HI

recently one of our Frame6 machine after Fuel nozzle replacement and 1 st nozzle replacement spread is going high on Baseload SP1 at 62degC.

the replaced fuel nozzles are brand new GE nozzles.

So we suspected of Bleed valve Passing and we took the temperatures of Bleed valve upstream and down stream of all GTs for comparision. this particular GT is showing 55degC and 88 degC where as the other GT is showing 130deg C and 142deg C. we also found water driping from flexible joint after the bleed valve in this particular GT.
Can some help us in

1. What could be the reason for High spread?

2. Why this particular GT is showing such a low temperature at Bleed Valve upstream and Down Stream?

3. What could be the reason for water dripping?

Regarding water dripping, due to slight passing of Bleed valve, due to nozzle effect is condensate forming near the valve?

 

You didn't say what fuel is being burned when you are experiencing the high exhaust temperature spread.

It's very difficult, and time-consuming, for us to try to write something that would cover gas and liquid fuels.

And, this is likely *not* a control system problem. Unless there is some reason to suspect that the exhaust T/Cs or the T/C inputs to the turbine control system (which you also didn't describe), then it's most likely a mechanical problem. Either the fuel nozzles aren't properly flow-matched, they're not properly installed, there is some kind of blockage of fuel or a check valve problem (if the problem is being experienced on liquid fuel), or there is some problem with the installation of the first-stage turbine nozzles, the transition pieces, the side seals (which can be a big problem on Frame 6Bs), or the transition pieces or the combustion liners are flowing too much air (cracks or improper installation). Improperly installed cross-fire tubes can also cause problems, but they should be glowing red hot if that's the case.

But, we don't have enough information, and it's highly unlikely that this is a control system problem. The control system is just telling you that there is some kind of problem with fuel flows to the individual combustors, or with the combustion cans and/or the transition pieces, or the side seals, or the installation of the first stage turbine nozzles (which, if improperly installed could allow compressor discharge air to pass directly to turbine).

So, prove there is no problem with the Exhaust T/C and the exh T/C input circuits to the turbine control panel, and then hand the issue back to the Mechanical Department (who likely handed it to you to begin with!).

And, when posting for help, please try to remember that we aren't at your site and don't have any information about the unit at your site or how it's being operated (what fuel(s) is(are) being burned; what the control system is; etc.). So, the more information you can provide, the more concise the response you receive will be.

Having said that this is likely NOT a control system problem, there have been cases where people working in the exhaust duct that have "played" with the exhaust T/C tips in the radiation tubes, bending them so that they are not in the center of the radiation tube and don't accurately measure the exhaust temperatures. Of, if now exhaust T/Cs were installed, sometimes they are not properly installed and either don't fully protrude into the radiation tube opening, or touch the metal at the end of the radiation tube and produce erroneous readings. So, this is worth a check when eliminating the control system as the source of the exhaust temperature spread.

 

I realized I didn't address the bleed valve questions you asked; sorry.

I would guess the temperature differential you are seeing from one machine to the other is the result of the length of time each unit has been running. If the unit which was just refurbished had only been running for several hours and the other machine had been running for several weeks, then the temperatures would have had more time to stabilize than on the recently started machine.

If a bleed valve is passing (leaking) then it will cause a spread if the discharge piping in the exhaust is directing the bleed air onto one or more of the exhaust T/Cs. Usually, the bleed valve discharge piping in the exhaust has a "diffuser" over the end(s) to prevent this from happening. But, it's conceivable that it could be happening; stranger things have happened.

Many Frame 6B units have their compressor bleed valves mounted outside the turbine compartment, on the roof of the turbine compartment. Just as many have their compressor bleed valves mounted inside the turbine compartment. And we don't know where yours are located.

I would have a difficult time believing that there would be enough of a pressure drop across a leaking compressor bleed valve to cause enough of a pressure drop to cause ambient humidity to condense, but, as was said before, stranger things have happened.

I've seen water from a leaking flame aft support leg cooling water jacket spraying on adjacent piping and causing it to appear as if water was leaking from the piping (which did not have any water flowing in it!). The spray was so fine it wasn't visible, but once it struck the pipe and "collected" it was visible.

When the turbine is running at rated speed, the compressor bleed valve should be closed. And the temperature on either side of the bleed valve should be relatively the same because there should be no flow. One would think that if one of the valves was passing, that the temperature upstream would be relatively high, because of the temperature of the extraction air from the compressor. And the temperature downstream of the valve would be slightly lower, because of the nozzle effect you cite. But both temperatures would likely be higher than normal because of the fact that air is flowing when there should be no air flowing through a closed valve.

Using an infrared temperature measuring device can yield very misleading temperatures. If there is a heat source close to where the measurement is being taken that is radiating heat, the device could be picking up false indications. It's certainly odd that the downstream temperature is higher across the valve you are suspecting of passing. But, we aren't there and we can't know how the measurements are being taken and if they are being taken at exactly the angle and the same locations on all the valves. And if the ambient conditions are the same, including exposure to sunlight or any breezes or air flows; there are just too many variables.

You could run a quick test by closing the manual isolation valve upstream of the compressor bleed valve you suspect is passing and monitoring the exhaust temperatures. If the spread goes down, that's a good indication the valve is bad. (You shouldn't leave the manual valve closed for very long; if the unit were to trip, that valve needs to be open!)

It's conceivable that water vapor in the exhaust is condensing in the piping immediately after starting and leaking from the expansion joint, but that doesn't seem likely if the valves are on the roof of the compartment. For there to be exhaust in the piping to be condensing, that would indicate there is some kind of "backflow" from the exhaust into the piping downstream of the bleed valve, and you are suspecting the valve is leaking from the compressor extraction which is at a much higher pressure than the exhaust. So, this is odd. If there was a leak in the expansion joint AND the bleed valve was *not* leaking, then exhaust could be flowing into the discharge piping and condensing.

Again, we just don't have enough information about the unit at your site and its configuration. And, it's not likely that this is a problem with the control system; a leaking compressor bleed valve is not a control system problem.

 

We are facing High spread on both the fuels Gas as well as Liquid fuel.
sorry for not mentioning it.

 

GT- FRAME6BEVK1360 - Bleed valves are inside the Turbine compartment.

We are facing problem of high spread on both Liquid as well as Gas fuel. We have measured the Bleed valve temperatures by using Temp Gun. we are not suspecting thermocouples because the thermocouples which are at high and low are shifting on Load variation.

Can any one give me a forum address where i can get ideas on mechanical side.

TY

 

HI CSM

TY for your some valuable information.

i got some more data may be by which u can help us a little more. We brought the machine to 4MW on Gas fuel and took thermocouples readings and opened the Generator breaker (you might be knowing that after generator breaker opens the Bleed valve will open) at FSNL (after Gen Breaker opening). we found particularly TTXD1_3, _4, _17 and _18 came down by 20degC and _10 & _11 came down by 10deg C. rest of the thermocouples around 4 to 6degC.

i am mentioning the baseload exhaust thermocouples reading have a look.

TTXD1_1 - 573, _2 - 554, _3 - 593, _4 - 541, _5 - 567, _6 - 558, _7 - 553, _8 - 579, _9 - 556, _10 - 572, _11 - 564, _12 - 579, _13 - 551, _14 - 567, _15 - 559, _16 - 578, _17 - 566, _18 - 570

waiting for ur valuable ideas
TY

 

chaitu,

You don't have to open the breaker to open the bleed valves; you can force open the bleed valves at low load (using Logic Forcing) without harming the machine; it will just not be as efficient because of the loss of mass flow and CPD.

I can't tell if you are trying to say that the leaking bleed valves are causing a hot spot in the exhaust or a cold spot. In my estimation, if one or both bleed valves were leaking that would cause a decrease in CPD (magnitude dependent on amount of leaking). Also, <b>depending on the location of the bleed valve discharge in the exhaust duct and whether or not there was any diffuser in place on the discharge pipe(s)</b> any nearby exhaust T/Cs which the bleed valve discharge might impact would be <b>lower</b> than the other T/Cs not impacted by the bleed valve discharge.

I also believe that if the bleed valve(s) were leaking and the cooler air from the bleed valve discharge was impacting on one or more exhaust T/Cs that the cold spot would <b>not</b> move with load.

As a unit is loaded, the mass flow through the machine increases as the IGVs open and as the fuel flow-rate increases. Because of the increase in mass flow, a cold spot caused by a plugged fuel nozzle orifice or a cracked liner or bad side seal will appear to move, with rotation. This is called "swirl angle", and it varies from machine to machine and configuration to configuration.

You have said that the spread is moving with load, so I would rule out the leaking bleed valve as the source. I believe that the extraction port in the compressor casing for the compressor bleed valves is a circumferential opening around the compressor casing, and for many other reasons as well, that it won't cause a spread in one or two combustors that will move with load. It will just decrease CPD which will decrease total mass flow (depending on the amount of leaking).

Nobody wants to shut the machine down to look for the cause of the high spread. Everybody wants to believe that it's just the control system not working properly, or that the control system can be manipulated to make the spread decrease or go away.

It was suggested that you close the manual bleed valve upstream of the compressor bleed valve to see if that had any effect on the spread. Did you do that? If so, what effect did it have?

It's happened many times here on control.com, that problems get blamed on the control system and it's subsequently learned that the problem was mechanical and was correctly being reported by the control system.

Have you found a possible cause for the water dripping from the bleed valve expansion joint? Has the dripping stopped or is it continuing?

The clincher is that if you are facing the same spread problem, in roughly the same areas on both fuels, that the problem is likely something amiss with the mechanical reassembly. I believe the transition piece side seals on Frame 6Bs can be very tricky to get properly installed and if compressor discharge air is leaking through one or more of those side seals then it's going to cause a cold spot, which will move with load.

I know this is not what you wanted to hear, but them's the breaks.

Good troubleshooting involves making reasoned decisions about what might be the most likely causes of a problem, then working to prove or disprove each of those reasons one by one until the problem is resolved. One can't will a solution to occur, or force a problem to be attributed to some condition. Logically work through the various possible reasons; unfortunately, some of them are mechanical and will require disassembly.

If you can obtain a swirl angle chart for your machine (no; I don't have one) then you can make and educated guess (and that's still all it is: an educated guess) about where the problem might be (which combustor or transition piece). But, this is how heavy duty gas turbines work and must be troubleshot.

Please write back and let us know what you find!

 

first I'd like to thank mr Chatu our plant engineer.

bleed valve checked and confirmed close but spreed problem still high. thermo couple checked and replaced what I feel and support CSA opinion about may be mechanical problem from liner our transient pieces. our view here at 31 MW spread is 30 c and next Combustion inspection will be after three month. do you think any problem to continue like this until CI after three month? and please I waiting for more reply
thank you

 

The responsibility for making these kinds of decisions rests with those on site who have knowledge of all of the variables and intangibles that go into operating a plant.

If the turbine is controlled by a Speedtronic turbine control panel, Mark IV or newer, then there is a signal called TTXSPL, the calculated allowable exhaust temperature spread limit. This number really represents the upper limit of exhaust temperature spreads that the machine can reasonably tolerate without causing damage to the machine over time. We have no idea what that number currently is for the machine at your site (nor has anyone ever provided the type of turbine control system at your site).

So, someone at your site needs to take into account all the information at your site about the need for power (and/or steam) and the economics of shutdown (loss of production; manpower for disassembly; budgets; upper management's concerns; etc.) and make this kind of decision. That's what supervision and management get paid to do: evaluate and manage risk and make the best decisions with the available information.

 

Hi CSA
TY for ur reply

There is no manual isolation for Bleed v/v's. so we couldn't do the test which you have suggested. But i forgot to mention in my earlier reply that the Bleed v/v which is having less temp reading is fine when the GCB opened & Bleed valve opened it has raised to 140degC. so it is clear that this particular Bleed v/v is not passing. During the same shutdown we checked all Thermocouples and changed the TTXD_3 and 4.

we have swapped fuel Nozzle 2 and 6 too but no major change in spread.

Regarding water dripping we couldn't find any exact reason for this and this water dripping is not persisting at present. May be the reason could be the exhaust air getting condensed in the line. i think it might be happening during high humid condition's (this i need to check).

I am giving you the least and highest TTXD which we collected at various Load on Gas Fuel.

20MW - TTXD1_16 & 5 - 390 (least)
TTXD1_12 - 397 (Highest)
25MW - TTXD1_11 - 433 (least)
TTXD1_18 - 456 (Highest)
31MW - TTXD1_6 - 500 (least)
TTXD1_18 - 524 (Highest)
35MW - TTXD1_4 - 531 (least)
TTXD1_3 - 571 (Highest)

37.7MW(Baseload) - TTXD1_4 - 545 (least)
TTXD1_3 - 595 (Highest)
TTXSPL - 71.7
TTXSP1 - 49.5

Our control panel is SPEEDTRONIC Mark V control panel and we have TTXSPL, can we rely on this signal, i mean if u see in Base Load (Gas Fuel) there was 20degC difference between SP1 and SPL.Even a small variation in Load SPL keeps varying.

Once again TY for ur Info, Ill update if we find any thing further.

 

Chaitu,

Exhaust temperature spreads are not always related to fuel nozzles. If you are experiencing spreads on both fuels, and the exhaust temperature profile is basically the same on both fuels, it's not likely the fuel nozzles that are the problem. It's something else like a cracked combustion liner, or a failing or bad hula skirt seal (the "spring" seal on the end of the combustion liner that fits inside the transition piece), or problems with the transition piece side seals where they fit into the first stage turbine nozzle ring, or a cracked transition piece or an enlarged hole in the combustion liner.

There can be lots of reasons for spreads, not just fuel nozzles. Again, I don't have a swirl angle chart, and there's just not enough information available to make an educated guess about where the problem might be.

You should be able to rely on the TTXSPL number presuming the axial compressor discharge thermocouples are reading properly and the other inputs to the block are working correctly (use Rung Display to check the inputs and the output).

The only thing that concerns me from the data that you provided is that at Base Load, the highest and the lowest exhaust T/C readings are adjacent. That's kind of not good, but the spread is still approximately 20 deg C below the calculated limit (presuming the inputs to the algorithm are working correctly).

Be aware also that TTXSPL is a number that is dynamic during load changes. In other words, when load is raised or lowered, TTXSPL goes up to a maximum and then decreases over time when the load change is complete. Same for when Base Load is selected, TTXSPL is driven to a maximum value and when the unit reaches CPD-biased exhaust temperature control TTXSPL starts decreasing down to the calculated value. So, be careful when looking at TTXSPL to ensure that it is relatively stable and that you have been sitting at a stable load for a couple of minutes or so before capturing a reading.

Best of luck with your issue(s). The "mechanical department" is the hardest group to convince that problems like this are *not* controls problems. The loss of generation and the costs involved with disassembling the machine to find the problem(s) are difficult to "justify" when that complicated control system with all those wires and LEDs is usually the problem. The "mechanical department" is very quick to jump on something when it's obviously broken and fix it, but with something like this, well, it's just too difficult to say for certain exactly what the problem is and where it is and so they just resist taking any action as long as possible.

Unfortunately, sometimes things break catastrophically if left unattended for long enough. And that's just part of operating machinery: taking "calculated" risks. We just have to hope those making the calculations are using the right information and have the right experience. Because that's what good supervisors and managers are supposed to have: experience that allows them to make calculated risks with the most favorable outcomes.

Don't fret; this happens, literally, tens of times every year, in locations all over the world. And there are thousands of gas turbines in service around the world (of every make and model).

And, controls always get the blame in the beginning and has to prove it's not the control system before anyone will take any further action.

It's unfortunate that there isn't more and better instrumentation for these kinds of situations, but, honestly, the costs involved are very high for the relatively few cases such as this. It doesn't make it easier when one of the few cases is occurring at your site, but, economics are economics. And, engineering is a series of compromises, including economic compromises. If the units are instrumented "to the hilt" then the salespeople would have a difficult time selling them. Even though when problems like this occur and people whine, "Why isn't there instrumentation to help us with this problem?" they wouldn't pay for it in the original purchase price!

So, cheer up! The spreads are high, and higher than you're accustomed to, but they don't seem excessive from the limited data you've provided (presuming stable and correctly working TTXSPL).

 

thank you very much for your great analysis, what we observe now that spread come down to 42 at base load which is normal also at 28 Mw become 36 c. I can say without the primary reading it looks now normal we will be in touch with this utilise site again.

thank you