Gas Turbine High Exhaust Temperature Trip

G

Thread Starter

Gunga Chea

For the last two days we have been struggling with a problem of high exhaust temperature trip on GT 2 , a frame six with Mark V controls.
First time it occurred I checked and found one check valve setting was at 60 psi and after replacing the problem got solved. It occurred again and this time I found two check valves had failed and I replaced them with new ones.

But this time I checked all the check valves and found them to be between 102 and 99 psi but it still tripped on the same problem. I have changed the Moog servo valve, serviced the fuel bypass valve but the problem is still persistent.

Yesterday evening it accepted to fire but could not accelerate beyond 2300 rpm, after which we found a union leak on the Compressor discharge pressure impulse line isolating valve. It was repaired and then the high exhaust temperature trip occurred on firing.

Conditions
1. The fuel pump discharge pressure was 104 psi and dropped to 60 after firing.

2. At start up the fuel percentage on the flow divider goes to 5.00% after firing it rises to 8.00% and trips.

If there is any information that I have left out and may be required to help in identifying the problem please let me know.

Please advise.
 
Hey I've seen this problem on a frame 5 a few times and found this symptom can have a few causes.

A) do you have a fuel stop valve in the system?
If so can you verify the correct operation of the fuel stop valve by enabling the Lube oil pump and picking up the coil responsible for the stop valve by forcing the logic for that coil? (verify the movement of the stop valve)

B) Also there might be a clogged filter located before the Moog valve preventing the moog from receiving adequate hydraulic press.(I have seen this problem before)

C) You can also check your igniters. Remove the igniters from the combustion cans set them on the floor, and stand clear. Go to the logic forcing screen, you will need to force your protective logic and also will need to force the logic for the igniter transformer. you should see a spark. If you see no spark then this a problem.

 
Have you just completed a maintenance outage (combustion inspection, hot gas path, major)?

What other Process- and Diagnostic Alarms are being annunciated during firing and acceleration? Precisely, what are all the alarms being annunciated say, 30 seconds before the trip, and up to 5 seconds after the trip?

What is the Liquid Fuel Supply pressure, upstream of the high-pressure Liquid Fuel Pump (from the Liquid Fuel Forwarding Pump(s)) during firing and acceleration and right up to the overtemperature trip? Most Liq. Fuel Forwarding systems have a pressure regulator, and if this isn't working properly then it can cause problems, also.

At what speed is the overtemperature occurring? Is it always the same speed?

Are you seeing steady flame indication in all flame detectors, and is the exhaust T/C feedback pretty stable and uniform after firing and during warm-up and acceleration, meaning that flame has been established uniformly in all cans?

Are you trending fuel flow vs. fuel flow reference? Is actual fuel flow (from flow divider feedback) matching fuel flow reference? If so, has someone tried to increase acceleration rate by increasing firing-, warm-up- or acceleration FSR values? If actual fuel flow is greater than reference, then it would seem that supply pressure is greater than Liq. Fuel Check Valve cracking pressure, causing fuel to flow even if the LFBV is fully open. This is a very dangerous situation.

Are you trending exhaust temperature during starting? Is it spiking when it trips or is it just increasing steadily until it reaches the trip setpoint? Have you compared the plot of exhaust temperature to a good plot? (This is why it's critical to make plots of all starts, because one can have records of good starts to compare against failed starts. One needs plots of all normal operations to compare to plots of abnormal operations.)

When you changed the servo (I'm getting closer to having enough money to make an offer to buy Moog!), did you verify the current polarity is correct for each coil? Servos are polarity-sensitive devices, and it's critical that the wiring and the polarity of the servo current being applied is correct (negative servo current closes the LFBV to increase fuel flow, and positive servo current opens the LFBV to decrease fuel flow). This is difficult for a Frame 6B (or Frame 5) as usually one can't see the valve stem, but it is no less important.

If you find one or two Liq. Fuel Check Valves which are opening at lower-than-normal pressures, there will usually be excessive white smoke coming from the exhaust (from unburned fuel) and usually high exhaust temperature spreads.

The fuel pressures you cite don't really make sense, because the discharge pressure of high-pressure Liquid Fuel Pump must be greater than the cracking pressure of the Liq. Fuel Check Valves. Are you citing the Liq. Fuel Fwd (Supply) Pump pressure or the high-pressure Liq. Fuel Pump discharge pressure? If it's the high-pressure Liq. Fuel Pump pressure, it's not clear how fuel can flow if the pressure is less than Liq. Fuel Check Valve cracking pressure (60 psig is less than approx. 100 psig check valve cracking pressure). If it's the Liq. Fuel Fwd. Pump (supply) pressure, then it would seem that there is some kind of restriction on the Fwd. Pump discharge such that when fuel starts flowing the pressure drops.

What are the Liq. Fuel Filter differential pressures? There is usually at least a low-pressure Liq. Fuel Filter (upstream of the high-pressure Liq. Fuel Pump) and sometimes there is one downstream of the high-pressure Liq. Fuel Pump. Also, there is usually a strainer upstream of the Liq. Fuel Fwd. Pump, with a differential pressure gauge. Have you checked/changed/cleaned the strainer recently?

If the unit fired but would not accelerate, it would seem there isn't sufficient fuel to allow the unit to accelerate (plugged Fwd. Pump suction strainer; high filter differential pressure(s); etc.).

If the unit is firing and warming-up and begins to accelerate, and actual fuel flow tracks the fuel flow-rate reference and then suddenly the actual fuel flow increases and the unit trips on exhaust OT, then something is causing a sudden high fuel flow (sudden jump in supply pressure or sudden increase in fuel flow). If this is happening, what will usually be shown on the trend of actual fuel flow versus fuel flow reference, is that the actual increases and the reference is decreasing, sometimes to zero.
 
G
Thanks for your advise.

> A) Yes we have a fuel stop valve in the system.<
The operation of the valve is smooth.

> B) Also there might be a clogged filter <
We have cleaned the filter and replaced the Moog valve with a serviced one.

> C) You can also check your igniters. <
The igniters are working because we can establish flame on all cans and both flame detectors are indicating.

We have limited the fuel from the Fuel skid and now the machine is not tripping on high exhaust temperature. The problem is that the Turbine is not accelerating beyond 2700 RPM. CPD is less than 1 at 2700 RPM.

Is the CPD affecting the acceleration command?
What could be limiting the turbine from accelerating?

Please advise further.

Gunga Chea.
 
G
Thank you for your reply.

The turbine has been in service but it tripped on high exhaust when it was being started and we have been working on it to solve this problem.

There are no other alarms that come before or after the trip. We have been able to restrict the fuel pressure now and the high exhaust temperature has been contained for the time being. The problem now is the Turbine cannot accelerate beyond 2700 RPM. CPD is below 1 bar at 2700 RPM.

The Liquid Fuel Supply pressure, upstream of the high-pressure Liquid Fuel Pump is 8-9 Bars during firing. The fuel pressure regulator is there and it was removed and inspected and found to be working.

There is steady flame indication in all flame detectors, the exhaust T/C feedback is stable and uniform after firing and during warm-up and acceleration.

Acceleration FSR values were not changed and the actual fuel flow is lower than the FSR value during firing.

The exhaust trend is the same, it just increases steadily until it reaches the trip setpoint.

Servo valve polarity is correct. We service the Moog valves and store them for later use when need arises.

The fuel pressure is enough and the differential pressure indicators are normal.

The unit is firing and warming-up and begins to accelerate, but stalls at 2700 RPM.

What could be limiting the accleration?

 
CPD has nothing to do with acceleration.

If the unit won't accelerate it's because either there's not enough fuel, or there isn't a sufficient torque assist from the starting means, or both.

8-9 barg is *MUCH* too high for liq fuel supply pressure at the inlet to the high-pressure liquid fuel pump. Look at the Device Summary and see what the setting for the Liq. Fuel Fwd. Skid discharge pressure should be; it *MUST* be less than Liq. Fuel Check Valve cracking pressure or fuel will flow uncontrolled into the combustors through the check valves, which is what appeared has happened before.

A Frame 6 starting means usually runs up to about 60% speed, so it should be providing some torque assist up to about 60% at which time the turbine shaft speed will over-run the torque converter output shaft speed and the clutch should disengage the torque converter output from the turbine shaft.

You haven't told us if the starting means is diesel engine or an electric motor or some kind of expander turbine or?

Also, we don't know if the torque converter has some kind of "spoiling" solenoid for reducing or cutting off the torque output at some speed.

You have not provided any of the information requested from trends. The Mark V operator interface has several trending programs, including one called 'Real Time Plot'.
 
On a machine running on Liq. Fuel, FSR and fuel flow are not proportional during firing and acceleration. This is because the high-pressure liquid fuel pump is driven by the Accessory Gear which is driven by the turbine shaft, which is not at rated speed during firing and acceleration.

You say the liq fuel supply pressure regulator was removed and found to be working, but they are usually set for something around 4-4.5 barg, not 8-9 barg.

If you are doing something to "restrict" liq fuel pressure to prevent the exhaust overtemperature, than are you also restricting liq fuel flow (flow-rate) at the same time? What is it that you are doing to "restrict" liq fuel supply pressure if it's not being done with the liq fuel supply pressure regulator?

It's *critical* to this issue to know what the liq fuel flow-rate is with respect to the liq fuel flow-rate reference. Usually, the liq fuel flow-rate reference signal name is FQR or FQROUT and the liq fuel flow feedback is FQL or FQLM or FQLM1. But, the servo-valve output regulator compares the flow-rate reference to the actual flow-rate and adjusts the servo-output current to try to make the actual flow-rate equal to the flow-rate reference (128 times per second!).

Unfortunately, most Frame 6s don't have LVDTs on the LFBV so it's not possible to see what the LFBV valve position is with respect to flow-rate.

But, you can monitor the servo-valve output current (usually signal name FAL) and while it's only updated at a rate of 4 times per second even though the servo current might be changing at the rate of 128 times per second, it would be the best indicator of whether or not the Mark V is actually trying to increase the actual liq fuel flow-rate to match the liq fuel flow-rate reference.

If the value of FAL exceeds for each processor exceeds *approximately* -4.0% (for a TMR panel; we also don't know if the unit has a SIMPLEX or TMR Mark V control panel!), then the Mark V is probably trying to increase the amount of fuel to match the flow-rate reference.

The Mark V will not increase fuel flow-rate if the actual average exhaust temperature (TTXM) reaches the exhaust temperature control limit (TTRX). So, if the Mark V is trying to accelerate the unit in conjunction with a torque assist from the starting means and the actual exhaust temperature (TTXM) reaches the exhaust temperature control limit (TTRX), it will not try to increase the fuel any further. And the speed will not increase.

So, if there is not enough torque assist from the starting means and the actual exhaust temperature has reached the "maximum" (TTRX), then it will not try to increase fuel, because the hot gas path components can be damaged if exhaust temperature goes any higher.

I think you might be fighting more than one problem. The first was that liq fuel supply pressure was too high and uncontrolled liq fuel flow was going through the bypass valve and through the liq fuel check valves into the combustors.

Now that you seem to have the liq fuel supply pressure situation under control, there may be a problem with the torque assist from the starting means. If this is being done through a torque converter, it may be that the torque converter is failing. Some torque converters have a strainer in the suction line which requires periodic cleaning and inspection. Some torque converters have spoiling solenoids (usually device name 20TU-2) which can fail and cause problems. Some torque converters have pressurization solenoids (usually device name 20TU-1) which can also fail and cause problems.

Some starting means employ a diesel engine, and if the diesel is not producing sufficient torque (because of plugged air- or fuel filters) or because of a dirty injectors or a misadjusted fuel rack, then they won't provide the assist necessary to help accelerate the unit.

If the diesel is working properly (rated speed; fuel rack not at maximum), then it's likely the torque converter is not working properly.

If the starting means employs an electric motor, one can monitor the current being drawn by the electric motor during starting and acceleration and compare it to current drawn during known good starts. If the current is less than during known good starts, then it's likely the torque converter is not working properly.

Lastly, do not increase fuel flow-rate to try to overcome a problem with the starting means torque assist. The unit will suffer damage to hot gas path components as a result.

So, you need to look at the required liq fuel flow-rate versus the actual liq fuel flow-rate and the actual exhaust temperature versus the exhaust temperature control reference (limit). If the actual liq fuel flow-rate is equal to the required, and the actual exhaust temperature is equal to the limit, then there is not enough torque coming from the starting means.

If the actual liq fuel flow-rate is less than the reference, and the actual exhaust temperature is less than the limit, then there is something restricting the liq fuel flow-rate.

I have been to several sites where filter differential pressure gauges all read zero (0), and when the filters were inspected, they were found to be ruptured because the differential had gone so high they had ruptured and there was filtration occurring. I have also been to sites where the differential pressure gauges weren't properly connected to the filters; sometimes the sensing lines were not connected, and other times the manual isolation valves were not in the proper positions.
 
I have seen the same problem for 2 6B GAS TURBINE that had completed a maintenance outage.

why the gas turbine speed keep at 2700. Because that the gas turbine is in the exhaust temperature control, and the FSRT will limit the fuel flow. there is enough fuel to increase the speed of gas turbine, the gas turbine will stay at 2700 rpm.

And i suggest that you can don't select auto mode to start the gas turbine,select fire at begin. after estabishing the fire, keep in fire mode and for about 5 minutes, change to auto to accelerate the gas turbine.

If you have completed a maintenance outage, I think that problem caused of the combustion componet.
 
O
Gunga,

You have got a lot of useful replies from the community. Probably you are too busy to reply back. While you are troubleshooting, we will not give up and overflowing you with our best advices! After this, you own us one!

I will have to guess that the your GT is equipped with diesel starting motor (maybe Detroit Diesel).If this is the case, there are two solenoids for acceleration and deceleration of the speed (20DAR1 and 20DCS or sometimes called 20DAR2). These solenoids are pulse controlled and sometimes there are issues related to this control. If the HP speed is not increasing at 2700RPM which is about 55% HP speed. It seems that your GT is hanging between L14HA and L14HC speed settings. Assuming that the site specified control constants are not changed from the control system, you may try to push manually with your hand the piston of the governor control and monitor the HP speed.

In case of having an electrical starting motor, ensure that the 20TU is not passing. Malfunction of the 20TU will give you also possible troubles.

Docendo Discumus
 
Speed limiting at 2700 rpm

Considering that you have a Detroit diesel engine, I feel in your case, 20DAR i.e. the logic L20DAR1 would be normal since the machine is already accelerating to 2700 rpm(@ 55% speed) but still you need to check the mechanical accelerator arm close to 20 DA-2. I suspect problem with the liquid fuel system. Normally for Fr-6B machines, TTXM values are at peak at around 50 % speed.

Have you checked the FQL1 value at 2700 rpm? I mean it can be due to flow divider problems as well. Check and compare these to your base values. A flow divider that is getting stuck could give these kind of problems. Also check the TTXM & TTRXB values at this time.
Also check the FSRACC value
 
First i wish to thank you all for your response to this problem. I heve been so busy trouble shooting and also attending to our diesel engine no.6, 12,000 HRS service.

The information you all gave has been very useful in determining the actual cause, read below.

Our starting means is by Detroit Diesel Engine. Initially the pressure setting for the liquid fuel PRV was 4.5 bars when it was first removed, i tested it. It was later changed to 6 bars.

We changed the flow divider with a new one and the turbine ran for four days and the same problem occured again. The Flow divider was eliminated as the problem.

Since 19th April 2009 the turbine operated for two weeks. We had a compressor failure and the turbine unloaded to zero load. The compressor blades were damaged and a new compressor was fitted.

During firing, after the compressor job the diesel engine was emitting black smoke from its exhaust and during acceleration the smoke increased.

Then the alarm "DIESEL ENGINE FAILURE TO BREAK AWAY TURBINE" operated. On inspecting we found the governor actuating arm was not in position. It was fitted back properly and the diesel engine started. The diesel engine accelerated and disengaged at 2500 rpm but the turbine managed to reach full speed and was loaded. It has been running non stop since that day.( 11th may 2009)

A diesel engine specialist is on sight and we are preparing to have the diesel engine inspected and serviced.
I will give a report after the inspection and service.

Below is data i gathered from the Mark V (Simplex Control) today 13th May 2009.

LOAD = 31.4 MW

FUEL PUMP DISC. PRESS. = 8.2 BARS

FUEL SKID DISC. PRESS = 6.5 BARS

LIQ FUEL PRESS. AFTER LP FILTERS = 6.2 BARS

TTRX = 544 DEG C

ACTUAL EXH TEMP = 544 DEG C

LIQ FUEL REF = 65.15

LIQ FUEL FLOW = 62.05 **

FSR = 64.83

LIQ_SERVO_CURRENT = -19.99%

Below find data for alarm and trip for 23rd April 2009.

STATUS FIELD - WARMING UP

FSR_CONTROL - TEMPERATURE

TTRX - 558 DEG C

ACTUAL EXH TEMP = 574 DEG C

TNH - 27.15 %

TNR - 100.30 %

TNH_RPM - 1380 RPM

MSG_FLD2 - SEQ IN PROGRESS

FQL1 - 7.64 %

DIESEL ENG SPEED - 2017 RPM

19:05:20 EXHAUST TEMPERATURE HIGH ALARM
19:05:23 EXHAUST OVERTEMPERATURE TRIP

Below find data after the trip.

TTRX - 558 DEG C

ACTUAL EXH TEMP = 581 DEG C

TNH_RPM - 1410 RPM

FQL1 - 0.00 %

DIESEL ENG SPEED - 2010 RPM

I hope the information is adequate but i can add more if need be.

I am looking forward to give more information on the outcome of the detroit diesel engine inspection.

Thank you very much.

Gunga Chea.



 
This is a truly distressing response, on several levels.

First, if by PRV you are referring to the pressure regulating valve on the liquid fuel forwarding system and you have adjusted that to 6 barg, then the supply pressure of the forwarding system is greater than the liquid fuel check valve pressure, which means that fuel can flow uncontrolled into the combustors if the unit has a liquid fuel bypass valve.

If the unit does not have a liquid fuel bypass valve and uses a variable displacement liquid fuel pump, and you have set the pressure per the Device Summary/Piping Schematic drawing, then that's okay--but we don't know enough about the liquid fuel system(s) at your site.

Second, from the data you supplied the liquid fuel flow-rate reference is greater than the liquid fuel flow-rate feedback. That is very, well, odd.

Third, and probably related, the servo valve output current is -19.99%--which is way out of tolerance for a typical GE-design heavy duty gas turbine servo valve. So, either it's not a typical valve or something is really wrong with the servo or the servo-valve output or the null bias current setting or the regulator gain.

Fourth, the exhaust temperature is equal to the exhaust temperature reference. Which is at it should be for a unit operating at Base Load, but with an actual fuel flow less than reference and a servo-valve output "less" than normal (more negative), this is just strange.

Fifth, I hope by compressor failure, you're not talking about the gas turbine axial compressor. Failing an axial compressor is pretty serious; even failing an atomizing air compressor is pretty serious. In any case, it seems unusual, and it's confusing.

Sixth, by diesel engine no. 6 at 12000 hrs, I presume you're talking about a diesel engine driving a generator, not the Detroit Diesel starting means diesel engine. Again, it's confusing.

So, I'm dropping out of this one. There's just too much I don't understand about the unit and it's auxiliaries and it's confusing to read the information. I can't add any more to the discussion without understanding and that's just not possible on a forum like this. There's just too much that's unclear and conflicting and we can't see the drawings or the unit configuration.

Glad to hear the unit's running okay. Best of luck.
 
Top