OPTIMIZATION OF PERFORMANCE PARAMETERS OF GAS TURBINE USING TAGUCHI BASED GREY RELATION ANALYSIS Saira Alam 1,* , Younis Jamal 1 , Asad Naeem Shah 1 Mustabshirha Gul 2 , and Asad Raza Gardazi 2 1 Department of Mechanical Engineering, University of Engineering and Technology Lahore-54000, Pakistan 2 Department of Mechanical Engineering, UCE&T Bahauddin Zakariya University Multan- 60000, Pakistan. Abstract: This study is mainly based on the technique of Taguchi method with grey relational analysis (GRA) for optimizing the multiple performance input parameters of Gas turbine. Such techniques are normally used for conversion of multiple- objectives problem into a single objective by using Taguchi based Grey relational analysis. The influence of process parameters such as inlet temperature, type of air inlet filters, and machine rotor rpm on power, heat rate, specific fuel consumption (SFC) and thermal efficiency is addressed in the study. Through this technique an optimal setting of input performance parameters was obtained as performance index. Orthogonal array (OA) L 9 (3 4 ) was selected as Design of experiment (DOE), so nine experiments were performed in this way. Subsequently, the signal-to-noise (S/N) ratio and the analysis of variance (ANOVA) were employed to find the best input process parameter levels, and to analyze the effect of these parameters on power, heat rate, SFC and thermal
35
Embed
research paper of saira alam-Revised by Dr. Asad N Shah.doc
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
OPTIMIZATION OF PERFORMANCE PARAMETERS OF GAS TURBINE
Grey-Taguchi optimization techniques have also been used in CNC turning process [27, 36],
casting process [28], machining and milling processes [23, 29-31], submerged arc welding [33]
and energy management systems [34].
Though Taguchi based methods have been vastly used in developing of industrialized
sectors [35, 37], but their applications to scrutinize the optimum parameters and thus to evaluate
performance of a gas turbine are in infancy. The purpose of this research is to reduce the specific
fuel consumption, heat rate and to increase thermal efficiency of the gas turbine. Presently, no
such technique except Grey-Taguchi method is available which may simultaneously be used to
solve and convert multiple-objective parameters into singular and optimized output parameters
of gas turbine
1. Materials and Methods
1.1 Experimental Setup:
In current study, experiments were conducted on two shafts Gas Turbine Model No: T-
4502.GT consisting of axial compressor, annular combustor liner, mixed flow two stage turbine
(GP), and mixed flow single stage gas turbine (PT).The axial flow air compressor consists of
eleven stages as shown in the Fig 1.
Figure 1. Schematic diagram of test bench
The GP and PT are not mechanically coupled to each other, GP runs independent of the PT,
and PT runs due to the exhaust of the GP. The other performance specification of the Gas
Turbine is given in the Table1.
Table 1. Performance specification of the Gas Turbine
Configuration Specifications
Fuel Gas Liquid
NGP RPM 15000 15000
NPT RPM 15500 15500
Minimum Horse Power 3902 3785
Heat Rate 9629 BTU/Hp-hr 9736 BTU/Hp-hr
Minimum compression Ratio 8.6 8.6
Compressor Mass flow 37.5 37.5
In Gas turbine both liquid as well as Gas can be used as fuel but in this study Natural Gas
was used as fuel is used at pre set temperature and pressure. The experiments were performed on
the performance testing facility of the SNGPL Multan, Pakistan. The PT was coupled with
dynamometer through flexible coupling fully equipped with the instrumentation gadgetry as well
as mechanical equipments. Dynamometer is used for the torque measurement which is then used
for the calculation of the horse power of the Gas Turbine. The load variation on the power
turbine during experimentation was done with the pressurized water controlled through pressure
control valve impinging on dynamometer. Fuel control valve was used for measuring, and thus
controlling the supply of the fuel during ramping and accelerating of the engine. For the
measurement of gas turbine speed in rpm, magnetic picks were hanged on OEM at recommended
location of MPU to send signals in terms of frequency to control the system for subsequent
calculation.
Figure 2. Schematic diagram of test bench
In Fig. 2 station 1 indicates the location of measuring the air inlet filters differential in
which there is an alarm setting mechanism through PLC, in order to observe the filter health. For
the control of the air inlet ambient temperature, air inlet cooling techniques were adopted and
thus evaporative cooling system was installed in the air inlet system. Station 2 shows the inlet of
the compressor, while station 3 is the exhaust of the axial flow compressor. The combustible
gases enter into the turbines (GP+PT) at station 4, and finally the exhaust of the gas turbine is
released at location 5. Table.3 shows the experimental activities carried out by varying the filter
type and rpm of machine which were increased and decreased continuously according to the
need for defined ambient temperature. Air inlet temperature was controlled trough evaporative
cooling system to make essential modification for hot climate, as air inlet temperature is the
primary parameter which affects the performance characteristics of the machine [38, 39].
Following instruments were used during the experimentation.
Pressure Transmitter
Thermocouples, RTD
Pressure Gauges
Lube oil Flow meters
Fuel flow meters
1.2 Grey-Taguchi Methodology
The Grey-Taguchi method categorizes the important factors which compose the best
contributions to the variation. The output parameters depend on several input factors. So Grey-
Taguchi is applied to design and perform the experiments for the assessment of output responses
without running the process at all possible combinations of given input values. The Grey-
Taguchi method determines the correlations among actual and desired experimental data and
converts numerous quality characteristics into single grey relational grade (GRG) [29]. It allows
knowing the optimal combination of parameters which may affect the performance of gas
turbine.
The optimal design parameters affecting the gas turbine performance are determined and
then are controlled easily. The no of levels of the input parameters to be varied is required to be
specified. Increase in no of levels also increases the no of experiments to be performed. Input
parameters used in this study are ambient temperature, gas turbine speed and types of air inlet
filters.
Table 2. Experimental Factor and Factor Levels
Control Variables Code
Level Output to be observed
1 2 3 HP
Temp(°F) A 84 60 56.6 ɳth
Speed(RPM) B 15000 14700 14400SFC
Filter Types C Conical Barrier Cartridge Heat Rate
The purpose to use the Taguchi method was to choose the suitable orthogonal array OA in
order to investigate the all parameter at three levels. Three factors were assigned to definite
column in the OA for analyzing the main effects. An OA was used for the DOE on the basis of
L9 (33). Table 3 shows the L9 (33) OA for the experimental work.
Table 3. Orthogonal Array L9 (33)
Run No A B C
1 1 1 1
2 1 2 2
3 1 3 3
4 2 1 2
5 2 2 3
6 2 3 1
7 3 1 3
8 3 2 1
9 3 3 2
Table 4 shows Experimental values for the four outputs and these are taken for
optimization in accordance with DOE.
Table 4. Design of Experiments
Run NoTemp Speed Type of Filters
OA T1 OA Speed OA Filter Type
1 1 84 1 15000 1 Conical
2 1 84 2 14700 2 Cartridge
3 1 84 3 14400 3 Barrier
4 2 60 1 15000 2 Cartridge
5 2 60 2 14700 3 Barrier
6 2 60 3 14400 1 Conical
7 3 56.6 1 15000 3 Barrier
8 3 56.6 2 14700 1 Conical
9 3 56.6 3 14400 2 Cartridge
1.3 Normalization of Experimental Data/Statistical Analysis:-
As the main objective of current work is to increase the thermal efficiency and horse Power
keeping the heat rate as well as SFC at their minimum values, so experimental data was
normalized by using higher the better & smaller the better criteria [36].
The considered criterion was completed with respect to better quality aspect of interest
[37]. So S/N ratio for the quality characteristics (higher the better) was formulated as follows.
Whereas the assessment signs of the Power and thermal efficiency to the ith time and n is the
repeated experiment number. As SFC and heat rate are the smaller the better quality characteristics hence calculation of the Smaller the better is formulated as
Actual origin of series and series for evaluation as So and Si (l) =1, 2, 3…, t where t= No of
experiment to be considered respectively Where l=1, 2, 3…….u and no of observation is u.
If the desired objective is to maximize the outputs then the normalization of original sequence is
given as follows.
For the objective to minimize the output then “smaller the better” criterion is as follows,
Now smaller the better is as under,
Target Value= OT.
Equation (iii) depicts original series sequence.
Where is the actual series the series after data giving out max the
biggest value of & min the lesser value of
As the experimentation was performed on the basis of Gray relational analysis, the GRG for
analyzing the most effective parameter formulation of grey relational coefficient (GRC) is
related as follows:
As ∆min and ∆max are the minimum and maximum values of ∆oj ( k ) sequences. ∆oj ( k ) difference
between the original values and the estimated normalized values of data are quality loss function.
The quality loss function is used to investigate whether a certain feature is lying among a given
exact limits or not.
Then GRG is the average of GRC corresponding to the given responses. The overall multiple
response characteristic of a process depends on the calculated GRG, given as follows
1.4 Analysis of variance (ANOVA).
The ANOVA was performed to find out the numerical implication of the performance
affecting optimizing parameters. The results were examined to establish the primary effects of all
the factors, and thus are tabulated by the sum of the squared deviations from the total mean of
the OGRG. It is revealed that input parameter with larger F-value has a significant effect on
multiple output responses.
ANOVA is performed after the analysis of results to approximate the most significant input
parameters which are tabulated on the basis of mentioned formulation. F-value shows the most
significant effect of input parameter on the outputs. [32].
Where SSA is sum of square of each of factor.
SSp is sum of square of process parameters
2. Results and Discussions.
2.1 Output Response based on OA
Table 5 representing the results obtained after experimentation carried out on the basis of
DOE Orthogonal Array. Nine experiments were performed by taking three input variable with
the three different levels to tabulate the output parameters i.e. Horse Power, Heat Rate, SFC and