Low temperature oxidation of n-butanol: Key uncertainties and constraints in kinetics

A recent chemical kinetic mechanism (Sarathy et al., 2012) describing the low temperature oxidation of n-butanol was investigated using both local and global uncertainty and sensitivity methods within the context of predicting ignition delay times in a rapid compression machine (T = 678-898 K, ϕ = 0.5-2.0, P = 15 bar) and species profiles in a jet stirred reactor (T = 800-1150 K, ϕ = 0.5-2.0, P = 10 atm) in order to determine the most important reactions driving the predictive uncertainty, and the constraints provided by the experimental measurements. A global sampling technique was employed for the determination of predictive uncertainties, and a high dimensional model representation (HDMR) method was further utilised for the calculation of global sensitivity indices following the application of a linear screening method. The calculated global sensitivity indices were used to identify and rank the rate parameters driving the predicted uncertainties across the conditions studied. Predicted ignition delay distributions spanning up to an order of magnitude indicate the need for better quantification of the most dominant reaction rate parameters. The calculated first-order sensitivities from the HDMR study show the main fuel hydrogen abstraction pathways via OH as the major contributors to the predicted uncertainties. Sensitivities indicate that no individual rate constant dominates uncertainties under any of the conditions studied, and that the target outputs are largely insensitive to the total rate of OH with n-C4H9OH. However, strong constraints on the branching ratio for H abstraction by OH at the α and γ sites are provided by the RCM measurements. In the JSR simulations, predicted n-C4H9OH and CH2O concentration profiles at T = 800 K, were particularly sensitive to H abstraction reaction by HO2 from the α site. Although abstraction by OH from the α site plays an important role for predicted n-C4H9OH profiles at higher temperatures, in general, better constraint is provided on the n-C4H9OH + HO2 abstraction rate by the measured concentration profiles of n-C4H9OH and CH2O at lower temperatures than for abstraction by OH.