Natural Gas Viscosity Estimation Using Density Based Models Ehsan Heidaryan 1 * and Azad Jarrahian 2 1. Faculty of Energy, Kermanshah University of Technology, Kermanshah, Iran 2. Department of Chemical and Petroleum Engineering, Sharif University of Technology, Tehran, Iran Accurate value determination of natural gas viscosity plays a key role in its management as it is one of the most important parameters in natural gas engineering calculations. In this study, a comprehensive model is suggested for prediction of natural gas viscosity in a wide range of pressures, temperatures, densities and compositions. The new model can be applicable for gases containing heptane plus and non-hydrocarbon components. It is validated by the 2011 viscosity data from 18 different gas mixtures. Compared to existing similar models and correlations, its results are quite satisfactory. Keywords: natural gas, viscosity, correlation INTRODUCTION N atural gas is a subcategory of petroleum that occurs natu- rally. It is composed of complex mixtures of hydrocarbons and a minor amount of inorganic compound. Natural gases’ physical properties, and in particular, their variations with pressure, temperature and molecular weight have a great impor- tance in petroleum and natural gas engineering calculations. The gas viscosity generally increases with pressure. The increase of temperature decreases the liquid viscosity, whereas it increases the gas viscosity at low and moderate pressures. At high pressure, the gas viscosity behaviour approaches that of liquid (Danesh, 1998). As far as most laboratories do not have required equip- ments for the measurements of gas viscosities, predicting gas viscosity is particularly important. Several correlation alternatives are available for estimating this quantity. However, wide-ranging and accurate correlations for predicting natural gas viscosity are not available, which are associated with an amount of heavier hydrocarbon components and impurity components, such as car- bon dioxide, nitrogen, helium and hydrogen sulphide. The present study introduces a new and reliable model to estimate the vis- cosity of hydrocarbon gas mixtures containing a concentration of nonhydrocarbon and heavy hydrocarbon components. The model is able to predict the viscosity over wide ranges of pressure, tem- perature, density and composition based on the modern theories of the general fluids (Born and Green, 1947), which was modified by Starling and Ellington (1964) and Lee et al. (1964). LITERATURE SURVEY Data In the present study, a large number of data for several natural gases has been used, including 2011 viscosity data points of 18 different gases (Lee, 1965; Gonzalez et al., 1970; Nabizadeh and Mayinger, 1999; Assael et al., 2001; Schley et al., 2004; Lange- landsvik et al., 2007). A description of the data composition is reported in Table 1. In the first try, Lee (1965) used a capillary tubing viscometer for providing a database of light hydrocarbon and natural gas mixture. In the next try, Gonzalez et al. (1970) used an absolute capillary tube viscometer to obtain natural gas viscosity data in a wide range of composition. Assael et al. (2001) using a vibrating-wire viscometer reported viscosity data in a wide range of pressure and temperature with measurement’s uncer- tainty of ±1% for a natural gas mixture. They reported maximum measurement uncertainties of 0.01 MPa and 0.1 K for pressure and temperature, respectively. In another study, a vibrating-wire ∗ Author to whom correspondence may be addressed. E-mail address: heidaryan@engineer.com Can. J. Chem. Eng. 9999:1–7, 2012 © 2012 Canadian Society for Chemical Engineering DOI 10.1002/cjce.21715 Published online in Wiley Online Library (wileyonlinelibrary.com). | VOLUME 9999, 2012 | | THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING | 1 |