Autoignition Characteristics of Gaseous Fuels at Representative Gas Turbine Conditions

Abstract

The autoignition properties of gas turbine fuels have been studied for many years and by numerous researchers. The advent of ultra low emission industrial gas turbines using lean premixed technologies has given rise to premixer designs with longer residence times. This, in conjunction with the ever-increasing pressure ratios of aeroderivative machines, leads to the potential for autoignition within premix ducts, and has therefore renewed the interest in this field. Although much has been published, data in the region of interest to high pressure ratio gas turbines is extremely sparse. Similarly, modelled autoignition delay times are not very accurate, as most reaction mechanisms were not generated to cover this range of conditions. Hence the uncertainties of autoignition delay times at gas turbine conditions are significant, thereby either imposing over-stringent design limitations or introducing risks of ignition occurrence in the early design process. A series of experiments have been carried out for methane and simulated natural gas fuels in the region of interest, using shock tubes as the test vehicle. The experimental technique was chosen to isolate only the chemical kinetic component of the autoignition delay time, without any additional delays due to mixing and heating of the test gases. Predictive correlations and a chemical kinetic model (the GRI mechanism) have also been used to predict autoignition delay times at the same conditions. The correlation between experiment and prediction has been shown to be poor at representative temperatures. This paper discusses some of the possible explanations for this poor agreement.