Comparison of a scroll, a screw, a roots, a piston expander and a Tesla turbine for small-scale organic Rankine cycle Parralel disk, viscous effects Hello! First I would like to thank the organizers to allow me presenting my work. Its is focused on the…Design, modeling and experimentation of a reversible HP-ORC prototype First Innogie ask us a theroretical modelisation of the prototype. As the results were promising we build a test rig, then whe the test rig will be fully operationnal, innogie will sell these Reversibles units O. Dumonta, L. Tallurib, D. Fiaschib, G. Manfridab and V. Lemorta aThermodynamics Laboratory, University of Liège (Belgium) bDepartment of Industrial Engineering,  Università degli Studi di Firenze (Italy)

Conclusions and prospects Table of contents Introduction Modeling Discussion Conclusions and prospects

Introduction State of the art Strong influence of expander on ORC performance (Almost) no experimental comparison of expander in litterature No single technology identified as optimal (cost, efficiency, compactness, working conditions..) 3 main technologies : Scroll, Screw, Piston (+roots+tesla) Parameter Scroll Piston Screw Roots Tesla Displacement [l/s] 0.76-32 [1.25-75] [25-1100] [12] - N/A Power [W] [0.005-10,000] [10] [0.001-10,000] [10] [2,000-2e5] [10] [1000-30000] 0.005 – 30.000 Max. rotational speed [RPM] 10,000 [12] 12,000 [10] 21,000 [10] 20,000 [10] Built-in volume ratio [1.5-4.2] [10] [2-14] [10] [n.a.-8] [12] ~1 Maximum pressure [bar] ~40 70 [12] Max. temperature [°C] 250 [12] 560 [12] Two-phase flow handling Yes [12] Low [12] Yes Yes [13] Isentropic efficiency [%] 87 [1] 84 [12] 47 [12] 59

Experimental setup Hydraulic scheme Test-rig : 3 kWe ORC (SUN2POWER) Fluid : R245fa and ~5% oil (but not for piston) Scroll: AC/DC convertor + variable resistor Piston : Asynchronous generator with four quadrants variable-frequency drive Screw : Air-cooled Eddy-current brake Roots : modified engine supercharger Tesla : Prototype from Firenze university

Conclusions and prospects Table of contents Introduction Modeling Discussion Conclusions and prospects

Important note Important note Perfectly objective comparison between different types of expanders not possible : Different level of maturity for expanders Not an expander sized for the test-rig (mass flow, pressure and temperature affect the performance of the expanders (not necessarily in the same way for each one). Sizing fluids for those expanders are not the one used in this ORC (R245fa). Nominal working conditions in terms of pressure and temperature are different for each technology (higher pressure and temperature for the piston for example). BUT: No such a comparison in litterature Same test-rig and fluid Models calibrated to predict optimal performance

Semi empirical model Advantages Low CPU time + robust + extrapolation + general formalism Supply DP, mechanical losses, leakage, under(over)-expansion, heat transfers Calibration of 7 parameters based on experimentation Performance extrapolation with speed optimisation Tesla deterministic model 2D code for the viscous flow solution, considering real compressible fluid properties

Conclusions and prospects Table of contents Introduction Modeling Discussion Conclusions and prospects

Model Assumptions Larger range of pressure ratio explored Larger efficiency for the screw and scroll Rather constant efficiency for the piston and Tesla So when the fluid is admitted in the cylinder, a part of the volume is already taken by some fluid trapped into this clearance volume. The effect of the clearance volume decreases when the pressure ratio increases because the mass flow increases and the then, the amount of fluid trapped into the clearance volume have less impact.

Model Operating maps Extreme conditions Flexibility Low temperature Optimal performance of a machine for a given condensation and supply temperature temperature !!!All axis are positive!!! Extreme conditions Flexibility Low temperature Efficiency

Conclusions and prospects Table of contents Introduction Modeling Discussion Conclusions and prospects

High Pressure and temperature Conclusion   High Pressure and temperature Wet expansion Compactness Flexibility Efficiency Piston + - Screw +++ Scroll ++ Roots Tesla turbine Screw expander minimum power ~10 kW  wrong! Scroll expander are the best because of higher isentropic efficiency  wrong! Piston expander not interesting because of low efficiency  wrong! Not a single volumetric expander is the optimal solution! Tesla optimal for low volume ratio Perspectives Maturity of technology Other technologies: Vane, Wankel, … Economic aspects

 The most volumetric of the turbines Conclusion - Tesla  The most volumetric of the turbines

Important note Important note Perfectly objective comparison between different types of expander not possible : Different level of maturity for expanders Not an expander sized for the test-rig (mass flow, pressure and temperature affect the performance of the expanders (not necessarily in the same way for each one). Sizing fluids for those expanders are not the one used in this ORC (R245fa). Nominal working conditions in terms of pressure and temperature are different for each technology (higher pressure and temperature for the piston for example).

Open source papers  http://orbi.ulg.ac.be Thank you! olivier.dumont@ulg.ac.be Open source papers  http://orbi.ulg.ac.be 15

Semi-empirical models Extrapolation Optimal conditions not reached on the test-rig  model to extrapolate Oh = 5 K (Tamb=25°C) RPM optimised Rp adjusted with Pexp,ex 10 bar 30 bars

(constant speed/ variable speed) Screw expander Piston expander Parameter Scroll expanders (constant speed/ variable speed) Screw expander Piston expander Swept volume [cm3] 20.2 12.74 19.96 195 Volume ratio [-] 2.2 2.19 2.5 4.74 Maximum inlet temperature [°C] 140 130 250 Maximum inlet pressure [bar] 28 25 16 40 Rotational speed range [RPM] 3,000 [800-8,000] 20,000 [1,000-4,000] Nominal shaft power [W] 2,277 2,000 4,000

Experimental results Isentropic efficiency Same trend for each technology Low pressure ratio  over-expansion losses High pressure ratio  under-expansion, mechanical losses and pressure drops Higher scroll efficiency Low efficiencies because of test-rig limitations (pressure drops, mass flow rate…) 𝜂 𝑠 = 𝑊 𝑠ℎ 𝑚 𝑟 ℎ 𝑒𝑥𝑝,𝑠𝑢 − ℎ 𝑒𝑥𝑝,𝑒𝑥,𝑠 So when the fluid is admitted in the cylinder, a part of the volume is already taken by some fluid trapped into this clearance volume. The effect of the clearance volume decreases when the pressure ratio increases because the mass flow increases and the then, the amount of fluid trapped into the clearance volume have less impact.

Semi-empirical model Operating maps Optimal performance of a machine for a given condensation and evaporation temperature !!!All axis are positive!!! Extreme conditions Flexibility Efficiency