GE Frame 9E Turbine Support Leg Bulge


Thread Starter


I would like to ask if anyone had experienced similar issue. When we turned on the cooling water pump, the center body of the turbine aft support leg is observed to be bulged slightly.

We do understand the material of the aft support leg is stainless steel and is thin (of about 3mm according to GE). Thickness gauging and metallurgy test found that the material of the aft support leg is indeed in healthy state.

We do experience water leak from the edge of the aft support leg in which we thought it is due to material fatigue. However, we are continuously experiencing water leak from the support leg and the cause of the leak always come from the weldment. Hence, we are suspecting that the weldment crack could be due to this slight bulge from the support leg as we only experience this weldment failure during start up (when cooling water pump runs when start up command is initiated), and not even at firing temperature of the gas turbine which removes the possibilities of failure due to material failure (high temperature).
So, this topic has come up before on

(Found using the cleverly hidden 'Search' feature of and the search term:

+"turbine support leg" ... )

Shutting off the cooling water circulation prematurely sounds like that's the root cause of the problem. Repair should be rather easily affected.
Yes i have came across this thread and read through the post. In fact, khairulsulaiman is my superior in my company.

But the main question that i am asking is whether the bulge on the turbine support leg is a usual sight? Or is it an abnormality that i am observing?

It is not normal for aft support legs to experience "bulging."

It is very difficult to imagine how a solid piece of steel (which is what all GE-design heavy duty gas turbine aft support legs I have ever seen are made of) could develop a bulge. The aft support legs I have seen which were water-cooled all had a welded jacket that water flowed through to cool the legs. Now, it's possible the design of the aft support legs has changed, but given that they have to be fairly rigid in order to support the weight of the turbine without flexing and require cooling to limit growth to prevent misalignment and vibration, and GE's penchant for keeping with proven designs it's difficult to think the design has changed much.

However, since Belfort now has responsibility for Frame 9 designs, and their proven record of needlessly over-complicating and changing decades-old designs and practices just because they can, well, just about anything is possible.

The real question, though, has not changed and remains the same: How did the bulge occur, what's being done to evaluate its effect--if any; and, how is it being prevented from occurring again? Because it's not normal, it could have adverse effects (reduced cooling leading to misalignment and increased vibration), and steps should be taken to prevent the bulging from occurring in the future.
Current procedure to reduce the "bulge" is by slow opening of the cooling water inlet valve to reduce the pressure from the pump. and fully opened it once the cooling water system is running. That is only the method i could think of as of now.

Replica test on the aft support legs shows that there is no issue with the metal and no metal thinning as well. So this really has baffled us.

I have probably been misunderstanding the query; correct me if I am wrong. You and your colleague are saying that the cooling water jacket, which is welded to the outside face of the aft support leg, is bulging when the cooling eater system is pressurized. And, when the cooling water system is depressurized the jacket material "de-flexing?"

Is that correct?

Without being able to see exactly what's happening my guess is that a very thin material has been used for the jacket, thinner than used in the past. This may, or may not, be a problem, only time will tell for sure.

How long has the unit been in operation before the bulging was noticed?

There was some mention of leaking weld seams; was the bulging noticed before the repairs were affected?

The situation bears vigilance, and perhaps some consultation with the unit packager.

If I now understand the problem correctly, if there has been no increase in vibration on the load coupling end of the turbine then it would seem the the aft support leg is receiving sufficient cooling and the situation just requires diligent monitoring to ensure the jacket doesn't rupture or develop leaks which reduce cooling effectiveness leading to misalignment and increased vibration and resultant damage.

If you are concerned the material is too thin and may weaken with continued flexing over time then you should work with the packager to determine a more acceptable material for replacement at a convenient time. In the final analysis, it's your Company's comfort level with the material's ability to not result in an outage causing lost production and reduced reliability. If the unit is not in warranty and the packager is unwilling to correct the situation to your satisfaction then you are free to make changes/repairs as necessary to achieve your desired level of comfort and satisfaction.

Hope this helps!
Hopefully, there are a response in this old thread.

I've GE Frame 6 with Turboline controller, I've encounter CW Leak at one of two turbine support leg, after this leak happened I found that Vibration BB1 and BB2 are slightly increase (1.8x to 2 - 3.4x mm/s) and CPD (compressor discharge pressure) gradually decrease from 9bar to 8.7bar in 4days. Also fuel consumption and exhaust temp are increase even though the load is still the same.

Is there any explanation how leak at turbine support leg can effect to CPD, or there are any possibility that causes CPD decrease and fuel consumption increase?

Thanks a lot for your response

Most of the GE-design Frame 6B heavy duty gas turbines I have worked on had water cooling jackets on the aft turbine support legs--nearest to the turbine and gas turbine exhaust. The purpose was to limit vertical growth of the support legs to limit misalignment between the gas turbine and the Load (Reduction) Gear. If one support leg wasn't receiving sufficient cooling I would expect to see higher vibration on the #2 bearing, the one in the Load Tunnel, and closest to the aft turbine support legs. I would also not be surprised of the input shaft of the Load Gear vibration increased at the same same time.

It's hard for me to find a correlation between lack of cooling water flow to one aft support leg and increased vibration at the #1 bearing area, though I guess if the alignment between the turbine and the Accessory Gear was marginal to start with this might occur. But, we really don't know the extent of the cooling water leak at the aft support leg.

Next, the fuel flow measurement systems usually installed and provided with GE-design Frame 6B heavy duty gas turbines are pretty crude--and I would be very surprised if you could see any significant heat rate changes using one of the typical fuel flow measurement systems.

The CPD decrease seems fairly large, but there could be a LOT of reasons for that. And, since I'm totally unfamiliar with the Turboline turbine control system I can't comment on how the load could remain the same with a significant change in CPD and exhaust temperature (unless the unit were NOT operating on true CPD-biased exhaust temperature control: Base Load AND the unit WAS operating on some kind of load control signal).

Many times multiple issues arise--or are discovered at the same time--and they are wrongly believed to be related. I suspect that's what's going on here. But, there are just too many things we don't know about how the unit is operated, what else might have changed recently, what the condition of the machine is, what the ambient conditions are, what the compressor cleanliness is, etc. There are just SO many intangibles and things we don't know it's very hard to say. AND, I have seen some non-OEM turbine control systems which were programmed to put out the same power (MW) regardless of ambient conditions or CPD because the Customer believed the turbine should ALWAYS produce nameplate rated power when operating at Base Load (and the PLC programmers didn't know any better, and everyone was just trying to get the Customer to accept the new turbine control system and move on to the next job). Again, we don't know if you're referring to Base Load, or some kind of load control reference--there's just a lot we don't know.

Hope this helps--but in my experience, a cooling water leak, unless it was pretty significant, would not result in higher-than-normal vibration at the opposite end of the unit AND increase fuel consumption for the same load AND a decrease in CPD. I think too many things are being wrongly believed to be related. (I have been wrong before, and I will be wrong again. I am human, after all.)

Please write back to tell us how you fare in resolving these problems!
Many Thanks CSA for your response,

First, for vibration, If I'm not wrong, the cooling in support leg is provided to avoid any thermal expansion at the support leg. I suspect, when one of support leg has cooling water leak, it's impact to un-balance thermal expansion between two support leg and will lead to slightly change in alignment. But, I'm agree with you the vibration that have possibility to increase is BB4 and BB5.
In my case, the BB1 and BB2 is just slightly increase, and there is still possibility the cause is not cooling water leak in support leg.

Second, the load is remain the same since we run in PRESELECT Mode, which is it's maintain load in certain set point by maintaining the speed of turbine (FSRN / FSRACC), but in some point, the temperature control will take over the control in case exhaust temperature (TTXM) near the TTXRB.

for heat rate change, I use the flow meter for fuel that outside of GE flow rate measurement system,

Currently, we already repair the leak at support leg.
Also, we replace the inlet air filter, check and test bleed valve open-close, and check IGV's opening.
The CPD seems already increase, but the heat rate still higher (compare at same load) compare to before leak or decreasing CPD happened.

Is CPD decreasing affect to increasing of heat rate?

Misalignment can be caused by thermal expansion (growth), yes?

And, thank you for that excellent tutorial on Droop Speed Control FSR and CPD-biased exhaust temperature FSR and the interaction between the two.

Heat rate is only guaranteed by the OEM at Base Load, which is when the unit is operating on CPD- (or CPR-) biased exhaust temperature control, AND when the IGVs are fully open to their maximum operating angle. If the IGVs are not fully open, then the unit is not at optimum heat rate. And, it should also be obvious that the calibration of the IGV LVDTs is also very important, especially at the maximum operating angle.

I do believe that for the same operating conditions if CPD decreases the heat rate will increase (more BTUs/MW). And, when the IGVs are not open to their maximum operating angle then CPD is being restricted, and the fuel consumption for the same operating conditions will increase. If a gas turbine exhausts into an HRSG (Heat Recovery Steam Generator; "boiler") then it's often desirable for the exhaust temperature to be as high as possible when the unit is NOT at Base Load. To do that, the IGVs are held closed longer during loading (and closed earlier during unloading) which maximizes the exhaust temperature which maximizes the steam production. The gas turbine is slightly less efficient (poorer heat rate) when the IGVs are being used to maximize exhaust heat, BUT, the overall thermal efficiency of the plant increases because of the increased steam production. For units which don't exhaust into an HRSG there is no need to maximize exhaust temperature at part load, so the thermal efficiency of the gas turbine at part load is slightly higher than if the IGVs were being used to maximize exhaust temperature.

Axial compressor cleanliness is also VERY important for maximizing efficiency, and an off-line axial compressor water wash is very helpful for restoring efficiency. High humidity and lots of dust (road dust; dust from a nearby industrial plant or cement plant) and airborne hydrocarbons (such as from a nearby refinery or a nearby road with lots of lorry traffic) can cause the axial compressor to foul which decreases CPD (for the same speed and IGV angle) and increases heat rate, I would imagine, at part load.
Thanks a lot CSA for your brief explanation,

Honestly, I'm new in gas turbine and my background is chemical engineering but I want to learn about gas turbine and its control.

In practical, how often we need to do offline cleaning for axial compressor with water wash?

during gas turbine running with gas fuel, is atomizing and air purge impact to combustion?

Thanks a lot

NO ONE likes my answer to the question of how often to perform an off-line axial compressor water wash. But, it is always the correct answer. Presuming the unit runs at Base Load the ambient temperature, load and CPD need to be plotted on a daily basis. (It's best to try to record the values at approximately the same time each day. It's also good to record the turbine inlet air filter differential, also (to monitor the effect filter cleanliness (or the lack thereof has on unit performance).

When the load for the same approximate ambient temperature drops to a level that management seems unacceptable--it”s time to shut down and perform an off-line axial compressor water wash). The time required to cool the turbine before water and detergent can be sprayed into the inlet, as well as the list revenue from electrical generation and steam production (if applicable). It all needs to be factored into the determination of when to perform the wash, because it's all about money: cost, revenue and payback from the increased power generation after the wash.

The decision is different for just about every site. You could just wash every week, or every third week. But, if you are diligent about recording the above data and comparing the results after the washes you will probably find you are either washing too often, or maybe not enough.

As ambient conditions change throughout the year (temperature, humidity, airborne contaminants (dust; soot; vapours from nearby industrial plants; etc.) the data will indicate when it's economically time to do the water wash--which means the intervals between washes will usually change throughout the year as conditions change. If the revenue for MWh changes, and/or steam production, that will also affect the decision and the intervals.

So, there is no easy or simple answer--because, And I guarantee this, if you just wash on a regular basis some bean counter (often dressed in engineer's clothing) will demonstrate to some higher-up how the site is losing money with this time-based method and/or could be making more money with a data driven decision--because there always someone looking to climb over you to get up the corporate ladder.

If you NEVER run liquid fuel, yes, you could marginally improve heat rate by eliminating atomizing air and purge air. But you cannot just temporarily shut them off when not burning liquid fuel without damaging fuel nozzles and/or combustors and or piping/tubing systems.

Hope this helps!