You are here

AN INVESTIGATION OF THE AUTOIGNITION OF POWER GENERATION GAS TURBINE FUEL BLENDS USING A DESIGN OF EXPERIMENTS APPROACH

Download pdf | Full Screen View

Date Issued:
2005
Abstract/Description:
Natural gas has grown in popularity as a fuel for power generation gas turbines. However, changes in fuel composition are a topic of concern since fuel variability can have a great impact on the reliability and performance of the burner design. In particular, autoignition of the premixed fuel and air prior to entering the main burner is a potential concern when using exotic fuel blends. To obtain much-needed data in this area, autoignition experiments for a wide range of likely fuel blends containing CH4 mixed with combinations of C2H6, C3H8, C4H10, C5H12, and H2 were performed in a high-pressure shock tube. However, testing every possible fuel blend combination and interaction was not feasible within a reasonable time and cost. Therefore, to predict the surface response over the complete mixture domain, a special experimental design was developed to significantly reduce the amount of 'trials' needed from 243 to only 41 using the Box-Behnkin factorial design methodology. Kinetics modeling was used to obtain numerical results for this matrix of fuel blends, setting the conditions at a temperature of 800 K and pressure of 17 atm. A further and successful attempt was made to reduce the 41-test matrix to a 21-test matrix. This was done using special mixture experimental techniques. The kinetics model was used to compare the smaller matrix to the expected results of the larger one. The new 21-test matrix produced a numerical correlation that agreed well with the results from the 41-test matrix, indicating that the smaller matrix would provide the same statistical information as the larger one with acceptable precision. iii After the experimental matrix was developed using the design of experiments approach, the physical experiments were performed in the shock tube. Long test times were created by "tailoring" the shock tube using a novel driver gas mixture, obtaining test times of 10 millisecond or more, which made experiments at low temperatures possible. Large discrepancies were found between the predicted results by numerical models and the actual experimental results. The main conclusion from the experiments is that the methane-based mixtures in this study enter a regime with a negative temperature coefficient when plotted in Arhennius form. This means that these mixtures are far more likely to ignite under conditions frequently encountered in a premixer, potentially creating hazardous situations. The experimental results were correlated as a function of the different species. It was found that the effect of higher-order hydrocarbon addition to methane is not as profound as seen at higher temperatures (>1100 K). However, the ignition delay time could still be reduced by a factor two or more. It is therefore evident that potential autoignition could occur within the premixer, given the conditions as stated in this study.
Title: AN INVESTIGATION OF THE AUTOIGNITION OF POWER GENERATION GAS TURBINE FUEL BLENDS USING A DESIGN OF EXPERIMENTS APPROACH.
0 views
0 downloads
Name(s): de Vries, Jaap, Author
Petersen, Eric, Committee Chair
University of Central Florida, Degree Grantor
Type of Resource: text
Date Issued: 2005
Publisher: University of Central Florida
Language(s): English
Abstract/Description: Natural gas has grown in popularity as a fuel for power generation gas turbines. However, changes in fuel composition are a topic of concern since fuel variability can have a great impact on the reliability and performance of the burner design. In particular, autoignition of the premixed fuel and air prior to entering the main burner is a potential concern when using exotic fuel blends. To obtain much-needed data in this area, autoignition experiments for a wide range of likely fuel blends containing CH4 mixed with combinations of C2H6, C3H8, C4H10, C5H12, and H2 were performed in a high-pressure shock tube. However, testing every possible fuel blend combination and interaction was not feasible within a reasonable time and cost. Therefore, to predict the surface response over the complete mixture domain, a special experimental design was developed to significantly reduce the amount of 'trials' needed from 243 to only 41 using the Box-Behnkin factorial design methodology. Kinetics modeling was used to obtain numerical results for this matrix of fuel blends, setting the conditions at a temperature of 800 K and pressure of 17 atm. A further and successful attempt was made to reduce the 41-test matrix to a 21-test matrix. This was done using special mixture experimental techniques. The kinetics model was used to compare the smaller matrix to the expected results of the larger one. The new 21-test matrix produced a numerical correlation that agreed well with the results from the 41-test matrix, indicating that the smaller matrix would provide the same statistical information as the larger one with acceptable precision. iii After the experimental matrix was developed using the design of experiments approach, the physical experiments were performed in the shock tube. Long test times were created by "tailoring" the shock tube using a novel driver gas mixture, obtaining test times of 10 millisecond or more, which made experiments at low temperatures possible. Large discrepancies were found between the predicted results by numerical models and the actual experimental results. The main conclusion from the experiments is that the methane-based mixtures in this study enter a regime with a negative temperature coefficient when plotted in Arhennius form. This means that these mixtures are far more likely to ignite under conditions frequently encountered in a premixer, potentially creating hazardous situations. The experimental results were correlated as a function of the different species. It was found that the effect of higher-order hydrocarbon addition to methane is not as profound as seen at higher temperatures (>1100 K). However, the ignition delay time could still be reduced by a factor two or more. It is therefore evident that potential autoignition could occur within the premixer, given the conditions as stated in this study.
Identifier: CFE0000817 (IID), ucf:46684 (fedora)
Note(s): 2005-12-01
M.S.
Engineering and Computer Science, Department of Mechanical, Materials, and Aerospace Engineering
Masters
This record was generated from author submitted information.
Subject(s): combustion
methane
natural gas
gas turbine
shock tube
chemical kinetics
Persistent Link to This Record: http://purl.flvc.org/ucf/fd/CFE0000817
Restrictions on Access: public
Host Institution: UCF

In Collections