OCR Text |
Show 3.2 Prompt-NO Nitric oxide can be formed from reactions of fuel-derived radicals with N2 which ultimately leads to what is termed "prompt-NO". The promptNO occurs at fuel rich condition, short residence time and low temperature region (i .e. regions near the flame zone). A number of species which resulted from fuel fragmentation have been suggested as a main source of prompt-NO in hydrocarbon flames (10), ego CH, CH2, C2, CzH, C etc., with the major contribution coming from CH and CH2 (10). CH + N2 ~ HCN + N (4) (5) The products of these react ions coul d 1 ead to format i on of ami nes and cyano compounds that subsequently react to form NO by reactions similar to the ones occurring in oxidation of fuel nitrogen, ego HCN + O2 ) NO + (6) HCN + N ) N2 + (7) It has been shown that total prompt-NO is very small in fuel 1 ean mixtures and is proport i ona 1 to the number of carbon atoms present ina molecule of hydrocarbon fuel ( 11 ) . Although the contribution of prompt-NO to total NO from stationary combustors is small, the concentration of prompt-NO in fuel rich eddies of turbu 1 ent d i ffus ion fl ames where ri ch combust i on takes place is significant. The prediction of prompt-NO formation within the flame requires coupling of the NO kinetics to an actual hydrocarbon combustion mechanism. Hydrocarbon combust i on process i nvo 1 ves many numbers of steps wh i ch can become very complex to solve and increase computer processing time considerably. However the quantities of prompt-NO formed in the practical combustion systems is small compared to the total NO formed. Therefore in the present invest i gat i on the gl oba 1 ki net i c parameter based on reduced chemical mechanism scheme (12) was employed to predict the prompt-NO emissions: d[NO]p dt = kpr [OJ a [N z] [FUEL] exp(Ea/RT) (3) where values of 1.2 x 107 (RT/P)Z, 0-1 and 60 kcal/mole are recommended for Kpr' a and Ea in C2H2/air flame . . In the present investigation the equation 3 was tested against the experimental data obtained by Bachmier et al (13) for different mixtures strength and fuel types. The predicted results indicated that the model varied significantly under fuel rich conditions and for higher hydrocarbon fuels. In order to reduce this error and predict the prompt-NO adequately in all conditions, the De Soete model modified by using the available experimental data. The effect of fuel type (i.e. number of carbon atoms) and air fuel ratio for gaseous aliphatic hydrocarbons are shown in figure 5 |