Technical fuels, such as natural gas, kerosene, gasoline or diesel, are complex mixtures of different hydrocarbons. Designing a surrogate for such fuels requires identification of these 5-8 components and appropriate mole fractions which best represent the thermodynamics, fluid dynamics and chemical properties of the real fuel. The design itself is an iterative process consisting of two global steps (see Fig. 1).
I. First Step
The composition of the surrogate firstly requires that each main group (paraffins, aromatics and cycloparaffins) is well represented. The final step is appropriate modelling of the surrogate’s physical properties, such as enthalpy of formation, heat release, boiling behaviour, two-phase diagram, critical temperature and pressure.The boiling curve, two-phase diagram and critical parameters are predicted by applying a specially developed in-house code. (Figs. 2, 3).
Once the surrogate properties have been numerically determined, they are compared with the experimental data obtained with the real fuels. Table 1 shows the composition of a kerosene surrogate which was obtained by applying the methodology and tools mentioned above; Fig. 4 shows its two-phase diagram.
II. Second Step
Modelling the chemical properties of surrogate mixtures with a kinetic model focusing on the chemical composition (C/H ratio), mole fraction profiles, laminar flame velocities, ignition delays, sooting tendency (TSI factor) and exhaust pollutant formation.Prior to verification of the chemical properties of the surrogate, a detailed kinetic model is required which describes the chemical properties of the individual hydrocarbon components and their mixtures (sub-models A, B, C, etc.; see Fig. 1) as well the properties of the real fuel. This chemical modelling is followed by another comparison with experimental data obtained for all the technical fuels mentioned above. Based on this comparison, the composition of the surrogate is improved, if necessary, followed by the procedures described in steps I and II until the comparison yields satisfactory results.