Overview of Gas Turbine Control philosophy


Thread Starter



I am enganged in Gas Tubine Control Application developement but never had any chance to visualize detailed Turbine operation and maintance at Power plant.

Its short of remote engineerin with half knowledge.

During my application developement, I had to rely on documentation and facing lots of problem for relating various control schemes with actual operation and requirements of turbine thouroughly.

I understood various controls of turbine in bits & peices but never realized overall picture of turbine control.

I would like to request documents or linkes which gives detail overview of overall Gas turbine controls involving following topics.

1. Tubine Sequence ( Startup & shutdown)
2. Basics Auxiliaries involved
3. Main control mode relating various stage of
turbine operation
4. Protection scheme
5. Cricital sensor and actuator mounting scheme
on turbine

I am very thankful if someone can share your knowledge at possible extent.


I don't know which company you are working for or under contract (seconded) to, but I can guess.

There is a feeling in the company I'm thinking of that heavy duty gas turbine control and application is all "cookie cutter" and "cook book" stuff and that it can be engineered by anyone with an engineering degree with minimal training and mentoring.

That's unfortunate. Because, it *is* all fairly simple to understand if one knows the basic philosophy behind the way their systems were designed and continue to be designed (for the most part).

After that, things really become a lot simpler and easier and one can generally pick things up very quickly when encountering something new or unusual.

That company has some information that is provided in their unit instruction manuals they call 'System Descriptions'. There is a written description for every system: L.O.; Hydraulic Oil; Starting Means; Cooling and Sealing Air; Fuel Gas; Cooling Water; etc. There are even descriptions of the control and protection systems: fire detection; vibration protection; speed control; temperature control.

One thing you need to know about these System Descriptions: They are very generic in nature, and they should be treated as generic. But, they are a really good place to start learning and understanding the individual systems, and learning how they work together.

Also, the Control Specification drawing provided with every turbine control system also has a lot of really good information, though it does have a lot of generic information, and does contain some really glaring errors and omissions. (Sorry; but it's true.)

These documents, along with the sequencing used to control the units, all have the information you seek, just not in the form you seek it. And not all in one place.

And that's what everyone has asked me for 'nigh on three decades now: What *one* document can I use to learn and understand heavy duty gas turbine control and operation?

I'm beginning to get the idea.

Please, if you have specific questions, don't be afraid to ask them here.

Having said that, please use the 'Search' feature of control.com. A lot of questions have been answered here over the last few years. If you want more information or clarification on a topic previously asked, please refer to the thread and tell us what you need help with.
Mr. Bhavesh
Different manufacturers have their own set of O&M procedures setup. If you have specific query you can visit respective manufacturer’s web site. But the information will be an in general concept. The details you are very exhaustive. Sawyers Turbo machinery Handbooks covers your queries. NAVEDTRA- Naval Education and Training O&M manuals will of some help. You can access them trough search engine.
Dear CSA
Apart from the manuals you mentioned above "where can i get GE heavy duty gas turbine basic control philosophy?"

if you guid us it would easy to learn things
Your frustration is showing.

I am very interested in learning how to present the kind of information people like yourself are looking for. I know that people interested in learning about GE-design heavy duty gas turbine control philosophy and operation and control systems come from all backgrounds and experience levels and needs. Some are technicians, some are operators, some are supervisors, some are plant managers, some are plant accountants.

This is not a subject which is extremely difficult or impossible, but it is very difficult to present the material in such a way as to satisfy all the various people who may be seeking it in one form or another.

So, with your help I would like to carry on a dialog on this site to learn what kinds of things you think you need to know. Not just you, 4_20mA, but anyone interested in this subject. I don't know precisely how to get this information, but I'm researching various methods as I write this.

In the interim, please, provide some information about what it is that is causing you to feel like you don't understand about GE-design heavy duty gas turbine control philosophy and control systems. Do you want to know more about the various Master Control modes (OFF, CRANK, FIRE, MANUAL, AUTO, REMOTE, etc.)? Is the unit at your site a generator drive gas turbine or a mechanical drive unit? Single-shaft or two-shaft turbine? What fuels does the unit at your site burn?

Do you want to know more about specific Process- and/or Diagnostic Alarms? Do you want to know more about how to troubleshoot Process- and/or Diagnostic Alarms?

Do you want to know more about servo-valves and actuators? If so, what specifically?

Do you want to know more about electrical power generation (something most power plant operators and technicians and managers really know very little about I find)?

Do you want to know about the starting sequence of your unit? Or the shutdown sequence?

Do you want to know what happens when you or the operator clicks on RAISE SPD/LOAD or twists the Governor Control handle in the RAISE or LOWER direction?

Please, help me to understand the concepts you are most interested in. We can work on answering your questions here, where others can also benefit from the questions and the answers.

I have felt your frustration, and I am sympathetic to it, also. I'm just not sure how to present the material in a meaningful, cohesive, and engaging manner on a site like this, which is basically geared towards questions and answers.

I learned the bulk of what I know from reading the System Descriptions in the GE Gas Turbine Service Manuals, reading the Control Specifications, reading the sequencing (for Mark IV, Mark V, and Mark VI), and studying and learning the system Piping Schematic Drawings and the Generator Control Panel elementaries, and putting all of those things together with what I learned in university and in my travels and readings. I had the opportunity to work on GE-design heavy duty gas turbines for more than 25 years, and I did come into contact with lots of people who know this piece of information or that piece of information that filled in a lot of the blanks, which I know a lot of people don't have. I'm trying to share some of that knowledge here, and to learn some things in the process.

As I alluded to in my previous post; I'm getting the idea that people want a document or a book, some kind of reference, that they can go to to understand the concepts that seem just beyond what's currently available (which ain't much). I've been watching the posts here for a couple of years now, and reading many of the previous related posts and beginning to get a sense of what is needed or required.

Anyway, write back and let's work on those frustrations on this forum while we can!
Dear CSA,

Thank u so much for your reply and kindness to help us

yes , i wanna know about MASTER CONTROL MODES(OFF, CRANK, FIRE, MANUAL, AUTO, REMOTE, etc.) How all these things derived and how inter related?

i wanna know about TRIP LOGIC circuits. how those PTR and ETR relays circuited? how that hardware voting happening? what is "SINK RELY PRINCIPLE"?

Yes i wanna know about "REGULATORS" and it is components as u told servo valve,feed back devices, and actuators. what basis this valve responding for RAISE/LOWER& RAISE SPD/LOAD? Why position control instead of flow or some other control method?

yes i wanna know about this question "Do you want to know what happens when you or the operator clicks on RAISE SPD/LOAD or twists the Governor Control handle in the RAISE or LOWER direction?"

I wanna know about START UP/SHUTDOWN Sequence with process relatd information such as like (why,where, when and what for?)

Most of out DIAGNOSTIC ALARM is voter mismatch, you pls explain us how to deal with it

Why do we go for "MINIMUM VALUE GATE", What is the basic principle behind it to select minimum of speed/temperature/acceleration/manual/shutdown/startup?

Our unit is frame 9E heavy duty gas turbine which is driving generator and ours is single shaft turbine.We are mainly using GAS as prime fuel some time NAPHATA also.

i think no rookie is going to get frustated........ thank you

<b>One at a time</b>, I think you should open different threads for each one of these questions.

The one I'd like to tackle first, though, is the Voting Mismatch Diagnostic Alarms. You say you are experiencing these frequently. Would you please provide the details? Are they primarily on contact inputs, or what?

If you have GE packaged units, have you read the 'Control and Protection' tab of the Service Manuals? There is some generic information there about minimum value selections, but the basic concept is that for most of the fuel control functions it's felt that by choosing the minimum value the unit is protected against over-firing. All values of FSR control are constantly being calculated and the minimum (lowest) value is chosen as the one that will control the fuel flow. There's not much more to it than that.

So, let's take it one topic at a time, in individual threads, and deal with your questions. Remember, though, that we can't see the sequencing being used at your site and some of the answers may not 100% applicable to your site and your sequencing.

Also, we are going to be asking you to refer to the sequencing/application code in use at your site. I don't recall if you have Mark Vs or Mark VIs at your site, so please remind us from time to time.

And please be patient. We will get through this, and it will not take all that much time, but let's just work through it at a reasonable pace.

And, remember, feedback is critical to this process. If you understand what's being said, let us know. If you don't, please let us know. But, let us know!

As a start, I'm going to ask you to look at the rung or function block diagram for logic signal L4, the Master Protective Logic signal. This signal must be a logic "1" to run, and when it goes to a logic "0" the unit will be tripped.

It's a very simple and very informative and very powerful rung. And, it's nearly identical on almost every GE-design heavy duty gas turbine using Speedtronic control systems since Mark IV, and some later Mark II systems.

It's not possible to "draw" this rung anymore on the new control.com, so I'm going to try to describe it. The first element in the rung is a normally open L4S contact, and in parallel with that is a normally L4 contact. In series with these contacts is a normally closed L4T and a normally closed L94T, all driving the psuedo coil L4.

L4S is a "momentary" logic signal which will go to a logic "1" when a unit START is initiated and all the start permissives, including sufficient L.O. pressure, have been satisfied.

L4T is the logic signal which will be a logic "1" when any trip (emergency shutdown) condition is detected by the sequencing or application code, such as low-low L.O. pressure, or high-high vibration, or exhaust overtemperature.

L94T is the logic signal which will be a logic "1" when a normal, automatic non-emergency shutdown is active and the unit reaches the point at which the fuel is to be shut off.

In order for L4 to be energized (go to a logic "1"), L4T must NOT be a logic "1" (meaning there must be no active trip conditions detected), L94T must NOT be a logic "1" (meaning that a normal shutdown is not active and fuel does not need to be shut off), and L4S must go to a logic "1" (meaning that all the start-check permissives must be satisfied, and a START must be active, and all permissives to run the turbine must be satisfied).

When L4S goes to a logic "1" and L4T and L94T are both logic "0", L4 will pick up, and the normally open L4 contact in parallel with L4S will close and "seal in" around L4S. This will allow the START sequence to continue and the turbine to run *as long as there is no trip condition detected and no shutdown is active where fuel is to be shut off.*

If you want to know everything that will trip the turbine in the sequencing/application code, you need to look only at L4T and determine what will make L4T go to a logic "1". Every condition that results in a turbine trip (L4T going to logic "1") is supposed to have a process alarm message associated with it, so you can identify every alarm message which is associated with a turbine trip (emergency shutdown).

This is an excellent way to begin understanding all the conditions which can trip the turbine, as well as learning how to navigate the sequencing or application code in the Speedtronic. One just has to dive in, and there's really no better place to start.

If you have questions about the trip logic, we can work on those in this thread, since it's related to gas turbine control philosophy.

Let's have some fun, then!
Can you pl. send me a copy of the Gas Turbine Philosophy. I have done many steam turbines, but absolutely not much idea of Gas turbines. I need it for some potential work

Appreciate it.

Moderators Note: There were all sorts of blanks and # signs in this post which I had to delete. If I messed it up, please let me know.


It’s long time since we met for this particular topic. correct? we had SHUTDOWNS so it was not possible to concentrate over here. as you told we are going to start with L4 logic.

L4--- master protection/control

Le me start with L94T which is one of NC contact in that L4 rung series with L4T and L4S and L4 is again parallel with L4S. Let us take it us topic and put our knowledge which we understood and doubts as questions.
As I know L94T should be logic ZERO while staring and running the unit. am I right?

If the logic of L94T becomes 1 while running it brings the machine to shutdown and it doesn’t give permissive to start the turbine while starting. pls explain it

If you look at the L94T coil (L94T logic) rung it has a big logic behind it which includes L94XZ fired shutdown which is series with L94T.

Parallel to this Serially connected L28FD( NC contact) and L83RB where L28FD flame detection control & L83RB Ramp blowout selection. Here L28FD logic I got it clearly that whenever flame detection logic is healthy while running L28FD will be logic 1 so it doesn’t give permissive to L94T to become logic 1 through L83RB.

If you come to the case of RAMP BLOWOUT selection logic coil (L83RB) which is connected serially with the series connection of L60RB(NC) and L94SD
Where L60RB above ramp to blowout speed logic
L94SD- Shutdown with breaker open

Here the L60RB logic is interlocked with L28FDY Time delay for the flame detection control ie this L28FDY becomes logic 1 just after 1 sec of loss of flame.

So when normal running condition this logic will be always ZERO. so it will be acting as ENABLE ramp blow out LOGIC. so speed will be above 95% hence L60RB logic will become 1 so it will not give permissive for ramp blowout selection logic become 1.

When shut down initiated your speed will also come down so after some RB speed constant( it vary with machines 35% TNH,40%,45%TNH) the L60RB logic becomes ZERO and L83RB will become 1 as we hav given shutdown command.

So as machine has come down below 40% speed and lost the FLAME(L28FD) and L83RB selected the shutdown will be initiated..
FROM THIS WE CAN EASLY GET THE LOGIC OF L2RBT which is trip of RB time out

i would like to discuss and clarify all the RB related logic then let us move on.

here my question is, why does RB speed constant is vary from machine to machine?

you please let me know how to discuss this thread if this kind of discussion is not good enough.

I am looking forward to read your replay

thank uuuu
L4S L94T L4T L4
------| |----------|/|----------|/|------------( )
L4 |
------| |----|

L94XZ (Excessive shutdown time) L94T
------| |------------------|-------------------( )
L28FD L83RB | (Flame is lost before
------|/|----------| |-----| L2RBT times out
| but below K60RB)
------| |------------------| (Chamber Flamed Out
| During Shutdown)
------| |------------------| ("Normal" shutdown)

L94T is the normal shutdown "trip" logic signal. It will be a logic "1" when fuel is to be shut off during a normal fired shutdown. A fired shutdown is a deceleration from operating speed while maintaing fuel flow and flame as low as possible. This is as opposed to an emergency trip during which fuel flow is immediately shut off and the unit decelerates from operating speed without flame.

The normal course of events during a fired shutdown from load for a generator drive unit with a Mark IV, Mark V, Mark VI, or Mark VIe control system is as follows:

TNR is reduced until reverse power is detected and the generator breaker is opened

Fuel is reduced as the unit decelerates

If flame is lost *before* the unit reaches L60RB then usually the process alarm "Chamber Flamed Out During Shutdown" is annunciated and if multiple flame indications are lost then fuel flow is shut off

If flame continues as the unit decelerates below K60RB (after L60RB is a logic "1") then the L2RBT timer starts timing

If flame is lost before L2RBT times out, L94T goes to a logic "1" and fuel flow is shut off by dropping out of L4

If flame is still present when L2RBT times out, L94T goes to a logic "1" and fuel flow is shut off by dropping out L4

Most units running on gas fuel should be able to maintain flame to below approximately 20% TNH (speed).

Most units running on liquid fuel cannot support flame below approximately 40-50% speed (varies depending on unit and atomizing air/booster atomizing compressor configuration)

Prior to Mark IV Speedtronic control systems, it was common to shut off fuel flow when the unit decelerated below 14HS drop-out (usually 94% TNH (speed)). This results in something of a thermal shock to the unit. Beginning with Mark IV (or thereabouts), a scheme called 'Low Cycle Fatigue' was implemented and one of the things it sought to do was to reduce thermal stresses on the hot gas path parts during starting and shutdown, primarily during shutdown by maintaining flame during coastdown much lower than approximately 94% speed. The logic you are wanting to discuss is all about this: trying to reduce the thermal stress caused by cutting off fuel flow at low speeds.

If flame goes out before the K60RB setting, then it's usually alarmed to the operator as it's not a desirable condition.

If the unit doesn't decelerate to the K60RB setting before L94XZ times out after the generator breaker opens, then L94T is picked up. (A cause would be incorrect settings of the FSKMIN control constant arrays, or excessive fuel flow during shutdown (control valve actuator or servo problem, etc.).)

Many liquid-fueled units cannot get below even 60% TNH without excessing smoking during fired shutdown, so it's necessary to raise the K60RB setting for liquid fuel firing just to prevent excessive smoking (it's a no-no in some parts of the world). Some dual fuel machines have one K60RB for gas fuel, and one for liquid fuel.

I hope this answers your questions and clears things up somewhat.
I would like to have details of overall Gas Tubine control philosophy including various control mode, switchover between modes, critication sensor and FCE location on turbine, its purpose etc.

It's heartening to see here some drawing which can make everyone understand. hats off to you dear CSA. that’s fantastic and many many thanks for your effort. i good understanding.

I have one question in this L94T rung. You have mentioned that "flame is lost before L2RBT times out but BELOW K60RB. Let us assume that flame lost above K60RB speed (it is in some unit 40%TNH and some cases 45%TNH) suppose above 60%TNH. Here the L60RB logic is generated from a CMP block where if TNH is GREATER THAN K60RB mean it generates logic 1 again this block SHOULD BE ENABLED then only L60RB will become logic state ONE. am I correct? so here flame lost with time delay signal L28FDY which is enable signal for L60RB logic by NORMALY CLOSED contact.

So it becomes logic ONE when speed is 60% so L83RB logic COIL is interlocked by serially connected contact of L60RB and L94SD where L60RB is NC. So L83RB coil will not become logic 1 and it doesn’t give permissive for L2RBT time delay timer logic so it should be a "CHAMPER FLAMED OUT DURING SHUTDOWN" alarm. It will not become a "NORMAL SHUTDOWN" either.

I understood that this happens ABOVE K6ORB constant speed also. Please explain me this.
You're going to need to refer to my previous explanation and this one to see all the relevant sequencing.

K60RB is the speed the unit is expected to be able to maintain flame down to during a fired shutdown. And if flame is lost any time during a normal fired shutdown, then L94T will be picked up (go to a logic "1") to drop out L4 (go to a logic "0"). GE just makes a distinction between whether the speed is more or less than K60RB. And for some unknown reason, they make a "distinction" of whether flame is lost before L2RBT times out or not, which is kind of odd.

<pre> L28FD L28FDY
1 sec
| CMP |
| |
L28FDY | Enable |
---|/|--------------------|---------- |
| _____ | |
TNH |a | | | |
--------|--| | | |
| |a>b| | | L60RB
| | |-| |-|------( )
K60RB |b | | |
--------| | | |
| ----- |

----|/|------| |------------------------------( )

----| |---------------------------------------(T)
30 sec

When there *is* flame (when L28FD is a logic "1"), L28FDY will be a logic "0".

When L28FDY is a logic "0" (when there *is* flame), the comparator for L60RB will be enabled, and it will be enabled until 1 second after flame is lost (when L28FD goes to a logic "0", one second later L28FDY will go to a logic "1" which will remove the enabling logic signal from the L60RB comparator).

This means there will be a 1 second time delay after flame is lost (L28FD goes to a logic "0") before the enabling logic for L60RB is lost.

L83RB says "the speed has dropped below K60RB (which means L60RB has gone to a logic "0") AND a normal shutdown is active". This will start the timer for L2RBT.

So, L94T will go to a logic "1" when L28FD goes to a logic "0" and L83RB goes to a logic "1" (which it will within one second of losing flame after TNH drops below K60RB).

If flame continues as speed drops below K60RB (L60RB goes to a logic "0") for longer than K2RBT, then L2RBT will go to a logic "1" which will pick up L94T to drop out L4.

I've always considered the L60RB comparator to be "inverted", meaning, that it should go to a logic "1" when speed is less than K60RB instead of being a logic "1" when speed is greater than K60RB. It just seems it would have made more sense (in GE-speak) to have L60RB go to a logic "1" when the speed was below "blow-out" and flame was still present, instead of L60RB going to a logic "0" when speed was below "blow-out" and flame was still present.

My use of "below K60RB" is probably misleading since in my mind the logic for K60RB is inverted. (The rung would have worked identically if they had used K60RB as the "a" input and TNH as the "b" input, and used a NO L60RB in the L83RB rung. And it would have made more "sense", at least to me. But they didn't and it is what it is.) They describe L60RB as "Above Ramp to Blowout Speed" and L60RB is a logic "1" when speed is greater than K60RB. It would also have worked if L60RB was a logic "1" when the speed was less than K60RB, which would have made rung L83RB easier to understand, as it would have been written:

<pre> L60RB L94SD L83RB
----| |------| |------------------------------( )

My apologies for any confusion.

GE logic for L83RB is read as: IF NOT L60RB AND L94SD THEN L83RB, or, IF Speed *IS NOT* above ramp to blowout speed AND a normal shutdown is active, then pick up L83RB (Ramp to Blowout Selected).

My logic for L60RB would be read as: IF L60RB AND L94SD then L83RB, or, IF Speed *IS* below ramp to blowout speed AND a normal shutdown is active, then pick up L83RB. Which reads much easier (at least to me).

If flame is lost when speed is above K60RB, then L2CANT will go to a logic "1" and L94T will pick up and drop out L4.

So, any time flame is lost during a normal, fired shutdown L94T will pick up and drop out L4. It's just that if flame is lost when the speed is above K60RB, that's considered "abnormal" (not dangerous, but it does cause higher thermal stresses on the machine than if flame is maintained to a lower speed) and L4 is dropped out, cutting off fuel (to prevent possible re-ignition). When the happens, an alarm, something like "Chamber Flamed Out During Shutdown", is annunciated.

If flame is lost below K60RB but before L2RBT times out, L94T is still picked up and L4 is dropped out and fuel is cut off.

If flame is lost below K60RB and L2RBT times out (the most desirable from a thermal stress standpoint), L94T is picked up and L4 is dropped out and fuel is cut off.

We're not going to go into the whole L2CANT string and all of its permissives (what you call "interlocks").

Isn't all of this logic stuff fun? (Sometimes, when it's written "logically".)