A Simple Statistical Model for Predicting Crude Oil API Values

Abstract

American Petroleum Institute (API) gravity value is the main indicator of crude oil quality and marketing value; hence, it must be simply and accurately determined. Existing crude oil API prediction models are complex and time-consuming because they use lots of parametric properties for predictions. Herein, we propose a simple two-variable (aromatic and naphthene content) statistical model for predicting crude oil API values. The statistical model in this study was developed using multiple linear regression techniques on about 80 crude oil samples from different locations. The study shows that the developed model in this work could accurately predict crude oil API gravity values with a standard error of 3.14 and a correlation matrix of 0.92. Also, the model confirmed that the use of crude oil aromatic and naphthene content could accurately describe its API values. The model could predict crude oil API better than some API models in literature by 38% - 62%. The findings in this work provide a simple and fast method of determining crude oil API for crude marketing and inspection.