GT Tripping High Vibration on Part Load


Thread Starter

Manoj Agrawal

We have Hitachi Based Gas turbine 25.6MW base load;

we have been having a recurring problem.

Once GT is started load variation is taking place freely without much change in vibration; we have a dusty environment; and have constant problem of compressor fouling.

In one of our Gas turbines which was running for 2 months load was reduced from base of 23 MW to 15MW; it was found that machine vibration increased above 155 and machine tripped; Bearing 1 vib (Turbine side) was higher than Bearing 2.

it is recurring; and once offline water wash is done; vibration is normal after startup.

did someone have a similar problem?
any suggestion for rectification.


This only occurs on one of several gas turbines? Or on all the gas turbines (how many does the site have)?

And, after the off-line compressor water wash the problem doesn't occur? Does that mean you have tried shutting the unit down the same way as before the wash immediately after the wash?

How often do you normally perform off-line compressor water washes? Every month? Every two months?

Does the site perform on-line compressor washes (which aren't really very effective, in my personal opinion, but make sense in some situations)?

There really isn't a lot of information to go on here. If the vibration lessens after the water wash it would seem the vibrations are related to compressor cleanliness. And, it would also seem that the root cause of the problem is the inlet air filters don't seem to be doing their job.

Higher vibration on Bearing #1 would seem to go along with a dirty compressor.

It's very common for owners and operators of gas turbines to find that the air filters originally supplied with the equipment are not adequate for the site's environment. They work with the OEM and/or filter suppliers to find a more suitable and better filter. I know of sites that have changed filter vendors several times over several years until they finally arrived at a workable and economical solution.

I also know that some inlet filter structures and inlet duct work only used stitch welding for construction purposes. Entering the structure and duct on a bright sunny day revealed LOTS of light between stitch welds--meaning that there was a LOT of area for dust and dirt to enter the inlet after the filters.

I have also seen inlet filter poorly installed causing dirt to have a path to enter the unit.

Do the units use evaporative coolers? Is the ambient normally humid? In other words, is there some "moisture" that makes the dust in the air stick to the compressor?

What is the nature of the dust at your site--is it from a dirt road, or from a nearby cement plant or other industrial process plant?

Again, it would seem there is something about the site conditions and the type of filters being used and possibly the construction of the filter structure and/or inlet duct that may be allowing a lot of dirt to get into the unit.

And, again, sometimes the originally-supplied equipment doesn't always meet the requirements of the site. And that means the owner/operators have to do some investigation and troubleshooting and work with the OEM/packager and vendors to find solutions that meet the site requirements.

But, based on the limited information provided it would seem you have made a link between compressor cleanliness and the high vibration. That would seem to lead to an effort to try to understand why the compressor is getting so dirty and if anything can be done about, which would alleviate the vibration issue. It may be as simple as washing the compressor more frequently until such time as the filtration issue can be improved.

Please write back to let us know how you fare in your problem resolution.
Dear Sir

Thanks for your reply. Here are answers to your questions

1. This only occurs on one of several gas turbines? Or on all the gas turbines (how many does the site have)?
Ans: We have 5 GTs onsite but this problem is occurring in only one of our GTs (GT no 1)

2. And, after the off-line compressor water wash the problem doesn't occur? Does that mean you have tried shutting the unit down the same way as before the wash immediately after the wash?
Ans: We had kept machine at varying loads just after water-wash to check whether this problem occurs or not; but just after water wash the problem did not occur. The Vibration phenomena has happened twice; in a gap of around 2 months; yes we are also considering compressor fouling has something to do with it; IGV manhole was opened and a lot of dust was found in the compartment and on IGV as well.

3. How often do you normally perform off-line compressor water washes? Every month? Every two months?
Ands: Vendor has suggested a gap of 1.5 to 2 months but normally we see the loss of load in the machine; which usually used to come around 2 months; but the loss of load has increased since we have installed new filters; Filters in GT-1 were installed in July but its DP has not increased since; We had called the filter company official for analysis and report is awaited. I do agree with you that we need to find a suitable filter for our GTs; we are in process of that; we have already tried 3 companies, AF/Cam fill/Donaldson.

4. Do the units use evaporative coolers? Is the ambient normally humid? In other words, is there some "moisture" that makes the dust in the air stick to the compressor?
Yes we do have evaporative coolers; we have checked RH after Evap coolers; which is around 70%; RH of ambient atmosphere is varying but rarely goes above 60%.; we still may consider your idea that moisture might cause sticking of dust in compressor.

5. What is the nature of the dust at your site--is it from a dirt road, or from a nearby cement plant or other industrial process plant?
Ans: Well that's our biggest problem; our plant is situated besides a coke-yard and lot of fine coke dust is present in the air. We have been struggling to find a decent filter but change to newer filter company has worsened our condition. Is there someway that a filter could let lesser amount of air into compressor without actually increasing DP across it? because once we had installed these filters our GT base load just after water-wash also reduced slightly. (we are suspecting that the newer filters are allowing less air to come inside and still not removing dust, doubling our troubles).

But still we have had issues of compressor fouling since many years; but this is the first time machine has been tripping on vibration due to this fouling.

Could you throw more light on our issues now; let me know any particualr data I can provide you on a personal ID
It sure sounds like there is a serious dust problem, and that can present unique challenges. I'm sure people at the site have been working on solutions for as long as the units have been in service (as well as living with the problem).

Is GT1 closest to the coke yard? Does it get more coke dust because of prevailing winds or traffic than the other units?

If you opened the Inlet access and found lots of dust in the compartment, then I would suggest the problem is not limited to the inability of the filters to remove the dust from the air stream.

On a clear day it should be possible to enter the inlet duct, downstream of the filters, and with the access door closed it should be fairly dark inside the ductwork. Look at all the seams (welded seams; bolted flange "seams"; etc.)--anywhere you can see daylight is a means for dust/dirt to enter the inlet. You should do this for horizontal and vertical duct sections.

Again, many welded seams were only stitch-welded--meaning, the weld wasn't continuous, but was intermittent. Over time, as sun and ambient conditions and stresses on the inlet duct from wind and sun and cold can cause the metal to twist and shift and can cause the welded seams to open, and even cause welds to crank and break.

You should also examine the inlet filter structure to see if there are any leaks--again, doing this in the daytime when it's bright outside and any leaks will allow sunlight through, is best. Even poorly installed filters can cause leaks through which dust/dirt can enter the inlet duct.

If you're only seeing a 10% increase in humidity it would seem the evaporative coolers aren't operating as efficiently as they should. Again, there is this false perception that evaporative coolers are "self-regulating"--and that's just patently false. The flow-rate across the media must be monitored, as well as excess water puddling downstream or in the bottom of the inlet duct, and water flow-rates adjusted. Changes in ambient temperature and -humidity will also affect evaporation rates, as will changes in IGV angle. So, water flow-rates across the evaporative cooler media need to be monitored and adjusted. There is no automatic control for this--not yet, anyway.

Increasing the humidity of the inlet air by any means is probably not helping the problem at all. This is what is allowing the dust to collect on the compressor leading to the imbalance.

Turbine packagers have almost NO real-world turbine operating experience. They design and build turbines; they don't operate them. They make estimates about operation, but those estimates are typically based on isobaric conditions--and do not take into account extremes of ambient conditions, such as the heavy dust at your site.

Off-line axial compressor washing, properly performed, is the best way to restore efficiency lost due to dirt/contaminants in the compressor. However, it comes at a cost to generation, and revenue.

On-line compressor washing is marginally effective at restoring performance, and at some point after multiple on-line washes there is no perceptible change in performance, at which time the unit must be shut down to perform an off-line compressor wash. And, on-line washing is really only effective on the first few stages of the compressor because as the air/water is heated during compression it evaporates reducing its effectiveness on later stages of the compressor. So, on-line compressor washing can only delay off-line washes.

No packager can tell any particular site exactly what the frequency of compressor washing should be. Every site has to determine for itself when the loss of efficiency is more than can be tolerated, and when the cost of shutting down and performing an off-line compressor wash is economically acceptable. It's not something which can be estimated by a packager, unless that packager has worked extensively with very similar sites and has relevant experience. About the best they can is provide guidelines and estimates for these kinds of activities (compressor washes) and site personnel have to decide for themselves when it makes sense to perform washes.

EVERY site is different--as much as gas turbine operators and owners don't want it to be so. To be specific, ambient conditions at every site are different--and that includes dust and contaminants. Some sites have coke dust; others cement dust; others road dust; others hydrocarbon contaminants from a nearby refinery or chemical plant; salt air from the ocean; and so on, and so on. Some sites can use pre-filters--like "socks" which can be placed over the outside of the main filter elements and replaced periodically increasing the life of the main filter elements. Other sites need "louvers" or "hoods" to help knock out entrained dust/sand or even rain water before it can get to the filters.

But, <i>based on the information provided</i> it would certainly seem that you are trying to find better filters, but you may not have examined the inlet filter house and duct work for areas where dust can enter the inlet. Again, we have no way of knowing that for sure--but it's a good place to start if that hasn't already been done.

I also believe that some GE-design heavy duty gas turbine axial compressor rotors have internal passages through which cooling and sealing air flows to various parts of the machine (wheelspaces; etc.). It could be that dust collects in those passages which contributes to the imbalance.

Finally, there are just some sites that have extremely unique circumstances and those sites have to develop their own operating procedures and guidelines. And, you should continue to work with filter suppliers to find a filter that meets your needs. You may have to live with a slightly higher than normal differential pressure and more frequent filter change intervals, but, again--sometimes site conditions warrant these kinds of changes to operating practices. You may even want to investigate some kind of "hood" to force the air to make several changes of direction before reaching the filter elements--like a cyclonic filter. Things like that make prove extremely useful--or not. Relying on the packager to help with this probably isn't going to be very productive, so you might want to find third-party consultants and contractors to provide you with several alternatives and make a choice.

I don't think there's any data you could provide us without being able to experience the site and it's conditions which could be helpful in analyzing the problem any further. Again, based on the information provided, it would seem the vibration is associated with compressor cleanliness. Keeping the compressor clean would seem to be the best way to avoid high vibration. So, working on the root cause would seem to be the best way forward.

Hope this helps!

Manoj Agrawal

Thanks a ton!!

we will definitely look into checking of welding seems on inlet side duct. we are in consultation with filter provider and will arrange a third party analysis for our problem;

lets hope we reach a solution soon.

Manoj Agrawal