Compressor Rubbing

L

Thread Starter

LKFAI

Dear all experts,

For your information, our unit is 350MW single shaft GE9FA coupled with D10 Steam Turbine. Our current operation mode is peaking plant for supporting national grid. Currently, we are in the midst of annual outage for GT borescope inspection. Upon inspection, we had observed several failures at compressor casing, rotor blades and stator vanes.

Details of the defects observed as below:

- Rubbing and mushroom tip at upper halve compressor vanes from S1 to S17

- Rubbing and mushroom tip at upper halve compressor vanes from R1 to R17

- Rub tip condition at R1.

- Casing rub condition observed at upper halve side. From S4 to S17, rub area was observed around 10 to 2 o'clock position. Casing rub area was reduced around 11 to 1 o'clock position at S3 and S2.

However, there's no rub observed at S0 and R0.

According to GE, there are a few possibilities contributed to this phenomena. It might be due to increment numbers of hot start in operation, improper cooling in GT compartment and hot gas leak from GT horizontal joint. However, we found there's no burning sign in GT compartment and no gap at GT horizontal joint. The unit was put on cyclic operation since last year and having frequent hot starts. Nevertheless, there's no major findings observed during 2014 borescope inspection.

They had proposed for compressor blades/vanes blending to remove and reshape the mushroom tip. Blending works will be carried out in-situ by removing Compressor and CDC casing. Thus, the work only involves rotor blades and upper halve stator vanes due to insufficient outage duration for the work on the rotor removal. This decision is made, considering our major inspection scheduled in May / June 2016.

I would like to know if any of the users here have info or experienced with this failure before. What you had done to remedy the failure and have you discover the failure root cause? Any other recommendations given by GE to remedy the failure?

Thanks.
 
Hello,

This is (primarily) a controls-related forum. To the extent that mechanical field devices and instruments are part and parcel of the control system we occasionally engage in discussions about mechanical issues--but this one is far outside the scope of control system-related field devices.

F-class turbines (including FAs) were not originally designed for cyclic, load-following, dispatched, or peaking operation. They were envisioned as Base Load generation assets--particularly single-shaft "STAG" units. As such, the OEM is encountering unanticipated problems and issues with these units which have been placed in modes of operation well outside their design scope. Further, 9F/9FA machines have come under the purview of the GE Belfort operation, and a lot of their engineers are still coming up to speed with these very large, and somewhat temperamental, machines and their operating characteristics.

While this particular OEM has amassed a lot of operational and historical data about F-class machines (again including FAs) they do encounter issues and scenarios and problems which they haven't in the past. While this OEM likes to say that F-class technology is a "mature" technology, one can safely argue that it's still relatively new and given the various iterations and issues the fleet has had over it's relatively short life that there are still conditions and issues which have not come to life--meaning the technology still has a ways to go to be truly considered a mature technology.

Further, over the years these units have been produced there have been a LOT (and I do mean a LOT) of changes to machine design and materials and manufacturing processes--a mind-numbing number of changes, to be polite. It's incumbent on the owner/operator of these machines, some of which have been retrofitted multiple times with different compressor components (vanes/blades; IGVs; EGVs; etc.) to keep very close track of the iterations and current state of their machines and to make sure the OEM is keenly aware of the current state of the machine and its components.

As OEM personnel responsible for various components and systems change there is an inherent learning curve for them just to come to grips with the large number of possible combinations of designs and components, much less to try to remember the particulars of every site. They need to be reminded from time to time of the current configuration of the machine--even if they are responsible for machine care due to some kind of LTSA or CSA (Contractual Services Agreement).

As for turbine compartment cooling, this has a ring of truth to it. This, too, has been an ongoing topic of discussion both inside and outside of the OEM. There are lots of things to consider in turning gear operation, not the least of which is the design of the compartment and the flow dynamics of cooling air circulating in the compartment.

Again, most of these machines were never designed for cyclic peaking operation, which finds them more frequently on turning gear than ever anticipated and for longer periods of time than ever anticipated. It might be prudent to perform a CFD analysis of the compartment to determine if the desired amounts of cooling air are reaching the appropriate areas--or, rather, if too much cooling air is reaching some areas.

This is not really a controls-related issue, though it might be possible that some modifications to the sequencing of compartment cooling fans may alleviate or eliminate the problem. It's really more of a controls philosophy problem--as well as an operational problem, that the operating mode of the unit is significantly different that originally intended.

(I imagine, from your post, and from past experience, that the plant was not originally intended to be a peaking plant but rather a Base Load plant, and that because of circumstances the mode has changed. The OEM does have some experience with these drastic operating mode changes, but it's not extensive--and again, given the permutations of components and compressor designs its virtually impossible to aggregate groups of machines because there's so much variability between individual machines.)

I wish the news was better. I suggest you continue to work with the OEM to understand the particulars of the configuration of your machine, the compartment cooling system, and the operating mode changes and to develop a plan for monitoring and mitigating the issues you are experiencing.
 
Thank you CSA for your reply,

We did boroscope our compressor last year in May and found no rubbing between the compressor casing and the blades with cyclic operation as well. However, there is one incident of CO2 discharge in May this year inside the GT compartment. Can it be one of the reason that could have cause the rubbing happen?

thanks and regards
LK Fai
 
LK Fai,

You say you borescoped the unit last year--but you didn't say how long the unit had been in this peaking mode of service before the borescope inspection....

If the turbine was running and hot when this happened and cold CO2 came into contact with the compressor casing, this could possibly happen (but you didn't tell us what the state of the turbine was when the CO2 was discharged....).

Usually, when CO2 is discharged when the turbine is running, the turbine trips and coasts down to zero speed. Usually, the unit doesn't automatically go on COOLDOWN (turning gear) because the Aux. L.O. Pumps are shut down to prevent L.O. from a leaking pipe from feeding the fire. So, it's incumbent on the operators to take appropriate action as soon as possible to either get the unit on COOLDOWN, or to leave the unit at zero speed to allow the turbine shaft to eventually straighten before putting the unit on turning gear and eventually attempting a START.

All of this would be important and relevant information if you wanted to analyze this as a possible cause of the damage you are reporting. We would need to know all of the steps that occurred and were taken when the CO2 discharge occurred--both automatic and operator-initiated, for a period of at least 36 hours. Including when and for how long turbine compartment doors were opened after the CO2 discharge was complete, what the internal turbine/compressor temperatures were at the time the compartment doors were opened.

It's most likely that a combination of things (change of service mode; operating changes; this CO2 discharge; etc.) is responsible for the damage you are reporting. Especially if this has gone on undetected for some time (lack of increased vibrations; lack of noticeable degradation of power output (decreased CPD); etc.).

It's natural to look for "THE" cause in cases like this, but, what should really be done is a thorough examination and analysis of operations logs and historical data for a long period, probably back to the previous borescope, to understand how the turbine has been operated and what conditions have changed during the time (such as vibration, CPD decrease, etc.).

Best of luck with your endeavour!
 
Top