Designed binary mixtures for subcritical organic Rankine cycles based on multiobjective optimization

The use of binary zeotropic mixtures as working fluids applied to Organic Rankine Cycles (ORCs) is investigated in this paper. In total, six (6) hydrocarbons and (2) hydrofluorocarbons are considered, leading to twenty-eight (28) possible binary combinations. The mixtures were tested with a basic Rankine cycle while using the heat source temperature as independent variable, which assumed six different values, ranging from 80 degrees C to 180 degrees C, in steps of 20 degrees C. The simulations aimed to identify the ideal mixtures that maximized the net power and exergetic efficiency, and minimized the heat exchanger's global conductance for a given temperature of the heat source. The optimization process relied on a genetic algorithm and the selection of the best mixtures, on a non-dominated sorting method (NDS), which returned Pareto fronts gathering the best solutions. While no one specific ideal mixture was identified, the results showed that the range of the so-called ideal mixtures narrows as the heat source temperature increases, with mixtures including fluids like R245fa and pentane being good options, whereas at low temperature, a larger number of fluid mixtures perform well. Finally, a scale analysis is proposed and shows that the maximal net power varies linearly with a Number of Transfer Units (NTU) factor while its slope depends on the heat source temperature. The latter analysis is compared with the results obtained with the Pareto front and NDS, showing that both sets of results agree well while correlated by a single constant for the entire temperature range covered in the present study.