gas turbine tripped on "loss of flame"

P

Thread Starter

power plant

after a gas turbine started the temperature on turbine compartment and load gear compartment increasing until reach the high temperature alarm, and after that the turbine tripped on "loss of flame". "this trip always happened after 1 or 1.5 hours". we check the flame detector loop, every thing is ok, and we measure the temperature on the turbine and load gear compartment it was >110 c. we tried to restart the unit many times, the same thing repeated.

why the turbine compartment and load gear compartment increasing?
why the turbine always tripped on "loss of flame"? is the flame detector will be as open on high temperature?

thank you for your help,

MAZEN A M SHANABEASH
instrument and control system engineer
[email protected]
 
Mazen,

While this problem may have been happening for some time (we don't know how long), the difficulty with trying to troubleshoot a problem like this is that it has been allowed to persist for some time and so the details of what happened or changed prior to the start of the problem are likely forgotten.

Many sites, when they encounter any difficulty, just keep "pushing the START button" hoping that the occurrence was a one-time event, or a two-time event, or a ten-time or thirty-time event. In other words, they just hope that the problem will go away.

We also don't know what kind of control system is used on this turbine so we can't ask for data or propose some troubleshooting/data-gathering.

Most of the time when I encounter high compartment temperature alarms it's usually found that the fire protection dampers are not working properly or are not latched in the proper running position. Also, I have seen compartment vent fans running in the wrong direction (the direction of rotation was not checked after some maintenance outage). A couple of problems have also been caused by loose fan couplings; the motors run but the fans don't turn as fast as they should.

Missing from this post is an indication of what the compartment temperatures are like when the unit doesn't trip (because I'll wager that occasionally the unit doesn't trip on this set of conditions, but does probably trip on others at some other point in time and then the problems with high compartment temperature and loss of flame occur while trying to get the unit back to generation). Are the compartment temperatures normal when the unit doesn't trip on loss flame?

Most compartments are under a slight positive pressure or a slight negative pressure. Opening the compartment doors of those with a slight positive pressure can be "alarming"--because the doors will usually fly out when unlatched. Opening the compartment doors of compartments with negative pressure usually requires a good amount of force because. So, has anyone noticed a difference in the door-opening of these two compartments?

High compartment temperatures can be caused by leaks which shouldn't be allowed to persist. In the turbine compartment, the leaks can be coming from the compressor casing joints, from the combustion wrapper joints, from many different locations. Has anyone looked for leaks which might be contributing to the high compartment temperatures?

In the load compartments of GE-design heavy duty gas turbines exhaust leaks can contribute to high compartment temperatures--but this should NOT result in a loss of flame trip because the flame detectors are located in the turbine compartment. Leaks in the load compartment can be difficult to find, but not impossible, because the compartment is usually very warm to begin with.

You say you checked the temperature of the compartments and no temperature was greater than 110 deg C. This is well below the typical high temperature switch setting for most GE-design heavy duty gas turbines. How, and where, were you measuring the temperatures? The only place that really matters is somewhere near the temperature switches used to indicate high temperatures. (Some newer GE-design heavy duty gas turbines seem to have thermocouples to indicate compartment temperature and may be used for alarm purposes as well. Does your unit have these?)

I have also seen temperature switch sensing bulbs not properly mounted after maintenance outages. When new, the sensing bulbs usually have one or two wraps of fiberglass tape under the clamp used to hold them in place (they are not just free to flap about in the compartment). And, many times these switches have not been checked for years, decades even.

Flame detectors--whether they are the Geiger-Mueller UV (Honeywell) type or the SiC (Silicon Carbide) Reuter-Stokes type can be negatively affected by abnormally high ambient temperature. Just about every electronic device has an upper temperature limit listed on the product's data sheet. GE-design heavy duty gas turbines with DLN (Dry Low NOx) combustion systems even employ water cooling to prolong the life and limit damage to flame detectors from high compartment temperatures.

Most turbines I have worked on have had at least two flame detectors, for redundancy. Those with only two flame detectors usually have to lose indication from both flame detectors before they are tripped on loss of flame. How many flame detectors does this turbine have, and are they all losing flame indication at the same time, or are one or more not indicating flame for some time before the loss of flame trip?

What's also missing from this original post is when this loss of flame trip is occurring (during start-up/acceleration, or when running on load). Also, presuming this is a GE-design heavy duty gas turbine, what are the exhaust temperature spreads prior to the loss of flame trip? I would have to believe that the exhaust temperature spreads are elevated at a minimum, and maybe even very high just prior to the trip.

I have never seen a GE-design heavy duty gas turbine tripped on high compartment temperature, but, then I haven't seen every control system configuration. These days, with different divisions of GE free to program the control systems as they see fit (in the past, all control system programming was done per a single, central engineering organization) there are all manner of different schemes being employed. I have seen some recent installations that tripped when a low level of hazardous gas concentration was detected AND a high compartment temperature was also detected. We don't what other alarms are present at the time of the loss of flame trip (and I have a suspicion there are many!).

There are a LOT of questions (some direct, and some indirect) in the above discussion. To be of any further help, we are going to need to have answers to most of the questions. I will list the most important questions to be answered below:

1) What kind of gas turbine?

2) What kind of gas turbine control system? Is it SIMPLEX, DUAL redudnant, or TMR (Triple Modular Redundant)?

3) How many flame detectors does the unit have, and what kind are they (UV or SiC; Geiger-Mueller or Silicon-Carbide)?

4) What are the indications from all of the flame detectors just prior to the loss of flame trip you are associating with high compartment temperatures?

5)Are you certain all of the compartment dampers are working properly and/or are latched in the proper running position?

6) Are you certain the compartment ventilation fans are running properly--in the proper direction of rotation--and that the fans are securely coupled to the motor shafts?

8) How are high compartment temperatures detected? By temperature switches, or by thermocouples? If by temperature switch, what are the high compartment temperature switch setpoints for each compartment? Have you verified that the switches are properly calibrated and are working
properly? Are the sensing elements (of the switches or the thermocouples) properly mounted in their original positions?

7) What is the actual current drawn by each fan motor (with the compartment doors closed!) and what is the nameplate current for each fan motor?

8) When does the loss of flame trip you are associating with high compartment temperatures occur--during starting/acceleration or when at rated speed?

9) What are the exhaust temperature spreads prior to the loss of flame trip? If the unit is a GE-design heavy duty gas turbine with a Speedtronic turbine control system, what are the values of TTXSP1, TTXSP2, and TTXSP3 just prior to the loss of flame trip you are associating with high compartment temperatures?

10) Presuming this is a generator-drive application, what is happening to the load just prior to the loss of flame trip you are associating with high compartment temperatures? Is it dropping quickly or is it stable? (This presumes the grid frequency is stable; if the grid frequency is not stable, please tell us.)

11) What other alarms are present when the unit is running at the time the loss of flame trip you are associating with high compartment temperatures occurs? (We need to know ALL of the alarms, Process and Diagnostic Alarms if the turbine control system is a GE Speedtronic.)

I have the feeling that there is more than one turbine at the site and that only one turbine is experiencing this particular problem associated with high compartment temperatures. Can you please tell us if this is correct or not?

When posting to a forum like this you need to tell us a lot more about the unit at your site. You should also tell us what troubleshooting steps you have taken and what the results of those actions were. We can help if we have information, but we only have the information you provide--and if you don't provide any information, we can't help very much.

Lastly, if you've read any of the other GE-design heavy duty gas turbine-related threads here at control.com, you will find that most people who post here for help with their problems or questions reply with feedback to say if they were able to resolve their problem with the help provided, or not. This feedback means that when others search for answers they can see if the information provided was useful or not--and everyone benefits from helping to solve one person's problems. Please be sure to write back to let us know how the problem was resolved.
 
Mazen,

Two more questions:

How long does the turbine trip after the high compartment temperature alarms are annunciated?

And, do the trips always occur at the same time after the high compartment temp alarms?
 
For the GT Alstom 13E2, fuel oil start up is using propane gas for the ignition stage. The reason is propane calorific value is much more higher than fuel oil and hence easy to catch the flame. Certain amount of propane will be determine depending on the GT condition (varies on difference temp after turbine). Once ignition successful fuel oil flow will be integrated for the speed acceleration. In this case it is good to consider for the ignition mass flow adjustment for the successful ignition.

Thanks
Alfa
 
Is there any leakage from the compressor casing? (i.e. can u hear noise)

Is the CPD lower than usual at this case?

Is the exhaust temperature higher than usual before the trip?

If so, the problem may be due to underfiring at low CPDs (if you have a cpd control curve), please keep in mind that the lower cpd occurs at higher ambient temps and in your case the lower cpd is at lower ambient temp than usual (since the cpd drop is coming from a leakage). so, you have 2 reasons for a weak flame. 1- lower firing temp at isotherm operation 2- lower ambient temp than usual. this for sure may lead to weak flame and then flame off.

The reason why the compartment temp is high is simply due to the hot air leakage from the compressor.

Please bear in mind that the discussion above is valid only if you have air leakage from compressor casing.
 
Top