Recently we had one of our gas turbine tripping on exhaust temperature trip though L86TXT was forced, the complete sequence of event is as below. Our's is frame 9 machine with mark v controls.

1.Exh.T/c no. 11 suddenly started giving fluctuating reading so it was decided to check the T/C at field junction box.(we have total 18 nos of T/C)

2.Prior to checking following tags were forced in Mark V
a>L86TXT - Exh.temp. high
b>L86TFB - Exh.T/C open
c>L30SPT- Spread trip

3.Field J.B. was opened and T/C no.11 was disconnected from the terminal strip and was checked for voltage, the reading was found to be fluctuating between 22 mV to 23 mV then resistance of T/C was also checked and it too was found fluctuating between 40 ohms to 50 ohms, as readings were erratic it was decided to make exh. T/C no.17 parallel with T/C.no.11.

4.T/C no.17 & T/C no.11 were made parallel in field JB and both became OK.

5.Prior to closure of JB by mistake the technician connected multimeter to terminals of T/C no.11 to check it's mV and as soon as it was connected machined tripped and on checking alarms on <I> station, high exhaust temperature alarm was first to appear and the reading of TTXM in trip log was also found to be more then 1100 Deg C thus it was concluded that machine had tripped on exhaust temperature high though L86Txt was forced.

6.On checking the multimeter, the selection of same was found to be on continuity checking instead of voltage selection which resulted in high temperature readings at exh. T/C no. 11 & 17 (both being parallel) which resulted in tripping of GT.

Now our query is why GT tripped on high exh. temperature though L86Txt was forced ? Will be a great help if you can share your knowledge/experience of any such incidence that might be in your knowledge or give your views on such an incidence.

 

You neglected to tell us if the Mark V (which is the only Speedtronic turbine control system to use <I>s) is a SIMPLEX or TMR system.

And I would *STRONGLY* advise you to ***NEVER, NEVER, NEVER, NOT EVER*** force L86TXT while the turbine is running. If there is a true exhaust overtemperature, the unit will disintegrate. It can happen in seconds with uncontrolled fuel flow, so if you think an operator can respond in time to protect the unit, you have made a huge error in judgment.

Just think about what led you to force L86TXT in the first place: A strong desire to not trip the turbine. In many parts of the world, operators, even if told to do so, will ask repeatedly if they should trip the turbine. Meaning, that if someone yells at them to "Trip the turbine!" they will ask, repeatedly usually, if they should trip the turbine. So, if an exhaust overtemperature was "detected" by the operator while the logic was forced or the turbine control panel he would still probably seek guidance from someone else about what to do even if previously told to trip the turbine if an overtemperature condition was detected while the logic was forced.

But, we really need to know if the Mark V is a TMR panel or a SIMPLEX panel.

 

Also, I neglected to ask: Precisely, what was the condition the unit tripped on? What was the Process Alarm indication that the unit was tripped for?

Please list the Alarm Drop Number and the Alarm Text message from the trip incident. Hopefully you printed a copy of the Trip History immediately after the trip, so you should be able to list all the Process Alarms at the time of the trip.

 

I would like to add some points
1. At the time of tripping Exhaust temperature recorded in Trip log was 1107 degC. As Exhaust temperature (TTXM ) is Sorted TTXD1_1 to TTXD18_18 to the array of descending order TTXD2_n then TC's less than 277.6 degC are deducted from TTXD2_2 and then Average of Remainder TC's are taken as TTXM. And as at the time of tripping of GT TTXM was recorded 1107degC then as per BBL all TC's should read @1100 degC ( If only TC11 & TC17 reads 1100degC ten average could have been @600degC something. so as per trip log all TC's could have gone in the range of 1100degC. How?

2.When GT tripped first out alarm was "Exhaust Temperature High" generated from L30TXA i.e. If TTXM is greater than TTRXB+16.7 degC or TTXM is greater than TTKOT2 i.e. 615 degC which ever is earlier. But this L30TXA is only Alarm. and Flollowed by this alarm "FLAME LOSS" alarm appered." Exhaust Over temperaure trip" alarm didn't appeared as l86TXT was forced to healthy state.

3.L86TFB "Thermocouple Open trip" was also forced in healthy state.

4.At the time of tripping FSR,FQL,FQLM1,DWATT all other parameters were normal as which were being monitored and also recorded in trip log.

5.First out Alarm was definately " Exhaust Temperature High" Second Alrm " FLAME LOSS" but "Exhaust Over Temperaure Trip" alrm did not appear ok will not appear as L86TXT was forced to healthy. Then on which logic GT trip command initiated?.Is there any internal block safety point of view because in Rung display only three outputs are generted are L30TXA, L86TXT, L86TFB. Then on what logic GT tripped on "FLAME LOSS"
GT is Frame6,MKV,TMR

Thank you.

 

In thinking about this problem, what probably occurred is the following: The average exhaust temperature went to such a high level because of the meter that the Mark V responded by reducing fuel so fast that the unit tripped on "Loss of Flame"--which is exactly the alarm condition which would be annunciated if the fuel were reduced very quickly in response to an exaggerated and abnormal increase in feedback, especially if was being operated at or near Base Load when the activity was taking place. Actually, I don't think even being at or near Base Load would have mattered in this case, since the exhaust temperature went so high the Mark V would have reduced fuel very quickly in response regardless of load.

I think the voltage from the meter's continuity function caused the T/C readings to go wildly out of range in the original post.

'Loss of Flame' is an alarm that is annunciated when flame is suddenly lost and no trip condition was active that should have caused flame to be lost. If the fuel were drastically reduced in response to a high exhaust temperature feedback, then the flame would be extinguished. If the overtemperature trip were forced and could not have tripped the fuel in advance of a loss of flame, then the loss of flame condition was the cause of the trip.

The original post was for a Frame 9 with Mark V (TMR or SIMPLEX was not specified). So, I'm a little confused about this post.

1. The Spread Monitor uses the *average* value after exclusion of low values; it doesn't force the values to be anything. You are confusing the Spread Monitor with Exhaust Temperature Control and Overtemperature Protection, which looks at the average value and has nothing to do with sorted arrays.

You ask how? If, as the originator said, the wrong meter scale was selected when the jumpered T/Cs were measured, then that could adversely affect the T/C input terminal board as a whole. What was done was not standard operating practice and the results could not have been predicted.

People seem to think that control systems should have built-in logic and protections to prevent abnormal operating and troubleshooting conditions from tripping the turbine. They don't, and if they did, they would complain about *that* preventing them from doing something else that wasn't recommended.

2. It would be necessary to examine the CSP in use at your site to understand why the turbine might have tripped under the conditions you describe it was being operated. It's just not possible to understand from what we have been told without being able to see the CSP and Trip History. The originator was asked if the unit was a SIMPLEX because the Mark V SIMPLEX has back-up exhaust overtemperature protection in <C> so if they had only forced the <R> value <C> could still have tripped the turbine.

3. Not recommended to run, especially at Base Load, with L86TXT and L86TFB forced to prevent tripping the turbine in the event of a problem for the reasons specified in my earlier response. But, it's your unit and you can operate it as you see fit; you pay the bills if something goes awry.

5. There is no hidden, internal safety to prevent against forcing L86TXT from tripping the turbine. Again, it would be necessary to see the CSP and Trip History for your unit to make specific comments. From what you describe, it doesn't seem like the unit should have tripped on high-high exhaust temperature, but we don't know what Process- and Diagnostic Alarms were present before or during the event.

But, I would also submit that the units (this one and the one in the original post) were being operated outside normal parameters with the logic signals described being forced and a mistake was made by the technician which had unanticipated results. How does one protect against this? How does one explain something that occurred when something abnormal was done? It was a risky operation, period. And, the fact that logics were forced to try to prevent tripping speaks loudly to just how risky it was.

I've seen this happen many times. Technicians tell their supervisors, "The unit shouldn't trip if we force this or that to do [something]," and when it does trip management is very unhappy and technicians are very upset. I think the originator described the events at his site pretty well and honestly, not that you haven't, but your argument is that the unit should not have tripped if we forced logics that are not recommended to be forced while running, especially at Base Load with no other data.

In the case of the original post, I can't say how the Mark V would respond to having two of the T/C inputs having voltage applied to them from the meter's continuity circuit. But I can say that I don't think the control system should be built and designed to prevent problems arising from questionable practices from causing turbine trips--which may have protected the turbine. I would rather the turbine trip when something like this happened than run and cause possible catastrophic damage. And if it tripped on loss of flame when the fuel was reduced because it thought the exhaust temperate was abnormally high, then I think the control system did exactly what it should have done to protect the turbine.

 

Well, either you guys have guts or you really don't know what the consequences could have been to your asset by forcing these primary protection signals! Let me tell you that, I have done similar forcing during commissioning of a FR9 in Malaysia. Again this was during commissioning, controlled environment and with assistance of competent E&I crew. Not sure were your plant is located but, hopefully you were not the person that has seen me doing this forcing and assumed that anybody can do it!

First lesson, never parallel TC's without understanding whether they are physically located at the same combustion chamber or not (number of faulty TC=>2)! Beside this you should understand that, paralleling TC's will never give you correct reading. Unless you would have disconnected the faulty one and jumpered the other TC terminal board.
Back to your question: Once you open a JB in the field and start checking things, you will introduce other problems such as this kind of "ghost" trips.

Remember that, the exhaust temp. is controlled by basically amount of fuel in the GT, which is the FSR parameter (CPD primary control). Due to the MKV is seeing nuisance "over firing" induced by your technician during trouble shooting activity, the FSR control will try to cut down the amount of fuel which might result in loss of flame.
Other possibility is that, your technician has touched and or shorted other instrument wires which are located in that particular JB. Sometimes a loose connection in other wires will cause this problem, once its cable is touched or pushed in order to close the JB cover.

Good Luck...Tempus Fugit...

 

Dear Mr.CSA
I understand its very clear as Temperature Control Setpoint was @570degC and TTXM measure shoot up to @1100 degC and resulted in drastic cutoff in fuel and GT tripped on " LOSS OF FLAME"
(It's so.. drastic in MKV speedtronic!! really beautiful!!!) 100% agree. No doubt at all.

From trip log also it is clear that GT tripeed on "LOSS OF FLAME"> GT is Frame6, MKV, TMR ( for information )

Thank You

 

First of all sorry for not being able to reply to CSA we too have concluded that as TTXM went to an abruptly high value which resulted in tripping of GT on flame loss. I would like to thank all who participated and helped us in finding the root cause as far as forcing of L86TXTxt is concerned, we will discuss it amongst our team and will take a decision on it because we as O&M personnel we always give priority to safe operations & safety of M/C.

Regards

 

Thanks to tclmarkv and DMello for the feedback. Others who read this post later will find this information useful. Feedback is the most important contribution.(c)

As A. Oztas says, if you are jumpering T/Cs (which is not really recommended, but commonly done, unfortunately), you need to do two things: choose a T/C in the same region of the exhaust as the failed or failing T/C, and completely disconnect the failed or failing T/C when jumpering another to its input terminals.

Failure to do the first can cause a catastrophic failure due to an undetected exhaust spread or can cause nuisance trips due to incorrect spread detection.

Failure to do the second can cause in intermittent failing T/C to improperly affect the readings of the two jumpered T/Cs.

If jumpering T/Cs is deemed "necessary" by supervision, then people need to be prepared for the possibility that a trip can occur. I've seen some very creative interposing T/C terminal block set-ups created and used for jumpering T/Cs, but even with one of these the possibility of inadvertently tripping the turbine can still exist, it just makes opening JBs unnecessary.

But, no matter: If exhaust T/C jumpering is going to be performed, the two factors above must be considered and followed to provide the most protection for the turbine and to prevent the possiblity of nuisance tripping.

Lots of supervisors seem to take great delight in "beating the system" by jumpering exhaust T/Cs. It can be useful in an emergency when power or steam production (for combined cycle applications) is absolutely necessary, but it's not without it's risks. Being a good turbine technician and supervisor involves risk management and good judgment; experience improves risk management, and judgment. And by understanding what happened in these scenarios you are building your experience and your judgment.

Experience generally means more compensation.

But remember: Good judgment comes from experience. Experience comes from bad judgment.

 

You neglected to tell us if the Mark V (which is the only Speedtronic turbine control system to use <I>s) is a SIMPLEX or TMR system.

And I would *STRONGLY* advise you to ***NEVER, NEVER, NEVER, NOT EVER*** force L86TXT while the turbine is running. If there is a true exhaust overtemperature, the unit will disintegrate. It can happen in seconds with uncontrolled fuel flow, so if you think an operator can respond in time to protect the unit, you have made a huge error in judgment.

Just think about what led you to force L86TXT in the first place: A strong desire to not trip the turbine. In many parts of the world, operators, even if told to do so, will ask repeatedly if they should trip the turbine. Meaning, that if someone yells at them to "Trip the turbine!" they will ask, repeatedly usually, if they should trip the turbine. So, if an exhaust overtemperature was "detected" by the operator while the logic was forced or the turbine control panel he would still probably seek guidance from someone else about what to do even if previously told to trip the turbine if an overtemperature condition was detected while the logic was forced.

But, we really need to know if the Mark V is a TMR panel or a SIMPLEX panel.

You neglected to tell us if the Mark V (which is the only Speedtronic turbine control system to use <I>s) is a SIMPLEX or TMR system.

And I would *STRONGLY* advise you to ***NEVER, NEVER, NEVER, NOT EVER*** force L86TXT while the turbine is running. If there is a true exhaust overtemperature, the unit will disintegrate. It can happen in seconds with uncontrolled fuel flow, so if you think an operator can respond in time to protect the unit, you have made a huge error in judgment.

Just think about what led you to force L86TXT in the first place: A strong desire to not trip the turbine. In many parts of the world, operators, even if told to do so, will ask repeatedly if they should trip the turbine. Meaning, that if someone yells at them to "Trip the turbine!" they will ask, repeatedly usually, if they should trip the turbine. So, if an exhaust overtemperature was "detected" by the operator while the logic was forced or the turbine control panel he would still probably seek guidance from someone else about what to do even if previously told to trip the turbine if an overtemperature condition was detected while the logic was forced.

But, we really need to know if the Mark V is a TMR panel or a SIMPLEX panel.

Dear CSA
when we have a failure in a exhaust TC (9FA) we do the exact same procedure that was describe in the post, and for do it in a "safe" condition we also force the L86TXT (to prevent a trip cause for a mistake of the technician)
Now after read your post i have a mayor concern about our procedures.
i know there is a trip for high exhaust temperature, but what is protected? is the gas turbine or the tubes of the HRSG
what could whappen is we have forced the L86 TXT and occurs a real high T°.

i hope you can read this

RATM

 

RATM,

Look--I get it. I do. It's a testosterone thing--men love to 'force' things, even on a digital, microprocessor-based turbine control system. Forcing makes them feel powerful--that they are not constrained by any inhuman control system or scheme. Except, that common sense and logic should tell them it's potentially a very dangerous and risky thing to be forcing things, ESPECIALLY L86TXT.

L86TXT is the logic signal that, when set to a logic "1" because of an excessively high exhaust temperature will cause the turbine to trip--which, means the fuel will be immediately (as quickly as is mechanically possible) shut off. Failure to shut off the fuel on a high-high exhaust temperature means the fuel will continue to flow into the combustors, meaning at a minimum, the exhaust temperature will remain at the same level--OR, it might even go higher. AND, for who knows how long if someone is not paying attention--very strict attention--to the HMI, which has probably a 1 second lag, more or less, in reporting exhaust temperatures.

If that doesn't scream DANGER!!! I don't know what does. The Isothermal exhaust temperature value (TTK_I[n] in a Mark VIe, if I recall correctly) is already set to protect the exhaust diffuser in the gas turbine exhaust from overtemperature which would result in deformation and even melting if left unchecked. There is a LOT of exhaust gas flow over the diffuser, and if the diffuser deformed it could do a lot of damage. And, if there is an HRSG and there are superheater tubes, and maybe even auxiliary duct burners--which might be burning natural gas--and they were struck by debris from the exhaust diffuser, well, you can guess what might happen.

Now, let's think about what's going on in the combustor and the turbine stages. For the exhaust temperature to get excessively high that means the combustion temperature in the combustor also has to be excessively high. Which means the temperature of the hot gases entering the turbine section (through the first-stage turbine nozzles) is also excessively high. This can cause cracking and/or exacerbate existing cracking on the first stage turbine nozzles, which, if they fail, will send pieces of the turbine nozzles downstream into the turbine section, which can be disastrous. This doesn't take long--especially if the nozzles are already near their end-of-life and needing replacement or refurbishment. We are talking about a LOT of high temperature exhaust gases--and at the first-stage turbine nozzles, it's also at a high pressure (essentially CPD).

But, that's what Managers get paid for--to evaluate risks and decide which ones are acceptable and which aren't. (Though, sometimes they don't really do a proper analysis--they are also men, after all, and also have testosterone and they like the word "forcing" even if they don't really comprehend all of the risks. If they don't have to shut the unit down, then just force it! "Damn the torpedoes--full steam ahead!")

Nuff said, RATM?

Really, there should just be some things which cannot be forced. Ever. Not even with a special password. But, then that would entail extra programming and forethought, and that costs money. And, if something breaks because someone forced something they shouldn't have and weren't paying appropriate attention whilst doing so, well, then, more spare parts and services will be sold. It's a win-win, right? Lesson learned, and parts and services sold. (Except it doesn't have to be that way.)

 

thanks a lot CSA for your really fast response, I appreciate so much

 

2. It would be necessary to examine the CSP in use at your site to understand why the turbine might have tripped under the conditions you describe it was being operated. It's just not possible to understand from what we have been told without being able to see the CSP and Trip History. The originator was asked if the unit was a SIMPLEX because the Mark V SIMPLEX has back-up exhaust overtemperature protection in <C> so if they had only forced the <R> value <C> could still have tripped the turbine.

Dear CSA

So if the system was a simplex have a back-up and will trip the turbine anyway, but what if they have a TMR system?
(we have a TMR in out turbine)

best regards

 

There isn't really the same kind of "back-up" exhaust overtemperature control in the TMR as in the SIMPLEX. In the SIMPLEX, there is only one control processor, so the designers of the Mark V decided to include (I'm sure at the urging of the turbine designers and warranters!) a back-up provision, and since there was only <C>, it was done in <C> (which technically isn't supposed to do any protection functions, but, one has to use what is available, and if <C> isn't working in a SIMPLEX panel, neither is the turbine (running, or working).

In a TMR control panel, there are three "groups" of exhaust T/Cs, one to <R>, one to <S> and one to <T>. And, while it was ultimately decided for each of <R>, S> and <T> to look at ALL exhaust T/Cs (not just the ones connected to each individual control processor), there is still some "back-up" protection. For example, if one of <R> or <S> or <T> is out of service, the other two still can adequately determine if there is an exhaust overtemperature, or if there is a combustion problem (spread). So, there is "back-up" but it's just done in a different fashion, and using control processors.

I would have to look at the CSP in question to understand what is happening. There is an exhaust overtemperature trip that is set to happen if the actual exhaust temperature is higher than 40 deg F above the exhaust temperature reference (TTRX), and there was often another one that trips the unit if the exhaust temperature ever exceeds 40 deg F above the Isothermal value (TTKI_n), which is kind of redundant some times, but, I don't know if that also activates L86TXT or if it activates a different logic signal that results in the same alarm, or a different alarm.... There have been a few (per)versions of this over the decades.