Efficiency Correlations for off-design Performance Prediction of ORC Axial-flow Turbines

orc2019 Tracking Number 70

Organic Rankine Cycle (ORC) power systems are efficient and cost-effective to convert heat from low/medium temperature heat sources into electrical or mechanical power. Depending on the time behaviour of the heat and cold sources, the ORC can be operated close to the design point or at part-load, which is generally affected by lower efficiency with respect to the nominal point. Axial-flow turbines are the dominant type of expander for large-scale ORCs and their behaviour is crucial for the overall ORC performance. While correlations for efficiency estimation of the design of ORC axial-flow turbines are already available, only very few works have studied the off-design of this component. Two MATLAB mean-line tools of ORC axial-flow turbines, one for the design, AxialOpt, and one for the off-design, AxialOff, are here presented and used to develop part-load efficiency correlations based on simple thermodynamic quantities. The codes have been validated against data and measurements available in literature. A total of eight turbines from different fields of application have been designed and performance maps have been developed to define the efficiency based on the relative difference in specific enthalpy over the turbine and the relative outlet volume flow rate with respect to the nominal point. The coefficient of determination for the fitting was larger than 99%. A test case not included in the curve fitting was studied to prove the prediction capabilities of the proposed correlations. They could predict the part-load behaviour of the turbine with coefficients of determination above 90% for one, two and three stages. The results of the work can be used in ORC system analyses to estimate the turbine performance at part-load conditions prior to its design.