DEVELOPMENT OF SAYBOLT NUMBER CORRELATION FOR CONDENSATES

One of the methods to determine the quality of petroleum condensates and crude oils before processing in refineries is by determining their Saybolt colour or number. The main focus of this work is to develop a statistically-significant correlation that relates Saybolt number with other pre-determined physical properties of condensates. The motivation is to obviate the time taken to determine the Saybolt number of condensates which is usually conducted through laboratory analysis. We apply analytics available in business productivity tools such as Excel Data Analysis and R to analyze the applicability of various multiple linear regression models in determining suitable regressors to predict Saybolt colour over a wide range of condensate types. Another outcome of the study is to identify issues to be addressed before conducting extensive statistical analysis.