An Application of Conventional First Guess Geometry Rules to Orc Radial-Inflow Turbine Rotors: A Critical Assessment

Climate change as well as more and more stringent emissions regulations along with the chance to recover waste heat from low to medium temperature have led to novel interest towards Organic Rankine Cycle (ORC) technology.

When the small scale is considered, the Radial-Inflow layout has proven to be the most suitable option thanks to high efficiency, compactness and larger achievable expansion ratio within a single stage.

The expansion process often takes place in the close proximity of the saturation line and it cannot be modelled as for an ideal gas. Additionally, organic fluids feature large molecular mass in comparison to air and exhaust gases. The latter, in combination with very large expansion ratio, may lead to transonic and supersonic flows within blade passages with intense shock formation. Consequently, conventional design guidelines developed for ideal gas might not generally hold for ORC applications, as well as, detailed and comprehensive design criteria for ORC turbines are still lacking.

To begin the detailed aerodynamic design phase, a first guess geometry is to be built based on the results of the preliminary design. Several methods to generate a first guess geometry of Radial Inflow Turbine (RIT) rotors exist for ideal gas, although a knowledge gap concerting their applicability to ORC applications still remains.

The present work aims, in a first phase, at assessing the effectiveness of conventional design rules for the first guess geometry of RIT rotors to ORC applications. In a second step, the most suitable geometries are validated by a CFD based analysis.

The preliminary design is accomplished by means of an in-house code accounting for real-gas properties, considering two test cases - high and medium temperature applications. An in-house implemented meanline program is used to carry out a parametric analysis to investigate the design space of the single stage ORC turbine.

Efficiency dependance on specific speed and specific diameter, as well as other design variables, is shown by means of contour plot.

Suitable design candidates are chosen for both test cases, i.e. high and medium temperature. Subsequently, several first guess rotor geometries are built for each design candidate by means of different first guess methods and compared by running 3D steady state CFD calculations. The calculations involve radial flow turbine stage simulation by means of classical methods for stator – rotor interaction.

Finally, design methods effectiveness is assessed comparing the achieved efficiency and losses of each design as well as against the prediction of the meanline program.

The proposed poster will display an overview of the methodology, together with the results of the parametric analysis and a selection of contour and streamline plots derived from the CFD computations.