GE Frame 9E DLN1 High vibration Trip during Combustion Mode change from LL to PM

Hi All,


We are experiencing a high vibration trip from Brg #2 seismic prob during the combustion mode changeover from lean lean to Premix. So far we have experience couple of trips and the vibration value goes up as high of 27mm/s to 35mm/s and at around 60MW. It does not reach trip limit all the time and during a successful changeover, vibration value goes up about 12mm/s. The Brg #2 is fitted with proximity probs and their values have spikes with slight time lag (around 14sec as per system 1 data and increase in value around 50-100%). Unable to capture the frequency domain of this fast spikes . rest of the turbine bearings seismic probs do not experience in any spikes, but Brg #3 shaft vibration experience in spikes with aforementioned time lag of 14 seconds. (change in value from 30-50%, not sure if the system 1 does have a time sync issue with trends)
I've attached two files of Trip event and successful mode changeover events showing gas valve positions/IGV and flame scanner intensities. Turbine is a peaking unit, approximately 5 starts per week.

Any of the forum member has experienced similar issue in the past? any view would be highly appreciated
Ken
 

Attachments

Yoganak,

Lots of questions come to mind; I don't have access to a PC to look closely at the attachments (yet).

Does the unit have a Gas Transfer Valve that is in use during the LL-PM combustion mode transfer?

What alarms (Process & Diagnostic) are present before and during the LL-PM combustion mode transfers? All of them; even ones you don't deem relevant.

When did this problem start? After a trip from load? After a maintenance outage? After some gas control valves had been "calibrated?"

Does this unit have Independent Gas Control Valves (meaning a SRV (Stop/Ratio Valve) and three gas valves (GCV1, GCV2 and GCV3)? Or does it have an SRV, a GSV (Gas Splitter Valve) and GTV (Gas Transfer Valve)? Or just an SRV and GSV? (All DLN-I gas valve configurations are not the same; it's important we understand what kind you have. A copy of the relevant Gas Fuel System P&ID would be very helpful.)

What have you done to try to resolve the problem, and what were the results?

Again, I haven't the proper screen to view the files, but what are the exhaust temperature spreads doing during the LL-PM combustion mode transfers that are not successful, and what do the exhaust temperature spreads look like during successful LL-PM combustion mode transfers?

What control system is in use on the turbine?

Does the unit have IBH (Inlet Bleed Heat), and if so , is it in service (enabled and operating) at the time of the LL-PM combustion mode transfers?

What is the ambient temperature during these failed combustion mode transfers? Is it cooler than normal? Hotter than normal?

Does the unit use heated gas fuel--either from a start-up gas fuel heater or a performance gas fuel heater? And, if so, is(are) the heater(s) in service?

I suspect there is some kind of issue with the fuel nozzles which is causing higher than normal dynamic pressures inside the combustion cans and this is resulting in the high vibrations. You say the #2 bearing has proximity probes--is that in addition to the seismic (velocity) vibration probe on the #2 bearing drain piping in the bottom of the Turbine Compartment? Has anyone had a look at the seismic vibration probe to see if it's still properly attached to the drain piping, and the cabling is also firmly connected to the sensor, and in the JB on the side of the Turbine Compartment base?

Have the exhaust temperature spreads during LL operation and PM operation changed since this problem started? In other words, what were the exhaust temperature spreads during LL and PM operation before this problem started and what are they now after successful transfers?

Does the unit also burn liquid fuel on occasion? If so, how often? And does it use Water Injection for NOx emissions reduction when operating on liquid fuel?

Again, the typical cause of high #2 bearing vibration during LL-PM combustion mode transfers is high combustor dynamic pressures caused by fuel nozzle issues. I have even seen where strainers put in the gas fuel pigtails after a maintenance outage have gotten plugged and caused similar issues--especially if the unit has a GTV or GCV3 that operates during LL-PM combustion mode transfers. Also, incorrect calibration of the GTV or GCV3 LVDTs can cause a problem with getting sufficent fuel into the transfer nozzles (or even too much fuel) which can cause high combustion dynamic pressures.

If the turbine control system is a GE Mark* turbine control system, do you have, and can you provide, the Trip History or Trip Log displays for the unsuccessful LL-PM combustion mode transfers?

Do the units have a Historian and can you get data from the Historian (PI-based, or PROFICY-based, maybe)? It's usually not very high-speed data, but it might be useful.

These are all very important questions that need to be answered. Especially the ones about what have you done to try to troubleshoot the problem--and what were the results, AND what Process- And Diagnostic Alarms are present and annunciated during the LL-PM combustion mode transfers (I'm presuming the turbine control system is a GE Mark* which will have Process and Diagnostic Alarms). The configuration of the gas fuel control valves is also very important. Some turbines which don't have a GTV or GCV3 that operates during LL-PM combustion mode transfers can have a problem known as "screech" during the transfers which can be aggravated by fuel nozzle issues or valve issues and cause higher than normal dynamic pressures in the combustion cans. So, understanding the combustors and the gas fuel delivery system is also important.

I'm traveling now (unfortunately--and not for much longer; people just aren't taking this COVID-19 thing very seriously (when traveling!)) and won't have a lot of time in the next couple of days. But, if you will answer the questions, as many as possible and to the best or your ability, there are others who can probably provide some help and suggestions also. But, we need more information.

Hope this helps!
 
Hi CSA, Thanks and appreciate your response. It took me sometime to verify few as you had requested.

I have copied your tread to answer the questions in bold.
All three turbines are in operation for around 10 years now and their majors had been done in 2017. The unit that trips has started this issue out of the blue without any maintenance or changes done to the unit. though it does not show any looseness within the prob, we have advise maintenance teams to change it out with new. Also we are looking into do a combustion tune up. Even though we are capable of comparing good mode changeover VS a trip as well as comparing with other units, no any abnormality has been found in any other process parameters. ( please note all the unit undergo a Brg #2 vibration level of ~12mm/s during a successful LL-PM transfer)

cheers,

Ken


---------------------------------------------------------
Lots of questions come to mind; I don't have access to a PC to look closely at the attachments (yet).

Does the unit have a Gas Transfer Valve that is in use during the LL-PM combustion mode transfer? Yes it Does

What alarms (Process & Diagnostic) are present before and during the LL-PM combustion mode transfers? All of them; even ones you don't deem relevant. We do not receive any except the trip

When did this problem start? After a trip from load? After a maintenance outage? After some gas control valves had been "calibrated?" this unit has done three years after major. Issue started 8 months ago and there had not been any changes done within or around GT

Does this unit have Independent Gas Control Valves (meaning a SRV (Stop/Ratio Valve) and three gas valves (GCV1, GCV2 and GCV3)? Or does it have an SRV, a GSV (Gas Splitter Valve) and GTV (Gas Transfer Valve)? Or just an SRV and GSV? (All DLN-I gas valve configurations are not the same; it's important we understand what kind you have. A copy of the relevant Gas Fuel System P&ID would be very helpful.) We have GCV 1, 2 and 3 system.

What have you done to try to resolve the problem, and what were the results? So far we have not done any physical work. Trying to get OEM to check the combustion tuning

Again, I haven't the proper screen to view the files, but what are the exhaust temperature spreads doing during the LL-PM combustion mode transfers that are not successful, and what do the exhaust temperature spreads look like during successful LL-PM combustion mode transfers?

We do not see any difference or high spread during the transfer

What control system is in use on the turbine? Mark VIe

Does the unit have IBH (Inlet Bleed Heat), and if so , is it in service (enabled and operating) at the time of the LL-PM combustion mode transfers? Yes. We have IBH and during the transfer it closes . there is no noticeable difference during a trip event or good run

What is the ambient temperature during these failed combustion mode transfers? Is it cooler than normal? Hotter than normal? No noticeable change. After first fail start, we have started the unit again within next hour with successful transfer

Does the unit use heated gas fuel--either from a start-up gas fuel heater or a performance gas fuel heater? And, if so, is(are) the heater(s) in service?

No changes had been made or noticed. We are further looking into see for any abnormality.

I suspect there is some kind of issue with the fuel nozzles which is causing higher than normal dynamic pressures inside the combustion cans and this is resulting in the high vibrations. You say the #2 bearing has proximity probes--is that in addition to the seismic (velocity) vibration probe on the #2 bearing drain piping in the bottom of the Turbine Compartment? (Yes) Has anyone had a look at the seismic vibration probe to see if it's still properly attached to the drain piping, and the cabling is also firmly connected to the sensor, and in the JB on the side of the Turbine Compartment base? We are planning the same checks on Seismic prob and even replacement

Have the exhaust temperature spreads during LL operation and PM operation changed since this problem started? In other words, what were the exhaust temperature spreads during LL and PM operation before this problem started and what are they now after successful transfers? Before and after remain almost same

Does the unit also burn liquid fuel on occasion? If so, how often? And does it use Water Injection for NOx emissions reduction when operating on liquid fuel? No liquid fuel injection

Again, the typical cause of high #2 bearing vibration during LL-PM combustion mode transfers is high combustor dynamic pressures caused by fuel nozzle issues. I have even seen where strainers put in the gas fuel pigtails after a maintenance outage have gotten plugged and caused similar issues--especially if the unit has a GTV or GCV3 that operates during LL-PM combustion mode transfers. Also, incorrect calibration of the GTV or GCV3 LVDTs can cause a problem with getting sufficent fuel into the transfer nozzles (or even too much fuel) which can cause high combustion dynamic pressures. We are having the similar impression on Acoustic dynamic issue. But the main issue is, it does not trip (high vibration) all the time.(approximately 1 out of 10 may trip)

If the turbine control system is a GE Mark* turbine control system, do you have, and can you provide, the Trip History or Trip Log displays for the unsuccessful LL-PM combustion mode transfers? I will check for this

Do the units have a Historian and can you get data from the Historian (PI-based, or PROFICY-based, maybe)? It's usually not very high-speed data, but it might be useful. What I have provided before is PI data. High speed data has been setup for any upcoming events
 
Yoganak/Ken,

Thanks for the complete answers to the questions!

This is odd, certainly. It's been a very long time since I looked at the vibration protection masks but I wonder if this has something to do with the problem(s) as well. (By masking I mean the Control Constants that are passed to the vibration block(s) that set up the redundancy correlations of all of the turbine vibrations sensors. Usually, one single vibration sensor, by itself, is not able to trip the unit and it requires more than one sensor to indicate excessive (high-high) vibration to actually initiate a trip--redundancy. Perhaps when I get home (traveling in the COVID-19 times is ... trying, at best) I can have a look at some Frame 9E software and get the Constant names to ask you for the values on your machine to compare against to see if there's an issue with redundancy configurations.)

I would also suggest you look into the potential System1 time synchronization issue. It would be interesting to know if the other units also experience the same spike in magnitude, at the same time differential, during their LL-PM transfers.

I'm not going to make any friends with this next question, but recently I have been to two sites and have received data from another site which had configured WorkstationST Alarm Viewer to ignore Diagnostic Alarms.... Yes; I know. Some units which were not configured or commissioned have LOTS of nuisance and intermittent Diagnostic Alarms, and the OEM is incredibly insensitive about the problem, basically ignoring it and saying, "That's normal; all Mark* turbine control systems do that. Diagnostic Alarms don't trip the turbine; just ignore them." Well, the first part of the last sentence is true--but that's the only part of the statement that is true. A SINGLE Diagnostic Alarm will not trip the turbine, but certain combinations of Diagnostic Alarms will result in a turbine trip. (No; there is not matrix or list of groups of Diagnostic Alarm combinations which will trip the turbine--there are literally thousands of possible Diagnostic Alarms (yes--thousands), so putting such a list together would be an enormous task which would never be fully complete given the lack of standards used when creating Diagnostic Alarms.) So, some enterprising Operators and I&C (Instrumentation & Control) Technicians have figured out how to filter out Diagnostic Alarms--which is a very dangerous thing to do.

I personally have never been to a Mark VIe-equipped site that doesn't have one or three (or more) Diagnostic Alarms. NEVER. It just doesn't happen even on the best maintained sites, even when there's a CSA or LTSA with the OEM. It's just not the norm for Mark* turbine control systems. So, I would be surprised--very surprised--to find there are NO active Diagnostic Alarms on the units at your site. Pleasantly surprised if there were actually zero Diagnostic Alarms--but I am rarely pleasantly surprised these days.

Filters on WorkstationST Alarm Viewer are a very useful tool--especially when troubleshooting. But, during normal operation when every Alarm (Process & Diagnostic) is important the use of filters can be very misleading. One of the sites I went to had been experiencing nuisance trips and had multiple Diagnostic Alarms from redundant critical sensors--and had no idea they were occurring--because one of the Operators had filtered out the Diagnostic Alarms (because he was "tired" of them--and they wouldn't trip the turbine, or so he was told...). And, there was no Log Book entry to alert others in the plant that the WorkstationST Alarm Viewer configuration had been changed.

There are high dynamic pressure pulsations during LL-PM combustion mode transfers. All of the fuel entering the combustors has to be "switched" to flow into the secondary combustion zone of the combustor--through the secondary fuel nozzles (and pegs) and the transfer passages/nozzles of the secondary fuel nozzle. (The Gas Transfer Valve, probably GCV3 in the Indepenent Gas Valve arrangement, will open to permit fuel to flow into the transfer passages and nozzles of the secondary fuel nozzle during the LL-PM transfer, while fuel is being shut off to the primary fuel nozzles and primary combustion zone of the combustors. This is necessary because it's the only way the diffusion flame in the primary combustion zone can be extinguished--by shutting off the flow of gas into the primary combustion zone, which means ALL of the gas entering the combustors has to be directed through the secondary, and transfer, passages and nozzles during the LL-PM transfer.)

If the unit has and is using IBH (Inlet Bleed Heat), that occurs at a lower load--and, hence, fuel flow-rate--than it would occur at if IBH was not in use, so this usually helps with reducing some of the dynamic pressure pulsations. I believe you said IBH shuts off during the LL-PM transfer; this would introduce some additional dynamics (air flow) to the combustion process which I can't think would be very helpful to the transfer....

I would expect if there were fuel nozzle (secondary and transfer) problems the exhaust temperature spreads would increase during the transfer, and that would cause the combustor dynamic pressures to also increase--which would increase the vibration in the combustor section of the turbine--which is closest to the #2 bearing. But, you said, the spreads haven't and don't change during the LL-PM transfer, so that basically rules out secondary fuel nozzle problems.

Many times the GTV (Gas Transfer Valve--or GCV3 in the Independent Gas Valve arrangement) doesn't actually go all the way open, so if it is opening fully when it hasn't in the past then that could be a problem. Or, if it isn't opening fully (when it has in the past) but is only opening partially, that could possibly introduce problems resulting in higher than normal combustion dynamic pressure pulsations during the LL-PM transfer.

I have seen many turbines, when the ambient temperature changes during the year (in the spring and then again in the fall) have trouble transitioning successfully from LL to PM. And, that is usually solved by combustion tuning.

My bet is on a combination of things. A failed, or failing, #2 bearing seismic vibration sensor, combined with some issues with IBH and/or IGV movement, along with the possibility that fuel flows through the secondary- and/or transfer passages and nozzles are quite the same as they were before. I'm also wondering about the vibration block configuration.... Seismic vibration sensors are pretty robust devices--as long as ones with the proper temperature rating are used in high-temperature locations (such as the #2 bearing area in the Turbine Compartment, and the #3 bearing area in the Load Tunnel). I have seen sites which installed seismic sensors from the #1 bearing area in the Turbine Compartment and/or the Load Tunnel which didn't last very long, and gave erratic readings until they finally failed. Sure; they both visually look the same, but they are not.... the same (they have different temperature ratings, most likely because of different materials used in the internal construction).

I would also like to know what the TTRF (or TTRF1) values were prior to the problems on this particular unit, as well as after the problems. And, if all of the "Performance Monitor Package" sensors are all in working order. This includes the ambient pressure transmitters (96AP's) and the two inlet flow measurement transmitters (96BD and 96CS). Also, are all of the axial compressor discharge temperature T/Cs working properly? AND, are the exhaust duct back-pressure transmitters working properly (if the unit uses exhaust duct back-pressure transmitters (96EPs, I think))? Because all of these could affect the TTRF1 value, which affects the LL-PM transfer point. (It would be interesting to know what load and IGV angle the unit was transferring to prior to the problem and after the problem, also.)

That's about all I can think of. Again, thanks for the complete answers to the questions. Please let us know what you find as you progress in troubleshooting and resolving this issue.
 
Hi CSA, Thanks for your reply. Once you get the computer access, please have a look on two trends attached in the first thread that contains IGV, Valve Positions, Load, Flame scanner intensities for trip event and successful changeover.
I'll look for other parameters as you have mentioned above.
Thanks,

Ken
 
Top