Entropy-Maximization Molecular Reconstruction Of Crude Oil

Molecular reconstruction techniques are methods that allow creating mixtures of molecules from partial analytical data. In this article, entropy maximization reconstruction algorithm applied in crude oil was presented. The content of carbon and hydrogen are more than 95% in most kinds of crude oil, but there are still small amount of non-hydrocarbon, such as molecules containing sulfur, nitrogen or metallic elements in it. Non-hydrocarbon molecules have a certain influence on the properties of crude oil. So in order to simplify the calculation process appropriately and have little impact on the accuracy of the results, only the form of hydrocarbon molecules and few molecules which contain nitrogen or sulfur elements in the crude oil are considered, and the form of molecules containing metallic elements are not considered. In this article, a library composed of 963 representative molecules is created by researching qualitative and quantitative analysis data of different component fractions of crude oil. The properties of molecules in the established molecular library are obtained by consulting properties handbooks or estimating with appropriate methods. Next, the objective function is established by incorporating kinds of linear constraints into the information entropy maximization function. The constraints are about kinds of properties comparison between calculated value from mixing rules and analytical data. Then the objective function is transformed into other form which is easy to apply BFGS algorithm to optimize. After optimization, the content of various molecules in the library is calculated, and then adjusting the content with different structure information in narrow fraction. Finally the entropy-maximization molecular reconstruction of crude oil is completed.