The problem of selecting an energy-optimal cooling system using natural refrigerants in a supermarket application in a humid continental and Mediterranean climate conditions.

• Cooling system is described by a vector. • Climate as a decision variable in refrigerant selection. • Energy demand was calculated for humid continental and Meditarranean climate. • An indirect expansion ammonia system and three R744 booster systems were compared. • Ammonia indirect expansion system uses 1.4–30.4% less energy annualy (Rome). The article presents an original solution to the problem of selecting an energy-optimal refrigeration system for a typical supermarket application. In this optimization problem carbon dioxide (CO 2) and ammonia (NH 3) have been assumed as natural refrigerants and the objective function is the minimum of the annual electricity demand. In the presented methodology, the refrigeration system is described by a vector whose coordinates are constant parameters and decision variables. A set of acceptable variants was determined using system analysis and matrix calculus as a tool, based on the limiting conditions. The following systems were determined as acceptable variants: classical CO 2 booster (1′ generation), 2′ generation CO 2 booster - with auxiliary compressors for parallel compression of flash gas from the liquid receiver, a 3′ generation CO 2 booster - with multiejectors in cooperation with parallel compression and flooded evaporators, and an indirect expansion (IX) system with ammonia. The objective function was determined from simulation models of the operation of the acceptable systems throughout the year. For the humid continental climate (Poznan, Poland) the energy-optimal variant is the 3′ generation booster system, in relation to which the annual electricity demand for the 1′ and 2′ generation CO 2 booster and the ammonia indirect expansion system is higher by 33.4%, 29.0% and 12.9%, respectively. For the Mediterranean climate (Rome, Italy) the energy-optimal system is the indirect expansion ammonia system, in relation to which the annual electricity demand for 1′, 2′ and 3′ generation CO 2 booster is higher by 30.4%, 23.1% and 1.4%, respectively. [ABSTRACT FROM AUTHOR]