Prediction of Flow and Combustion Characteristics for a Gas Turbine Combustor Burning Low Heating Value Fuel

In this study, a numerical model has been developed to simulate the flow and combustion in a gas turbine combustor of type (Winnox-TUD-Combustor), which burns low heating value gas. The model relies on the computational code “FLUENT”. This code has been used to solve the governing equations. The characteristics of the model are; steady, turbulent, two dimensional, axisymmetric and swirling flow. The combustion process has been simulated as non-premixed combustion. The study includes the impact of several design and operational parameters on the characteristics of the flow and combustion inside the combustion chamber. These parameters include; ratio of secondary to primary air, ratio of tertiary to primary air, swirl ratio, number of inlets of the secondary air and their direction. Four performance indicators have been used to evaluate the impact of the aforementioned design and operating parameters. These indicators include; average temperature of the exhaust gases to the turbine, specific NO_x emission, pattern factor and combustion efficiency of the combustor. In order to identify the optimum values of the aforementioned design and operational parameters, Artificial Neural Network (ANN) technique has been utilized to enrich the output results. This facilitates searching for the optimum values of the aforementioned parameters. Furthermore the effect of the variations in fuel composition on the combustion characteristics and accordingly on the performance indicators has been studied. It has been found that, all the studied parameters affect the performance of the combustor to a certain extent. However, fuel swirl ratio, primary to secondary air ratio and tertiary to primary air ratio as well as carbon monoxide to hydrogen ratio in the fuel are the controlling factors. Optimizing these parameters can lead to a substantial reduction in specific NO_x emissions down to 4.0 gm/kg of fuel. Also an improvement in pattern factor to values below 0.3 has been achieved.