IGV Hand Wash Effect on GT Output and Heat Rate


Thread Starter


how to quantify the Improvement in output and HR of gas turbine solely after compressor blades (and vanes) by hand wash cleaning?

could any one provide the calculation methodology to quantify the improvement in output and Hr of the gas turbine if we do only igv and inlet plenum by manual hand wash only...(without considering offline wash)

It would be the same calculation methodology as for an off-line compressor water wash. You're quantifying the same parameters but only cleaning part of the axial compressor.
Dear CSA,

>It would be the same calculation methodology as for an
>off-line compressor water wash. You're quantifying the same
>parameters but only cleaning part of the axial compressor.

could you please explain a bit elaborately.

1). how to calculate offline compressor water wash and IGV hand wash together effect on gas turbine output and heat rate?

2). alone IGV hand wash effect on gas turbine output and heat rate?

3). is there any possibility to take out the fractional effect by IGV hand wash only on output and heat rate when offline wash and IGV hand wash done together?

Both operations are essentially the same thing: cleaning the axial compressor (inlcuding the axial compressor inlet (IGVs and inlet bellmouth). Some cleaning of the IGVs and inlet bellmouth are accomplished with the Off-Line Compressor Water Washing, but if the IGVs and inlet bellmouth are particularly dirty then manual washing augments (supplements) the effects of Off-Line Compressor Water Washing and usually results in a larger performance improvement than an Off-Line Water Wash by itself (if the IGVs and compressor bellmouth are really dirty--that is, if they IGVs/inlet bellmouth are not that dirty and the type of dirt isn't "sticky" or "gummy" (such as hydrocarbon-based contaminants then hand-washing dry dust doesn't usually add much to the operation).

Myself, I take a data set before Off-Line Compressor Water Washing (CPD, DWATT, FQx (Fuel Flow-rate for the fuel being burned)) at Base Load. Then after washing, I look at the effects on the same parameters at Base Load. The data sets are hopefully taken at about the same atmospheric conditions (temperature; humidity). They will usually show an increase in CPD and DWATT for a similar fuel flow-rate, indicating an increase in air flow through the axial compressor for the same operating conditions.

By taking data periodically between Off-Line Water Washes one can see a decrease in CPD and DWATT at Base Load as the compressor (including the IGVs and inlet bellmouth) gets dirty. Typically, most sites choose a percentage decrease in output performance (say about 2-3%) at Base Load at which they will shut the unit down, cool it, perform an Off-Line Water Wash after inspecting the inlet (IGVs and inlet bellmouth) to see if the amount and type of dirt/contaminant warrant a manual cleaning in addition to the water washing. Then, after the washing they start the machine, run it up to Base Load and take another data set to determine the effectiveness of the compressor cleaning (including the manual IGV/inlet bellmouth washing, if performed). If the cleaning was effective in removing most of the contaminants and restoring air flow through the machine and CPD and power output at Base Load then the wash is deemed good. If the percentage increase after washing/cleaning didn't restore machine output to near previous pre-cleaning conditions then the washing/cleaning probably wasn't done very well.

If you want to gauge the effectiveness of manual cleaning of the IGVs and bellmouth in addition to the Off-Line Compressor Water Wash, you can shut down, perform either operation (the manual washing of the IGVs and inlet bellmouth, OR the Off-Line Compressor Water Wash), then re-start the machine, load it Base Load and take a data set to see the effects of the operation performed.

Then, shut down the machine and perform the other operation (Off-Line Compressor Water Wash, OR manual IGV/inlet bellmouth cleaning, respectively), then re-start the machine, load it to Base Load, and take another data set to see the effects of the second operation.

Heat rate calculations are, well, similar. You calculate BTUs/KWhr at Base Load before the cleaning, and then after the cleaning. (Most sites calculate BTUs/KW or BTUs/MW, but regardless if it's KW- or MW or KW/hr or MW/hr either should be an indication of performance.)

There's no rocket science involved in what you're trying to do. Again, manual cleaning of the IGVs and inlet bellmouth would be an enhanced Off-Line Compressor Water Washing. If you want to measure the effect of either, you just need to gather the data and compare it (always at Base Load, and at similar ambient conditions/inlet filter dp, exhaust duct back-pressure, etc.). It's a rough calculation, admittedly, but it's sufficient for most purposes--unless there is some kind of performance guarantee involved, in which case there would be more detailed procedures and instrumentation (which is beyond the scope of this forum and would be particular to the situation and application).

Hope this helps!
Dear CSA,

Recently we had a package#4 upgradation to the compressor. our unit is 9fa gas turbine. package# 4 upgradation and HGPI done together. after outage we did offline wash and igv handwash manually...so we would like to separate each of its effect on gas turbine output and heat rate. GE given the effect of all together. offline wash for sure will boost the GT output. so our payment is only concern to package 4 upgradation as HGPI they have to as per the CSA (contracual service agreement).

Yeah; I had an extremely strong suspicion it was something exactly like this.

It's not possible after the fact to separate them. Any time an Off-Line Compressor Water Wash is performed the IGVs and inlet bellmouth should be inspected before and after the procedure to determine if they are acceptably clean after the procedure. If not, then they should be manually cleaned before returning the unit to service. The type of contaminants and dirt (including any L.O. from the #1 bearing) should always be considered when washing the compressor--which includes the IGVs/inlet bellmouth.

Just because the site procedures don't include inspection of the IGVs/inlet bellmouth before or after compressor washing doesn't mean it isn't necessary or shouldn't be done. When you wash dishes after a meal don't some meals require extra washing, and don't some dishes not get as clean as others sometimes--even requiring re-washing? Even automatic dishwashers have problems with some foods, especially if left on the plates/silverware to dry before washing. Washing an axial compressor is no different--some dirt/contaminants cause more of a mess than others, and require more effort.

If it was felt that manual washing of the IGVs/inlet bellmouth were warranted after the HGPI by GE (who don't usually take extra time for niceties if it's not truly warranted) then it's highly likely that they were excessively dirty--and we can't know that without having been told or without having been on site during the outage. If the IGVs/inlet bellmouth warranted manual cleaning then it's part and parcel of axial compressor washing to return the unit to the highest performance possible--which is what <i>most</i> Owners/Operators want after a maintenance outage.

Live and learn.

To expound on the 'Live and learn' statement. If, on some future maintenance outage, there is concern about performance increases after the outage versus before the outage an idea would be to shut the machine down a day or two before the outage, thoroughly clean the inlet and compressor (including the IGVs), start the unit and run it up to Base Load and gather data (pre-outage with a clean compressor and inlet), shut the machine down, complete the outage, start the machine back up, load it to Base Load, and gather data.

Another thing which can more greatly impact performance than IGV/inlet bellmouth cleanliness is IGV LVDT calibration accuracy. If the IGV LVDTs were calibrated after the maintenance outage and before the unit was re-started and performance data was gathered it's very likely that differences in calibration/scaling before- and after the outage could account for differences in performance. If proper LVDT calibrations were done, the as-found calibration accuracy would be noted on the Calibration Data Sheet, along with the calibration/scaling factors of the Speedtronic, and if a recalibration was deemed necessary the as-left calibration/scaling factors of the Speedtronic would also be noted on the Calibration Data Sheet for the IGVs.

Also, it's pretty common for the CPD- and inlet flow measurement transmitters to be "calibrated" during such a big maintenance outage. Differences in as-found versus as-left calibration/scaling should also be noted on the appropriate Calibration Data Sheets--which would be helpful in determining where some of the performance was gained/lost.

In fact, in writing this, I would also recommend checking/verifying the accuracy of the calibration of these very important instruments (IGV LVDTs; CPD transmitters; inlet pressure measurement transmitters) during the compressor/inlet washing procedure prior to the outage--just to be sure that the same instrument calibration factors were used before and after the outage.... to prevent inconsistency of readings and calibrations from affecting the data before and after the outage. In other words, calibrate all the appropriate devices before the outage--to the extent possible--and use them to gather pre-outage performance data as well as post-outage performance data, so eliminate them as possible sources of inconsistencies.

These are the kinds of things one learns as one lives with heavy duty gas turbines. With heavy duty gas turbines, it's not just one thing or another. And, since we're not there we don't know what you're considering, what you're not, what the conditions were before the outage, nor the guarantees involved and the data gathered before the outage. We can only answer the questions you have asked with the limited data you have provided, and try to anticipate other questions and situations based on our experience.

I know this doesn't help with your immediate problem, but this should be considered a useful learning situation for future outages.
Dear CSA,

Thank you very much for your prompt reply. I am sure we will follow your guidelines for then next machine...

Now we have another performance issue.

Our plant is configured like 3 blocks..for each block 2 gas turbines (9fa) and 1 steam turbine (GE D11). 2014 year end for one block 2 gas turbines comp pkg 4 upgradation and hgpi done. post outage results shown both gas turbines generating almost same output. but after 6 months during summer season we did base load test.. that time we experienced one unit is generating 3 MW less than other Unit. Now recently one week back we did baseload testing and we experiencing the difference is increased to 8 MW.

this was really confusing us...which parameters making this 8 MW difference.

because all the parameters for the unit which is generating less output are much better than the unit which is giving more output.

in our daily operation our machines are under AGC control(remote grid control) they are giving the load setpoint and the load is equally sharing between both the GT's

one GT can able to meet the desired set point with good operating parameters while the other cant..

some of the parameters are<pre>

DWATT 195 mw 195
IGV 66.8 63.2
Inlet filter dp 48.1 62.4 mm of h2o
AFPCS 64.7 79.1 mm of h20
exhmass 547 535 kg/sec
CTD 385 381
cpd 12 12
cpr 13.7 13.5
FSR 73.1 72.5</pre>
you can observe GT 2 at less IGV and even more filter DP and at less CPR can able to generate the same load set point.

i am thinking may be GT 1 requires tuning. can you please post your ideas and views how can i troubleshoot this problem?

You .... assume that all the instrumentation is identically calibrated, all the instrument tubing and isolation valves are identically clean and in the same positions, and that all Control Constants and I/O Configuration Constants are exactly identical, that all the internal clearances of both machines are exactly identical, the exhaust duct back-pressures are exactly identical, the fuel control valve LVDT calibrations are exactly identical, all the inlet filters (and an pre-filters, if present) are all exactly, identically installed with the same amount of leaks, the

If yours is the site I'm thinking it is, there has been so much switching of rotors and stators and compressor components and IGVs over the years, it's really hard to know what unit has what components.

I've seen insects build a nest in the atmospheric sensing line of inlet filter differential pressure transmitters and inlet air flow differential pressure transmitters and cause errors in measurement. The transmitter calibrations were correct, but using a water-tube manometer revealed the transmitter readings were incorrect, and that's when a more thorough investigation revealed the blockage caused by insect nests.

IGV LVDT calibrations can be very difficult to get done accurately, especially when multiple people/crews have responsibility for multiple units. Time allotted for LVDT calibrations can also be an issue, and extremely few LVDT calibrations (of any LVDT-equipped device) are done correctly (by actually measuring the physical angle and comparing to the LVDT feedback); far too many people believe the Speedtronic control systems are much more intelligent than they are, and don't understand what's required for an "Automatic Calibration" to be done accurately.

And, I would say the exact same thing about gas fuel control valve LVDT calibrations--extremely few are ever done correctly with too much reliance on AutoCalibrate. Further, some gas fuel control valves can be improperly assembled after refurbishment, and some refurbishers don't always use the proper parts with the proper tolerances. And, whenever there's flow through a system there is also wear--and lots of natural gas suppliers these days don't do much in the way of removing sand, rocks, liquid (like lubricating oil and even diesel and gasoline!) from the gas they supply to sites. So, wear on internal valve components can be hastened by poor quality fuel.

Exhaust duct back-pressure doesn't get as much attention as it should when considering performance. Think of trying to breath out with a surgical mask fitted to your face; it's not easy. And, if you're exerting yourself and breathing harder that makes breathing out even harder. But, many's the HRSG (Heat Recovery Steam Generator) I've climbed into which had insulation from behind plates blown out and found to be clogging superheater tubes, which impacted both the back-pressure on the gas turbine as well as the heat transfer and efficiency of the HRSG.

AND it's VERY common for the tubing/sensing lines for exhaust duct pressure (back-pressure) measurement to not be sloped and routed properly, allowing moisture to condense in the lines.

Inlet filters have the same affect on unit operation--it makes it difficult for the unit to breathe (in, in this case). And, again, tubing/sensing lines can also be different between units, and can have moisture condensate in them if not routed properly and periodically checked/drained.

It's not common for all units in a block to have the same maintenance outages at the same intervals; so the internal components can have varying amounts of wear and tolerance differences. Also, it's not very common for water washes to be performed at the same time on multiple machines, and with the same degree of cleanliness. Rinsing is important, and is not usually done properly (either because or time constraints or inattention to detail--usually time constraints).

Finally, you can have two cars, made one right after the other in the same factory which have varying degrees of performance and operation. There are a LOT of parts, all of which have to have the same tolerances and be put together with the same attention to detail--and that just rarely happens. The same is true for heavy duty gas turbines--or any major piece of equipment.

I liked your joke about tuning GTs--can I use it? Many people think there is some kind of air/fuel monitor which continuously checks for proper ratios and adjusts fuel/air as needed to maintain maximum efficiency. Sorry; the Speedtronic isn't as good as some inexpensive automobiles or scooters with fuel injection systems. They don't monitor air/fuel and adjust the ratio to maintain maximum efficiency. (That infuriates some Operations- and Plant Managers who just assume it's a digital computer-based control system which costs as much as a Ferrari or Aston-Martin and therefore it MUST be more sophisticated than a Tata or a Proton.)

Look, Mate, differences in machine operation are normal. We haven't even talked about how "coarse" the fuel flow-rate measurement system is on GE-design heavy duty gas turbines. They are very crude, and not meant for high-accuracy heat-rate calculations. And, we don't anything about how the machines are operated, how frequently they are started, stopped and tripped; how they are maintained; where the parts and service are provided from; etc.

We just don't have enough information to be of much help. It's entirely subject to plant- and ambient conditions, parts and assembly conditions and attention to detail, instrumentation quality and maintenance and calibration practices. There are just too many intangibles for us to be of much help with a problem like this. Yes; it would seem two machines of identical size and manufacture and installation and similar maintenance and operation would be more alike in terms of output and performance. But, it's just not typically the case. Unless the Speedtronic turbine control systems use what's called MBC--Model-Based Control. And then there will be all kinds of additional instrumentation and controls and logic and alarms to let you know if things are as they should be. And, STILL that all relies on attention to detail on instrument calibration and maintenance practices.

It's entirely possible that the performance degradation isn't even a gas turbine issue--it could be the exhaust duct pack-pressure.

As for troubleshooting--start with verifying the accuracy of LVDT calibrations, and checking all tubing lines for transmitter sensing to see they are free of condensation, insect nests and/or dirt, and all manual valves in the sensing lines are in the proper positions. There are so many possibilities and variables it's not likely it's just one issue--but several, which, when combined, are all working to reduce machine performance. (I've even seen problems with PT (Potential Transformer) and CT circuits which provide the crucial inputs to MW transducers have bad contacts and wiring and shorted turns.)

Wish the news were better--but that was a good joke you told!
Dear CSA,

My concern is why other GT can able to produce the same power output at lesser IGV. I hope following IGV opening remaining all following.

You are still assuming that all instrumentation and configurations and internal clearances are identical. <b>If</b> all things were equal, it would certainly seem there is something amiss.

Until you have hard data on which to base your analyses--which means a systematic and logical verification of operating parameters until you have arrived at the root cause of the problem--it's all just assumptions and speculation.