This paper describes how automatically generated detailed kinetic mechanisms are obtained for the oxidation of alkanes and how these models could lead to a better understanding of autoignition and cool flame risks at elevated conditions. Examples of prediction of the occurrence of different autoignition phenomena, such as cool flames or two-stage ignitions are presented depending on the condition of pressure, temperature and mixture composition. Three compounds are treated, a light alkane, propane, and two heavier ones, n-heptane and n-decane.