Abstract: Fluid modeling methods and systems are provided for predicting phase behavior of petroleum fluids using a computational model that employs pseudo-components of varying carbon numbers that represent asphaltenes, asphaltene-like hydrocarbon, and non-asphaltene hydrocarbons. The computational model models the phase equilibria of the mixture of components with mixing rules that employ binary interaction parameters determined from correlations that involve certain properties or parameters associated with respective pseudo-component pairs as input variables. The binary interaction parameters are tuned by adjusting at least one parameter of the respective correlations based on matching at least one measured property of one or more petroleum fluids to a corresponding property predicted by the computational model for the one or more petroleum fluids. The tuned values for the binary interaction parameters can be stored as part of the computational model for subsequent fluid modeling.