Optimization of an organic ranking cycle radial turbine using a reduced-order mode coupled with computational fluid dynamics

Abstract

This paper presents the geometry optimization of a single stage radial turbine for an organic ranking cycle (ORC) system operating over a pressure ratio of 9. The specific fluid used in this investigation is R1233zd (E), but the methodology applies to other organic fluids as well. The ORC system is used to recover excess waste heat from the operation of an offshore oil and gas platform in the gulf of Thailand and its conditions will be replicated at pilot plant level. The geometry is optimized for the highest total-tostatic efficiency using nongradient based algorithms to allow for wide design space. Firstly, a one-dimensional meanline geometry is optimized, which is followed by a computational fluid dynamics (CFD) optimization in three-dimensional using a parameterized model. CFD is used to validate and calibrate the meanline model as well as to understand the flow and the sensitivity of the design parameters not captured by the low-order model. Moreover, the flow field of the successful designs is analyzed by CFD to identify the main flow structures that explain the difference in performance among the designs. The nonideal gas thennophysical properties of R1233zd (E) are calculated using equations of state to account for the nonideal gas behavior.