shaft of the liq fuel flow divider.

I would ask, "Why is it necessary to perform Diagnostic Alarm resets during normal operation (while running or when shutdown)?" Because the panel should be Diagnostic Alarm-free, Diagnostic Alarms being an indication of some kind of panel issue (hardware or software). If the operators are just resetting Diagnostic Alarms for no other reason than they think they should do so before ever start or after every shutdown, then that's poor operating procedure. If you are experiencing nuisance Diagnostic Alarms, the GE field service person should work with you to eliminate them to the extent possible.

 

Dear CSA and Sanderson,

We had the same issue after retrofit with MarkVie. I read your article. I solved the issue by providing rubber packing underneath flow divider legs while unit running on gas. Now the fql_pr1 signal level is almost zero and no more diagnostic alarms.
Our flow divider pick remains same as before at 0.02".
Thank you guys.

> Good news on the smoking issue.
>
> As for adjusting the constants, it's part and parcel of field engineering.

---- snip ----