Modelling of gas coolers for CO2 refrigeration systems

Consulta en la Biblioteca ETSI Industriales (8973)

[EN] The CO2 cycle easily becomes transcritical due to the low critical temperature (31.1ºC). Large changes in carbon dioxide properties characteristics occur in transcritical area. Dividing the gas cooler into many thermodynamic segments was shown as a good method to study it in transcritical performance. This partition method was assumed valid when a maximum relative variation in heat specific capacity of 2.19% between two consecutive gas cooler segments was obtained. A transcritical gas cooler model for CO2 refrigeration systems was developed with specific software: Engineering Equation Solver (EES). A subcritical model was created as well. The lowest possible high-side pressure is normally not the most energy efficient. The optimal pressure varies depending on the fluid temperature before throttling: the warmer the fluid, the higher the optimal pressure. This was reflected in optimal gas cooler parameters found for Valencia (Spain) and Copenhagen (Denmark). In Copenhagen it was found an optimal outlet pressure of 9223 kPa at design air temperature of 30ºC, while in Valencia it took the value of 12177 kPa at its design air temperature of 40ºC. Due to setting the same temperature difference between gas cooler outlet and ambient temperatures and the higher pressure in Valencia, a shorter total gas cooler length was obtained for Valencia ¿ 343 m ¿ than for Copenhagen ¿ 405 m ¿. Other parameters between both optimal solutions were also compared. It was studied the annual operation depending on the climatology, comparing gas cooler performances both in Copenhagen and Valencia (Spain). It was pointed out the convenience, in terms of thermodynamics and fluid mechanics, of having distribution valves at the beginning of gas cooler tubes in order to control how many tubes should be working ¿permitting the refrigerant flowing through them- and to achieve acceptable values for liquid velocities. The scope of this project did not include experimental verification of the results presented in this report.