Estimating flammability limits through predicting non-adiabatic laminar flame properties

Lower and upper flammability limits (LFL and UFL) are widely and effectively used as simple safety boundaries for preventing gas-mixture ignition, and predicting them for new compounds would be valuable, especially the LFLs for low-flammability non-C/H/O species. Prediction with mechanistic flame models uses the idea that as these limits are reached, the flame speed and temperature drop below a threshold of flame feasibility because of radiant heat loss. Using methane as a reference study because its mechanisms and flammability are well characterized, non-adiabatic laminar flame speeds and profiles of temperature, flame structure, and chemiluminescent OH* and CH* are calculated using a modification of the hydrocarbon kinetics model of Hashemi et al. (2016), executed in CHEMKIN and Cantera. LFL prediction is emphasized here; large-hydrocarbon and soot radiant losses have been proven to be necessary for accurate UFL values (Bertolino et al., 2019) but were not included in this work. Property dependences on concentration are compared to published flammability limits of 5–15% methane concentration for methane/air mixtures at T 0 = 298 K and 1, 5, and 10 atm. At 1 atm, the LFL occurs at a laminar flame speed of 2.7 cm/s, and at the UFL, flame speeds between 4.5 and 6 cm/s correspond to the limit. Similarly, adiabatic temperature of 1450 K in a fuel-lean environment and 1560–1670 K in a fuel-rich environment correlate with the flammability limits. The ASTM standard test for flammability uses visual detection of a flame; ultraviolet chemiluminescence of OH* radical at limits of less than mole fraction of 10−3 is shown to reflect the methane LFL, and the equivalence-ratio dependence of OH* should resemble that of the visible C2* emission. Effects of pressure and challenges for modeling are discussed.