Our GE 7FA Gas Turbine often trips on Servo Valve Suicided.

Is there any risk involve if the SERVO SUICIDE FUNCTION is DISABLED?

 

I doubt whether it is a sound idea to disable "Suicide". Servo can suicide due to one or more faults. Call up the diagnostics on the SVSO to identify the problems.

It will be better to consult the vendor, before you decide to disable this feature.

 

Patro,
The picture is becoming a little clearer; but only slightly. Seems you have an F-class Frame 7 at your site (possible more than one, but you're having some problems with at least one, right?)

Yes, there is risk involved any time one tries to bypass a protective function without resolving the condition which is making it active. You need to read the Mark VI System Guide section on servo-valve outputs and understand what is causing the servo-valve suicide. Is it a difference in the LVDT feedbacks? Or the Liquid Fuel Flow Divider feedbacks? You haven't told us which servo-controlled device is experiencing the problem (though it might seem from your other post to an old thread that it's the IGVs).

If there's a problem that's causing the suicide, it's generally a prudent course of action to resolve that issue rather than to disable the suicide. It is not recommended to disable any suicide (or any protective function for that matter) at any time without fully understanding what's causing the function to be active and making sure everyone at the plant (operators, maintenance, and management) understands the risks involved with disabling the function *temporarily* until such time as the underlying condition can be corrected and the suicide re-enabled.

What are the LVDT feedback values being seen by <R>, <S>, and <T>? What are the servo currents being seen by <R>, <S>, and <T>?

It would also be helpful if you would report any and all alarms which are active when this condition occurs, including any Diagnostic Alarms. (I know; *nobody* pays any attention to Diagnostic Alarms as they're (wrongfully considered to be) just nuisance alarms. "Heck; as long as the turbine didn't trip, the Diagnostic Alarms can't really mean anything, right? I mean, *nobody* understands those cryptic messages anyway, right? Why waste our time with them; we should be surfing the Internet to see if anyone can help us with our Process Alarms and trips without giving them the whole story and any meaningful details, right? Never mind that the same Diagnostic Alarms are being annunciated every time the unit trips or just before the unit trips or every time the unit starts; they're not the real problem, right? They couldn't be an indication of the underlying problem, could they? I mean, they couldn't be an indication of the underlying problem, could they? I mean, they're just nuisance alarms, right; as long as the unit continues to run they don't really mean anything, right? Right? Right? Right?" I hear it all the time. From just about every site I've ever been to. From field engineers and Technical Advisors, to E&I and I&C techs, to operators, to plant managers.)

 

Let it be CSA. Many of the problems presented on this forum are people wanting quick/easy answers to big problems.

I have thought of no longer reading the posts; but I believe that REAL help can and will be provided to those that need it and will accept it. Even on this forum, which consists of questions that most of the time reside in manuals easily accessible and not read.

Disable master protective functions/alarms/trips and commit turbine suicide! GE needs the money.

May be my last entry?

 

Not everyone can RTFM (Read The "Fine" Manual) and understand what's written there. It can take a lot of effort to try to associate what's written to something that we are experiencing (actually, it's damn near impossible for the most part!). Also, most of the GE Manuals are less than adequate (anyone care to argue with me??). It's very difficult to understand a Speedtronic turbine control system unless one has some kind of feel for or basic understanding of GE's heavy duty gas turbine control philosophy since the system is built based on that philosophy, which isn't like that used for the majority of other similar (microprocessor-based) control systems. The control philosophy and the control system design go hand in hand. Unfortunately, for GE field engineers and Customers alike, that information (the GE control philosophy) isn't really properly documented anywhere. (Hmmm, makes one wonder about the opportunities....but not for very long!)

Yes; some days it's harder than others to answer questions like what does CPD mean. And it's getting harder and harder to keep asking what Diagnostic Alarms are or were annunciated prior to the turbine trip, because *nobody* understands the importance of Diagnostic Alarms.

And the Droop Speed Control questions; I don't know how they've continued for so long. (Actually I do; without the ability to post graphics and pictures, it's really hard to get a point across sometimes.) At least we're not still trying to explain what a negative slope means when talking about exhaust temperature control (at least it hasn't come up for a while). The patience of some of the former respondents was impressive.

But, everybody learns at a different pace and from a different perspective. It's very gratifying to see when someone "gets it". Myself, I'm slow in the beginning, but look out after one or two of the little pieces fall into place. Yes, reading the manuals (no matter how boring or unsatisfying they were) was difficult; but it was a beginning. For those for whom English is a second language, reading those manuals must be very difficult (I grew up speaking English, and it was still tough to choke that stuff down, and it's still tough to force myself to read the new manuals, which are much better than the old ones!).

So, please don't stop posting. We need all the help we can get here. The number of questions far outstrips the respondents. We need all the perspective we can get. We just need it to be patient, but we should also have a bit of fun here. (I, for one, miss the ribbing markvyguy used to give Texans!)

Anyway, hope to see you around in the future. You were pretty active for a while a few months back, then seemed to drop off for a while. It was good to see you back. We are a community here, and we hate to lose anyone.

 

CSA;

Yes it true we have a multiple F-class GT; and majority of our units experienced a failure in the IGV servos.

The diagnostic alarms are all the same; FAULT CODE 47-51 "Servo Current # disagrees with reference, suicided." Probable cause "a cable/wiring open circuit, or board". All the possible causes were checked by our team, along side other GE TAs, but to no avail. We have had this problem since 2005, that GE cannot resolve until now. They are just offering us an electro-static oil filtration unit (which cost USD100K), suggesting that our oil is contaminated. Then why is it only the IGV servo that are affected, and not the other GCVs?

The idea of disabling the Suicide Function was suggested (unofficially by a GE TA). This is the reason why I posted this question; thinking that this practice was done by the other sites, and not because we are looking for a quick fix.

There was an independent company which analyzed our "defective" servo that was taken out from the site. They have found out that the spool inside the servo is having a restictive slot which prevents it from moving freely. This company suggested a remedy, in which GE is all against it.

So until now, we have a frequent unit trip in between 6~8 months of base load operation.

Thank you all for replying to my post. I hope you can still send me other relevant issues regarding this issue.

To all others who posted their reply; take it easy mate!

patro

 

Agglomerator is the GE technical term for a electrostatic precipitating coalescing filter.

Check the small pencil filters for each servo. Oil contamination can usually be detected in those filters. If detected, it is usually to late for the servo. Replace with new servos (expensive) and always check null bias settings.

Change the large hydraulic filters and small pencil filters often.

We endured months of trouble because SOMEONE decided to RECYCLE flush oil during turbine commissioning and contaminated the oil system causing some kind of varnish. It appears that the varnish only occurs in the newer types of synthetic oils.

I cannot recall all of the details now, but I will address my notes and reply. The IGV and the GSVs have different servo setup issues. I will also address my notes and reply.

We have addressed the problem with positive results and the brand new Agglomerator (40K not 100K) still resides on it's brand new trailer with it's brand new hoses in our storeroom.

 

For CSA:

I have lost my patience in the past and regret some of my responses.

The sudden "off the hip" posts cost me; credibility.

I understand "GE's heavy duty gas turbine control philosophy" because of past and continued experience with the Mark I (yes that is a ONE) turbine controller. The philosopy has not changed and makes the Mark V an VI nothing more than a digital representation of Mark I.

The manuals have not really changed, you have just acclimated. I will quote you; "Yes, reading the manuals (no matter how boring or unsatisfying they were) was difficult; but it was a beginning.

For all GE turbine operator, techs, mechanics.
Learn the manuals for ease of operation.

 

Patro,
Including information like this in your original post would have helped many people understand your point of reference and situation.

Yes, it's true that new Lube Oil formulations are resulting in increased problems for servo-valves (as if they didn't have enough "problems" already what with being the most misunderstood and often-blamed devices on GE-design heavy duty gas turbines). It doesn't help that the oil companies don't seem to be recognizing what their new formulations are now lacking, or that for many people just changing L.O. filters is all they believe is necessary to maintaining L.O. suitability for the application.

I believe the IGVs are something of a unique situation in that during operation the air flow through the unit is trying to open the IGVs and the actuator is always trying to close them. The IGVs are modulated these days on DLN combustor-equipped units to try to control air flow to maintain flame stability and emissions and to keep the compressors operating within their safety limits. The axial compressors are really the limiting factors these days in heavy-duty gas turbine design and output; the designers have reached the limits of current axial compressor design (as evidenced by the new compressor designs of all the turbine manufacturers on their H machines).

If they could get more air flow through the units they could increase output, but compressor design is the limiting factor these days and hot gas path materials and cooling have greatly improved. So, the IGVs, and their "antiquated" actuator and rack and pinion mechanisms (the H-machines have scaled-up aero-derivatve style IGV actuators) are being asked to do a lot more these days than in previous generations of gas turbines when they were either "open" or "closed" or only modulated very minimally. IGV exhaust temperature control is yet another mode of operation that feeds into the selection of several IGV control references during unit operation. The IGVs are really the only "knob" for adjusting some operational parameters and trying to prevent compressor failures and several control functions are all trying to "turn that knob" during unit operation.

It's important to note that the IGV LVDTs are NOT connected to the IGVs, but to the IGV actuator. They're not measuring actual IGV position, but relating actuator position to IGV position. As anyone who has climbed into a compressor bellmouth and twisted an IGV will tell you, the blade will move as much as 1 to 1.5 degrees even though the actuator remains stationary (this is due to "hysteresis" in the rack and pinion mechanism on the IGV shaft).

It seems (and this is a personal observation) that blade movements during unit operation of this magnitude can have an effect on air flow and exhaust temperature and compressor discharge pressure and all of these things can have an effect on the IGV control references which can cause the servo-valve outputs to be constantly changing.

There's also been talk about whether or not servo-valve output "dither" is at least partially to blame for some of the servo-valve "problems" or if it's even required for servo-valve control applications, and the proponents of using dither and the opponents of using dither seem to be about equally divided. Dither is a control function which applies a very slight and high frequency "modulation" to the servo-valve output current, ostensibly to help prevent actuator sticking during extended periods of steady-state operation. It's a configuration setting in the Mark VI, and there's some question about whether or not it was included by default by the Mark V designers (meaning it's always been there without anyone really knowing about it).

Personally, I believe that with the hysteresis in the IGV actuator rack and pinion mechanism and the multitude of control functions each "trying" to control IGV position can cause the servo-valve outputs to fluctuate "excessively" which is contributing to the oil flow through the servo-valves on the IGV actuators which is compounding the problem of L.O. formulation.

Throw some servo-valve output control dither "action" (intentional or not) on top of the whole mix and I think the servo-valves on the IGV actuators are just taking a beating. Again, I have no data or evidence to back up my belief, but with all the things the IGVs are being asked to do these days, particularly on DLN combustor-equipped units (look at all the IGV control references feeding into the IGV output control function) combined with LVDTs mounted on actuator that don't actually measure IGV position which can "flutter" as much as 1 to 1.5 degress even if the actuator is stable and there's a lot of potential for servo-valve operation. Even the IGV actuators have been experiencing a higher-than-desired failure rate of late and at least one of the manufacturers has a new and improved design.

Even $100,000 US dollars can be considered cheap if the production of steam and electricity is critical (though that does seem a little excessive). Sometimes, GE doesn't always do a lot of homework on these issues, and just picks the recommendation of a trusted vendor to use. Do some homework of your own.

How about working with the L.O. supplier/manufacturer to try to understand if there are any additives that can help improve the situation or other formulations that are better? There are trade journals (magazines) devoted entirely to L.O.! For them to exist there must be people willing to advertise their products and services; contact some of the advertisers and do some research on the multitude of machines and methods that people are using to varying degrees (reference CSAs comments about their equipment purchase) to determine if there are ways to improve the L.O. performance. Contact the IGV actuator manufacturers (I thing Young & Franklin in the US is one of them) to see what they're offering and what their experience is and has been.

There's even Moog which might be interested in helping to work on this issue; their servo-valves certainly have been taking the brunt of the criticisms and blame of late, which seem to be escalating in recent years (since the change in L.O. formulations and the increase in IGV control schemes). Hmmmmm, kind of makes one wonder, doesn't it????? Add an antiquated actuator mechanism (the ring and pinion portion, specifically) and some control dither (whether or not its present, or required; another huge discussion all by itself!), and that's an interesting mix; a *very* interesting mix.

It would be very interesting to see a team of people from various various components and disciplines work together to understand this issue and all its permutations. It will never happen, but it might have had some very interesting conclusions and recommendations and even design changes on the part of several of the groups involved.