Gas Turbine High Gas Consumption RCA

I am a graduate engineer working on a project to carry root cause analysis on high gas consumption of a GE MS5001P Gas Turbine.
Please, what parameters/factors do I need to look into, to figure the root cause?

The gas turbine is a refurbished one but recently overhauled of which the thermal block was replaced with another refurbished one.
But there's no improvement in the efficiency, the same level of gas consumption before it was overhauled.
Hi, Lusco,

Gas turbine efficiency is difficult to measure, and very often difficult to troubleshoot, also. There are many, Many, MANY factors involved in gas turbine efficiency.

First, it's really only possible to properly measure and troubleshoot efficiency when the gas turbine is operated at Base Load (also known as CPD- or CPR-biased exhaust temperature control), which is how the manufacturer guarantees performance (efficiency). It requires some sophisticated equipment for measuring fuel flow-rate and power output and axial compressor inlet air flow as well as ambient temperature, barometric pressure and ambient humidity. The gas turbine manufacturer publishes documents for measuring performance and correcting for ambient temperature, barometric pressure and humidity. This makes it somewhat possible to take measurements at different times of the year (when ambient conditions may be different) and compare them to each other, because a measurement today compared to a measurement three months ago or six months ago can seem very different when in fact they are very similar when corrected back to manufacturer's conditions.

It's important to gather the generator electrical output data properly, too, which is also documented in the manufacturer's publications. (Sorry; I don't know the GE publication number for the performance measurement procedures and calculations; but it is readily available many places on the World Wide Web.)

Axial compressor cleanliness is VERY important to gas turbine efficiency; a dirty compressor--and inlet guide vanes--and inlet air filters can have a very bad effect on performance/efficiency. A dirty axial compressor, and/or inlet guide vanes and/or inlet air filters wastes energy. It's not just turbine parts which affect performance, though they can also have an impact--it's very much about the cleanliness as well as the condition of the axial compressor (clearances; etc.).

So, how you measure the various parameters, when you measure various parameters, and total machine condition (including inlet air filters) can all have a significant effect on performance.

Another factor which can greatly affect performance is the source of the parts used in the refurbishment. If they are purchased simply from the least cost supplier (of which there are many for GE-design Frame 5 heavy duty gas turbine parts) the tolerances and clearances may not always be exactly like the original equipment manufacturer's (OEM's). Also, how the parts are assembled (the care with which they are put back together and the consumables used to assemble the parts) can also have a large effect on performance.

You really need to detail how the performance is being measured (at what operating conditions--for both the "original" data AND the post-refurbishment data), the condition of the other parts (time since last replaced; time since last off-line compressor water washing), and how the data was taken (hopefully using the same instrumentation, with no change in calibration/scaling).

Hopefully this will give you some idea of how the data is (should have been) gathered, and whether or not you have good data to begin with. You can usually find the OEM's performance correction curves in the Operation & Service Manuals provided with the turbine and accessories. (Usually, the performance monitoring procedure is also in the manuals; maybe not the latest revision, but still a usable document.)
Thank you so much for your response
For the compressor there was an offline washing during the overhaul and air inlet filters were also changed
Same operating ambient conditions for the post refurbishment and original data although the air humidity has never been monitored

But I will have to make more findings about other concerns raised like the sources of the parts used in refurbishment, instrumentation and scaling used in gathering both data, performance correction curve and others
I will also try to get GE documents for measuring gas turbine performance and correcting for ambient temperature, barometric pressure and humidity.
Hi Lusco,

Totally agree with CSA, as there can be so many factors affecting the performance of a Gas turbine.
Consider the Fuel specifications also one of them.
you can check if there is any change in the fuel composition as well since the efficiency of the turbine is degraded.

and keep posting your findings.

Happy learning.
Thanks Dodo.

The fuel composition is 93% Methane and other constituents take 7%.

It is a bit difficult troubleshooting the cause of the turbine high gas consumption.
1. The plant was bought over and there was no proper handover of the documents. As a result of that, we could not established the performance at commissioning which i felt was quite important.
2. No available correction curve to reference performance at various operating conditions
3. The expected performance of the newly assembled refurbished parts were not in anywhere given by the TA

Now, from my close monitoring of the unit I observed that there are many issues with the unit like the high load gear vibration, high bearing drain temperature, lube oil header pressure of 1.85bar(avg), and uneven exhaust spread. But we have recorded an optimal average efficiency of 26%(92% attainment of ideal heat rate) on a favourable day with an ambient temperature range of 24.4degC - 33.5degC against ISO of 28.4% at 15degC. However, comparing this unit with two of our units which have not undergone any major overhaul, it seems to perform poorly as they have better gas consumption, heat rate and efficiency. And so the argument has been why are unoverhauled units outmatching an overhauled unit in performance and efficiency.

Of course, i understand that solving the different issues associated with gas turbine will optimise the performance but what could be the reasons it has a lower efficiency as compared to an unoverhauled unit is still a puzzle i am trying to solve most especially now that i do not have necessary data to work with.

You are in a very difficult predicament; you are a graduate engineer, so you are expected to have many, if not all of the, answers and methods to solve just about any problem. BUT, you (probably) have little practical experience (which is difficult to obtain in university for a field such as gas turbine operation and design) and--it would appear--even LESS data with which to work.

Usually when people experience low performance after a maintenance outage it's because of:

1) Improper reassembly of the hot gas path parts,


2) Incorrect parts, poorly refurbished parts, or parts manufactured with poor tolerances and/or poor manufacturing methods.

If you combine these two possibilities, the potential for low or poor performance increases, sometimes greatly.

If you, or your supervisors, want to be serious about this issue you need to install a second fuel gas monitoring method to confirm the data you currently have. Most GE-design heavy duty gas turbines are supplied with a precision metering tube and orifice to measure fuel gas flow rate. One can install a manometer in parallel with the existing differential pressure transmitter(s) to confirm the differential pressure being detected by the differential pressure transmitter(s).

And, since the accuracy of any orifice-derived measurement is totally related to the condition (and even the orientation and position) of the orifice in the metering tube, it is a good idea to remove and inspect the orifice, making accurate measurements of the orifice internal diameter AND inspecting the condition of the edges of the orifice plate (they should be "sharp" and not worn or smooth). And, the orientation of the orifice plate should be noted WHEN IT IS REMOVED, and when it is re-installed. I have seen MANY orifice plates improperly installed over the years--too many to recall. People just make assumptions about how the orifice plate is to be installed without any basis in fact, and without asking or confirming how it's to be installed. And, so, they get installed incorrectly, and they also get installed such that they are not well centered in the flange faces of the precision metering tube. This is THE critical component of the fuel gas flow measurement method (for MOST GE-design heavy duty gas turbines--yours may be different) and if there is a concern about performance it needs to be confirmed in order to begin a proper analysis and review.

Next, you need to be getting quality data about the output of the generator--and that means verifying the meter or transducer(s) used to indicate the amperes and volts and real power (Watts; kW; MW) out of the generator--as well as the reactive power (VArs; kVArs; MVArs). About the best way to do this is to learn how to use the watt-hour meter in the generator control/protection panel to calculate real power. It's a very simple formula, and it can be found in the GE documentation for performance measurements. (I don't have quick access to the document or I'd provide it here. I'll check and post it if I find it.) One needs a stop-watch and a patient individual to record disc revolutions. (I'm presuming the unit has an electromenchanical watt-hour meter, with a rotating disc movement/mechanism--sounds archaic but it's highly accurate and time-tested. New digital meters are also pretty accurate, and they can be verified using sophisticated equipment, and in some cases, PC-based software.) One needs to try to gather real power data during performance measurements when the power factor is as close to unity (1.0) as possible. So, if that means the site needs to coordinate with the grid/utility to operate for a few hours at unity power factor, so be it.

Also, the back pressure on the gas turbine exhaust needs to be measured--and that's not the easiest of tasks. But, if the unit exhausts into a heat recovery steam generator (HRSG), or a "boiler," then the back pressure could be higher than it should be. Thing of trying to breathe out wearing a dust or surgical mask while you're running--it's not easy. If the mask is dirty or contaminated then breathing out (and in!) could be very difficult, and it will affect your ability to run at your best. It's the same for a combustion engine--it has to have clean inlet air filters to draw air in through, and it has to have a reasonable (normal) exhaust path, unobstructed and clear, in order to move air most efficiently. And moving air is key to gas turbine output--so the exhaust duct back pressure has to be low and reasonable.

And, IGV angle. The definition of Base Load (which is where the manufacturer guarantees heat rate/performance) is when the IGVs are at the maximum operating angle AND the unit control system is operating on CPD-biased (or CPR-biased--which is a fancy CPD measurement) exhaust temperature control. So, the actual position of the IGVs is critical. The scaled ("calibrated") IGV LVDT feedback can be grossly in error with respect to the actual IGV position, and so it's critical that the scaled ("calibrated") IGV LVDT feedback be confirmed to be nearly equal to (as close as possible) to the actual IGV angle at the maximum operating angle reference. This can only be done with a machinist's protractor and trained personnel--trained on how to position the IGVs to take the necessary measurements and trained on how to use the protractor to obtain the actual IGV angle to compare to the scaled IGV LVDT feedback. Not an easy task, but not impossible, either.

So, the place to start is to confirm the accuracy of the measurements being used to make the performance calculations. There is no other way to proceed. AND, this is NOT just for the machine in question--it's to be done for ALL the machines which are being compared to each other.... ALL the machines. You have to compare apples to apples (like to like)--so quality, vefiriable/verified data is critical.


After that, if it is determined the performance is less than it should be and less than acceptable, then a group of people need to decide how to proceed. This means someone--a group or committee, likely--has to sit down and discuss how to proceed. If it's decided that some action needs to be taken, then someone has to be designated to start gathering data about the condition and provenance of the equipment which is currently installed in the machine. I'm referring to axial compressor blades (including the IGVs and IGV bushings)--both rotating and stationary, the first- and second-stage turbine nozzles, the first- and second-stage turbine buckets (that's what GE calls turbine blades), and the condition of the gas turbine exhaust diffuser. Parts condition, and the quality of the parts used (and there are a LOT of people and companies manufacturing hot gas path parts for GE-design heavy duty gas turbines because there are SO MANY of them around the world) is very critical, as well as the assembly of them. The source (provenance) of the parts--not JUST the parts installed in the machine during the last outage, but any hot gas path parts (liners; transition pieces; turbine nozzles; turbine buckets; axial compressor parts; exhaust diffuser parts--to name most but still just a few)--must be taken into account.

MANY Frame 5 parts vendors claim, "Our parts are just as good as GE's--at a much lower price!" That's true in some cases, but not all cases. I've been to sites inspecting problems just like this one where I learned (after much investigation and after gaining the confidence of some of the site personnel) that the mechanical department had firmly persuaded the sourcing department to purchase parts from a particular vendor, and a couple of the mechanical department personnel had received monies in return for the business. It was impossible to prove, but because the mechanical department was ADAMANT that there were NO issues with the parts and the supplier REFUSED to provide specifications and materials lists AND the mechanical department said even if the supplier provided the information they wouldn't review it--it was very clear that something was amiss and everyone agreed to do a better job of sourcing on the next maintenance outage, getting spec's and documents BEFORE any purchases were made or contracts awarded. If you're new to the site, no one is going to hold that against you. You have to learn about people (70% of any technical job is people-related--maybe more; the best technical people are also the best at understanding their situation (managers and subordinates) and suppliers and constraints, and working with everyone to achieve the best result.

You say you don't have performance correction curves. If you have a typical GE-design MS5001P heavy duty gas turbine there are likely thousands of copies of performance correction curves in existence around the world. Even if you can't find them in the Operations & Service Manuals (you DO have unfettered access to the Operations & Service Manuals, RIGHT???), you should be able to find them on the World Wide Web or from a service provider, or even from GE or the packager of the turbine-generators at your site. They are NOT very turbine-specific, especially for 51P units--it's the WORKHORSE of the GE fleet; hundreds of them were built and sold and installed around the world. They are beasts--meaning they are well-designed and capable of producing power under some of the worst operating conditions (not always with the highest efficiency--but pretty reliably, and that's the name of the game in power generation: reliability).

Your situation is not clear, and I may be making some assumptions (I work VERY hard at NOT making assumptions--but I fail sometimes; I am, after all, human). But, if you and/or your colleagues and superiors really, Really, REALLY want to get to the bottom of this--you have to start somewhere. And, ensuring the accuracy of your data is the first place to start. You may find, you've been working with bad data. You need to ensure all of the control system instrumentation--in particular the CPD transmitters and and barometric pressure transmitters (if the unit uses them)--are all working properly. You want to have an even exhaust profile--meaning you don't want to have high exhaust temperature spreads (the differences between the highest and lowest exhaust temperature readings). You want to make sure the IGVs are truly at the maximum operating angle, and that the LVDT feedback is properly scaled ("calibrated"). You want to make sure you are using accepted industry practices for measuring performance metrics, and you want to have proper correction curves. (If you don't have unfettered access to the Operations & Service Manuals, I would say something is amiss on the site. Again, you can find a lot of generic Frame 5 information on the World Wide Web if you can't get access to the Operations & Service Manuals. AND, you need to sit down with the Manuals and slowly go through each and every one of them. You want to have a pad (or two) of sticky notes ("Post-It notes")--and you want to paste them on pages and information you want to return to. Go through the manuals, skimming them for information, from beginning to end. There's a LOT of information in them--it's just not all in an intuitively obvious place. Plan to spend at least an entire day going through them (maybe you want to do it two or three or four hours at a time)--but DO familiarize yourself with them. You will be much more informed and knowledgeable if you do. You can make copies of things you want to keep close at hand and study and make notes on. But, you need to use the manuals, and to use them properly you need to know what's in them, and the only way to know what's in them is to page through them slowly, one page at a time, or in some cases, one section at a time (there are some very thick instruction manuals from equipment manufacturers that may or may not be new and informative to you)--that's the ONLY WAY to know what's in there and to be able to USE what's in there.)

Okay; I'm signing off this thread. I can't add anything more to it based on the information provided. ONLY YOU can gauge the seriousness of the effort to truly understand what may or may not be the problem, and take appropriate action. If I find the GE publication about performance testing, I will post the publication number here. And the formula I referred to for using the watt-hour meter disc revolutions to calculate real power output is in that publication.

Best of luck! It would be great if you could let us know how you progress in this effort, and what you find, if anything. But, I suspect you are going to (ultimately--maybe in a few months, or a few years) that you were given an impossible task and that not everything was as it was explained or you understood. It never is in cases like this.
To the Moderator...
I apologize for posting to the wrong thread. My eye-sight is not what it once was.

To Lusco...
Are you looking for a Figure-of-Merit, i.e., compare a machine's performance before and after some overhaul ? Or compare one machine's performance to another ? Or compare a machine's summer to winter performance ?

To Mr. CSA...
Very few words !

Regards, Phil Corso!
If CSA has convinced you I'm too incompetent to provide an answer to your problem, then so be it. But, if you want to evaluate my competence yourself, then contact me at [email protected], and I will send you my resume !
Regards, Phil Corso
1) My History With Areo-Derivative GTG!
In 1964, as an employee of Esso Research & Engineering (now Exxon-Mobile), the 1st Aero-Derivative GTG in an Esso refinery was being commissioned. We were tasked to test flare-gas, and other gas-fuels. A newly acquired IBM-360 was to crunch very complicated numbers. Unfortunately, it was assigned a higher priority project! So, a simpler method, called ”Figure-Of-Merit” was developed. The “rest of the story” is detailed in the following paragraphs.

2) ‘Key Word’ For Solution to Lusco’s Problem Is "Energy"!
I think most on this forum will agree that ''Performance" is related to “Efficiency''. Defining Output-Energy as NRG.OUT, and Input-Energy as NRG.IN, then the efficiency formula in percent is % = 100x(NRG.OUT/NRG.IN). But, keep in mind numerator and denominator must have the same unit-dimension, i.e., kW·hr!
Obviously, NRG.OUT is readily known, even metered. But, determining NRG.IN isn’t simple, as was painfully expressed in CSA's 4-page treatise!

3) Figure-of-Merit (FoM) As Measure Of Performance!
The Figure-of-Merit doesn’t need the same unit-dimension for both numerator and denominator. Just something simply related to the amount of fuel-consumed (F), or Hours of Operation (H)! Either one, or both, are then individually compared to output kW·hr (W), yielding two ways to measure performance, one from knowing Fuel-Consumption… FoM.FUEL= F/W, the other from knowing Operating-Hours… FoM.HRS = H/W. Both will reveal an up or down trend in Performance!

No matter the method used, the data has to be verifiable. I may be painfully treatisizing the importance of gathering verifiable data, but it's still important--nay, critical--nonetheless.

Here's a good resource I found:

Here's another with a publication number of an ASME document/procedure:

In what appears to be yet another change to their websites, GE has seemingly made some of their publications unavailable on the World Wide Web. I have searched high and low and using many different search terms and at least two search engines--and I can't locate the basic GE gas turbine performance test procedure publication. I won't stop looking, but if you have a unit that was originally provided by GE, you should be able to find it in the Operations & Service Manuals originally provided with the unit(s). It's a generic document which is applicable to any Frame size, so it's ubiquitous. If anyone reading this thread has it in their Manuals and can attach a copy to this thread, that would be great!