Superstructure-based mixed-integer nonlinear programming framework for hybrid heat sources driven organic Rankine cycle optimization

Abstract

Organic Rankine cycle (ORC) is a promising technology capable of harnessing low-grade energy, e.g. renewable energy and waste heat, and converting it into electricity. Conventional ORC often operates with single heat source, which can be unfavorable to its efficiency due to the poor matching between its working fluid and heat source. Implementing hybrid heat sources is generally more energy-efficient as multiple heat sources can effectively match with the working fluid, however, the design can be a challenging task. This study proposes a superstructure-based method for the ORC design that simultaneously synthesizes heat exchanger network for hybrid heat sources and working fluid, and optimizes the ORC and heat sources' parameters. A mixed-integer nonlinear programming model is formulated to achieve the simultaneous optimization. We also develop a tailored multi-step initialization algorithm to facilitate the optimization. The model is applied to the design and analysis of ORCs driven by solar energy and waste heat with four different working fluids. Results demonstrate that the hybrid heat source-driven ORC improves the system performance. Cyclohexane is found to be the optimal working fluid for the proposed ORC system with a 48.19% increase in net power output compared with the single heat source-driven ORCs running separately.