Natural gas hydrates are considered as a promising fuel source in the relatively near future. Kinetic modeling of various steps of the natural gas hydrate formation process, such as dissolution, nucleation, and growth processes, has received numerous attentions. A novel mechanism is introduced for the entire nucleation and growth steps, and a proper mathematical model is presented to estimate the gas consumption rate in a constant temperature and pressure process. The proposed model covers the entire dissolution, nucleation, and growth stages. The combined Lax–Wendroff/Crank–Nicolson method is employed to solve the population balance equation for estimation of total surface area of evolved hydrate particles and corresponding particle size distributions. A special class of artificial neural network (known as the Regularization Network) is used to predict the solid–liquid equilibria. The proposed model is successfully validated using experimental data borrowed from the literature for both methane and ethane hydrate formation processes. The simulation results indicate that, for both methane and ethane species, the mole fractions in the bulk of liquid are often close to the corresponding concentrations at solid–liquid interfaces and decrease over the time during hydrate growth processes. It is clearly demonstrated that the overall resistance shifts from the nucleation reaction to mass transfer as the hydrate formation progresses.