We are operating GE Frame 9E open cycle Gas Turbine for a peak load. The GT now is experiencing temperature spread problem caused by thermocouple at location P,Q,R,S location (lower temperature than the rest) that caused high temp diff. This triggered the alarm (about 140 deg C diff between lowest and highest temp of thermocouple raedings). The GT cannot go to base load as it may trip, so it is now limited to about 85MW.Suspecting it could be the burners problem we replaced about 5 burners corresponding to the thermocoule position (referring to the GE swirl chart). The problem solved for about 2 weeks but since has recurred. Now we are suspecting the GCV as it shows opening at 55% at about 85 MW loading (normally it should be 50%). There is also observation of a brownish colored smoke coming out during the GT loading at 85 MW, which could explain of incomplete combustion. Hope you can give some pointers. Thank You.

 

Usually, if a "cold spot" is detected by several T/Cs as you're saying it's genuinely a combustion problem. And, you say you've replaced some fuel nozzles (not "burners"; you have burners on your stove, not in a 120 MW gas turbine; Texans have burners in their gas turbines, but nobody else does, and let's keep it that way--Texans are spatial, er, uh, special!) and the problem "went away for a couple of weeks."

You didn't say by how much the spread decreased when you replaced the fuel nozzles--was it just enough to clear the alarm or did the spread drop by 20 or 30 deg C or more before it started creedping up again?

Did the temperature spread start creeping up or did it jump up when the trouble began again?

What about the first time the spread was detected/alarmed--did it happen all of a sudden or gradually?

Cold spots are not always related to plugged fuel nozzles. Cold spots can be indicative of excessive air flow into the combustor, or transition pieces (the part that takes the round end of the combustor and flattens it out into an arc that's 1/14th of a circle (in the case of a Frame 9E) to flow into the first stage nozzle). A cracked liner, a broken transition piece seal (sometimes referred to as a "hula skirt"), a cracked transition piece, a bad first stage seal--any number of problems which are not easily detected without disassembling the unit. Which requires a shutdown, which means no power generation, which means no income....

You didn't mention if you saw any evidence of plugging on the fuel nozzles you removed; was there any? Plugged fuel nozzle passages cause reduced fuel flow into the combustor which causes cold spots--but cold spots, again, are NOT always caused by plugged fuel nozzles.

Smoking is not a good thing, and this author has only seem smoking on DLN heavy-duty units and on liquid fueled units. A light brown haze can almost ALWAYS be detected on units which are running liquid fuel (distillate or heavy fuel). An even lighter haze (a "NOx haze") can almost always be detected on DLN combustor-equipped units when they are NOT operating in Premix combustion mode (especially "transferless" units when they are transitioning from Lean-Lean to Premix!), and on conventional combustor-equipped units which are not using any kind of Wet Low NOx abatement (water- or steam injection for emissions reduction).

It has been this author's experience that, quite often, nobody really sees the "NOx haze"--until there's a combustion problem, and then EVERYBODY sees it, and sees it for the FIRST time (when it's been there all along).

A light brown haze or tint on a conventional combustor-equipped unit without water- or steam injection is probably normal. Incomplete combustion on a heavy duty gas turbine is pretty difficult because of the amount of excess air flowing through a gas turbine--as much as three times the amount of air required for combustion of the fuel. (The excess air enters the combustion liner of conventional combustors through slots and holes in the body of the liner.)

It would be highly unlikely for the GCV (Gas Control Valve) to cause combustion spread problems. The GCV discharges into a ring manifold, to which all the fuel nozzles are connected. It would be very difficult for a GCV problem to manifest itself in one or two combustors.

Depending on the answers to the above questions, it sounds like you may have some problem(s) other than fuel nozzle problems. The fuel nozzles provided with GE-design heavy duty gas turbines are usually flow-matched--but only to approximately within 10% of each other. Many owners have specifications which call for flow-matched nozzles SETS of 5% or less--and these owners usually find they have less spread problems over time.

How long has it been since the last maintenance outage? Have the units been operated in peaking application for a long period of time? The thermal stresses of starting and stopping a peaking application unit do cause more problems with combustion hardware.

markvguy

 

Hi Mr. Markvguy;

Thank You for your response. BTW I am not a Texan. I am about halfway around the globe from Texas.

The GT is a daily start/stop peak loader since 1994 and the last inspection (HGPI) was carried out about 6 months ago.

The first time it happened, the GT tripped on temperature spread (setting of alarm at 180 deg C and tripping at 210 deg C). The GT load vs. spread during the loading up was recorded as following; 70MW/60 deg C spread, 80MW/86 deg C, 90 MW/131 deg C, 100 MW/195 deg C (GT tripped at about this point). I can conclude that it was a sudden occurrence, because on the previous day running, the GT recorded a normal load vs. temp spread readings: 70 MW/22 deg C, 80MW/37 deg C, 90MW/53 deg C and 100 MW/42 deg C.

The funny thing was (well, it is always funny doing troubleshooting) right after the tripping, we stroked the GCV and restart the GT until at base load and the problem disappeared (recording 70 MW/21 deg C, 80MW/34 deg C, 90MW/46 deg C and 100 MW/36 deg C -this temp spread is normal by our operating standard). That was when we discounted the combustor problem as the caused of the spread, and thought maybe the GCV opening caused the imbalance in fuel distribution.

The problem came back on the next start-up at GT running to 80 MW (spread as high as 121 deg C with exactly the same cold spot area observed as previous finding). We shutdown the GT and replaced two suspected burner, err I mean nozzle, and run back the GT to base load with the spread return to normal (70 MW/33 deg C, 80MW/44 deg C, 90MW/58 deg C and 106 MW/52 deg C.) This normal condition lasted us for another 6 operating days.

The third time it happened, we carried out fuel change over to distillate and the observation was the same as of the fuel gas (high temperature spread). We replaced 5 more nozzles corresponding to the lower temp (cold spot) at thermocouple O,P,Q,R,S. There were no significant findings on the replaced nozzle. No evidence of plugged nozzles from visual inspection. After the nozzle replacement the GT could run with normal spread (only about 40 deg C at highest temp difference at base load) but it only lasted for about 7 operating days.

Which brought us back to the GCV. We replaced the servo valve, and the power piston & servo block of the GCV, ran the auto calibrate and restarted the GT. The GT this time could go to base load with normal temp spread at base load (also about 40 deg C).

The next day the problem came back. Now tell me why I shouldn’t go crazy?

If it is a real combustor problem (transition piece, cross fire tube, liner), why is the problem coming on and off especially after we attended to the GCV and the nozzle? It is like a flip-flop/switch on-off things, no gradual deterioration which could suggest it is caused by the static mechanical parts. The GT was showing a perfectly normal temp spread (when it is OK) or extremely high temperature range (with yellowish smoke) when it gave the problem. Our inspection on the TP, CFT and liner during the nozzle replacement also revealed there was no abnormal condition.

You mentioned about excessive air flow to the combustor. Does each combustor receive air from individual route or is it supposed to be common path for all combustors? How about the purge valve or the bleed valve system. Could they have an effect on the temp spread?

Based on the above information, I hope you can have a clearer picture and give more suggestions on how to solve the problem (since my boss is on my back).

Thanks for your help. I appreciate it.

 

Hi markvguy,

Thank You for your response to my question. I am reposting again my earlier reply which was not posted here due to some error on this website (I guess).

The GT is a daily start/stop peak loader since 1994 and the last inspection (HGPI) was carried out about 6 months ago.

The first time it happened, the GT tripped on temperature spread (setting of alarm at 180 deg C and tripping at 210 deg C). The GT load vs. spread during the loading up was recorded as following: 70MW/60 deg C spread, 80MW/86 deg C, 90 MW/131 deg C, 100 MW/195 deg C (GT tripped at about this point). I can conclude that it was a sudden occurrence as on the previous day running, the GT recorded a normal load vs. temp spread readings: 70 MW/22 deg C, 80MW/37 deg C, 90MW/53 deg C and 100 MW/42 deg C.

The funny thing was (well, it is always funny doing trouble shooting) right after the tripping, we stroked the GCV and restart the GT until at base load and the problem disappeared (recording 70 MW/21 deg C, 80MW/34 deg C, 90MW/46 deg C and 100 MW/36 deg C - this temp spread is normal by our operating standard). That was when we discounted the combustor problem as the caused of the spread, and thought maybe the GCV opening caused the imbalance in fuel distribution.

The problem came back on the next start-up at GT running to 80 MW (spread as high as 121 deg C with exactly the same cold spot observed as previous finding). We shutdown the GT and replaced two suspected burner, err I mean nozzle, and run back the GT to base load with the spread return to normal (70 MW/33 deg C, 80MW/44 deg C, 90MW/58 deg C and 106 MW/52 deg C). This normal condition lasted us for another 6 operating days.

The third time it happened, we carried out fuel change over to distillate and the observation was the same as of the fuel gas (high temperature spread). We replaced 5 more nozzles corresponding to the lower temp (cold spot) at thermocouple O,P,Q,R,S. There were no significant findings on the replaced nozzle. No evidence of plugged nozzles from visual inspection. After the nozzle replacement the GT could run with normal spread (only about 40 deg C highest temp difference at base load) but it only lasted for about 7 operating days.

Which brought us back to the GCV. We replaced the servo valve, and the power piston & servo block of the GCV, ran the auto calibrate and restarted the GT. The GT this time could go to base load with normal temp spread at base load (also about 40 deg C).

The next day the problem came back. Now tell me why I shouldn’t go crazy?

If it is a real combustor problem (transition piece, cross fire tube, liner), why is the problem coming on and off especially after we attended to the GCV and the nozzle? It is like a flip-flop/switch on-off things, no gradual deterioration which could suggest to be the static mechanical parts. The GT was showing a perfectly normal temp spread (when it is OK) or extremely high temperature range (with yellowish smoke) when it gave the problem. Our inspection on the TP, CFT and liner during the nozzle replacement also revealed there was no abnormal condition.

You mentioned about excessive air flow to the combustor. Does each combustor receive air from individual route or is it supposed to be common path for all combustors? How about the purge valve or the bleed valve system. Could they have an effect on the temp spread?

Based on the above information, I hope you can have a clearer picture and give more suggestions on how to solve the problem. Thanks for your help. I appreciate it.

 

It certainly is odd that it's intermittent. Does the unit start/stop every day? Are there some days when it doesn't run? If so, do the spreads occur on days when the unit is started after being run the previous day or when it hasn't been run for two or three or more days?

Does this unit have Water- or Steam Injection for NOx emissions reduction? Is either running when the spread is present? Or Steam Injection for Power Augmentation?

Are the wheelspace temps higher or lower than before the inspection outage?

Are there any Diagnostic Alarms being annunciated? Especially when the spreads are being experienced?

The best news is you're not a Texan; there's too many of them outside the great (sic) state of Texas as it is, and there aren't any Frame 9s running in Texas so it wasn't presumed you were in or from Texas. Okay; some Texan might have used a reduction gear to couple a Frame 9 to the 60 Hz grid... They are known for having more dollars than sense (pun intended). And, do you know what's the best thing about being in the very middle of Texas? No matter which direction you choose to go, you're leaving! So, being halfway around the globe from Texas is almost as good as it gets.

The really interesting piece of information from your reply is that the same cold spot appeared when you ran on liquid fuel. Well, that's not exactly clear, but you said the same temperature spread appeared, so it's presumed that it was at the same location--which also says it's not related to the GCV.

The Gas Fuel Schematic Piping Diagram (if you have a GE-packaged unit; or the similar P&ID (Piping & Instrumentation Diagram) if the unit was packaged by someone other than GE) will show that the output of the GCV (Gas Control Valve) flows into a manifold. The manifold has fourteen (in the case of a Frame 9E) connections, one to each fuel nozzle. Each fuel nozzle has orifices to limit the flow of gas fuel into the combustors. It's not clear how a problem with the GCV could manifest itself in one or two combustors, always in the same location... especially if the same spread occurred on liquid fuel in the same location.

If the spread is combustion-related, it should be "moving" with load. Does the cold spot "shift" with load? Say, at low load it's at locations K, L, M, N, O, and then at high load it's at O, P, Q, R, S? It's believable that the differential will increase with load, but it should shift location with load also. If it doesn't, then it's a problem with those exhaust T/Cs, their radiation shields, or their terminations. Since they are not all connected to the same processor in the Mk V, it's probably not a problem with the TCQAs or the T/C TB.

Every combustor receives air from one area. The axial compressor discharges into a common area, and that flows into the fourteen combustion cans.

Which purge valves are you referring to? Gas Fuel or Liquid Fuel? Again, by looking at the Schematic Piping Diagrams or P&IDs you can see how the air flows through the purge valves to the combustors. It's not likely that these would be contributing to this problem--especially since it always seems to be at the same place (if the posts are understood correctly).

Presuming the fuel nozzles which were replaced were from the area which would cause the spread in the location reported, and the spread has always occured in the same location--especially if it doesn't move with load--then the problem may be something blocking the affected exhaust T/Cs or a problem with the Exhaust T/Cs or the radiation shields of those T/Cs.

Have you gone into the exhaust area and examined ALL the exhaust T/Cs to be sure the tips are properly located in the radiation shields? Generally, the T/C tips are NOT to be touching any metal of the radiation shield. There are usually instructions for installing T/Cs in radiation shields on the Exh. T/C Arrangement drawings in the instruction manuals. It's presumed these T/Cs are at the 11-12 o'clock position, at the top of the exhaust. What kind of exhaust do you have--straight up into the stack or sideways into an HRSG (Heat Recovery Steam Generator, a boiler)?

Are all the Exhaust T/C sheaths properly clamped down in the Load Compartment? (They should not be loosely routed in the Load Compartment between the radiation shield and the four JBs.) Have you been in the Load Compartment to look at the T/Cs? Have you opened the four Exhaust T/C JBs on the two sides of the Load Compartment and examined the terminations to be sure they are correct and tight?

Again, since the same spread (especially if it's always at the same location and doesn't move with load) appeared when you transferred to liquid fuel, it's either something wrong with the combustion hardware or first-stage nozzle seals or something wrong with the exhaust T/Cs or the wrong fuel nozzles were replaced. If the unit has Wet Low NOx (Water- or Steam Injection for NOx emissions abatement) there may be a problem with the injection nozzles or the Water Injection Shear Valve (if so equipped). Just because the hardware was fine six months ago, doesn't mean it's good today. Also, whenever things are disassembled and then reassembled there's a chance for problems, even for things to get broken during reassembly.

Have you considered just swapping fuel nozzles between combustion cans--say moving two or three to another location and then putting those two or three nozzles in their place? This would be to see if the spread moved, or decreased, or got worse.

markvguy

P.S. Remember: Bosses can be like diapers: All over your arse and full of shite!

 

I have the experience of exhaust temp high spread trip due to the N2 leaked in the gas fuel, and the GCV stroke is higher than normal operation.

 

Muggsy brings up a very real possibility--you have not said if your unit uses nitrogen (or some other inert gas) to purge the natural gas lines/piping when the unit isn't running; if so, it could be leaking into the gas line. As Muggsy says, this would explain the higher-than-expected GCV (Gas Control Valve) position because, for the same load, the GCV would have to open more to pass the same amount of natural gas because of the inert gas mixed with it.

It might also explain the intermittent nature of the problem. The nitrogen valve(s) leak more some times than others or don't seat as well sometimes as they do other times.

If the unit doesn't have nitrogen/inert gas purging of the gas lines/piping, can you tell us what the indicated swirl angle is from the swirl chart at the highest load at which the exhaust temperature spread is experienced?

Also, please tell us specifically which fuel nozzles were replaced--both times the fuel nozzles were replaced.

markvguy

 

Thank You guys for your response. As for Muggsy’s suggestion, unfortunately the GT does not use N2 (or any other inert gas) in its purging system. So we can eliminate that possibility.

The GT now is back to normal (normal temperature spread) on the last run-up which was successful (but we have not run the GT ever since due to no request from the dispatcher). The latest work that we did on the GT before the successful start are as below;

1. Replaced two thermocouples at the coldest spot
2. Swapping four other thermocouples' terminal (the affected with the healthy ones) at the Mark V panel terminals
3. Check (and clean) the speed sensors

As for the no.3 above, the reason we did was because we suspected the GCV abnormal opening during the problem was due to the speed sensor. ( Speed sensor input to FSRN which goes to FSR and FSR2 (to open GCV). I have much doubt this is the cause of the problem but when you have intermittent problem like this you tend to do whatever that cross your mind.

After that we ran the unit on gas to about 30 MW, c/o to liquid fuel and increase the load to base load. We reduce the load again to 30 MW c/o to gas and load up again to base load. The temperature spread recorded throughout the test was perfectly normal (<50 deg C at the highest).

If it can tell something, the above test proved the thermocouple is not a problem since the cold temperature recorded earlier did not follow to a new spot after we swapped the location. When we experienced the problem, the trending showed a parabolic-type curve at T/C location N, O, P, Q, R, S, T with the lowest spot at P and Q. (We replaced T/C P and R and swapped four adjacent T/C with other T/C located across).

Let me answer markvguy questions here;

1. The GT does not start/stop everyday. Maybe about 3-4 times a week (not necessarily consecutive days). The GT will run on average between 5-10 hours per start.

2. The spread happened any time at will (sometime on the next day after it was started OK)

3. No water or steam injection.

4. When the problem occur, no other alarm except for “Combustion trouble” alarm.

5. The cold spot remain at same location (does not shift to other T/C) throughout the loading from minimum until maximum load. The differential of temp however increase with load increase.

6. The exhaust system is a kind of straight up into the stack (no HRSG).

7. The lowest cold spot is located at about 10-11 o’clock position if you look at the circumference from the compressor side.

I would like to tell something about the nozzles. On the first occasion we changed the nozzle (no. 4 and 5) we noticed there was a slight indication of liquid fuel mark at the tip of the nozzle no. 5. The nozzles sequence arrangement is counterclockwise (nozzle no.1 at 12 o’clock position).

On the second occasion we replaced nozzle 6,7,8,9,10 and noticed the same on nozzle no.8.

On the third occasion we replaced nozzle 3,11,12,13,14 and noticed the same on nozzle no.8.(We did some nozzle swapping as well at this stage)

Where did the oil come from since we were always on gas firing? FYI we also replaced the 14 check valves (liquid fuel) to eliminate the possibility of back flow.

In our last start which was successful, was it because we did c/o to liquid fuel first that may have “flushed” the nozzle and provide a clear passage for the gas when we c/o back to gas firing after that?

I hope the above will enlighten you further. I am typing this from home (it is weekend break-I need some break from the "diapers"). I will get more data when I get back to work.

You always have something up on those poor Texans

 

Are there any Diagnostic Alarms present before the T/C spread occurs? Any Diagnostic Alarms at all?

Does the exhaust spread occur at exactly the same spot while running on liquid fuel and gas fuel????

One should be really careful when moving Exhaust T/Cs (wiring) at the Mark V panel.... They are terminated in a specific sequence to divide the T/Cs evenly between the three processors and such that one T/C is connected to <R>, the next to <S>, the next to <T>, the next to <R>, the next to <S>, and so on. Swapping the T/Cs at the Mk V TB can cause the Combustion Monitor function to miss a true spread because hottest and coldest T/Cs might be separated because of the swapping. It's advised to return the exhaust T/Cs (wiring) to their normal positions as soon as possible.

You really need to look at the Schematic Piping Diagrams for the Gas Fuel, Liquid Fuel, and Atomizing Air systems. Without the ability to place drawings in this forum it's very difficult to explain in writing how the fuel and purge air flows through the fuel nozzles. You also need to look at the cutaway views of the fuel nozzle.

When the unit is running on gas fuel, there should be a continual flow of purge air through the liquid fuel passage of the fuel nozzle (which is usually in the center of the fuel nozzle tip). The purge air comes from the compressor discharge through the Liquid Fuel Purge Valve and then through a purge check valve at each of the fuel nozzles. The purpose of the liquid fuel furge air is, initially, after a transfer from liquid fuel to gas fuel, to purge the liquid fuel passage of the fuel nozzle of liquid fuel so that it doesn't become carbonized and hard due to the heat, and after that to maintain a flow of cooling air through the liquid fuel nozzle passge. So, to answer your last question, the liquid fuel nozzles are purged when the transfer from liquid to gas fuel is complete (there should be no purge air flowing when running on liquid fuel).

There's such a small quantity of liquid fuel in the nozzle downstream of the liquid fuel purge air check valve that it's not likely it could be the source of the smoking (if the smoking persists for some time). And extra fuel would cause a hot spot--not a cold spot.

When the unit is running on liquid fuel, the liquid fuel purge valve is closed (actually it's a three-way valve and it's shifted to allow any liquid fuel which might leak through any of the purge check valves to flow through a drain known as the "tell-tale leak-off", so the an operator can check for leaking purge checks when running on liquid fuel). But, the flow of air to the liquid fuel purge check valves is shut off.

When the unit is running on liquid fuel, a valve in the atomizing air system closes to increase the atomizing air pressure ratio--and the flow of air through the atomizing air passage in the fuel nozzle (usually the "ring" around the liquid fuel cartridge in the center of the fuel nozzle. When the unit is running on gas fuel, the valve is open to decrease the atomizing air pressure ratio because atomizing air is not required for gas fuel operation--and to reduce the power consumed from the turbine shaft to drive the atomizing air compressor and make more available to the shaft output driving the generator (it's only kw, but every little bit helps the efficiency).

When the unit is running on liquid fuel, the gas fuel purge valves are open to prevent combustion gases from backflowing into the gas fuel passages of the fuel nozzles and into the gas fuel manifold around the compressor. Another reason for purging the gas fuel passages when running on liquid fuel is to provide some cooling for the gas fuel passages (the gas flows around the outer portion of the fuel nozzle tip, through the small passage which are almost perpendicular to the axis of the fuel nozzle and into the rectangular passages of the "swirl" portion of the fuel nozzle tip.

If the gas fuel purge valve limit switches are adjusted properly, if the valves don't close fully when they should while running on gas fuel they should indicate a problem. Also, there should be a pressure switch sensing the pressure between the two gas fuel purge valves which should generate an alarm if there is pressure between the valves which there would have to be if purge air was leaking into the gas fuel system. Further, gas fuel pressure has to be higher than compressor discharge pressure--so purge air can't leak into the gas fuel piping.

If you're experiencing liquid fuel "dribble" (carbonization) on the fuel nozzle tip, that's indicative of a purge check valve that's not flowing properly and purging the liquid fuel from the cartridge. But that would normally only cause a cold spot when running on liquid fuel (due to plugging of the liquid fuel nozzle caused by carbonization of the liquid fuel because it wasn't completely purged while running on gas fuel).

During the hot gas path inspection, was the upper half of the exhaust duct, above the #3 bearing, removed? If so, the exhaust T/Cs on the upper half of the echaust (where you are indicating the cold spot/spread is appearing) had to be disconnected and reconnected....

Also, this author has gone into exhaust ducts to check T/Cs after an outage prior to a re-start to find several of the T/C tips bent over inside the radiation shields--by "idle" maintenance workers not away of the importance of the position of the T/C tips in the radiation shields. Also, at the same time one or two exhaust T/C tips were found to be touching the end of the radiation shield--also not a good thing. It is believed there is supposed to be a .gap of approximately 0.250 inches between the T/C tip and the end of the radiation shield.

You can take a small heat source (such as a cigarette lighter) and wave it in each of the radiation shields while someone watches the exhaust T/C display to see some temperature increase. DO NOT put the flame directly on the T/C tip for any length of time! If you do this for each of the T/Cs in order around the exhaust and the indicated temperature increase on the display is indicated at the same position the heat is being applied, it's a quick check that the exhaust T/Cs were reconnected properly. (Any heat source will do; an electric heat gun can be used--you're not trying to sense a known temperature, just see a temperature rise on a particular T/C.)

But all of the above is still just "thinking out loud on html." You have said that the cold spot doesn't change with load--and it should if it's combustion-related. First, while the IGVs are not fully open, the flow through the machine is different than it is when the IGVs are full open. And the flow through the machine at part load with the IGVs full open is different than it is when the unit is a full load--hence the development of the swirl chart. So, if the cold spot is at the same location regardless of load (and regardless of which fuel the unit is being operated on), then there's something else wrong--and it's probably in the exhaust duct or with the exhaust T/Cs.

One thought could be if the compressor bleed valve which discharges closest to the upper half of the exhaust duct is leaking intermittently the flow of cooler air from the leaking compressor bleed valve could be "impinging" on the upper exhaust T/Cs. There is usually no limit switches on the compressor bleed valves to indicate they are fully closed (only fully open). This might also explain why the GCV (Gas Control Valve) is open more when the spread is high than when it's normal for the same load--because a portion of the compressor discharge is being "bled" off and the unit isn't as efficient as normal so more fuel is required for the same load. One could use a non-contact (infra-red) temperature sensor to check the piping downstream of the check valves when the spread is high to see of the valve(s) is(are) leaking. This would also explain why the spread is also intermittently present on liquid fuel and in the same area (if it is--is it???).

Because the spread doesn't move with load (swirl), it's likely not a combustion hardware problem. But that's the key here--the cold spot is not moving with load, it's always in the same spot even if you change/swap fuel nozzles. And, it may be in the same location regardless of which fuel the unit is running on. A problem with the combustion hardware would not likely be intermittent, and the spread caused by such a problem would change location with load.

markvguy

 

Hi Markvguy,

Today the unit has misbehaved again, after running fine and dandy for about a week. To answer your questions, yes the cold spots appear at the same t/c be it on liquid or gas fuel, and at whatever loading value. The only difference is that the spread becomes bigger at higher load, hence load derate. We are still confused on why there are traces of liquid fuel at the recently replaced nozzles, even when the unit was not running on liquid at all. We are positive that the forwarding pumps were dead (literally) and the line was none pressurized (even the flow divider showed 0 bar pressure). However, we noticed that the FSR2 value remained at 50% and 80 MW load, when the value was always 55-56% at 80 MW. We also found the IGV opening was at 86% (calibrated by GE TA recently during a HGPI+ recommissioning). Isn't it supposed to be 84%?

Care to shed some lights to these unenlightened?

 

The residue you are seeing doesn't have to be liquid fuel.... This author has seen oil carry-over in the gas fuel from gas compressors, as well as gasolines, hydraulic oil, and other refined- and unrefined liquids found entrained in gas fuel.

Is the unit at the end of the gas fuel supply line?

Does the unit have coalescing filters, or a "knockout drum" or a "cyclone separator" in the gas fuel supply line where you could look for oily liquids and have them analyzed?

Have you looked at the low point of the gas fuel manifold to see if there are any oily liquids there? Have you checked every low point in the gas fuel supply piping for oily liquids?

Are the nozzles where you find the oily residue close to where the line from the GCV (Gas Control Valve) enters the gas fuel manifold?

Now, there's always atomizing air flowing in the atomizing air lines even when the unit is running on gas fuel. Could there be lubricating oil entrained in the atomizing air? Perhaps there is a problem with atomizing air compressor and oil is collecting in the low points of the atomizing air piping, and when enough of it collects and the flow gets high enough a slug of oil goes through the piping and carbonizes on the fuel nozzle tips.

Either scenario could also explain the smoking you have observed....

When you replace the fuel nozzles do you see any evidence of oily residue on any of the flanges/piping? If so, which system?

What is the value of CSKGVMAX for your unit? That value should define the maximum opening while running. What is the value of CSGV (Control Stroke - Guide Vanes) while the unit is running? Were the IGVs recalibrated as part of some performance upgrade? Was the 86 DGA (DeGree Angle) measurement just the full, mechanical stroke value from the report?

Calibration of the GCV (which you say you've done) could account for some of the differences, as could ambient conditions, though 5% is a large amount to be attributed to ambient conditions alone. Without knowing the other operating conditions (IGV angle, ambient temperature and humidity, etc.) it's really difficult to comment on the FSR value difference being reported.

This author is running out of ideas here.

But the really strange thing for this author is that the low temp doesn't move with load or fuel. It could be a control system issue. This author has seen air conditioners blowing directly on the turbine control panel cause problems with T/C cold junction compensation. Some days it's not running and blowing on the turbine control panel; other days it is when you're experiencing the problem. This has happened on a couple of DLN combustor-equipped units and caused intermittent problems with emissions.

It's really difficult to troubleshoot a problem like this from afar, but at least it's not from Texas!

markvguy

 

It has recently come to this author's attention that while operating on gas fuel some operators have been able to detect compressor discharge pressure working back through the liquid fuel check valves, through the liquid fuel flow divider, around the high-pressure liquid fuel pump and into the high-pressure liquid fuel filter. The pressure builds up in the top of the high-pressure liquid fuel filter(s) and when the unit is shut down, the pressure is high enough to force liquid fuel past the check valve and it dribbles into the combustion can, causing coking in the fuel nozzle and would cause the residue you are seeing.

So, put a pressure gage on the vent connection of the high-pressure liquid fuel filters and monitor the pressure when running on gas fuel and especially while shutting down on gas fuel. If the pressure builds up to be nearly equal to CPD (Compressor Discharge Pressure) and then decreases when the unit shuts down, you can consider that you may have a liquid fuel check valve problem.

Another check you could make would be to trend or monitor the liquid fuel flow divider feedback (usually signal name FQL1 or FQLM or FQLM1) while shutting down on gas fuel. If you see flow while shutting down and you had pressure in the high pressure liquid fuel filter, it's probably a safe bet you have a liquid fuel check valve problem.

You could then remove the check valves and test them by reverse pressurizing them and checking for leakage.

You could also try switching check valves around and seeing if the dribbling/residue follows the check valve and if the spread moves with the check valve. But it's still odd that the low temperature never moves with load--it should change with IGV angle changes and load....

Please write back and let us know what you find!

markvguy

 

Hi MarkVguy,

Sorry for not updating you on this problem. We have solved the problem but the root cause is still not known. We opened the combustor and the findings were; Combustor liner no. 6 and 7 damaged (cracked and torn), cross fire tube 6-7,7-8 and 8-9 damaged and purge air piping blocked with oil sludge.

Regards.

 

Well, hope you guys will help a Texan. I have a GE 1962 year, Frame 5 Model E, 14 MV Gas Turbine that is having the same type problems you have been discussing since Feb. I just started this project in Oct, and we have been having problems all along.

First of all, the unit has only been run on diesel before we got it. We started on diesel and the problem is that we get high EGT spread (400 to 600 degrees) on startup from lightoff (930 rpm) through ignition and on up to around 3300 rpm. Then the EGT spread disappears and all readings are within 30 degrees. We haven't put a load on the unit yet, because we can't get it to run steady without this high EGT spread.

We put an ESD on the system that will shutdown the unit if the spread is over 400 degrees. We have checked all T/C's in the EGT steam, we have 10 cans and 12 EGT sensors. When the EGT spread occurs, two or three of the sensor will peg out at close to 1100 degrees on one side of the unit and the other side will be around 400 degrees.

We have changed,cleaned and replace fuel nozzles and have replaced all the check valve. On most starts the ESD occurs within 1 minute of ignition and at other times it will occur after 3 or 4 minutes. We have borescoped the cans, the crossover tubes, we have changed the flow divider and have changed all filters in the fuel system. We have moved T/C's at the j-box to see if the hot spot moves with the T/C and it does.

The only thing we have found that is out of the ordinary is that the compressor bleed valve on the lower right side does not close until the CDP reaches 100 psi, which it should close at 60 psi. Sometimes the startup will have all the cans full flame and then within a minute, cans on the cold side will flicker and flame out. This unit was taken apart in Florida and put back together in Texas.

A lot of wiring was not properly tagged or labeled. Once the unit reaches operating speed (4860 rpm), everything is fine. Any help would be greatly appreciated and emails to my email address would be the best contact. Thanks to one and all. Don't hold it against me for being from Texas. (HA)

 

What control system is being used on the unit?

The really interesting thing you have told us is that when you move "... T/Cs at the j-box to see if the hot spot moves and it does...." That would suggest a problem with the control system and/or wiring.

Usually, if one experiences cold spots during acceleration while operating on liquid fuel, there will be white smoke coming from the stack (white smoke is "vaporized" liquid fuel that is unburned). Are you seeing any of this during acceleration?

It's not uncommon for some units to experience loss of flame in one or more combustors during acceleration, especially if the fuel is cut back after firing (for "warm-up"). Usually as the unit approaches rated speed, all of the combustors will eventually light, and you have said this does happen.

The other interesting thing you have noted is that one of the compressor bleed valves isn't closing when it should, or, specifically that it is closing later than the other bleed valve. Does this unit have a 20CB-1 solenoid to close the bleed valves, usually at 95% speed or so? But, you also say you haven't loaded the unit or run it past 100% speed. Have you rebuilt/replaced the compressor bleed valve actuator that is closing so late?

In thinking about this, is it possible that the one compressor bleed valve is closing too early which is causing increased air flow through the unit which is causing flame-out(s)? Usually CPD doesn't build up until the unit gets close to rated speed.

1100 deg F is a common maximum exhaust temperature limit for CPD-biased exhaust temperature control. But you say that the 1100 deg F temperature follows exhaust T/Cs.

There have been problems with broken flow divider shafts causing problems, but you've replaced the flow divider.

You've got an interesting problem there in Texas.

 

The control system is a Allen-Bradley solid state system. We did find two T/Cs that were bad, replaced them and still have hot spots but not as before.

We have white smoke coming out during acceleration. We lose one or two cans during start up and eventually they will all light off, the problem is with the EGT spread, we can't run long enough or the can to light off. The spread is so high that the EGT Limiter on the unit will not let the fuel control valve open up and we are timed out by the computer.

We rechecked the CBV and one is closing at 50 psi and the other is closing at 100 psi.
Today we checked them again and both close at 50 psi.

We do not have any solenoid operating the CBV, it is strictly CDP straight off the transition duct.

We put another flow divider in yesterday and it started up okay, still have high EGT spread, but it wasn't way out of limits as usual. Thanks for the inputs.

 

I won't say it's *normal* to have high exhaust temperature spreads during start-up, but, when it does happen the spreads can be 400-600 deg F until all the cans finally light off. The Combustion Monitor in a Speedtronic turbine control system is not enabled until the unit is at rated speed.... In other words, GE doesn't trip their machines on high spreads during starting.

Now, since you're talking about a vintage Frame 5, one of the workhorses of the world, those things can usually take quite a beating and continue to produce power like nothing ever occurred. Of course, one doesn't want to unnecessarily do harm to the machine and it should theoretically be possible to tune the fuel during starting and acceleration to keep all the cans lit during acceleration.

What have you done in the way of adjusting fuel flow-rates to try to alleviate the problem? Or, have you been increasing fuel to try to overcome the loss of energy in the cans that lose flame in order to be able to sustain acceleration? (That might explain the 1100 deg F temps you've seen, if so.)

Sure seems like a couple of things are contributing to this problem.