Very Interesting! Have you compared your results to the hydrate predictors in the top commercial process simulators? I never found a method I trusted completely. I think the biggest problem I had was accounting for the effects of individual components other than the simplified component lumping. In designing process temperatures, I tried to err on the conservative side by using more than one prediction method. In diagnosing plant problems for suspected hydrate formation, it was difficult to completely trust any calculation result that gave assurance the temperature was above the hydrate formation temperature.