Beyond a heuristic analysis: integration of process and working-fluid design for organic Rankine cycles

Abstract

Industrial activity produces an enormous amount of waste heat energy in the form of low-temperature sources, the utilisation of which remains challenging due to low efficiencies and poor economic viability. Organic Rankine cycles (ORCs) offer an appealing solution to this challenge because the ORC efficiency can be drastically increased by tailoring the organic working fluid to the heat source. Due to the large number of potential fluids available, however, it is difficult to identify those fluids that give the highest ORC performance. Here we present a robust computer-aided molecular and process design technique for the determination of pure-component working fluids based on ORC performance metrics. In particular, factors that lead to process infeasibility are recognised, thereby enabling a diverse exploration of the working fluid design space while reducing the computational cost of the method. These algorithmic developments, in combination with the use of the predictive SAFT-γ Mie group-contribution methodology, provide a reliable platform for designing a wide range of ORCs. Excellent capability for determining optimal ORC designs from a vast number (58 960) of prospective fluids is demonstrated through three case studies. The optimal fluids identified include low molecular weight alkanes, alkenes, and ethers, such as propane, pent-1,4-diene and methyl ethyl ether. Importantly, we find that the performance of these fluids is not always transferable across applications. An analysis of fluid properties further highlights the limitations of using a heuristic approach for fluid selection. Instead, a holistic consideration of the working fluid in the cycle can lead to increased ORC performance.