We have carried out a major inspection on the PGT10 gas turbine. After startup, we observed that the second-stage nozzle does not close. We have stroked the nozzle from 0 to 100% back and forth several times, and everything appears to be functioning correctly. The turbine is currently operating under temperature control. Is this behavior considered normal under these conditions?

 

Under temperature control, it is normal behavior for the second-stage nozzle not to fully close if the control logic determines that such a position is not required.

 

The questions should be:

At what speed is the HP shaft? (RPM and percent of rated speed)

At what speed is the LP shaft? (RPM and percent of rated speed)

What is the gas turbine exhaust temperature?

What is the valve opening of the SRV (Stop-Ratio Valve)?

What is the valve opening of the GCV (Gas Control Valve)? (Assuming the machine runs on natural gas or LPG--some gaseous fossil fuel).

Do you know if the LP shaft is running slowly because the load on the driven device (centrifugal compressor or pump) is very high? (Presuming the turbine drives a centrifugal compressor or pump)

IN GENERAL, GE turbine control philosophy uses the second-stage nozzles to control HP speed. LP speed is primarily controlled by fuel (yes there is some interaction between HP and LP speed when fuel flow changes) but that's the GE controls philosophy. Usually, the second-stage nozzles are open during starting and acceleration of both the HP and the LP until the HP reaches around 80% (varies with machine and frame size) and then the second-stage nozzles start modulating to limit HP speed. At this point the fuel flow-rate controls LP speed, as it should be at or very near the normal operating range. So, if the HP shaft is not near normal minimum operating speed range and/or the LP shaft is not at or near rated operating range and the exhaust temperature is higher than normal causing the machine to be on exhaust temperature control then there is something wrong. Could even be the compressor discharge pressure transmitter is not working properly and causing problems with the exhaust temperature reference value.

It could even be that the axial compressor is dirty (limiting air flow through the machine), and/or the turbine inlet air filters are dirty (also limiting air flow through the machine), or if the machine exhausts into an HRSG (a "boiler") that the exhaust duct back pressure is excessively high (also limiting air flow through the machine).

But, we don't have enough information to be of very much help based on the information provided.

 

The questions should be:

At what speed is the HP shaft? (RPM and percent of rated speed)

At what speed is the LP shaft? (RPM and percent of rated speed)

What is the gas turbine exhaust temperature?

What is the valve opening of the SRV (Stop-Ratio Valve)?

What is the valve opening of the GCV (Gas Control Valve)? (Assuming the machine runs on natural gas or LPG--some gaseous fossil fuel).

Do you know if the LP shaft is running slowly because the load on the driven device (centrifugal compressor or pump) is very high? (Presuming the turbine drives a centrifugal compressor or pump)

IN GENERAL, GE turbine control philosophy uses the second-stage nozzles to control HP speed. LP speed is primarily controlled by fuel (yes there is some interaction between HP and LP speed when fuel flow changes) but that's the GE controls philosophy. Usually, the second-stage nozzles are open during starting and acceleration of both the HP and the LP until the HP reaches around 80% (varies with machine and frame size) and then the second-stage nozzles start modulating to limit HP speed. At this point the fuel flow-rate controls LP speed, as it should be at or very near the normal operating range. So, if the HP shaft is not near normal minimum operating speed range and/or the LP shaft is not at or near rated operating range and the exhaust temperature is higher than normal causing the machine to be on exhaust temperature control then there is something wrong. Could even be the compressor discharge pressure transmitter is not working properly and causing problems with the exhaust temperature reference value.

It could even be that the axial compressor is dirty (limiting air flow through the machine), and/or the turbine inlet air filters are dirty (also limiting air flow through the machine), or if the machine exhausts into an HRSG (a "boiler") that the exhaust duct back pressure is excessively high (also limiting air flow through the machine).

But, we don't have enough information to be of very much help based on the information provided.

BTF thank you for your prompt response
The running data are:
TNH 100%
TNL 86%
TNR 87%
AMBIANT TEMP 36C
CPD 11.5 Barg
SRV 65%
GCV 57%
TTXM 512C
It's true that the driven load which is centrifugal compressor needs more energy to be delivered by the turbine that's why thise machine is in control temperature all the time winter or summer, but I kow this machine sic 20 years, and this is the second time happen this phenomena, after start up and the machine reach the minimum operating load., the the operator start loading the machine in the meantime closing the anti surge valve of the centrifugal compressor, usually the second stage nozzle start closing when TNH reach 100% , I can' remember the reading of the Lp speed but it's near 80%, than normally the second sage nozzle continous to close, beacause the fuel demand to respond to the load should increase too, the nozzle keep TNH speed over100% and the excess energy gone to the LP, Reducing the exhaust temperature, than the turbine can be loaded more, so we get far from the temp control cuve,
But this does not happen after the ast MI, Even mechanically every parts is new, axial comp, transition peace, 1st and second stage nozzle LP wheel buckets.
The bleed valve is closed, the CPD is fine, what else....
I hve a many images saved that shows nozzle in half cours i mean between 22 degree ad _8, Usually at 8 degree.
I want to know whene should the second stage nozzle stzrt moving what the algorithm to control it

 

@belmehda,

We can't tell you how to find what is the "logic"/algorithm driving the second-stage nozzle servo output because we don't know what control system is being used for the turbine. It seems to be a digital Mark* turbine control system (from the signal names) but sometimes when a third-party control system is used to replace an older Mark* turbine control system the programmers use the same signal names GE used in the original control scheme.

Has this happened more than once after the most recent MI (Major Inspection) during starting and loading of the machine?

If you have a GE Mark* digital turbine control system (Mark* IV or Mark* V or Mark* VI or Mark* VIe) then you should be able to find the signal name that is the second stage nozzle servo-valve output and work backwards in the logic/sequencing or application code to find the information you seek. (Actually, this should be true for just about any control system--find the output that drives the second stage nozzle servo-valve and work backwards in the logic/sequencing/programming to determine what makes the signal change value to affect a movement of the second stage nozzle ring.) In a GE Mark* turbine control system there is usually a second stage nozzle reference position that is calculated and that signal eventually becomes the second stage nozzle servo output current. I would imagine the second stage nozzle reference position would have DGA (DeGrees Angle) engineering units, but it might just be percent, also.

What is the reference value for the second-stage nozzle position when this problem is occurring? (In a GE turbine control system there will be a second stage nozzle position reference signal and that gets converted into a servo current output signal, usually something between approximately 8.0 and -8.0 (sometimes the engineering units are percent, and sometimes it's mA), and when the second stage nozzles are at the reference position (actual position equals reference position) the servo current value should be around -2.67 or -4.00 mA, approximately, for steady-state position (when the actual position equals the reference position).

What is the value of the servo output current signal to the second stage nozzle servo-valve when it is not moving? Could it be something that is causing the nozzle ring/actuator to "stick" or "bind up" when the machine gets warm? (You say you can move the second stage nozzles through the full range of motion when the machine is not running, but if something was reassembled improperly and gets bound up or sticks when the machine gets warm that could be causing what's happening. In this case the servo output would be putting out a lot of positive servo current to try to close the second stage nozzles if it wasn't able to move. (Positive servo current usually closes the servo-operated device (gas valve; IGVs; second stage nozzles; and negative servo current usually opens the device. If the magnitude of the servo current value(s) are very high (around -8.0 or +8.0) that would typically mean that the second stage nozzles have not moved to the reference value--meaning they might be "stuck" or "bound up" and the hydraulic actuator doesn't have enough force to make the nozzle ring move.

So, that's about all I can add at this point based on the information provided. I know--it seems like I ask a lot of questions, but the intent of the questions is not to confuse anyone it's to get the reader to think about what might be causing the problem, possibly something they haven't considered before or didn't investigate properly if at all. Digital control systems are usually pretty stable and don't have problems with "drifting" over time as components (resistors; capacitor; etc.) age and begin to fail. If this problem started after an MI (Major Inspection) then it would seem it's one of two things: either the mechanical reassembly wasn't done correctly, OR, the polarity of the servo current(s) being applied to the second stage nozzle servo valve coil(s) was not properly verified and is incorrect. We haven't talked about servo current polarity verification, but if the second stage servo-valve was replaced then it could be the colored coil leads aren't the same as the ones on the servo-valve that was replaced (it happens A LOT, actually--even with new Moog servos and more often with refurbished servo-valves from third-party servo service/repair facilities). The good news is: It's an "easy" problem to solve, because it just involves swapping two wires for coils that are not receiving the proper polarity. Verifying the polarity is not so easy (actually, it is--but it's usually more work than most people want to do when changing a servo valve), especially if one has never done it before. [CAUTION: The procedure in the Mark* Control Specification for verifying servo current polarity is usually WRONG. But, the process has been covered MANY times before on Control.com; use the Search feature to find threads which discuss the proper procedure.]

Remember: We are not there standing or sitting next to you. We can't see what you can see; we don't know what you know. Help us to help you because we are "blind" unless you tell us.

 

Thank you BTF for your prompt replay
The system in use is a GE Mark VIe turbine control system.
Yes, this issue has occurred previously after a major inspection. During that time, the root cause was mechanical—the first stage nozzle installed was a new version, but its support structure was from an older version. This mismatch led to changes in the thermodynamic behavior of the machine during startup and loading, affecting its performance.
The closed loop between the second stage nozzle servo-valve output and the LVDT feedback is functioning properly. When the TNH exceeds 100%, the nozzle starts closing as expected. This confirms that the feedback and servo-valve output signals are operating correctly within the control loop.
The machine starts and completes the full startup sequence without any problems. It is loaded until TNR equals TNL, and during operation, the machine runs very smoothly with all parameters within normal limits, except for the nozzle position. In my opinion, temperature control should not activate before the nozzle reaches 50% of its travel range. The nozzle ring is new (installed during this MI along with the new LP turbine stator and second stage nozzle). Simulations of the nozzle movement do not indicate any sticking or binding issues.
I agree that digital control systems are generally quite stable and reliable, and they typically do not cause such issues.
We did not touch the Moog servo or the LVDT connections individually. The entire assembly (Moog piston and LVDT) is always removed and reinstalled as a complete unit to avoid any wiring mismatches.

I would like to get your opinion regarding the behavior observed during the machine loading process.
Is it normal for the machine to enter temperature control mode while the nozzles do not move even a single degree?

 

@belmehda,

Even if the servo currents are not all correct (to each coil of the servo-valve) the second stage nozzles will still move. There will be a larger than normal difference between the second stage nozzle reference position (which you still haven't told us what it is when this problem is occurring) and the actual second stage nozzle position.

As @bidrohini writes, if the turbine control system determines (for whatever reason--true or false (meaning the operational data may be correct or it may not be incorrect, such as a large exhaust temperature spread or other issues with mismatching or erroneous inputs) the second stage nozzles should not be moving/opening then the command to move the second stage nozzles won't change. What we still don't know is if the command to the second stage nozzles is changing during loading or not. And, the calculated reference position could be the result of erroneous or incorrect inputs or even calculation--most of which would be made known through alarms or be conscious and experienced operators taking note of unusual values/data.

I'm not directly familiar with PGT10 machines, but I have seen my share of two-shaft Frame 5 machines driving centrifugal compressors and GE is know for scaling machines up or down when designing and building larger or smaller machines and using much of the same control philosophies and schemes in the new machines. (That's changing with new technology machines and high technology parts and the use of practices which were only previously used in aircraft turbine engines (such as variable axial compressor stator vanes).) Re-read your posts; I think you will find some of your statements to be conflicting about how the machine is currently operating. I also wonder about your statement that TNH rises above 100% and how that is affecting machine operation, because to me that says that TNL is not what it should be for the given operating conditions, perhaps because the centrifugal compressor is overloaded or something like that.

Digital control systems are pretty stable and the operating parameters and limits don't drift over time like they did when the circuits were analog and used many potentiometers to set parameters and limits.

GE Mark* VIe turbine control systems also use something very commonly ignored in the industry--Diagnostic Alarms. Diagnostic Alarms are meant to alert a conscious operator, operations supervisor or control technician to a problem with the "health" of printed circuit cards or input or output channels. MANY people consider Diagnostic Alarms to be nuisance alarms--often because they may be so many of them and the alarm text messages can be so cryptic. Also, there is a predominantly true perception that a Diagnostic Alarm can't trip the machine--and while that is technically true, there are combinations of Diagnostic Alarms that CAN trip the machine. (Before you ask, there may be hundreds of Diagnostic Alarms in a particular Mark* turbine control system, and it's impossible to detail what combinations of Diagnostic Alarms will always result in a turbine trip.) It's usually in GEH-6721, Vol. II, where there is some troubleshooting information for Diagnostic Alarms. I think newer versions of ToolboxST also have the same information more easily accessible than paging through the manual, but it's still there and it can be very helpful at times.

The point of this description is that left unattended Diagnostic Alarms can--and have frequently--resulted in serious problems, up to and including machine trips. They can also lead to low efficiency of operation, low power output and incorrect data/calculations. Contrary to wildly popular belief--even amongst GE technicians--IT IS POSSIBLE to resolve and clear ALL Diagnostic Alarms, but it can take some effort and knowledge. But, leaving Diagnostic Alarms unresolved can lead to bad outcomes--maybe even like this one in this thread.

It may also be that there are Process Alarms which are active which are being left unresolved and unattended (because the machine is still running and hasn't tripped or can't be started). GE doesn't make it very clear that even if the turbine control system will allow a machine to continue operation with some Process Alarms it's NOT recommended and can lead to machine damage and unavailability. GE can't trip the machine, or even send it to a controlled, automatic shutdown, every time the turbine should be stopped and problems investigated and resolved or owners and operators would be complaining very loudly about the machines being unreliable and unavailable. So, they leave some conditions to be investigated and resolved by on-site personnel. Alarms and Alarm Management is a very important aspect of GE-design gas turbine operation and maintenance. As much as it would be great if the turbine control system could use machine learning, or AI (Artificial Intelligence), to determine and alert operators to potentially serious conditions if left unattended, that capability hasn't developed to a usable point yet. So, there are still human operators and operations supervisors and Plant Engineers and Plant Managers to review operations and data and make decisions based on acceptable risk levels.

Anyway, best of luck with your issue. Troubleshooting is often a process of elimination--that sometimes involves making a list of as many possible causes as can be understood, prioritizing which ones are the most likely to be the problem, and then working through the list methodically and logically using proper test AND writing down the results (other than "That didn't fix the problem"). Sometimes the machine has to be shut down. But, it's always a logical process and rigorous testing that yields the best results. In the least time.