Optimization of the Composition of a Gas Mixture in a Joule-Thomson Cycle

Abstract

Mixed gas working fluids can be used within Joule-Thomson devices to achieve a greater refrigeration effect than is possible with a pure substance. Lower temperatures and higher cooling powers can be obtained with appropriate mixtures, resulting in lower operating pressures and mass flow rates. This paper describes a computational tool that allows the composition of gas mixtures to be optimized for a particular operating condition. Automated optimization algorithms are described, evaluated, and validated for two- and three-component mixtures consisting of nitrogen and various hydrocarbons. A genetic optimization algorithm was found to be the most robust and reliable technique and was adapted to this application. Subsequently, the optimization space was extended to a larger number of components for a range of load temperatures and operating pressures. The performance of optimized hydrocarbon gas mixtures is compared with that of nonflammable hydrofluorocarbon mixtures for a range of load temperatures and supply pressures.