Recently after completion of combustion inspection, one of GE Frame 9 Gas Turbine during shutdown time showing high vibration and high vibration alarm is appearing in Mark IV control system. Total Nine nos seismic vibration pickup (Manufacturer- Metrix instrument co., USA) installed- Bearing no -1: BB1 & BB2, Bearing No-2: BB3, Bearing No-3: BB4 & BB5, Generator driven end Bearing: BB7 & BB8 and Generator Non driven end Bearing: BB9. Specially BB3 is going up to 19mm/sec during shutdown around 89% TNH (speed) and vibration exist up to 60% speed. Other Turbine side bearing vibration also going high but not that much. For clarification i am citing vibration level during 66% (TNH): BB1=6.2, BB2=6.4, BB3=10.7, BB4=4.8, BB5=5.1, BB7=1.2, BB8=1.2 & BB9=1.8 mm/sec. During 90MW load the vibration are: BB1=2.5, BB1=2.1, BB3=3.4, BB4=3.2, BB5=3.4, BB7=3.4, BB8=3.2 & BB9=2.4mm/sec. Before combustion inspection at 90MW load the vibration are: BB1=2.8, BB2=2.2, BB3=3.5, BB4=2.7, BB5=3.1, BB7=3.1, BB8=2.9 & BB9=2.3mm/sce.

We, i & C, changed the vibration pickup BB3 and check cables & connector and found okay. During overhaul we do not change any calibration setting of Gas Control Valve, Gas Ratio valve, IGV & LFO bypass valve. Mechanical arguing as starting and running time vibration are ok and so they are thinking they do not have problem.

Could anyone tell me as soon as possible what are the causes for the above high vibration during shutdown. and what could be checked. Then it will be a great help.

 

It seems you changed the vibration pick-up and cable after the high vibration incident. Did the high vibration occur again on a shutdown after the pick-up/cable replacement?

Was the 39V-3 (BB3) pick-up firmly bolted to the flange when you removed the original one? Could it have been loosened during the maintenance outage, or even before during normal operation?

When you say the pick-up and cable were okay, did you test it on a shaker table for some period of time? How did you test the pick-up and cable?

Was the unit being shutdown after running for some time (several hours) under some load (say 50-100%)? If so, the internal turbine temperatures should all have been fairly stable. If, however, the unit was just started and run up to FSNL or lightly loaded for a short period of time then shut down, the internal turbine temperatures (shaft, bearings, etc.) may not have reached normal, stable conditions, and this could contribute to a high vibration condition.

Another factor which greatly affects vibration is lubricating oil header temperature. If it's too cold (less than approximately 80 deg F) then higher-than-normal vibrations will frequently be experienced, but that's not usually the case during shutdowns after the unit has been running under load for some time--unless there was some kind of failure of the L.O. temp regulating valve which resulted in a very cold L.O. header temperature. Or, if the unit wasn't really run for very long before being shut down the L.O may not have gotten up to near normal temperature; normal L.O header temp should be approximately 125-130 deg F.

The "unfortunate" part about incidents like this is that there is no redundant sensor on the #2 bearing of Frame 9s, so there's no way to tell if the reading was caused by a faulty sensor. Except, that usually if a high vibration is experienced on the #2 bearing there is usually some vibration "transmitted" through the shaft to the #1 and/or #3 bearing(s). You said the other turbine bearings indicated a slightly higher than normal vibration, so it may have truly been a high vibration condition, though it doesn't seem to have been excessively high.

Also, the #2 bearing vibration pick-up isn't really mounted on the #2 bearing--it's actually mounted on the #2 bearing L.O. drain piping which is bolted to the #2 bearing housing.... So, usually the #2 bearing vibration has to be fairly excessive to register on the #2 bearing vibration pick-up. Also, if the piping or the pick-up is loose it can accentuate any actual vibration experienced at the #2 bearing.

Remember that GE heavy-duty gas turbines pass through two "critical" vibration points during each start-up and each shutdown. The exact RPM ranges for these criticals are never "published" by GE for a particular engine, but one can usually "see" them as peaks, however slight, during acceleration and deceleration. Usually, they're so slight that most people never notice them. However, if there are other factors which can contribute to vibration (as above) when passing through a critical, they can and will accentuate the vibration experienced during the critical.

It's not uncommon for instrumentation (including cables and tubing) to be damaged, even slightly, during maintenance outages (combustion inspections, hot gas paths, major inspections, etc.). The best maintenance outages begin with the removal of all instrumentation which could possibly be damaged by mechanics, millwrights, and/or pipefitters; it's a time-consuming process, but this author has seen thousands of dollars spent replacing damaged instrumentation and tubing--and that doesn't include the lost generation!

If instrumentation is not removed prior to the start of mechanical disassembly, it invariably "gets in the way" of the wrench- and hammer-wielding laborers, and it will not be spared. (It's funny how instrumentation just jumps out and gets in the way of mechanics, millwrights, and pipefitters. It's even funnier when none of them knows how the instrumentation got damaged.... The #2 bearing vibration pick-up is in a very vulnerable spot, as it's in the bottom of the turbine compartment where wrenches and flanges and pipes get dropped, and people have to crawl to move around, sometimes kicking the pick-up and it's cable. ) As you cite, there is always a competition of sorts between the instrumentation technicians and the mechanical "technnicians"--some of it is good-natured fun, and sometimes it's near all-out war!

markvguy

By the way, arocon, you're getting better at providing data with your initial posts.... You should find after a while, that writing posts requesting help or information will help you to anticipate questions and provide clearer descriptions and better data. It's not easy, but it's good exercise for the mind. Don't give up!

 

we replaced the vibration pick-up, cables but still same result. Could anyone help on this matter?

regards,

arocon

 

You've replaced the pick-up, the cable, and still have the same result. Seems to be pointing at a true vibration problem.... But, if the problems have persisted on subsequent shutdowns after replacing the pick-up and cable, then you have a persuasive argument for a real vibration problem.

How about switching one of the #1 Bearing vibration pick-ups with the #2 Bearing pickp-up in the Mk IV (exchanging the wiring at the AIO-xx terminal board)? This would tell you if the problem is in the Mk IV. Or, you could just replace the NVCx cards in <R>, <S>, and <T> to eliminate them as being the source of the "problem."

What were the wheelspace temperatures before the Combustion Inspection and after the Combustion Inspection at Base Load? Could it be possible that one of the orifices in the Cooling & Sealing Air system was not re-inserted, or was exchanged with one of a different size?

One last thing to check is that the Compressor Bleed Valves are opening fully during start-up and shutdwon. The valves usually have 11 seconds to open when told to or an alarm is annunciated, and they must be open during start-up and shutdown. Many turbines have manual isolation valves upstream of the Compressor Bleed Valves which must be fully open for proper operation during start-up and shutdown.

The NVCx cards read the voltage produced by the suspended magnets in the velocity (seismic) vibration pick-ups. The voltage is generated in the pick-up, not by the Mk IV, when the unit is running. During shutdown, the Mk IV impresses a test voltage across the pick-ups to ensure they are working properly prior to starting; but during running the signal is generated by the pick-up(s) and read by the Mk IV (the NVCx cards).

Although not a simple task, there were usually cables and a test card provided with Mk IV panels. By connecting the proper cable to the card and following the procedures in the Mk IV Maintenance Manual, one could test the various cards. This would be an option if you don't have three spare NVCx cards to exchange with the ones in the Mk IV currently.

markvguy

 

Could you tell me vibration pick up how much voltage generate? and what is the range of voltages? Is it possible there is any relation between voltage & vibration reading?

 

Most velocity-type (seismic) vibration pick-ups used by GE generate a millivoltage that is directly proportional to the movement, e.g., 150 mV/IPS (Inch-Per-Second). There are too many vibration pick-ups used by GE to compile a list of them all without access to GE's electronic file system.

Depending on which SpeedTronic turbine control panel is used, you can look at the scaling Constants for the vibration pick-ups to determine the sensitivity of the vibration pick-ups connected to the panel. In the Mk IV the scaling constants are Control Constants; in the Mk V the scaling constants are I/O Config. Constants. (Be forewarned about the Mk IV vibration scaling Control Constants--often the scaling is expressed in IPS/Volt...which is the inverse of V/IPS, which is how most manufacturers specify the sensitivity of their sensors...)

markvguy

 

Mark V guy, can I have your email. I too look after gas turbines and u seem to be an expert at it. I will really like to interact with u. or u. can mail me on [email protected]

 

One of the reasons for contributing to control.com is to SPREAD information to as many people as possible. An email address is provided for exchanging files and other information which can't be shared or transferred via control.com.

Answering one person's questions via email helps only one person. One of the nice things about this forum is that many people can read other's posts and learn from other's questions and the responses posted to them.

Note that everyone is free to contribute answers as well as questions, and it would be nice if others who have experience with some of these problems would share their knowledge. Many people will have experienced many problems and can help others to understand some concepts (droop speed control comes immediately to mind, eh?).

It has been this author's experience time and time and time and time again that quite often through the act of writing a request for help and including all the information which might be required including all the troubleshooting done up to that time that the answer has become quite obvious (hint, hint, hint; great big HINT).

One more thing to remember about contributing to Control.com--there's no charge for contributing questions, and no one receives compensation of any kind for responding. This is becoming a very time-comsuming endeavor, and one of the things which makes this satisfying is feedback from people who have been helped--or not. Hopefully, quite often we provide meaningful responses; sometimes it's a complete miss. It would be nice if people would write back and say, "That was helpful." or, "That wasn't even friggin' close!" Everyone can benefit from that--seeing what works and what doesn't. The lack of feedback is extremely frustrating as its difficult to tell if the problem has been solved or the issue is explained satisfactorily.

Remember: There's no such thing as a dumb question--just a dumb answer. There are incomplete questions (not enough information), and there are questions which some might consider less than "worthy" but every well thought-out question and request is welcome here.

One can almost be guaranteed, though, that others have had similar questions, or that your question can help or inspire others.

So, please don't be shy. Ask your questions here; help others with you answers and feedback. If you've learned from the questions and responses here, be assured others have--and can--also.

markvguy

 

Could you please tell me name of the control constant?

 

Later Mk IV turbine control panels used Control Constants to scale the input from the velocity (seismic) vibration sensors (pick-ups). If the panel(s) in use at your site have the feature, you will find it in the vibration sensing algorithm where the inputs enter the algorithm. It will also be listed in the 'Control Specification, Control System Settings' document, Sect. 09.

Some Mk IV systems may have used hard-coded scaling; again, the information was usually specified in the 'Control Specification, Control System Settings' document.

markvguy

 

Please aware of the natural frequency of the rotor. I mean wheather the equipment is running above the first critical speed. so always there may be high vibration while the machine croses its natural frequency.

 

Sometimes appears on GE frame 9FA Gas turbine high axial vibrations at 0,313X (other times appears at 0,375X) during a piloted-premix combustion mode. It's a rumble problem? how can it be solved?

 

You should contact the OEM/packager for assistance with problems related to DLN combustion systems, including any "unusual" vibrations experienced during combustion-mode switching or continuous, steady-state operation.

Only the OEM/packager knows for sure.

markvguy