We are having GE Frame 9E (PG 9171E) gas turbine, operation controlled by MARK VI technology.M/C running on Natural Gas & operating at 70 MW.

All 24 nos of thermocouple working normal but one day we got alarm "Combustion Trouble" we checked flame intensity, Spread & fuel pressure but every parameter found normal then we checked all 24 nos of thermocouple and we found that TTXD_23 thermocouple showing less value then other thermpcouple (60 degree Celsius less). So after discussion we forced that thermocouple at equivalent value which other adjacent thermocouple have and like that we run for so many days. But that day as Shutdown was planned on that m/c and operator forgot to change the force value (TTXD_23) m/c tripped at 50 MW on high exhaust spread.

I want to know following things in this regard..

1. What to do when any thermocouple got open or showing less value then other thermocouple.

2. It is possible to link that faulty thermocouple value with nearby thermocouple (IN JB or MARK VI back plane)

3. Can we force the value at lowest value then all the other remaining thermocouple have.

4. Possible to do any kind of work on that faulty thermocouple? If yes then how?

5. precaution to be taken to avoid such trip in future.

 

If you had "Combustion Trouble" followed by "High Spread Trip" it sounds like you have a genuine high combustion Spread. 60 Deg. C low on one thermocouple does not sound like a thermocouple fault, when that thermocouple was showing low, what temperatures were you experiencing on the thermocouples adjacent to it? If you really have one bad thermocouple you will get an "Exhaust Thermocouple Trouble Alarm" Usually also showing the thermocouple No. Temporarily, if you can't shutdown, you can try jumpering one adjacent thermocouple output to the faulty thermocouple input. Only temporarily until you can plan a shutdown and only one thermocouple. "forcing" is not really recommended.

 

These are my answers:

1. If only 1no TC is reading low, this can be looped to a good one in another quadrant.

2. I don't think TC value can be logic forced.

3. You cannot do any work on the faulty TC cause of heat.

4. Looping better done at the JB.

 

Forcing thermocouple inputs is really a very bad practice and is to be discouraged at any time. It's really unfortunate that GE made that possible with the Mark VI and many operators have learned, just as this original poster learned, that forcing them and then forgetting to unforce them--or, worse, finding when they do unforce them that they have failed low or high!--usually leads to an unplanned trip (which is self-induced).

Spreads are almost never caused by the exh T/Cs or the control system, and using the control system to mask a spread problem usually leads to other unforeseen problems (just as happened in this case).

The original poster has not provided enough information for us to be able to say what might have happened.

And looping to another T/C in another quadrant is nearly as bad as forcing the input. If looping is deemed appropriate or necessary one should loop to an adjacent T/C in the same quadrant so as to try to maintain some semblance of protection with the Combustion Monitor.

 

In most cases where you find one thermocouple reading lower than others, the possibility is that the thermocouple is unable to reach its saturation point. That is you need to check first the calibration span before concluding to loop it.