The liquefied natural gas (LNG) industry utilises process simulators for plant design and optimisation. Unfortunately, their predictions are known to become increasingly inaccurate with increasing pressure and decreasing temperature. Improved phase equilibrium predictions to determine the amounts of vapour, liquid and (unwanted) solids present at these conditions are needed. This thesis presents new data and improved models for phase equilibrium in mixtures and conditions relevant to LNG production. The models are implemented within a new software tool "ThermoFAST" which improved vapour-liquid and solid equilibrium predictions for multicomponent systems by factors of 1.9 and 6.9, respectively, relative to standard industry tools.