OCR Text |
Show INTRODUCTION Fluid dynamics, transport, and chemistry all play significant roles in combustion processes, and all must be considered in tackling complex combustion problems. Meeting the challenges of ever-tightening environmental regulations on combustion effluent while maintaining combustion efficiency, safety, and overall system costs is a crucial issue taking the gas industry today. Chemical processes are paramount in the production of pollutant emissions, in particular N O x , and in controlling flame speed, which influences ignition and blowoff, through the production/destruction of hydrogen atoms. However control of these chemical processes in practical combustion requires properly designed fuel/air mixing. This design can of course be complex, for example, staging the introduction of fuel to reduce N O x emissions via re-burn. A solidly founded understanding of the chemical processes involved, and their dependence on temperature and local fuel/air mixing ratio, is of considerable help in design considerations. O n the other hand, a complex system such as an actual burner is far too complex to attempt to extract credible detailed chemical information. This paper describes experiments on laboratory flames specifically designed to investigate the chemistry, testing our understanding through the measurement of free radical compounds that are reactive intermediates in those processes. In particular we focus here on low pressure, one-dimensional flame studies of the CH and H C O radicals, and include measurements of O H and N O as further tests of that understanding. The problems of interest are the formation and abatement of N O x pollutants in natural gas/air flames. At flame temperatures below 1800K, C H radicals play a crucial role in N O x formation in the chemical mechanism known as prompt-NO. C H radicals are also very important in re-burn strategy. H C O can be both a sink and source of H atoms, which govern the flame speed and hence ignition and blowoff properties. The measurements w e describe are designed as targets for testing a comprehensive mechanism of natural gas combustion chemistry. THE ROLE OF MODELS AND TARGETS Many chemical reactions are involved in the pyrolysis and oxidation of methane; when minor processes such as pollutant formation are included, the number becomes huge. For a detailed understanding the information must be incorporated into a model of the combustion chemistry. In the past, the approach generally taken was the assembly of a set of chemical reactions, with rate coefficients chosen from a survey of the literature. In many cases experimental rate coefficients were not available and needed to be estimated. W h e n the overall predicted result did not agree with measurement, one or more of the rate coefficients were adjusted to achieve agreement. Because of the complexity of the chemistry involved, this could be quite unsatisfying unless the experiments were very carefully designed to isolate one specific reaction. A new approach, known as GRI-Mech™ (Gas Research Institute Mechanism), has recently been introduced.1 It begins with a carefully chosen set of rate coefficients, taken from individual kinetics experiments, for the chemical network(s) of interest. (For example, an early version considered only methane pyrolysis and oxidation, whereas the current version also includes N O x formation chemistry.) With such a set of rate coefficients, one can predict any given combustion property (e.g., flame speeds, shock tube ignition delay times, N O concentrations) with an error associated with the uncertainties in the rate coefficients. With GRI-Mech, one takes a new direction by optimizing the entire system against measured "targets," i.e., system properties such as those just listed. This is done by adjusting all the relevant rate coefficients, within the error limits imposed by their independent determination. 2 |