The main objective of this research is to study the phenomenon of composition gradient in hydrocarbon reservoirs and to develop its thermodynamic modeling. With this point of view, this dissertation has been developed in 5 phases which first 4 phases provide the fields to have more reliable results from modeling, while in phase 5 using the results from preceding phases, modeling of compositional grading has been developed. In the first phase, screening and quality control of available samples have been investigated. To this end, using the data of more than 40 gas condensate samples, the range of acceptable fluctuations of parameters affecting the stability of well conditions is determined using statistical concepts. The results show that only for 11.5% of the samples all parameters were within the acceptable range.
In the second phase of this study, using the WinProp software and its correlations, the physical and critical properties of plus fractions of gas condensate samples are calculated and the most suitable group among 18 available correlation groups is selected. The criterion for choosing the appropriate group of correlations is the minimum amount of relative error in calculating the saturation pressure as well as the suitability of the 2-phase diagram. The results show that the samples containing plus fraction with lower carbon number has a lower relative error. Edmister correlation for calculating accentric factor with Riazi-Daubert and Goossens correlations for calculating physical and critical properties have the best results in gas condensate samples. On the other hand, Edmister correlation with Lee-Kesler correlation have the minimum relative error in volatile oil samples.
In the third and fourth phases, compositional gradient modeling has been conducted using WinProp software for gas condensate and volatile oil samples. The purpose of this phase is to identify the factors that can affect the modeling outcomes such as type of fluid (number of components