A predictive correlation for the constant-pressure specific heat capacity of pure and ill-defined liquid hydrocarbons

Abstract

A predictive correlation for the constant pressure specific liquid heat capacity ( C P ) valid for pure organic compounds and ill-defined mixtures such as heavy oil and bitumen is reported. With six universal coefficients, the correlation is a function of temperature and a similarity variable, α, which is proportional to the number of atoms per unit mass. Molecule structure is shown to be a variable of secondary importance and only the elemental composition of a liquid is required to apply the correlation. A training data set comprising 150 experimental heat capacity values for 19 liquid organic compounds (including paraffins, naphthenes, aromatics, sulphur/oxygen/nitrogen derivatives) and 3 molten polymers in the temperature range (207–590) K was used to regress coefficients. The correlation was evaluated using a test data set comprising 111 heat capacity values for 12 liquid organic compounds and 3 molten polymers. The average absolute deviation for the test data set was found to be 0.067 J K −1 g −1. The correlation appears to be preferred over the Lee–Kesler correlation, a common reference equation for engineering calculations, both with respect to over all accuracy, and range of application. The correlation was also applied to two ill-defined hydrocarbon liquids, namely Athabasca and Maya pentane maltenes. For such ill-defined hydrocarbon liquids neither molecular structure nor critical properties are available. Their liquid heat capacities were predicted to within 6.0 and 2.8%, respectively from 325 K to more than 500 K. This is the first such prediction, and represents a significant advance over prior practice. With this new predictive correlation, liquid phase heat capacities for ill-defined hydrocarbon fluids, such as bitumen, and pure compounds and polymers can be predicted with equal ease. The correlation may also be used for the development of high-accuracy fluid-specific correlations where warranted by customizing one or more of the universal coefficients.