HI,

I would like to seek your expertise and advise on how to logic work on the exhaust temperature trip logic.
I would like to understand how the philosophy and algorithm of the average calculation as given on the TTXSPV4 block .
The trip happen when both TTXD 9 and 12 value spike in millisecond as good value to 1300F. No diagnostic recorded for any failure of TC probe nor the MarkVIe module. No high spread trip appear just Exhaust over temp trip directly.

Alarm event.
G1.L4T Master Protective Trip
G1.L30TXA_ALM EXHAUST TEMPERATURE HIGH
G1.L86TXT_ALM EXHAUST OVER TEMPERATURE TRIP
G1.L30TXA_ALM EXHAUST TEMPERATURE HIGH

Looking on the block EX_TMP_MD_7E, it is masked but the block TTXMV4_1 is visible though to understand the algorithm of it is confusing. I will attached the active block and trip log trend for your reference.

Question is:
1. How the TC affected the trip if calculation is average, does it have a separate logic that detect the TC high temperature against its threshold.
2. If only one (1) TC went on 1300F, does the machine will trip as well?

Looking forward to expertise reply. Thank you.

 

I only have a few minutes, and without being able to see the Item Help or Block Help for the blocks I can't really comment precisely. (The OEM has also gove away from creating Help for many blocks, preferring to convert the algorithmic blocks into macros, which is very difficult for me to decipher sometimes without a pencil, a big eraser and a lot of paper.)

Anyway, in the past, the exhaust temp block used the AVERAGE exhaust temperature after removing the highest and lowest exhaust T/C values. (This is different from the high-high exhaust temperature spread trip calculation which (must) use all the exhaust T/C values.)

I have seen failing exhaust T/Cs cause spiking values, spiking high and spiking low. And, this isn't always caught by the diagnostic health checks. In my experience, the health checks for T/Cs are usually looking for open or shorted T/C circuits.

We don't know how old your machine is. We don't know if this happened once or has happened multiple times. We don't know how the exhaust T/Cs are wired to the T/C I/O Packs, or TCAA modules (I think that's the multi analog I/O modules). We don't know if this happened after a recent maintenance outage. We don't know how this machine is operated (load following; Base Load; daily start/stop--maybe even multiple starts/stops per day).

We also don't know what Diagnostic Alarms are present for the exhaust T/C sensing modules/packs, if any.

I have also seen wiring problems, especially after maintenance outages when the exhaust T/Cs were disconnected to remove the exhaust frame and not reconnected properly. I've seen exhaust T/C tips bent in the radiation shields by workers in the exhaust duct with nothing better to do during a maintenance outage.

So, don't overlook the easy stuff (wiring; terminations; T/C tip placement in the radiation shields; Diagnostic Alarms on the modules/packs processing the exhaust T/Cs; 125 VDC battery grounds; etc.).

 

Dear @WTF?, I would like to express my gratitude for your prompt response. In regards to the Exhaust block calculation, I have noted that the low temperature setting to be disregarded is 500F (am I right?), however, I am unable to locate the constant for the high temperature value. Despite this, if the high value is included in the calculation, the calculated average value remains satisfactory based on the 18 TC,

The two exhaust TC (9 and 12) are connected in same I/O nodule together with the other TC used in different area on monitoring.
There was plan to replace the TC but up to now all TC are in good value.

I am still unclear on how the block calculates it. Additionally, I am concerned about the scenario where only 1 TC exceeds the high temperature threshold of 1200F. Does anyone experience that scenario before?

 

When exhaust T/Cs are beginning to fail, this is one of the classic behaviors--spiking. I am quite busy this week and next and won't have time to look over the Trip Log for a few days.

A COMMON misconception is that ANY single exhaust T/C over the alarm or trip setpoint will trip or cause a exhaust overtemp alarm--and that's JUST NOT TRUE. (It's the STOOPUD way they configure the bar graphs on the exhaust display to change colors when any single T/C is above the setpoint(s)--the OEM FAILS to understand how confusing those bargraphs can be.

It's when the average of all of the exhaust T/Cs--rejecting the highest and the lowest--and rejecting any value less that 500 deg F when the unit is running and above 14HS (if I recall correctly) that the alarm- and trip will be annunciated and cause a unit emergency stop (trip).

If you right-click on the TTXMV4 block in ToolboxST a dialog box should pop up and one of the possible selections should read "Item Help" or "Block Help". Sometimes those selections are greyed out--meaning there is no Item Help or Block Help for the function/block (which is another serious failure of the OEM for these multi-million USD machines to not have proper documentation for all blocks (macro or otherwise); they treat the blocks as proprietary information, and many times there are warnings to that effect even when the Help feature is available for the block--but it's beyond anyone's comprehension why the number of functions/blocks with Help are dwindling and being replaced with macros, without Help!).

If you have access to any Mark* V CSP (Control Sequence Program) files for GE-design heavy duty gas turbines you will find MANY of the functions/blocks used back then are still used today (with some added bells & whistles (such as I/O health diagnostics). AND, they are in graphical format so are easier (I know that's a relative term for some people who haven't been exposed to and used the OEM's RLD (Relay Ladder Diagram) graphical displays) to read and follow. In the back of later versions of the Mark* V Application Manual, GEH-6195, there is a section that explains how to read and understand the graphical representations, sort of a "legend description" (to use old-school designation) for the symbols in the blocks and CSP. Again, one of the REALLY GREAT things about GE's evolution of the Mark* turbine control system beginning with the Mark* IV is that they stuck with virtually the same graphics and representation and symbols for decades across multiple evolutions of the Mark*. For those of us who worked on Mark* II (which was a period of serious departure from that practice--and one that was abandoned, thankfully!), Mark* IV, Mark* V, Mark* VI, and Mark* VIe it was something of a short and easy learning curve to learn the new graphics and such, which made the work of commissioning, troubleshooting and maintaining easier over the decades.

So, I'll have a look at the TripLog when I have a chance. But, it's noted that you didn't really answer any of the "questions" (which don't always end with a question mark--but are questions, nonetheless, and will help our (and your) understanding of the situation/event and the machine and how it's operated, etc. You don't have to answer any questions, but you would be happier with answers/responses if you did. Guaranteed. (The more information you provide, and in the initial post!, the more pleased you will be with the responses and suggestions you receive (not always, but as a general rule--that's been true here on Control.com for more than 20 years of GE Mark* turbine control system posts and responses).

It seems to be the way of the newer generations that they believe--fervently and steadfastly--that they only have to provide the briefest of descriptions of a "problem" (often which isn't even a real problem, just a perceived one...) anonymously to a World Wide Forum of "experts" and they will be rewarded with complete and concise and detailed instructions on precisely how to remedy the problem. GE-design heavy duty gas turbines are made up of many components--and they're not always the same even on the same Frame size machines!--and there are a LOT of intangibles which can go into a particular event. I, for one, loathe guessing at the precise cause with very little information to begin with, because I have learned (painfully) over decades the root cause of a problem with symptoms similar to other problems can be VERY different. I often suggest multiple possible causes (and remedies in some cases) just because I don't want to play 20 Questions trying to get answers to questions (which I still usually pose, but few people respond to).

One last thing I will add is that if the machine was operating on gaseous fuel (Natural Gas; etc.) if the dewpoint of the incoming gas fuel supply is very close the temperature of the gas fuel that on passing through two or more control valves AND the gas fuel nozzles there are multiple pressure drops which cause temperature drops--which can cause moisture entrained in the gas in the form of hydrocarbon chains to condense in the system, and sometimes that condensation gets pushed into one or more combustors and cause a VERY high firing temperature for that particular combustor (or combustors). This is usually accompanied by a spike in load, but if the machine trips the load spike can be very short-lived and not very large. So, that's yet another intangible in this possible causes. But, my money is on failing exhaust T/Cs, or wiring problems, or improperly installed exhaust T/Cs. BASED ON THE INFORMATION PROVIDED (that I've had time to read and analyze).

 

Apart from calibration drifts of thermocouples and RTD's, temperature sensors fail as an open circuits either gradually or instantly.
Open circuit thermocouples read low, open circuit RTD's read high. As WTF mentioned the vendor is not using effective logic when multiple sensors are involved.
good luck.

 

When exhaust T/Cs are beginning to fail, this is one of the classic behaviors--spiking. I am quite busy this week and next and won't have time to look over the Trip Log for a few days.

A COMMON misconception is that ANY single exhaust T/C over the alarm or trip setpoint will trip or cause a exhaust overtemp alarm--and that's JUST NOT TRUE. (It's the STOOPUD way they configure the bar graphs on the exhaust display to change colors when any single T/C is above the setpoint(s)--the OEM FAILS to understand how confusing those bargraphs can be.

It's when the average of all of the exhaust T/Cs--rejecting the highest and the lowest--and rejecting any value less that 500 deg F when the unit is running and above 14HS (if I recall correctly) that the alarm- and trip will be annunciated and cause a unit emergency stop (trip).

If you right-click on the TTXMV4 block in ToolboxST a dialog box should pop up and one of the possible selections should read "Item Help" or "Block Help". Sometimes those selections are greyed out--meaning there is no Item Help or Block Help for the function/block (which is another serious failure of the OEM for these multi-million USD machines to not have proper documentation for all blocks (macro or otherwise); they treat the blocks as proprietary information, and many times there are warnings to that effect even when the Help feature is available for the block--but it's beyond anyone's comprehension why the number of functions/blocks with Help are dwindling and being replaced with macros, without Help!).

If you have access to any Mark* V CSP (Control Sequence Program) files for GE-design heavy duty gas turbines you will find MANY of the functions/blocks used back then are still used today (with some added bells & whistles (such as I/O health diagnostics). AND, they are in graphical format so are easier (I know that's a relative term for some people who haven't been exposed to and used the OEM's RLD (Relay Ladder Diagram) graphical displays) to read and follow. In the back of later versions of the Mark* V Application Manual, GEH-6195, there is a section that explains how to read and understand the graphical representations, sort of a "legend description" (to use old-school designation) for the symbols in the blocks and CSP. Again, one of the REALLY GREAT things about GE's evolution of the Mark* turbine control system beginning with the Mark* IV is that they stuck with virtually the same graphics and representation and symbols for decades across multiple evolutions of the Mark*. For those of us who worked on Mark* II (which was a period of serious departure from that practice--and one that was abandoned, thankfully!), Mark* IV, Mark* V, Mark* VI, and Mark* VIe it was something of a short and easy learning curve to learn the new graphics and such, which made the work of commissioning, troubleshooting and maintaining easier over the decades.

So, I'll have a look at the TripLog when I have a chance. But, it's noted that you didn't really answer any of the "questions" (which don't always end with a question mark--but are questions, nonetheless, and will help our (and your) understanding of the situation/event and the machine and how it's operated, etc. You don't have to answer any questions, but you would be happier with answers/responses if you did. Guaranteed. (The more information you provide, and in the initial post!, the more pleased you will be with the responses and suggestions you receive (not always, but as a general rule--that's been true here on Control.com for more than 20 years of GE Mark* turbine control system posts and responses).

It seems to be the way of the newer generations that they believe--fervently and steadfastly--that they only have to provide the briefest of descriptions of a "problem" (often which isn't even a real problem, just a perceived one...) anonymously to a World Wide Forum of "experts" and they will be rewarded with complete and concise and detailed instructions on precisely how to remedy the problem. GE-design heavy duty gas turbines are made up of many components--and they're not always the same even on the same Frame size machines!--and there are a LOT of intangibles which can go into a particular event. I, for one, loathe guessing at the precise cause with very little information to begin with, because I have learned (painfully) over decades the root cause of a problem with symptoms similar to other problems can be VERY different. I often suggest multiple possible causes (and remedies in some cases) just because I don't want to play 20 Questions trying to get answers to questions (which I still usually pose, but few people respond to).

One last thing I will add is that if the machine was operating on gaseous fuel (Natural Gas; etc.) if the dewpoint of the incoming gas fuel supply is very close the temperature of the gas fuel that on passing through two or more control valves AND the gas fuel nozzles there are multiple pressure drops which cause temperature drops--which can cause moisture entrained in the gas in the form of hydrocarbon chains to condense in the system, and sometimes that condensation gets pushed into one or more combustors and cause a VERY high firing temperature for that particular combustor (or combustors). This is usually accompanied by a spike in load, but if the machine trips the load spike can be very short-lived and not very large. So, that's yet another intangible in this possible causes. But, my money is on failing exhaust T/Cs, or wiring problems, or improperly installed exhaust T/Cs. BASED ON THE INFORMATION PROVIDED (that I've had time to read and analyze).

Hi @WTF? , our machine is running at HOG fuel gas supply, no other activities done previously specially on Exhaust TC though we made HGPI and done Mark VI to Mark VIe controller upgrade 6 month ago but Exhaust TC was not touched, also this is the first time we experience this kind of behavior on our ExhTC which spike on 1300F in millisecond in good value causing the machine to trip .

I really appreciate your expertise information and suggestion on this matter. Still I'm trying to understand how the logic works on this kind of behavior. By the way, is anyone here having the GE manual GEH-6195