Prediction of the properties of brines using least squares support vector machine (LS-SVM) computational strategy

Natural brines occur underground or in salt lakes are commercially main sources of common salt and other salts, such as sulfates and chlorides of potassium and magnesium. This paper reports the implementation of a novel least square support vector machine (LS-SVM) algorithm for the development of improved models capable of predicting the properties of reservoir brine properties i.e., liquid saturation vapor pressure, density and enthalpy. The validity of the presented models was evaluated by using several statistical parameters. The predictions of the developed models for determining the liquid saturation vapor pressure, density and enthalpy were in excellent agreement with the reported data with an average absolute relative deviation (AARD) of %0.069, %0.033, %0.072, respectively and coefficient of determination values (R²) 0.999. According to the results of comparative studies, the developed models are more robust, reliable and efficient for calculating properties of oil field formation water during crude oil production than other techniques.