ABSTRACT
This paper presents the application of bivariate analysis to the thermodynamic equilibrium parameter (response). The predictors of the equilibrium parameter consist of correlations that could cloud it with mathematical equations leading to more difficult solutions. Bivariate analysis is used to mathematically remove correlations in the predictors such that the effects of these predictors on the response can be observed more clearly. A new set of independent predictor random variables is then developed with a reliable regression model whose R-square remains the same before and after removal of correlations. Conclusions are made highlighting the importance of these procedures in petroleum engineering modeling.
Acknowledgement
Thanks to the Pan African Materials Institute (PAMI), and World Bank for providing financial support towards this work. The department of Food, Agricultural and Biological Engineering of the Ohio State University for providing the facilities used.