Today is...
Tuesday, September 17, 2019
Welcome to, the global online
community of automation professionals.
Featured Video...
Featured Video
EtherCAT with CTC’s master lets your multivendor network play well together...
Our Advertisers
Help keep our servers running...
Patronize our advertisers!
Visit our Post Archive
Fire Protection System of a Gas Turbine
Looking for an explanation of the fire protection system of a gas turbine

Hell everyone,

Myself engage in operation of GE make gas turbine generator (19.6 MW, Frame-5, Mark-VIe control). kindly explain the fire protection system of the turbine-generators. As per the schematic, there are UV detectors & thermal detectors in aux compartment, turbine compartment, L/G compartment, and also in generator compartment. Just need to understand how it functions.


1 out of 1 members thought this post was helpful...


The fire protection system uses various sensors to detect a fire in any of the compartments where the sensors are located. When a fire is detected, the turbine is usually tripped (emergency shutdown--fuel flow is stopped immediately), including the compartment ventilation fans (because the fans would supply air (and oxygen) to the fire).

After some time, usually 15-30 seconds or so, the extinguishing agent is discharged into the compartment(s) (or zones) where the fire was detected and might spread to. Usually there are two sets of discharge nozzles--one set which very quickly "floods" the compartment/zone with extinguishing agent to quickly build up the concentration of extinguishing agent and help put out the fire, and another set, with small orifices, which attempts to maintain the concentration of extinguishing agent in the compartment for some time (often as long as 30 minutes) to make sure the fire doesn't erupt again.

Many of the ventilation dampers in compartments are "gravity-operated." This means they are closed by gravity, and most are opened by air either drawn into the compartment or pushed out of the compartment by the ventilation fans. This is another reason for the time delay between turbine trip and extinguishing agent discharge: to allow the ventilation fans to slow down sufficiently to allow gravity to close the dampers. Some dampers are also latched open--meaning they have to be manually latched open after a discharge, and when the extinguishing agent is discharged the pressure in the pipe releases the latch to allow gravity to close the damper, keeping air out and extinguishing agent in. (There were some GE-designed heavy duty gas turbines which also used AC motor-operated dampers for compartment ventilation, but I believe the design was quickly deemed not very ... good.)

Hope this helps!

Thank you so much CSA!

I have seen the drawing and was trying to understand the function of 45CR1 & 45CR2 and also 45CP. Can you help out?

Recently one of our GT tripped on false fire signal. When we analysed the alarms, we found the alarm "Manual CO2 release/turbine area zone-1"(black letter) came first, and later "Forced signal detected alarm" (red letters) came later, and the GT tripped. However, actually CO2 was not released and found that there was moister ingress was there in 45CP switch. What is the meaning of black & red letter?


1 out of 1 members thought this post was helpful...


45CR-1 & -2, if I recall correctly (and I don't have any drawings to refer to at this writing), are the release solenoids for the Initial- and Extended extinguishing agent discharge systems. The release solenoids are energized after the time delay (usually 15-30 seconds) between when the various fire system sensors tell the fire protection system that a fire has been detected and when the extinguishing agent is to be released.

45CP-1 is a pressure-operated switch that changes state when the Extended extinguishing agent discharge occurs (I believe the Extended extinguishing agent discharge starts at the same time as the Initial extinguishing agent discharge starts). It's purpose is to "latch in" the extinguishing agent discharge until the Extended discharge is complete (usually approximately 15-30 minutes)--meaning that if the sensors were to detect there was no longer a fire, that might de-actuate the fire protection system, and the fire could flare up again. So, 45CP-1 is actuated by pressure in the extinguishing agent piping (which there will be after the extinguishing agent is release) and as long as there is extinguishing agent pressure in the pipe 45CP-1 will remain actuated.

When the Extended extinguishing agent discharge time is complete, the Extended discharge extinguishing agent release solenoid will be de-energized and a few seconds after that the Extended discharge extinguishing agent piping will be de-pressurized. At THAT point, or sometime AFTER that point, a human will have to manually reset 45CP-1, in order to be able reset the fire protection system and re-start the turbine.

Another method of tripping the turbine and immediately discharging extinguishing agent is for a human to manually actuate 45CP-1 (there is usually a black knob on the bottom of the switch which can be pulled down by a human to actuate the switch, trip the turbine and discharge extinguishing agent--all virtually simultaneously. You mentioned someone found water in the 45CP-1 switch housing, which could result in a short of the contacts in the switch and cause a trip/release. If the CO2 at your site DID NOT release, then I would suspect someone has the manual discharge piping isolation valves closed (to prevent a CO2 discharge caused by an false discharge event); that happens at all too many sites where multiple false discharge events occur, wasting time and extinguishing agent. HOWEVER, that means if there IS a REAL FIRE, that some will have to go out and manually open the extinguishing agent discharge valves to allow extinguishing agent to be discharged.... And unless ALL the operators and site personnel know and understand this, and it's sanctioned by management/supervision and/or ownership, this is a dangerous way to operate the unit.

MOST--but NOT all--GE-designed fire protection systems have enough extinguishing agent for two fires so that the extinguishing agent supplier can be contacted to deliver more extinguishing agent as quickly as possible, but if the turbine can be re-started after the discharge event (which may be erroneous) to produce power. Some of the fire protection systems have to have the bottles replaced or recharged immediately after a discharge because there is insufficient extinguishing agent to supply a second discharge event should it occur before the supply can be replenished. You will have to determine the capacity of your system at your site to know how much capacity the system has before it needs to be replenished after a discharge event.

[NOTE: I have never personally seen a fire protection system on a GE-design heavy duty gas turbine actually detect a real fire, trip the turbine and discharge extinguishing agent. In more than 30 years of working on GE-design heavy duty gas turbines. It's almost ALWAYS an operator who notices billowing smoke coming out of a compartment who runs over to the fire protection system and manually initiates a trip and discharge. If there are two or more zones of fire protection, they almost ALWAYS discharge extinguishing agent into the wrong zone--because the just reach for the first button/switch they can find, and it's usually for the wrong zone. The turbine is tripped, and extinguishing agent is discharged--but very frequently into the wrong compartment, which does nothing to extinguish the actual fire. The other way a turbine gets tripped during a real fire is when the fire causes the insulation of wiring in the area of the fire to melt and the wires short and/or touch ground--which can be very damaging to the turbine control system and battery powering it. The turbine gets tripped, but usually there's no indication to the operator why the unit tripped--until someone goes to the unit and sees flames and/or smoke and manually discharges extinguishing agent or calls the fire brigade/department to put out the fire.]

Now, as for the "black and red letters." I believe you are referring the color of the alarm message text on the GE HMI Alarm display. ?.?.? Usually, red letters mean the alarm condition is current--it exists as long as the text letters are red. And, black letters usually mean the alarm condition no longer exists and can be reset to clear the alarm text message from the display.

FORCED SIGNAL DETECTED means that an operator has forced some signal(s) in the turbine control system. Red alarm text message letters mean the signal is still forced...! The alarm text message will not go to black letters until ALL forced signals have been unforced.

From your previous description of the fire protection system at your site, it has--in addition to heat detectors--UV flame sensors which augment the heat detectors to try to prevent false discharges. But, quite often, the UV sensors are not wired to the fire protection system properly and so problems with heat detectors can cause false trips/discharges.

It sounds as if the turbine control system at your site has been upgraded to a Mark VIe, and at some point someone added the UV flame sensors to the fire protection system. Older fire protection systems (pre-Mark IV production) would trip and discharge if any single heat detector sensed a flame--which caused many nuisance and false trips/discharges. Beginning with Mark IV turbine control system, it required any two heat detectors in any zone of fire protection to trip and discharge extinguishing agent, which significantly reduced the number of false trips/discharge but didn't eliminate them completely. It's impossible for us to say how the fire protection system at your site is configured, and how it seems to have been modified (with the addition of UV flame sensors) without being able to see the electrical schematic drawing for the fire protection system.

Again, it's very disturbing that if 45CP-1 initiated a trip that extinguishing agent (CO2 at your site) wasn't discharged. As was said above--when 45CP-1 is actuated it serves to "latch in" the fire protection system and ensure the supply of extinguishing agent continues for the time set by the system controller (or until someone manually stops the discharge and can manually reset 45CP-1). 45CP-1 is EXACTLY like one or more fire sensors detecting a fire--it's actuation (or false actuation in this case) SHOULD result in a turbine trip AND a discharge of extinguishing agent.

Either there's a problem with the extinguishing agent release solenoids/mechanisms, or manual isolation valves weren't open when the unit was running, or the extinguishing agent system pressure was too low (low or no extinguishing agent). Many GE-design frame 5 heavy duty gas turbines used multiple high-pressure CO2 bottles for the extinguishing agent, and the solenoid-operated extinguishing agent discharge mechanisms use CO2 pressure in the bottle to actually open the main discharge valves on the bottle heads. (It's called "piloted" operation.)

As for why there was a FORCED SIGNAL DETECTED alarm shortly after a trip/discharge event--that, too seems odd. My guess is that the operator or supervisor on shift when the event occurred just forced a logic signal (or multiple signals) because that's what he (or she) had been taught to do in the event of a trip (and/or discharge) event. You would need to ask whoever was operating or assisting with operating the unit when the trip/discharge event occurred to know for sure. This kind of "operation" (forcing signals after a trip) should not normally be necessary, and can lead to unintended consequences if not done correctly or without understanding all the consequences of forcing signals. (But, having said that, it happens all too often at too many sites around the world.)

I hope this helps! I wish you would ask all your questions in your original post(s); it's easier that way. Especially when you have specific questions about a system or its operation, or turbine operation, or alarms associated with an event.

As for the HMI alarm window display alarm text message colors, you need to ask an experienced operator exactly what the different colors mean. If your GE Mark* HMI has different colors for alarm text messages in addition to black and red it can signify different things and require different and specific actions. There should be a training manual somewhere at the plant for the Mark VIe training which was most likely provided at the time the turbine control system was upgraded; finding that manual could be very helpful to you.

Again, if GE provided the package (turbine, generator and auxiliaries) when it was new and the unit is more than approximately 20 years old, it's very likely you will find a copy of the 'Device Summary' in Vol. III of the Operations & Service manuals provided with the unit and auxiliaries. The 'Device Summary' has the descriptions of MOST (but never ALL) of the devices supplied with the turbine and auxiliaries, and sometimes the generator and its auxiliaries, also (or at least some of the auxiliaries). There may also be (should also be) a copy of the 'Piping Symbols' document in the same tab in Vol. III which can be useful when trying to decipher the symbols on the P&IDs.

Hello CSA..
Thanks a lot for your time and sharing your vast knowledge with us. This really helps.

(1) Regarding the Forced signal, actually the R(Instt) was handling the issue immediately and will discuss with him about it, if something was done like this.

(2) Why CO2 was not released? Regarding this I would like to know what is the industry practice to know the truthfulness of the system. Like CO2 will release or not in case of fire.

You have also put up an excellent probable explanation. Anyway...waiting for reply of the committee.

thanks again

1 out of 1 members thought this post was helpful...


>(2) Why CO2 was not released? Regarding this I would like to
>know what is the industry practice to know the truthfulness
>of the system. Like CO2 will release or not in case of fire.

I really don’t understand the sentence about “... what is the industry practice to know the truthfulness of the system.” CO2 SHOULD release when there is a fire! But at sites where they have experienced a lot of false trips and discharges when there was not an actual fire sometimes the fire protection system is..., modified so that it won’t actually discharge unless a human goes out and does something to allow CO2 to discharge.

I’m NOT saying that’s what happened at your site, but I am saying that happens at some sites. It’s expensive to refill the CO2 system after it is discharged. And sometimes it takes a day or more to get the CO2 supplier to come to site to replenish the CO2 that was discharged. And if it happens often that the unit trips and discharged CO2 when there wasn’t actually a fire, well, the easiest thing to do is to modify the fire protection discharge system so that it can’t automatically discharge CO2 and it requires someone to manually allow CO2 to discharge when there is a fire.

What should really happen if there are repeated cases of trips and discharges when there was no actual fire is that site management should insist that the fire protection system be made reliable. Which is possible. But, many times sites have called fire protection suppliers to come to site to review the fire protection system provided with the unit to make it more reliable and many (most, unfortunately) will say, “We can’t make THAT fire protection system reliable. You need our new and improved fire protection system!” And, the new fire protection system is VERY expensive.

You mentioned the fire protection system at your site has both heat sensors (which is what is typically provided with mos GE-design heavy duty gas turbines) and UV sensors. It’s possible that the UV sensors were added to the original fire protection system to try to make it more reliable. But if it wasn’t done properly then the system is not more reliable.

Many sites have insurance on the equipment and the insurance company would NOT be very happy to learn that the fire protection system is not reliable and requires manual intervention to discharge CO2. Not happy at all.

This is a very tricky situation. You need to ask people at your site questions about how often the turbine trips because of a fire, and how often there was actually a fire. You may learn some very interesting things. And you may not. But you need to know why CO2 didn’t discharge. Someone knows, and it shouldn’t be a secret. Secrets at power plants can be very dangerous.

Look, it just might be that the reason CO2 didn’t discharge is there was a problem with the discharge solenoids and it’s been fixed. But, that shouldn’™t be a secret. And we can’t tell you why the CO2 didn’t discharge. Because it should have. If the cause of the trip was that 45CP-1 shorted out because water got into the switch then CO2 should have discharged on a typically-provided fire protection system.

The fire protection system doesn’t just protect the unit; it also protects the people working at the site.

If the system tripped as you described then CO2 should have discharged on a typically-provided system. As an operator, you need to understand why CO2 didn’t discharge. If the fire protection system at your site is not the typically-provided fire protection system you need to know that, too.

But, be aware. Sometimes asking these kinds of questions can get a person in trouble. If the fire protection system has been modified and it should not have been modified that’s not something that the person or people who are responsible for the modification want to have everyone know (even though everyone should know!). So, ask carefully, but you, as an operator, need to know why CO2 didn’t discharge if the reason for the trip was that 45CP-1 shorted out because of water getting into the switch.

It may be that in order for CO2 to discharge the UV fire sensors need to also see fire. That could be the reason CO2 didn’t discharge.

But you need to know. I don’t know what an R(Instt) is, but I suspect it might be the instrumentation and control technician and that person should be able to tell you the reason CO2 didn’t discharge.

Don’t get yourself in trouble. Some day you’ll learn what happened and why. Be patient. I wish we could tell you the reason why CO2 didn’t discharge, but we can’t. We’re not there and we can’t look at the schematics and the P&IDs and we don’t know the “history” of fire-related trips at your site.

Be patient, and most things will become clearer. Just like in life; things can suddenly become clear that were not clear for a long time.

What I have written is just from my experience; sometimes my experience was good, sometimes not. I have seen some good things, and I have seen some bad things. But, mostly I have seen sites that have investigated the causes of nuisance and erroneous trips and have taken action to make the unit and the various components and systems reliable. Over time, you will eventually discover many things at the power plant where you work, and in your personal life.

Be inquisitive; but also be patient.

By gustavo_marcelo on 2 August, 2019 - 4:22 pm


Since you mentioned about Emergency ShutDown when there is a fire, what about the manual Emergency SD? Does it also do the same sequence as when it detects fire? In other words Emergency SD main purpose is for fire incident, triggered automatically or manually?


I think I used the term "emergency shutdown" in parentheses, immediately following the word "trip." In GE-design heavy duty gas turbine parlance, a trip is an emergency shutdown--which means the fuel stop valve will be immediately closed to shut off the flow of fuel to the combustors. And, there are LOTS of things that can trip a GE-design heavy duty gas turbine with a Mark* turbine control system. There are trips generated in application code (for a Mark VI or Mark VIe), or the CSP (Control Sequence Program in a Mark V), or sequencing (in a Mark IV). There are hardware trips--some are connected to discrete inputs of the Mark*, some are connected to the <P> Protective processors (of a Mark VI or Mark VIe), some are Emergency Stop Push-buttons. Units which still use a mechanical overspeed bolt will be tripped because hydraulic supply to the fuel stop valve actuator will be lost when the overspeed bolt is actuated. There are MANY ways a GE-design heavy duty gas turbine can be tripped. Some are done in <Q> (a collective term for <R>, <S> and <T>); some are done in <C> (for Mark IV and Mark V SIMPLEX turbine control systems); some are done in <P> (the Protective processors in a Mark VI or Mark VIe, or the Protective core of a Mark V). A turbine can be tripped because of low-low L.O. pressure; or for high-high L.O. Temperature; or for high-high vibration; or because of a HRSG (Heat Recovery Steam Generator) boiler drum level problem; or because of loss of fuel flow (from the supply--which will cause a "LOSS OF FLAME" trip); and on, and on, and on.

It's necessary for you to be very specific when you refer to Emergency ShutDown or Emergency SD. Are you referring to the Emergency Stop (or ShutDown) Push-button which would have to be manually pressed by a human?

Trips can be manual (such as the E-Stop P/B), or automatically initiated by software, or because of loss of fuel or hydraulic supply pressure, or high-high exhaust temperature, or high-high exhaust temperature spread (the last two being done in software using the exhaust T/C inputs and a set of parameters).

The fire extinguishing agent is only going to be discharge when the fire protection system detects a fire OR a human manually initiates a discharge of the fire extinguishing agent. EVERY GE-design heavy duty gas turbine I have ever worked on would trip the turbine if a fire was detected, OR if a human manually initiated a discharge of the fire extinguishing agent. (As I wrote, I have never seen a fire protection system on a GE-design heavy duty gas turbine actually trip the turbine because it detected a fire and discharged the fire extinguishing agent. Never. Every discharge of fire extinguishing agent from a fire protection system on a GE-design heavy duty gas turbine I know of occurred because a fault with one or more of the fire sensors, or a manually-initiated discharge.)

Pressing the turbine E-stop P/B in the control room or on the turbine control panel door will only trip the turbine (by immediately shutting off the flow of fuel to the combustors). Pressing the E-Stop P/B will not automatically discharge fire extinguishing agent--even if there is a fire. ONLY the fire protection system or a human initiating a fire extinguishing agent discharge will cause a discharge of fire extinguishing agent. (I haven't yet seen a fire protection system with artificial intelligence to know the operator was pressing the E-stop P/B because a fire was suspected or reported or observed.)

An emergency shutdown (a turbine trip) can be the result of many different conditions, including a fire.

I was only trying to let the reader know that a turbine trip was an emergency shutdown, and I think I even said an emergency shutdown (trip) resulted in an immediate closure of the fuel stop valve (or something similar but with the same intent, because closing the fuel stop valve shuts off the flow of fuel to the combustors--immediately).

Hope this helps!


I think with your strong experience you can give us more calrifications about this systeme.

We have a real repetitive problem:
partial discharge of the CO2 bottles (3/8) of the dedicated skid, it becomes repetitive with each trip by a flame detection alarm.

I inform you that the detector has been changed and the CO2 discharge presostats are controlled and the gas turbine (model 5002C-GE) continues to work until this problem (2 MONTHS).

Another problem in the same turbine is that the temperatures in the combustion chamber are not homogeneous (i.e., a difference of 50 ° C - confirmed by the intensity of UV flames) and sometimes the turbine trip by loss of flame.

thanks for your support.

I have no knowledge of or experience with systeme 613/5000, so I can't provide any help with this.

Your other problem is very serious! You have high exhaust temperature spreads, and it seems your unit may not have combustion monitoring, which checks for exhaust temperature spreads and adjacency and alarms and trips the unit to protect it from the damage which can be caused by high exhaust temperature spreads, or the damage which is causing the exhaust temperature spreads.

Exhaust temperature spreads and the causes have been covered MANY times before on There is a 'Search' feature on, cleverly hidden at the far right of every Menu bar of every webpage in the desktop version. (If you're using the mobile version of, 'Search' is found under the tab.) Best to use the 'Search' Help because the syntax of search commands for this search engine is very different from your preferred World Wide Web search engine.

Exhaust temperature spreads are nothing to ignore; one ignores them at the peril of the machine and anyone working nearby.

Please write back to let us know what you find. If you want further assistance you are going to have to provide a LOT more information about your unit (manufacturer; size; type of combustor (DLN or conventional diffusion flame; time since last maintenance inspection; fuel being burned; number of flame detectors; number of exhaust thermocouples; type of control system (manufacturer and model number), etc.). The GE Mark* turbine control community and GE-design heavy duty gas turbine community is very active here on, but other turbines and control systems not so much.

Hope this helps!

So we have a MAX combustion imbalance allowed on the gas injectors of the turbine (deviation of the flame intensity). An inspection is made to see the state of the burners (always presence of black deposit).

Gas turbine model : MS5002C / POWER / 26MW at 45°C
manufacturer : NUOVO-PIGNONE
type of combustor: conventional diffusion flame
time since last maintenance inspection : 3 years
fuel being burned : naturel gas
number of flame detectors : 04
number of exhaust thermocouples : 12
type of control system : GE - Mark VIe
Air compressor stages : 16 / type : axial / pressure ration : 8.2

Thnaks for your support