Assessment of the Combustion Process in Ultra-Lean (λ>1.8) Natural Gas Engines

The majority of today’s natural gas fired engines are applying a premixed combustion concept, which is commonly assumed to be based on the turbulence-enhanced propagation of a thin flame separating the burnt and unburnt fractions of the mixture volume. This concept has been confirmed by means of comprehensive experimental investigations on passenger car engines operating at air/fuel ratios close to stoichiometry; however, for larger industrial engines (4-stroke and 2-stroke) designed for ultra-lean (λ >1.8) operation in order to achieve highest efficiencies, this assumption is no longer valid, as will be shown in the following. On these engines, the combustion process is largely controlled by the reaction kinetics of the chemistry and hence exhibits more similarity to homogeneous charge or spark assisted compression ignition (HCCI or SACI) combustion concepts. This is substantiated by a detailed review on theoretical and experimental investigations of ultra-lean combustion processes, a characterization of the combustion regime and by extending the earlier analysis of combustion in ultra-lean medium-speed 4-stroke gas engines to large two-stroke engines, which are run at yet higher air/fuel ratios. Tests have been performed on a lab engine allowing variations of key parameters such as exhaust gas recirculation rate, charge conditions, mixture quality and ignition intensity over a large range. The results obtained clearly support the hypothesis that combustion on such engines is actually governed by volume reaction instead of the propagation of a turbulent premixed flame.