An Investigation of Using Isochoric Data Points in the Development of Natural Gas Equation of State

Access to energy is essential for the survival of humans, and the need for energy rises continuously because of population increase and economic progress. Fossil fuels continue to play the major role in satisfying energy demand. Among the fossil fuels, natural gas is the cleanest, most available, and most useful of the energy sources. It finds extensive use in residential, commercial, electric power generation and industrial applications. Moreover, the international energy outlook report released in 2011 indicates that yearly world natural gas consumption should increase from 111 trillion cubic feet in 2008 to 169 trillion cubic feet in 2035. Recently, new natural gas reservoirs have been discovered in many places throughout the world. In 2012, the total world supplies of proved natural gas reserves were estimated to be 6,746.8 trillion cubic feet. Thus, studies on natural gas are significant to advance the technique of natural gas processing, transportation and storage. In these three sectors, an accurate knowledge of the thermodynamic properties of natural gas is essential for engineering and technical processes. Developing accurate equations of state is important, and can provide us with accurate thermodynamic properties for natural gas. In addition, developing new techniques to produce mathematical models is important to create more accurate results and to enrich this field with new ideas, which might provide progress in the future. The aim of this thesis is to demonstrate a new approach for developing an equation of state.

This technique relies upon isochoric data of carbon dioxide pure component to develop mathematical models. This thesis contains nine models based upon experimental and generated data. The generated data come from REFPROP, which also provides an accurate means to adjust experimental data to true isochores. Within this thesis, a regression analysis was performed - using Polymath 6.1 - to provide mathematical structure of the equation for carbon dioxide. Results indicate that models covering vapor phase has less deviation than models covering liquid or both phases, and models developed by the generated data has less deviation than models developed by the experimental data. The deviation obtained by most of the models was less than the random error imposed upon the data. In this study, we conclude that modeling an equation of state from isochores appears to provide sufficient advantages to encourage additional studies on pure fluids and multi-component mixtures.