Experimental Investigations of Spark Ignition in a Model Combustor With Synthesis Gas

Abstract

The components of syngas derived from coal, biomass, and waste are significantly different from those of typical gas turbine fuels, such as natural gas and fuel oils. The variations of hydrogen and inert gases can modify both the fluid and the combustion dynamics in the combustor. In particular, the characteristics of spark ignition can be profoundly affected. To understand the correlation between the varying fuel components and the reliability of ignition, a test system for spark ignition was established. The model combustor with a partial-premixed swirl burner was employed. The blending fuel with five components, hydrogen, carbon monoxide, methane, carbon dioxide and nitrogen, was used to model the synthesis gas used in industry. The ignition energy and the number of sparks leading to successful ignition were recorded. By varying the fuel components, the synthesis gases altered from medium to lower heat value fuels. The ignition time, ignition limit, and subsequent flame developments with variations of air mass flow rates and fuel components were systematically investigated. With the increase of airflow, the syngas with a lower hydrogen content has a shorter ignition time compared with higher hydrogen syngas in the lean condition, whereas in the rich condition, syngas with a higher hydrogen content has a shorter ignition time. The effects of the hydrogen content, inlet air Reynolds number and spark energy on the ignition limit were investigated. The ignition limit was enlarged with the increase in the hydrogen content and the spark energy. Meanwhile, three distinct flame patterns after ignition were investigated. Finally, a map for the characteristics of the ignition and subsequent flame development was obtained. The results are expected to provide valuable information for the design and operation of stable syngas combustion systems and also provide experimental data for the validations of theoretical modeling and numerical computations.