We have two identical gas turbines for power generation in our captive power plant for our fertilizer complex.

Both the machines are GE frame 5 packaged generators (PG5361) manufactured by Hitachi. Both are running on MArK VI control system.

Let's call them GT#1 and GT#2.

GT#1 is not able to attain 100% rated RPM (5105) when run in Isochronous mode. While it runs at 100% rated RPM when run in Droop mode and synchronized with the greed. Due to this the frequency of the generated current by the generator remains slightly less than 50 HZ (standard in India). There is no such problem in GT#2 (GT#2 is installed 10 years later than GT#1, which is installed in 1985).

How can we rectify the problem? Please advice.

 

Sanjay Solanki,

It's most likely that someone has disturbed the Isochronous speed reference: TNRI. It should be 100.0%, and it can get changed when someone tries to adjust the load on an Isochronous machine by clicking on RAISE- or LOWER SPEED/LOAD. (It's not possible for an operator to control load on an Isoch machine--not by clicking on the RAISE- or LOWER SPEED/LOAD targets.)

So, check TNRI and report back the value. If it's not 100.0%, then when the unit is running in Isochronous mode next time, someone will have to use the RAISE- or LOWER SPEED/LOAD targets to get it as close to 100.0% as possible.

If TNRI is not 100.0%, please write back and let us know and we can look at some other parameters.

In any case, write back to let us know how you fare with resolving the problem.

 

Thanks for the quick response CSA!

This turbine is under shutdown for maintenance now. I've checked and confirmed that TNRI was actually less than 100.0%, and I'll check and report again when we run the turbine.

I've a small doubt, if I increase TNRI this way, will it have any undesirable effect on other parameters?

 

Sanjay,

We like the word 'question' instead of doubt (check you Oxford's and you'll see why), and to answer your <i>question</i>, no; you won't affect any other parameters or operation if you change it. Again, it can only be changed by some operator clicking on RAISE- or LOWER SPEED/LOAD.

You can likely select Isoch now (or force the Isoch permissive for enabling TNRI) and change TNRI with the unit off. Just remember to remove the force when you're done.!.!.!

It would have been nice had you told us how much the frequency was low when in Isoch and what the value of TNRI is.... Enquiring minds want to know.

Please write back--with the requested information, and then when the unit next runs in Isoch to let us know if all your <i>questions</i> have been answered.

 

The frequency was 49.8~9 Hz. I'm not able to tell you the value of TNRI (as I'm not allowed to open toolbox. I've to ask our instrumentation engineers and they haven't noted the value.)

A field support engineer from GE recommended changing GCV servo and we did that. Today we checked the operation of this machine on Iso and the frequency was 50 Hz (TNRI is also 100% now).

I've one more QUESTION. As per my reading of GE manual, FSR should increase exponentially in acceleration mode (14 HA was displayed on the startup screen) but it was increasing very slowly when we started the machine today. Please point out possible causes and for this.

PS: I checked Oxford's

 

Thanks for the quick reply again!

GE field support engineer recommended changing GCV servo, we changed it. Now the turbine is running at rated RPM in Iso mode.

Frequency was 49.8~9 Hz before this. I'm not able to tell you the value of TNRI (as I'm not allowed to access toolbox, I've to ask our instrumentation engineers for it and they've not noted the value).

I've one more QUESTION!
Today we tested the GT on Iso mode. As per my reading of GE manual, FSRACC should increase exponentially when 14HA senses suitable speed in the start up sequence. But FSRACC was increasing very slowly. Please point out possible causes.

PS: I checked Oxford's.

 

Sanjay,

I won't even comment on the recommendation to change the servo-valve--except to say that I wish I owned Moog. You just replaced a perfectly good servo-valve, which just increased Moog's profit. (And if the servo-valve was purchased from GE, then you also increased GE's profits--and for that I thank you as an owner of a lot of GE shares of stock.) However, if the servo-valve was provided under warranty from GE (in other words, at GE's expense), then I'm not very happy because there went some of my dividends.

As for your reading of the manual, if you could copy the section that leads you to believe that FSRACC increase exponentially after 14HA, it would be helpful in answering your question.

14HA is an "intermediate" speed level that indicates the unit has reached reached a speed that is usually around 60% (the value of TNK14HA1) during starting and acceleration to FSNL (Full Speed-No Load). For some machines, 14HA is said to be "self-sustaining" speed--meaning that the unit doesn't require any further assist from the starting means to continue accelerating to FSNL.

FSRACC doesn't have anything to do with 14HA. FSRACC is a fuel stroke reference value that is intended to be just slightly higher than FSRN and "track" FSRN up and down when operating on speed control (specifically, Droop speed control). FSRACC has two basic functions; first, on newer Speedtronic turbine control panels it is used to control the fuel during acceleration to maintain an acceleration rate. TNHA (the actual acceleration rate) is compared to the acceleration rate reference (TNHAR) and fuel is adjusted to make TNHA track TNHAR. (I have to note here that early digital GE Speedtronic turbine control systems didn't use FSRACC in this way, and even some Mark Vs were very poorly programmed and FSRACC for acceleration control didn't work very well--if at all.)

Second, when the turbine is at rated speed FSRACC still "tracks" FSRN and is just slightly greater than FSRN. If something were to suddenly cause TNHA to increase then FSRACC would act to limit FSR to try to prevent an overspeed condition. So, if FSRACC were to be increased "exponentially" (I'm presuming to 100%) the over-acceleration rate protection would be lost.

It's extremely rare to see FSRACC be the controlling value of FSR during normal operation at rated speed--except on machines where the grid frequency is oscillating wildly.

FSRSU is usually increased to 100% after the unit reaches rated speed, to keep it from adversely affecting unit operation (since start-up is complete and the unit has reached rated speed).

Hope this helps!

 

Thanks CSA for your valuable reply!

I put my question the wrong way. Here I'm copying the text from GE manual,

"At the completion of warm up period, startup control increases FSR exponentially to ACCEL setting for "ACCELERATION LIMIT". As fuel is increased, the turbine begins the acceleration phase of startup. When the turbine overrunsthe starting device, the clutch will be thrown out. Speed relay 14HA indicates the turbine is accelerating."

And regarding the servo valve replacement, I've started working on a GT just 2 months before and I'm under training. No one will listen if I tell them not to change the servo, but I agree with you that we changed a perfectly working servo (and increased MOOG's profit).

 

Sanjay,
It's very difficult to understand how reasoning, logical people can decide that if the turbine runs fine in Droop speed control that because it doesn't maintain desired frequency in Isochronous speed control that the problem could be the GCV servo-valve--because that same GCV servo-valve is controlling the same GCV regardless of the governor mode.

The GCV servo-valve get's it signal from the Speedtronic, and if the reference for the Speedtronic isn't correct then the GCV servo-valve isn't going to control to the desired parameter.

If the GCV is stable when operating in Droop speed control mode, and it's also stable when operating in Isochronous speed control mode, why would someone think that changing the servo-valve would make the turbine speed change in Isoch mode versus Droop mode?

Anyway, I can sympathize with your dilemma. There is too much tribal knowledge and too many myths and falsehoods surrounding servo-valves, and even GE field service personnel (one can hardly call them engineers when they can't reason through a situation like this and come to a more logical conclusion) believe a lot of the misinformation.

I think what you're referring to is a "sub-FSR" value called FSRSU_IA--but that's just a guess, and probably not a very good one, either; sorry. I really don't understand exactly what the quoted paragraph is trying to say; it's very nebulous.

I usually caution everyone reading GE written documentation to carefully read everything--as you have--and think it through and question it. But, the ONLY document that really describes exactly how any turbine operates is the application code or sequencing running in the Speedtronic panel--because that's what the Speedtronic is using to control and protect the turbine, not what's written in some document in some manual, which is really just someone's attempt to describe the INTENT of what's happening, not exactly what's happening.

 

CSA,

You explanations are always enlightening,always!

But one problem occurs to me:
As both of IGVs and FSR (right?) have ability to regulate exhaust temperature,how to make sure when the IGVs are not at maximum operating angle,the exhaust temperature is controlled by IGVs exhaust temperature control not CPD-biased exhaust temperature control?

Best regards
Neo

 

Thanks for continuing the conversation CSA.

I personally believed that there is some problem somewhere in the Speetronic system with some logic or variable (as you pointed out TNRI) that is why I started the thread here.

We can close this thread now. But I've a request, I want to learn everything about gas turbines (I may be too much ambitious) so that I can solve problems like this and others in future.

I've studied and learned the P&IDs for our machine. I've seen most of the equipments in the field too. I can show you all the devices from which the Speedtronic system takes its input. I know the locations of all the pressure and temperature switches, transmitters, vibration probes, thermocouples, etc and I know their working principles.

As stated above I want to learn everything (and I know the field is very vast), please suggest me how should I proceed for efficient learning. Should I start a new thread or can we continue here?

I'll be very much thankful to you.

 

Neo,

I'm giving here the link to a thread in which CSA has very well explained the concept of exhaust temperature control. Have a look at it.

http://www.control.com/thread/1403944262

Hope this helps!

 

Neo,

Wrong thread, perhaps?

 

Sanjay,
You are well on your way to learning how GE-design heavy duty gas turbines operate, and how to troubleshoot them if you've already started studying the piping schematics (P&IDs) and gone out and located all of the control system field devices! Good on you, mate!!! I just can't seem to get people to understand just how critically important the P&IDs are, and how important it is to know where the devices on the P&IDs are in the field.

I think you should be asking your questions in individual threads, rather than piggy-backing on this one. Try to choose a good Subject for the thread, and we'll try to do our best to answer your questions.

It will help us if you will tell us a little about the machine at your site. What kind of combustors (conventional (diffusion flame), or DLN) does the machine have? What fuel(s) does the machine burn? How old is the machine? How is the machine operated (you've already told us that the two machines can be operated in Isoch mode, but is your plant always separated from the grid (which might be a blessing in some parts of the world!)? Do the machines exhaust into a boiler--and does the configuration have a bypass stack?

Anyway, ask away!

 

Thank you very much CSA!

I'm away for a week on mini vacation. I'll ask my questions in separate threads as and when I feel the need to ask. I'll also write my plan on how I'm going to divide my learning of GT in various sections and subsections. It'll be of great help for me if you spare some time to give your valuable suggestions then.

I'll give all the details you asked for in my next thread.

see you soon..

 

CSA,
Hello! (after a long time though)

> It's most likely that someone has disturbed the Isochronous
> speed reference: TNRI. It should be 100.0%

Today we had to run Unit 1 on ISO due to Unit 2 tripping. Unit 1 is not operating at 100% rated RPM. As per our previous discussion you suggested that TNRI should be 100% for the turbine to run at 100% rated speed.I checked TNRI when the unit was running on ISO and it was 100%. Still the unit is not running at the full rated speed.

Please give your valuable suggestions.

P.S. This is an old thread so let me remind you that Unit 1 and 2 are identical GE Frame 5 Gas Turbines without DLN combustors. I think I'll have to provide more info, please let me know.

 

SB,

When you write to tell that the unit isn't running at 100% rated RPM, <b>please do tell what RPM the unit IS running at when it's NOT running at 100% RPM.</b> It would also be helpful if you told the load the unit was running at when it wasn't running at 100% RPM. And what the value of TNH was when it wasn't running at 100% RPM.

There's a huge gap in knowledge about 100% RPM for GE-design heavy duty gas turbines which use reduction gears to drive synchronous generators. 5105 RPM is the design speed for the gas turbine axial compressor. Whether it runs at 5094 or 5105 or 5129 o4 5134 RPM doesn't matter much to the efficiency of the gas turbine--those RPMs are just slightly off the 5105 RPM design value and, again, they really don't affect efficiency or output very much. The same is true for the turbine section--whether it runs at 5094 or 5105 or 5129 or 5134 RPM doesn't affect efficiency or output very much.

The people who take a chunk of metal and cut it into a toothed gear assembly have the ability to analyze the grain structure of that chunk of metal and make determinations about the width of the teeth to cut to maximize the strength of the gears. Some need to be a little wider; some need to be a little narrower. Knowing that the gas turbine and axial compressor don't need to run exactly at 5105 RPM and can run perfectly acceptably in a small range around 5105 RPM they then choose the best tooth width which results in a particular gear ratio which may make the gas turbine and axial compressor run at 5094 0r 5105 or 5129 or 5134 RPM. <b>The reduction (load) gear box nameplate defines what is 100% RPM--not the Control Specification, the reduction gear box nameplate.</b> Full stop. Period. End of discussion.

The reduction gear box nameplate defines the gas turbine (and axial compressor) speed which corresponds to the synchronous speed of the generator. If the generator is a two-pole, 50 Hz machine, the output of the gear box needs to be 3000 RPM. And the nameplate says what the input RPM will be for the necessary output RPM. Not the Control Specification--the Control Specification will almost always cite 100% speed as 5105--because the Control Specification is related to the gas turbine and axial compressor, and we (now) know that the axial compressor (and gas turbine) don't need to run at exactly 5105 RPM, but can run just fine at 5094, or 5105, or 5129 or 5134 RPM.

I have even seen two Frame 5s, installed at the same time, sitting side-by-side, commissioned simultaneously, have two different 100% RPM settings in the Speedtronic control panel. Why? Because two different field service persons commissioned the units, and in GE, a commissioning person gets to do whatever he wants to do and his way is the right way and anyone else's way is the wrong way. So, one person decided the value should be 5105, and the other person looked at the reduction gear box nameplate and put the proper value in the Speedtronic for 100% RPM--the value of turbine/compressor speed that corresponded to 3600 RPM (that site was in a country that had a 60 HZ system).

So, it would be very helpful if we also knew what frequency the generator output was at when the turbine wasn't running at 100% RPM. <b>AND</b> you need to locate the reduction gear box nameplates for BOTH machines, take pictures of both of them, and then write back with the reduction gear speeds (input/output) for both machines.</b>

About the only other thing I can suggest which might be the problem is that there is a small deadband for Isochronous speed control, +/- 0.13%, if I recall correctly. I don't recall the exact name of the Control Constant value (I think it's TNKRn, where 'n' is some value between 0 and 7), but it's possible that it's slightly different on one machine than the other.

Remember--the 100% speed value of the turbine (and axial compressor) is 100.00000% dependent on the reduction gear box nameplate value--nothing else. If the reduction gear box requires 5094 RPM to produce 3000 RPM to drive the generator at 50.0 Hz, then the 100% speed setpoint for the gas turbine (and axial compressor) driving that two-pole, 50 HZ generator is 5094 RPM--and that's the value that should be in the Speedtronic for 100% speed. The Speedtronic is going to convert the frequency from the turbine (and axial compressor) shaft speed pick-ups to percent based on the value in the 100% speed setpoint field.

Looking forward to the answers to all the requests for information in order to be able to provide more information.

 

CSA,

Thank you for the quick reply.

>please do tell what RPM the unit IS running at when it's NOT running >at 100% RPM. It would also be helpful if you told the load the unit >was running at when it wasn't running at 100% RPM. And what the value >of TNH was when it wasn't running at 100% RPM.

The Unit was running at 5085-5089 RPM. The load was around 15 MW. TNH was 99%

>what frequency the generator output was at when the turbine wasn't >running at 100% RPM. AND you need to locate the reduction
>gear box nameplates for BOTH machines, take pictures of both
>of them, and then write back with the reduction gear speeds
>(input/output) for both machines.

Frequency was 49.8-9. The reduction gearbox nameplates for Unit 1 and 2 are identical. Input speed=5106 RPM and Output speed=3000 RPM

>About the only other thing I can suggest which might be the
>problem is that there is a small deadband for Isochronous
>speed control, +/- 0.13%

I'll have to ask the instrumentation personnel to open Toolbox and check this. I'll report back.

The unit is now running on Droop mode and is perfectly following the ISO machine and is running at 100% speed.

Please let me know if I need to provide more information.
Thank you for your time.

 

SB,

When the unit that doesn't have speed issues when running in Isoch mode is in Isoch mode, what--exactly--is the speed, TNH, TNRI and frequency? Please provide data with at least one, preferably two, decimal places.

We're splitting hairs here, which means the accuracy of the running conditions seems to be more critical than it need be. 49.9 Hz is better than a lot of grids even on a good day. And we don't know what level of accuracy you're applying to the second unit--except to say it runs at "100%", which may be 99.96 or 100.05.

Another thing I've seen is that some HMI displays are configured to display more decimal places than others.

AND, if you could provide the Isoch deadband values for both units that may be informative. (Again I can't recall the exact name of the Control Constant, but I think it's one of the TNKRn array values. The Constant should have a value of something like 0.13%.)

The "problem" may not be in the unit that doesn't run as well as perceived but in the other unit that seemingly runs better. I believe the "problem" is a mismatch in some configuration value--but it is frustrating not being able to get information in a more timely manner.

Please write back with the requested information.

 

CSA,

Thank you for your reply.

>When the unit that doesn't have speed issues when running in
>Isoch mode is in Isoch mode, what--exactly--is the speed,
>TNH, TNRI and frequency? Please provide data with at least
>one, preferably two, decimal places.

TNH = 100.0 Frequency = 50.0 Hz. I'm not able to get the value of TNRI as of now.

>We're splitting hairs here, which means the accuracy of the
>running conditions seems to be more critical than it need
>be. 49.9 Hz is better than a lot of grids even on a good
>day. And we don't know what level of accuracy you're
>applying to the second unit--except to say it runs at
>"100%", which may be 99.96 or 100.05.

Actually slightly lower frequency is not a problem for us, but I want to know why it is so. And as you may remember I'm asking questions to gain some knowledge. I'm not assigned the task of solving this problem, even people don't want me to try to understand this problem.

>Another thing I've seen is that some HMI displays are
>configured to display more decimal places than others.

Our HMI displays only one decimal place.

>AND, if you could provide the Isoch deadband values for both
>units that may be informative. (Again I can't recall the
>exact name of the Control Constant, but I think it's one of
>the TNKRn array values. The Constant should have a value of
>something like 0.13%.)

I'll try to provide the data asap.

>The "problem" may not be in the unit that doesn't run as
>well as perceived but in the other unit that seemingly runs
>better. I believe the "problem" is a mismatch in some
>configuration value--but it is frustrating not being able to
>get information in a more timely manner.

I know that I'm not able to provide the required data in timely manner, but I can not help that. Sorry for that. Thanks for being patient. If you wish to discontinue replying to this thread I'll not insist you much to do so. I wish you'll provide your guidance and experience in my learning.