Hi there,

I am trying to find out how the air flow into a gas turbine is measured? I have looked through other threads on this website and have been unable to find the answer I am looking for.

I am specifically looking for the method that the OEM (preferably GE) uses to measure air flow during site visits, such as the site acceptance test or the performance guarantee done at commissioning.

In my research so far I understand that a calibrated bell mouth is really required? I understand this is done during the factory tests but that the bell mouth installed on site is not a calibrated one? If this is correct, how is the airflow measured/calculated on site during performance testing? And, if anyone knows, how is the airflow measured/calculated when a calibrated bell mouth is used?

If anyone can help with my inquiry it would be greatly appreciated.

Kind regards.

 

Apollo,

I've commissioned scores of GE-design heavy duty gas turbines, and participated in many performance tests of both new and existing units. Granted, I've been out of the field for a few years now but I don't things have changed very much.

In my experience, the larger "Frame" units don't usually have very precise air inlet flow measurements; they are really fairly crude approximations just used as checks against other performance criteria. Accurate fuel flow-rate is much more important, as is exhaust flow. And, accurate compressor discharge pressure measurement is critical to determining the performance of the Frame units. When inlet air flow is measured, it's usually done with water-tube manometers connected to special sensors (called "Kiel probes") inserted into opening in the inlet duct work. And by measuring the site ambient air pressure and static inlet air pressure and the pressure drop at the IGV inlet, along with ambient pressure they can roughly calculate inlet air flow.

I believe the aeroderivative units (LMs and PGTs) do use something called a calibrated bellmouth set-up, but I've never actually seen one in use. For these machines, the Variable Stator Vane angle is measured to assure it is adjusted ("calibrated") correctly and because these are basically aircraft engines (the gas generator sections, anyway) there are much more detailed methods of calculating air flow from VSV angle and compressor discharge pressure.

Inlet air flow is a function of many things, including inlet air filter cleanliness, bellmouth cleanliness, axial compressor cleanliness, axial compressor clearances, IGV position. There are a lot of variables, and I don't know exactly how one would accurately measure inlet air flow at any given point in time.

A lot of these kinds of tests and measurements are really intended to be viewed over time, in other words they are to be trended with time as the horizontal axis. Presuming the instrumentation and sensors are maintained in good condition and calibration then what can be observed is a change over time, which would indicate a dirty compressor or dirty inlet air filters or some other condition which is reducing inlet air flow versus a previous measurement some time prior to the current measurement.

GE offers what they call a Performance Monitoring Package, but, again, it's primarily a group of sensors used to gather data to be analyzed over time--not an instantaneous measurement of inlet air flow.

What I'm trying to say is that without a lot of other machine information I don't know of a way to accurately calculate instantaneous inlet air flow to any degree of accuracy.

But it's not clear exactly what you want the measurement for.

 

I found myself nodding along to everything CSA said (no suprise there).

Actually I found the Operations/Performance Team at my plant asking the same question some time ago. I also reached the conclusion that the Flow rate is indicative at best, and should be considered to be more of a qualitative - even though the units and the number of significant points in the decimal number would have you believe otherwise. GE isn’t very good at documenting many things, this happens to be one of them. I wonder if the Control Specification document on your site gives some details.

The flow rate is basically a function of
A- the delta pressure between the upstream and downstream of the belmouth (This is the primary contributor)
B - Mass flow rates of fuel/water injection etc (minor contribution)

The flow rates are compensated for ambient conditions.

The Delta P is created by the geometry and flow in the bellmouth <pre>
~~~~~~~~
        ~~~~~~~~
                ~~~~~~~~


AIRFLOW --->



                ~~~~~~~~
        ~~~~~~~~        |
~~~~~~~~                |
|                       |
|                       |

|                       |
|<------Delta P---------></pre>
I distilled the logic for the flow calculations in the logic into an excel file so I could play around with the numbers and try to find the "perceived" source of the error. I found that the calculation is EXTREMELY sensitive to the difference in pressure. In other words, small changes in the delta P (on the order of the measurement error of the Delta P itself!) produce large swings in the reading.

I am attaching the excel sheet. It uses the native toolbox engineering units. This will help you understanding the relationship of the various variables to each other.

https://www.dropbox. com/s/uitcw81wtk2v5kk/FLow_calculation.xlsx (remove the space after dropbox.)

 

Hello,

try reading "ISO 2314:1989, Gas turbines - Acceptance tests" or "ASME PTC 22:1997, Performance Test Code on Gas Turbines".

In my 6FA gas turbine, compressor inlet air flow is a calculated value obtained from the compressor bellmouth differential press, the inlet air total press, the compressor temperature-inlet flange, the specific humidity and a series of constants.

 

REI,

How does your site use the calculated air flow from the instruments in the inlet?

Also, has your site ever verified the accuracy of the air flow measurement (not the individual inputs--but the calculated result of the inputs)? If so, how?

 

I don't understand the relation of CPD and exhaust temperature. My concept is, when cooling air flow will increase then firing temperature will reduce and to make it constant as it was before. we will increase the fuel to get more power. Then how exhaust temperature in reduced????
e.g<pre>
Ambient temp Firing temp Exhaust temp Fuel
20 1000 550 FSR:50%
15 990 540 FSR:50
15 1000 550 FSR:52%</pre>
these are assumed values. can u please elaborate CSA?

 

CSA,

first of all sorry for my English. It's not my native language and it has been a long time since I last wrote in English. I'm not quite sure what your question is, hope this answers it.

The compressor inlet air flow (AFQ), from what I've seen in the CSP, is just used to calculate de exhaust mass flow from a simple mass balance (EXHMASSOUT &#8776; AFQ + FQL + WQ), where:

- FQL is the measured liquid fuel mass flow from the flow divider magnetic pickup.
- WQ is the measured water injection flow.

To calculate AFQ the CSP uses the performance monitoring instruments (compressor bell mouth differential press. transmitter [96BD], the inlet air total press. transmitter [96CS] and the barometric press. transmitter [96AP]) and the compressor temperature-inlet flange [CT-IF]. Approximately, AFQ &#8776; SQRT[[(AFPAP-AFPCS-AFPBD)/(AFPAP-AFPCS)]•[(AFPBD/CTIFR)•(AFPAP-AFPCS)]]. A series of constants intervene in the sequencing I believe some are tuned during commissioning, others for engineering unit conversion. By looking at freelanceng’s excel sheet I see his unit uses the compressor bell mouth area for the air flow calculation mine doesn’t.

The instruments are verified at least with every scheduled maintenance inspection. If you’re asking if the different constants that are used to calculate the air flow have been changed after commissioning, I don’t think so. If you’re asking if we have used a dedicated instrument for measuring the air flow and then compared it with the calculated value of air flow, the answer is no.

 

REI,

The question was how does your company use the air flow calculation, if at all. Is it just used by the Speedtronic for its purposes, or does your company use the value for any other purpose?

What you have described is GE's typical method of calculating air flow, and many times buyers pay extra for this information in what's called a 'Performance Monitoring Package' with the turbine. I have only seen a couple of sites that actually attempt to use the information but have found that is difficult to verify and only useful for monitoring performance degradation over time. They usually find that monitoring CPD versus MW proves just as useful and far more verifiable.

 

To measure Compressor air inlet flow, Along with these above parameters we have to consider the filter DP (96TF). as we know if the filters are chocked with dust, air flow will be reduced. And i think angle of IGV should also consider, as the angle increase more air enters into Compressor and AS IGV angle reduced air flow will reduced.

 

REI,

you can’t evaluate Air flow, AFQ with just pressure transducers. you have to include the bell mouth area KAREACBI, Coefficient for Compressor Inlet Air Constant, KAREACQC , ratio of sp heats AFRSH and gas constant, AFRSHGC in the equation. which I think you mean by (A series of constants intervene in the sequencing I believe some are tuned during commissioning).

another method to estimate the air flow is the energy balance around the GT (like in ASME and ISO codes). I tried that and there is always a difference between the two values the energy balance involve a lot of measurements, the fuel flow the LHV compressor discharge temp, exhaust temp ... i.e. a lot of uncertainties.

as CSA says the problem of accuracy is still there I think if one have this calculated AFQ and an accurate flow meter in one machine can tell us how accurate this calculation is.

but I believe also that trending this value with time can be used as a compressor fouling indicator together with power output and cpd as in online performance monitoring modules.

 

Ok,

no, we don't use the air flow calculation for any other purpose.

 

Salmanism,

taking a look at the main constants that intervene in the sequencing of air flow measurement in the different units we have on our site I see these differences:

1º Our 6Bs don't calculate the air flow into the compressor.
2º Our 6FAs only use AFKQG (compressor inlet flow constant).
3º Our 6FAs+ use KAREACQC (coefficient for compressor inlet air constant) and KAREACBI (compressor bell mouth area).

The logic I described was from the 6FAs. I don’t know if GE uses the bell mouth area to determine AFKQG. If it does, it is outside the CSP.

Take care.

 

Hello friends;

Good day. I read this post and also download the .xlx sheet which an expert upload.

We r using solar INC. gas turbines. what are the units is using in the excel sheet. Is there any document in which GE instrumentation position shown by its tag discription like 96CS-1 transmitter and etc., Because i want to check where from the data collect?

Regards

 

Hello Experts,

i want to calculate air inlet mass flow in pps, of Gas turbine specifically solar inc.

while i have
T1=84.3F
P1 14.40 psia vacuum is excluded from it,
T2= 639.6F
i don't know bell mouth area

can anyone help me out this or also for other parameters which are useful for this calculation?

 

Hello
i am talking about solar inc. centaur GT curve model 4702 between Hp(SL)& inlet air temperature in deg F i have seen in this curve that Exhaust temperature is is plotted in inversely proportional, while Exhaust flow, fuel flow are in straight line.i want to know the reason of the behavior of exhaust temperature in the curve?

 

> can anyone help me out this or also for other parameters which are useful for this calculation?

Hello Saam;
The only way to calculate the air flow of the gas turbine by using energy balance if you have not know the bell mouth area.

Regards

 

Hi,

I want to know how calculation the inlet air flow for GTG FRAME 9E type DLN, because we haven't any calculation measure formula for this type.

I want to said, we have GTG type FRAME 9E without DLN and we have formula for this type. I think is not corrected for DLN type, so help me for this problem.

For example, we have 96BD, 96AP, 96CS for GTG non-DLN, but in DLN type we have just 96CS, 96AP, but we haven't any transmitter for 96BD and there isn&#8217;t any calculation measure formula for air flow calculation.

Thank you for reply.