Please all, what could be wrong. We were about to bring up a Frame 5 Nouvo Pignone (Now GE) gas turbine and suddenly almost all the alarms go on red. Please help. We are using a Mark 5 panel with Cimplicity.
Moderator’s Note: I dithered about posting this. I know that those of you who answer such questions want more information about alarms. I just didn’t know what to email back to this poster to ask him to do. So I’m hoping you can refine things and give him a direction to go.
Actually, there's a LOT of information that's required.
Type of Mark V: SIMPLEX or TMR
Was the Mark V powered down and then powered-up just before all these alarms became red?
Is there more than one HMI, and if so, is the same thing happening on all HMIs?
Do the CIMPLICITY displays show live data, or hash marks and black fields?
What have you done to try to troubleshoot the problem--and what were the results? (For example: We tried re-booting the HMI, but it didn't have any noticeable effect.)
Have you tried disconnecting the BNC connectors and reconnecting them several times (this includes the termination resistors on the BNC tee connectors)?
Does the StageLink use a fiber optic link and MODHUBS to convert from coaxial to fiber and from fiber to coaxial? If so, is it working correctly?
Is there an EX2000 exciter that is also on the StageLink? If so, can you data from the exciter on any CIMPLICITY display?
Are you sure all of the "alarms" are actually Process Alarms (the Alarm Drop Number often begins with the letter P; or one of the columns displays "Process" or "PROC")? If the description of the red "alarms" is EVENT or SOE, those are NOT alarms, but events (there are several types of events, including EVENT, SOE and PROCess- and DIAGnostic Alarms). It would make more sense if they were all SOE events, and that would mean there is something wrong with the DENET between processors and one or both of the Discrete I/O Modules (<CD>, <QD1>, and <QD2> if so equipped).
SOEs are actually changes of state of contact (discrete) inputs. If something happened to the DENET or the 125 VDC power to the discrete I/O modules (<CD>, <QD1>, and/or <QD2> if present) then that would cause a LOT of SOEs to be annunciated. But, those aren't alarms--they are changes of state of the discrete inputs. If the "wetting" voltage (interrogation voltage) to the discrete I/O module(s) was lost, that would make many contact states to appear to change--and that would result in a lot of SOE event messages, which (in GE's infinite wisdom, are shown on the Alarm Display, even though they aren't really alarms). Also, a problem with one or more of the IONETs which connect the discrete I/O modules to the processors could cause something similar.
As the Friendly Moderator says, we need more information. And better information, to be of help. We need to also know what you've done to try to troubleshoot and resolve the problem--AND what the results of your troubleshooting were.
This really shouldn't be too difficult to troubleshoot and resolve (it usually isn't, anyway). The one thing to remember is that the Alarm Display is a special type of CIMPLICITY object (an AMV object) and it kind of has it's "tunnel" to/from the Mark V for displaying and managing alarms and events. If they are all really alarms (Process- and/or Diagnostic) that would be very unusual. If they are SOEs, that would be more likely. You need to tell us what type of messages are all red in the Alarm Display (because all messages in the Alarm Display are NOT alarms!). And, it would probably help if someone would go through the Diagnostic Alarm display and look for any relevant alarms (such as loss of communications, power supply voltage problems, etc.).
Help us help you by providing more information. PLEASE answer ALL the questions to the best of your ability--because even if one or more may seem irrelevant, or you've already tried this or that, you haven't told us anything other than it's a Mark V on a Frame 5 and there's a lot of red messages on the Alarm Display. And, that's not really a lot of information.
Thank you for shedding light and asking such relevant questions. I have been unable to log in because of the troubleshooting processes. The management have brought in experts and the issues have been fixed.
However, it is proper that I reply some of your questions. It is a single HMI and the trouble was with most of the active alarms. The unit has been out of operation for about year since it's last tripped on start up overheating before reaching FSNL. Recently, a trouble shooting maintenance operation was carried out and the GCV and SRV were re-calibrated. It was afterward, while we were attempting that start up, that the alarms came on. However, they have been resolved now. The hardware was a TMR Mark V with Cimplicity software using a single HMI.
However, the unit still overheated before 100%TNH, precisely 56.8%TNH (580C exhaust temperature at this point) and as of now, we have been unable to resolve this issue. Please if you can be of further assistance, I would appreciate. Thanks.
This topic (failure to accelerate to FSNL because of high exhaust temperature) has been covered many times before on control.com. (There is a 'Search' field at the far right corner of the Menu bar of every desktop control.com webpage. The search syntax is not like most World Wide Web browsers, so you should probably use the Search Help the first few times.)
The Mark V (most of them; some very early versions did not) uses an acceleration rate reference after warm-up to get the unit to FSNL (Full Speed-No Load). This means that the Mark V is going to increase fuel (up to a limit) to try to assist with accelerating the unit to FSNL. Axial compressors are relatively inefficient at low speeds and this means that the exhaust tends to get hot fast. And in addition the compressor bleed valves are also open which is taking a portion of the air flow through the compressor and directing it to the exhaust--meaning the amount of air going to the combustors and through the turbine to the exhaust is reduced, which is also working to make the exhaust temperature a little higher than it would otherwise be.
Now, up to about 60% speed the starting means is providing a torque assist to the unit to try to help it accelerate to FSNL. BUT, if the torque from the starting means is low then the Mark V is going to try to compensate (up to a limit) with more fuel to help maintain the desired acceleration rate. If the torque from the starting means is low and the Mark V is having to increase the fuel flow-rate to help with acceleration it's very possible that the exhaust temperature will hit the maximum allowable limit--which will cause the Mark V to stop increasing the fuel flow-rate. This doesn't reduce the exhaust temperature, because the Mark V is keeping the fuel at the maximum allowable level to try to help accelerate the unit but not exceed the maximum allowable exhaust temperature.
Some times when this happens, the unit actually stops accelerating and begins decelerating, and after a while the Mark V will usually trip the turbine to prevent damage to the hot gas path parts. Or, the unit just accelerates VERY slowly.
You say the unit is "over-heating." EXACTLY what do you mean by that? Is the unit tripping? If so, what are the alarms that are being annunciated when the unit trips? Exhaust Overtemperature? Over-heating is not a reason for tripping--unless something is causing the exhaust temperature to spike during acceleration, and that shouldn't happen (unless someone is trying to force things to happen that shouldn't be happening; or, the gas fuel supply pressure is unstable; or, the P2 pressure (the pressure between the SRV and GCV) is not stable).
The Mark V should limit the exhaust temperature (signal name TTXM) by limiting gas fuel flow, to the maximum allowable limit (signal name TTRX). If the exhaust temperature exceeds the maximum allowable limit by more than 25 deg F the Mark V will alarm (something to the effect of high exhaust temperature, or exhaust overtemperature). If the Mark V still can't control the exhaust temperature for some reason and it exceeds the maximum allowable limit by more than 40 deg F, the Mark V will trip the unit and alarm something to the effect of 'Exhaust Overtemperature Trip.' BUT, the Mark V should--under normal conditions--not have trouble maintaining the exhaust temperature to not exceed the maximum allowable limit.
So, the most likely problem is that the amount of torque being provided by the starting means through the torque converter is low, either because the starting means is in need of refurbishment or the torque converter is in need of refurbishment, or the amount of oil flowing into and through the torque converter is insufficient (in this case the torque converter will get very hot to the touch!). If the starting means is a diesel engine with a hydraulic rack and solenoids for fuel control it's possible the fuel rack control mechanism is out of adjustment or was not adjusted properly after some service on the diesel. Or, one of the solenoids on the hydraulic fuel rack is leaking, or one of the actuator rods on the hydraulic fuel rack is not working properly. I have also seen dirty diesel fuel filters restrict the fuel flow to the diesel and cause problems. I have also seen dirty diesel air filters be choked and cause problems with the diesel being unable to develop full torque. It could be a combination of some or all of the above.
If the starting means is an electric motor, they, too, need periodic maintenance and refurbishment.
"Over-heating" of the unit during acceleration is not a technical reason for failure to reach FSNL. You need to be more specific about what exactly is happening. Is the unit "stalling" and someone is just pushing the Emergency Stop Push-button to shut the unit down, or clicking on STOP to initiate a controlled shutdown from the Mark V? Or is the unit tripping, and if it is tripping, what exactly are the alarms before and when the unit trips? If you can't decipher what is actually tripping the unit, then post the alarms from the 30 seconds prior to the trip, in chronological order, and we can help understand what is happening.
Dirty turbine inlet air filters can also cause problems with acceleration, as can a dirty axial compressor. If the turbine has not experienced a maintenance outage such as a hot gas path inspection or a major inspection in some time (years) then it's very possible the internal clearances of the turbine parts are excessive and/or the turbine parts are worn and one or both are contributing to a low torque production by the turbine--which the Mark V is trying to compensate for by increasing the fuel flow.
But, we need more, better information to be of any further help. And, it would be nice to hear back if you resolve the problem to know how you resolved it.
[I can also say that calibrating the "SRV and GCV" was probably NOT in any way helpful to the situation. One does NOT calibrate valves (or the IGVs)--one only calibrates LVDT position feedback from devices equipped with LVDTs. And, LVDTs are very reliable devices that very rarely drift and require calibration. It's a FALSE MYTH that "calibrating" devices has anything to do with the servo-valve or response rate of the device (fuel control valve or IGVs). It's just a waste of time and resources, and it makes it seem like someone is "doing something" even when they are not. BEFORE anyone EVER "calibrates" a device, the first thing which should be done is to check to see if the LVDT position feedback needs calibration. But no one EVER does that--they just "calibrate" without checking to see if calibration was required, or what the calibration results were compared to the as-found calibration values. Sure; "calibration" can sometimes point to a problem with a servo valve or a hydraulic actuator, but, again--the ONLY THING calibration does is re-scale LVDT position feedback. NOTHING MORE and NOTHING LESS. Full stop. Period. And, LVDTs are extremely reliable devices--they are used on aircraft and rockets of all types, and you don't hear of the pilot having to "re-calibrate" the flap position sensors during mid-flight. They are used because they are so reliable for such critical applications. It's just a myth promulgated by inexperienced and unknowledgeable people that "calibrating" the SRV or the GCV or the IGVs or the LFBV will solve a problem with the device or the servo valve of the device. A huge myth that just will not go away.]
So, please write back with more detailed information as requested. And let us know what you find. Happy to hear the Mark V "alarm" problems were resolved and the unit was able to be started. Hopefully, if you can provide more information, we can be of more help. Or, that the suggestions we provided are helpful in determining what the problem is with "over-heating" (an imprecise and indefinite term for a GE-design Frame 5 heavy duty gas turbine). And, it would be really helpful if you do solve the problem using the information provided if you would write back to let us know what the problem(s) was(were).! A LOT of people read these threads on control.com and learn a little from them. If you can take a few minutes to post for help, you can take a few minutes to let us know if the help was useful--or not--and how you resolved the problem.
Looking forward to hearing back from you!
The GT unit has not start. I am a very junior staff at this plant, and apparently, nobody is listening to me. So, they have worked on the exhaust insulation and did some manipulation to increase airflow, after the LVDT re-calibration I informed you about.
Again, we attempted start up, and this time, we had to shut down on low lube oil discharge pressure. I would continue to update you as the process goes on. I do wish they would start looking at the starting means and torque converter.
By the way, this is the second time an alarm with a '$' prefix has been posted to control.com. I personally have never seen such an alarm before.
I'm continuing to look into this, but I suspect this is something that has changed with a new HMI upgrade. Or, possibly, with a new Toolbox (or ToolboxST) upgrade. You say you are not familiar with Toolbox; it's the original software application used to configure, program and troubleshoot Mark VI turbine control systems.
GE is "migrating" sites to new HMIs for sites which still use Mark VI. (GE are also "migrating" sites with Mark VI to Mark VIe with what's called the 'Mark VI-to-Mark VIe Platform Upgrade' ... but that's another story. Or, it may be that your site has been upgraded to Mark VIe with new HMIs running ControlST and WorkstationST. It's very hard to keep up with GE as they make all their changes. Worse, different divisions of GE are "doing their own thing" with Mark VIe and HMIs, so that complicates matters even further.)
If your panel has been upgraded to the Mark VIe Platform, or the Mark VI control processors were upgraded when the HMIs were upgraded it's possible that ToolboxST (the newest version of Toolbox) is now being used to configure, monitor and troubleshoot the Mark VI (or Mark VIe). GE has very many different possible configurations. It's getting harder and harder to keep up with them.
About the TRLY problems you are experiencing, the 125 VDC that feeds the TRLY cards for use as "solenoid" outputs (meaning the Mark VI provides the 125 VDC power for solenoid connected to the output terminals) is usually "daisy-chained" from the <PD> to one TRLY, and then to another TRLY, and so on. Sometimes three or four TRLYs are all connected together in this fashion. I'm wondering if something hasn't happened to one or both of the TRLYs to cause this problem for the six (6) VCRC cards.... It just seems odd that six VCRCs are experiencing similar problems, seem to have started experiencing the problems at the same time.
You say 4X-4 is one of the VCRC outputs. Do you know if the output is configured as a solenoid output (with the hardware jumpers), OR if the power for the solenoid comes from outside the Mark VI panel (and the TRLY output is simply switching the power on and off to the solenoid)?
Do you know how large the ESD is that the 4X-4 output powers is (in terms of amp draw)? Was the ESD provided by GE as part of the gas turbine package, or as part of the plant controls, possibly by an EPC or contractor?
If you can provide more information about the exact configuration of the "Mark VI" (or Mark VIe) and HMIs at your site, we might be able to be of more help.