Exhaust Temperature Spread MS5002D Gas Turbine


Thread Starter


Dear All,

We have a frame5 GT, D model (MS5002D ) controlled with Mark V and used to drive 3 BCL Centrifugal compressors (HP Train).

The GT is a 12 combustion cans and running with Natural Fuel Gas; Right now, we have a slight exhaust temperature spread (âT= 51°C) close to the allowable spread (Trip at 55°C).

Hot and Cold Spots are located at the TCs number 4 and 10

Kindly someone help me to identify the cans causing the troubles on the TCs 4 & 10. I need the Swirl Angle of this machine. Does someone know it?

One more thing; We already checked the TCs and replaced with new ones the TCâs 4 & 10. The hot and cold spots stay at the same TC position either when TNL increase or decrease.

Thanks in advance & Regards

Exhaust temperature spreads (which are the indicators of combustion troubles) are the result of:

1) too little fuel getting into one or more combustors;

2)too much air getting into one or more combustors;

3) too much fuel getting into one or more combustors.

The causes are listed in order from most common to least common.

Most exhaust temperature spreads are caused by "cold spots"--too much air or too little fuel. It's rare that more fuel can get into a combustor (or combustors) than into other combustors--but it does happen (usually when operating on liquid fuel when there is a problem with the liquid fuel flow divider).

If the highest exhaust temperature T/C reading is approximately equal to the other three or for highest exhaust T/C readings (or the average reading), and the coldest exhaust T/C reading is much lower than the average reading then the spread is the result of a cold spot caused by little fuel or too much air getting into one or more combustors.

Exhaust temperature spreads are determined to be excessive when, for example, the difference between the highest- and the lowest exhaust T/C reading <b>AND</b> the difference between the highest- and the second lowest exhaust T/C readings exceed setpoints <b>AND</b> the highest- and lowest and second lowest exhaust T/C readings are adjacent (next to each other).

So, the condition you describe where the highest and lowest readings are separated by six exhaust T/C locations is not a problem. Even if the difference between then is high--they are not adjacent. And, a true combustion problem would be indicated by two or more of the lowest exhaust T/Cs being next to each other (adjacent) for conditions 1 and -2, or by two or three of the abnormally high exhaust T/Cs being next to each other. Your post is not clear on this subject. And, again, just because the magnitude of the highest and lowest exhaust T/C readings is at or near the allowable that does NOT mean a true combustion problem exists--or that the turbine is about to trip.

When monitoring any temperature (or pressure, or any variable)--one of the most important things to consider is how fast is the variable changing. If it's increasing or decreasing very fast, then there is likely a problem. If it jumps significantly and suddenly then instrumentation should be investigated first. If it's increasing or decreasing slowly but continually, then there is time for investigation and resolution.

The subject of exhaust spreads has been covered MANY times before on control.com. If you have a Mark V turbine control system you can create a Demand or User-Defined Display of all the relevant T/C values.

The incoming exhaust T/Cs are numbered, with rotation, and named TTXD1_nn, where 'nn' is a number, say from 1 to 18. These values are then sorted into an array (a list of values) from highest to lowest, named TTXD2_nn, where 'nn' in this case is usually 0 to 17 (eighteen values). At the same times, the locations of the highest-to-lowest exhaust T/Cs are also stored in another array named JXD1_nn (where 'nn' is usually 0-17). The algorithm then looks at the values <b>and</b> their locations to determine if a true exhaust temperature spread (combustion trouble) condition occurs by setting six logic signals: L60SP1 through L60SP6. These six logic signals are then used in three or four rungs of logic which look at magnitude <b>AND</b> adjacency and annunciate an alarm or trip.

So, even if the magnitude of the difference between the highest and lowest exceeds the allowable (TTXSPL) the a 'Combustion Trouble' alarm and an 'Exhaust Temperature Spread Trip' won't be annunciated <b>UNLESS</b> they are adjacent.

I'm <i>not</i> saying the condition you are describing isn't an indication of a problem that shouldn't be investigated and resolved. I am saying that <b>based on the information provided</b> a true combustion trouble problem (too little fuel in one or more combustors; too much air in one or more combustors; too much fuel in one or more combustors) does not exist <b>at this time</b> and a turbine trip is not imminent.

I am also suspect about a true combustion problem because an exhaust temperature spread almost always moves (with rotation) when loading a unit. An exhaust temperature spread that doesn't move during loading and unloading is usually a problem with either the thermocouple(s), the radiation shields or something blocking the radiation shields (again, causing a cold spot), improper insertion depth of the T/C in the radiation shield, or a problem with the wiring/circuit from the exhaust T/C all the way back to the Speedtronic (sometimes a problem can exist in intermediate junction boxes between the exhaust and the Speedtronic turbine control panel).

I would imagine that a swirl angle chart for a MS5002D does exist--but you would do best to obtain it from GE Oil & Gas (the former Nuovo Pignone). I would also imagine that because the machine is a two-shaft machine with variable second-stage nozzles that tracking an exhaust temperature spread back to a particular combustor or combustors is going to be difficult, if not impossible.

If the unit is running on liquid fuel, it's relatively simple to track a combustion trouble using the manual selector valve and gauge at the liquid fuel flow divider. Any combustor fuel pressure that's 10% more or 10% less than the others is suspect--doesn't tell you exactly what the problem is, just which combustor to start looking at.

Please write back to let us know how you fare with your troubleshooting.
First many thanks for the quick responses MANY MANY THANKS for the clarifications and for all what you are doing on this forum :)

Here are below the TCâs readings from MarkV when loading the GT (100% TNH & 24 °C ambient conditions) <pre>
CPD (bars) 7.77 8.19 8.55
CTD(°C) 307 313 312
IGV 83.9 83.7 83.4
NOZZLE 12.4 6.3 3.5

%TNL 84 95 102
TTXD1_1 416 456 516
TTXD1_2 418 455 511
TTXD1_3 422 464 525
TTXD1_4 424 469 536
TTXD1_5 422 469 531
TTXD1_6 416 462 523
TTXD1_7 418 462 521
TTXD1_8 409 463 517
TTXD1_9 391 444 513
TTXD1_10 388 420 485
TTXD1_11 409 442 486
TTXD1_12 413 456 515
TTXD1_13 415 458 520
T Average 414 457 516</pre>
With 24°C ambient Temperature we reached the 102% TNL. A couple weeks ago with 28°C (and we are in the North Africa Desert), the GT is under temperature control at 93% TNL and no way to load it anymore. Usually, the frame 5D is a robust TG and reaches easily 105%TNL even with hot summer.

Regarding TCs reading, Is there a Combustion Trouble ?

Right Now; the Pulse Jet Air Filter (Self-cleaning) is working all time under âP= 10,5 mbars (alarm at 12 and Trip HighHigh at 15 mbars). Does this filter âP generate a poor or non-homogeneous combustion?

What about when Cooling air passes through the hot gas path (bad Liner sealing or crack on the exhaust diffuser,..) causing hot/cold Spots ? What should be the TCs readings in this case?

Again many thanks for all what you are doing..
For the problem exhaust temperature t/c #4&10, hotspot is because of excess fuel is going and burning at particular point and fuel air ratio is to be checked.

However air intake filter differential filter self cleaning has to be done.
Combustion chamber lining problem

My first inclination is to say there is a combustion problem because of the cold spot around exh T/C #10, but I hesitate because: 1) I have no personal experience with MS5002D machines and their swirl patterns, and, 2) because the highest exh T/C is not close (adjacent) to the lowest T/Cs.

In my experience with other GE-design heavy duty gas turbines a high inlet filter differential will make an existing spread worse by a few degrees but will not--by itself--cause a combustion problem in one or more combustors. Combustion problems and the resultant high exhaust temperature spreads, again, caused by one or more of the reasons listed above.

The high inlet filter differential can lead to ruptured inlet filters releasing large amount of dirt into the axial compressor. Or, the opening of implosion doors which will allow unfiltered air into the axial compressor. Or a shutdown or trip, depending on how the machine is configured. It's not a condition to be overlooked or taken lightly.

I recommend monitoring the exhaust T/Cs over time and if the spreads increase for similar operating conditions then some investigation should be performed.

I have seen insulation from exhaust duct panels and flex seals block exhaust T/C radiation shield openings and cause cold spots such as your values indicate. Simply replacing the exhaust T/Cs from the outside won't dislodge the insulation.

Finally, you haven't said when this problem started or how long it has existed or how quickly the spread has increased to the current state. all of this can be very helpful to know when planning a course of action to investigate and resolve the conditions.

Please write back to let us know how you fare in resolving the problems.!
Dear CSA

Regarding Air Filters change out, the GT OEM recommends 18 months and the Filters maker recommends 2 years of use and 3 years for storage. And regarding the end user (We) experience, the strategy followed is to change them out every 5 years (usually during MI) with keeping an eye on them (Pulse Jet system, monitoring diff pressure, PMâ¦). Right now, we never had any incident or trouble related to this strategy.
My questions: What's your opinion about this strategy? and what around the world GT users are doing?

And what about manual cleaning of the air filters when pressure diff is little bit higher to extend filters life?

One more time... Thanks a lot..

SOOO MANY questions.

I've been to sites where the filters had to be replaced every six months, and others where they lasted for a decade or longer. It's <b>NOT</b> about time---it's about site conditions (how the turbine is operated; how much humidity is present; what the nature of the dirt/contaminants is (hydrocarbon vapors; cement dust; road fumes; the type of dust; etc.)). Oh, and by the way, operating and site conditions do change over time.

I've been to sites with multiple turbines and the site management is whinging madly and screaming about why the inlet filters on one machine are experiencing a higher dp than the others. And every day his car drives by the machine with the higher dp--on a dirt road, which is heavily traveled by lorries and other vehicles, while the other machines are not located near a road.

Trying to compare site inlet filter experiences is like trying to compare hazelnuts and apples.

If the cost of labour will allow for manual cleaning, then I guess it would be okay. But when I've seen it tried the labour did not use any care and a lot of the filter elements were damaged--which was, of course, the OEM's problem for not providing better quality filters (it actually was learned the site had purchased and installed non-OEM filters!).

Sounds like you've been quite fortunate up till now.

How is inlet filter cleanliness related to the control system?
Dear CSA

Sorry to be so long for replying.. I'm just back on the plant site.

About you inquiry: "How is inlet filter cleanliness related to the control system?".

The response from our control engineer regarding your question (and what I understood) is as follow.The DP in the plenum, relative to ambient, will increase as contaminants build up on the surface of the filter media. When DP reaches the preset limit (12mbars), a pressure switch activates the timer/sequencer board and the filters pulse-cleaning starts until the DP is drooped to the preset limit. The sequencer is linked to Mark V control system and on its screen is viewed the Filters DP reading and setting (High alarm and H.H.Trip). The machine will trip at 15mbars.

Hope I replied to your inquiry..

About the Combustion Trouble.. It looks worse than last month.. Yesterday; "COMBUSTION TROUL"shown on MarkV monitor when TNL reached 96%. No way to shut down the GT is saying the management even for deep investigations.

Right now the machine is running under 93%TNL and Next inspection which is a MI is due for the second quarter of 2015. What do you think about keeping GT under 93% TNL until the MI? What's the worst thing could happen?

Again many thanks and regards;

Continuing to run the unit with a high exhaust differential does not bode well for long-term parts life. But, if they need the machine to drive the compressors then sometimes the loss of parts life is worth the revenue--and that's what Managers get paid to do: analyze risk versus revenue and make the decision about what to do (or not do, as the case may be).

My main reason for asking how inlet filter cleanliness was related to control systems was that this is a controls-related forum. Several of my colleagues--and some other contributors to control.com--have asked my why we spend so much time on GE-design heavy duty gas turbine machine hardware-related issues. I generally respond that control systems have to interact with physical devices and field instruments to control and protect the turbine and auxiliaries. But, I do tend to agree, in general, that some questions are less relevant to control systems than others.

I do recognize that there really isn't any other open, public forum where these kinds of more general questions related to GE-design heavy duty gas turbines and auxiliaries can be asked and answered.

It's kind a quandary, don't you think?
Pretty well you are going to do some serious and expensive damage to this machine if you keep running it as it is. I know it can be difficult to convince management to allow a shut down but sometimes financial arguments can be best. Inform them of the potential damage in writing and get them to agree that they understand the implications of keeping running in this condition.

If you run at reduced TNR, you are just avoiding the issue, your spread is still high, just not high enough to trip.

Stand up for yourself!!
Dear All;
Here I'm writing back to let you know what's ongoing to solve this problem...

About <b>Swirl angle Charts:</b> I made a formal request to the GT OEM for the GT we have and the response I received is that the Swirl Charts are confidential and <b>not sharable with customers</b>. Does this make sense? I could not understand why it's NOT sharable with the machine Owner?. Swirl Chart is a great troubleshooting tool and it helps to localize the chambers having troubles. I made another request at higher level this time through management; I hope I would succeed this time!

Back to our topic: We stopped the GT and we did deep investigations.

<b>First</b>, we found that the Air filters Pulse jet auto cleaner was not working properly (may be at about 20% of its efficiency). So we fixed it and we cleaned cartridges filters (some cartridges were a little bit too dirty so we replaced them).

Note: Do you know the Air Filters Pulse Jet Cleaner is more efficient when activated by plant compressed air and GT turned off (filters are cleaned in place, no need to dismantle them).

<b>Second</b>, with instrumentation engineers we did double check and nothing found wrong (wiring, connections, T/C readings, etc..); We also opened the exhaust trapdoor to access to the exhaust diffuser and we checked the T/C radiation shield. All Fuel Strainers were checked and found clean.

<b>Third and this point I would like to share with you</b>; When Examining the machine history I found that the combustion troubles have always existed on this machine since its installation on 1996. And people here decided and <b>bypassed</b> the COMBUSTION ALARM.
What I also found that during 2 of the 3 Major Overhauls done on this machine, cracks were found on the exhaust diffuser (upper half) and if it's the case this time that means cooling air is mixing with hot gas near the suspected T/C (9 and 10).

Right now, the GT is still used at partial load (95% TNL) to avoid any Combustion Alarm. And the spread seems to be the same (no increase).

My concern is if the <b>main reason of the cold spot</b> (which causing the exhaust spread) is due to a crack on the diffuser is it safe to leave the GT running until next MI (end of this year)?

Dear CSA;
I confirm all what you wrote below..

>In my experience with other GE-design heavy duty gas
>turbines a high inlet filter differential will make an
>existing spread worse by a few degrees but will not--by
>itself--cause a combustion problem in one or more
>combustors. Combustion problems and the resultant high
>exhaust temperature spreads, again, caused by one or more of
>the reasons listed above.
>The high inlet filter differential can lead to ruptured
>inlet filters releasing large amount of dirt into the axial
>compressor. Or, the opening of implosion doors which will
>allow unfiltered air into the axial compressor. Or a
>shutdown or trip, depending on how the machine is
>configured. It's not a condition to be overlooked or taken

When Air Filters Auto-cleaner fixed (Low Inlet Differential Pressure) and GT started; the spreads and exhaust T/C readings stay at the same levels. That means <b>high inlet filter differential has NO (or just a little) impact on the "Combustion Trouble"</b>

For this concern, I printed out all the discussion happened on this topic, I highlighted each phrase you wrote and I did a list of actions to be performed. Another way to tell you thanks CSA!.

Best Regards

I'm very happy to hear back from you about this problem; thank you.

If the problem is the cracked exhaust diffuser--and it certainly sounds like it is--then I would say there's no problem running until the next MI.

Be prepared, though, for the "clever" people to tell you that they knew all along that the cold spot wasn't a combustion problem. (But you and I know otherwise!) They were taking a gamble based on their gut--and sometimes that works out. And they should just be man enough to admit that's what it was--a lucky gamble. But, they won't, and unfortunately that's human nature so we just let them have a pass on this. No harm; no foul.

As for the OEM not wishing to share the swirl chart, well, let's just say we'll remember this and there is karma in this world.