For many aspects of engine development, 0D/1D-simulation has evolved into an important tool to obtain reliable results at passable effort, especially for transient operations. Based on the neces-sary simplification of the three-dimensional reality to one-dimensional models, 1D-simulation heavily depends on the quality of the used sub-models. For internal combustion engines, adequate modelling of combustion chamber processes is of essential importance. Quasi-dimensional approaches to describe SI-engine-like burn rates of natural-gas engines base mostly on the modelling of laminar flame speeds. However, direct measurements of laminar flame speeds are usually taken in the air-fuel equivalence ratio range of 0.7 to 1.7 and pressures of only several bar. Generally used approaches then extrapolate to unsurveyed ranges, which causes contradictory data for laminar flame speed values. To avoid problematic extrapolations into engine-related boundary conditions, reaction kinetics calculations have been carried out to determine laminar flame velocities. The therefor used reac-tion mechanisms follow know, physico-chemical principles which allow a mechanism usage out-side of its measurement-based validation range. Consequently, calculated laminar flame speeds can be approximated with a correlation to be used computing-time optimal in 0D/1D-simulation. Besides flammability limits, influences of fuel composition can be account for to eventually model binary CNG substitute fuels up to low methane numbers. At this juncture, the influence of hydrogen admixture can be of interest, since rising percentages of hydrogen in the gas distribution system, produced with excessive renewable power, can be ex-pected in the future. With changing fuel compositions or lean operating conditions changes, on the one hand, the burn rate with direct influence on knock tendency, on the other hand the reaction kinetics of self-ignition itself. Therefore, these influences have been investigated using reaction kinetics calculations as well to develop an enhanced knock model for 0D/1D-simulation. The model quality is confirmed by comparing with engine measurements. The combination of both models allows the prediction of knock limits. To determine lean misfire limits, the new approaches can be additionally combined with a cycle-to-cycle variation model.