Proceedings of the
5th International Seminar on
ORC Power Systems
9 - 11 September 2019, Athens Greece
Home Program Author Index Search

Analysis of Transonic Nozzle Loss Generation in Organic Rankine Cycle (ORC) Turbines


Go-down orc2019 Tracking Number 105

Presentation:
Session: Session 3A: Turbine Nozzles Simulations
Room: Olympia
Session start: 16:00 Mon 09 Sep 2019

Tao Chen   tao.chen17@imperial.ac.uk
Affifliation:

Peter Newton   peter.newton03@imperial.ac.uk
Affifliation:

Miles Robertson   miles.robertson09@imperial.ac.uk
Affifliation:

Ricardo Martinez-Botas   r.botas@imperial.ac.uk
Affifliation:


Topics: - Turbines (Topics), - Oral Presentation (Preferred Presentation type)

Abstract:

The Organic Rankine Cycle (ORC) is a technology facilitating the generation of useful mechanical power, from a wide variety of low-temperature heat sources. The specific application considered in this paper is heavy-duty vehicles, as part of a secondary (or bottoming) cycle. A single stage high-expansion-ratio turbine is a suitable type of expander for this application, the nozzle would be transonic given the high expansion ratios. This paper analyses the loss generation mechanisms within the transonic nozzle, with a view towards improving the design of high-pressure ratio turbine expanders for an increasing stage efficiency. This paper simulates the transonic flow of R1233zd(E) in a linear nozzle blade cascade by means of Computational Fluid Dynamics (CFD). The cascade is designed to achieve a uniform outlet flow Mach number of 2. An initial analysis uses irreversible entropy generation rate to study and explain the loss generating mechanisms in the nozzle under on- and off-design conditions, and the entropy loss coefficient is used to quantitively analyse the loss distribution. The loss generated in the expansion process takes 28.2 % of the overall loss, and the loss in the region after expansion takes the remaining 71.8 %. The impact of real-gas properties is then considered by comparing with an equivalent air cascade. The loss generated in the expansion process for the R1233zd(E) cascade is 19.7 % larger than that for air due to a larger non-ideal expansion loss. The loss generated in the region after the expansion process for the R1233zd(E) cascade is 11.4 % lower than that for air due to a lower loss caused by trailing-edge wave and trailing-edge wake. The overall loss is 4.4 % lower in the R1233zd(E) cascade; conversely, the total pressure loss coefficient is 13.8 % larger in the R1233zd(E) cascade than that in the air cascade. This is explained by an analysis of Gibbs entropy formula for a real-gas which shows that the stagnation pressure loss is attributed to both an entropy gain and a real gas effect.