City Research Online

Abstract

Supercritical carbon dioxide (sCO2) blends have been found to be promising for enhancing the performance of power cycles for concentrated solar power (CSP) applications; allowing for up to 6 percentage points enhancement in cycle efficiency with respect to a simple recuperated CO2 cycle, depending upon the nature of the used blend and the choice of cycle configuration. Despite this promising potential, there have been limited studies into axial sCO2 turbine design in comparison to growing interest in the design of radial turbines for small-scale applications. This thesis focuses on the mean-line flow path design of a multi-stage axial turbine operating with sCO2 blends for installation in a 100 MWe CSP plant.

A multi-stage axial mean-line turbine design tool is first developed. The tool is coupled with multiple loss models which are classically used for axial turbine designs to predict
the performance over a range of operating conditions. Following this, the aerodynamic design tool is constrained with both mechanical and rotordynamic design criteria to allow
for developing feasible flow path designs from an industrial standpoint. The mean-line design methodology is verified against multiple case studies from the literature in addition
to 3D CFD simulation results.

The prediction capability of the loss models is investigated for conventional air, sCO2 and ORC turbines over a range of scales given that these models were originally developed for conventional working fluids such as air and steam. Following the loss model comparison, multiple flow paths are designed for sCO2 based blends, namely CO2/TiCl4, CO2/C6F6 and CO2/SO2. Ultimately, similitude theory is used to generate the turbine offdesign performance maps and evaluate the turbine performance over a range of operating
conditions.

The Aungier loss model was found to be suitable for predicting the performance of large-scale sCO2 based turbines whilst the Dunham and Came and Craig and Cox models
were found to over-predict and under-predict the turbine performance with respect to the Aungier model respectively. Using the Aungier loss model, selected blend and cycle configuration for the 100 MWe CSP plant, a 130 MW 14-stage axial turbine is designed for a precompression cycle operating with CO2/SO2 blend. This design is capable of achieving a total-to-total efficiency of 93.8%. A good agreement is achieved between the mean-line design tool and CFD results with a maximum difference of 0.5% in the total-to-total efficiency. Ultimately, the off-design performance of the turbine showed large deviations in the predicted total-to-total efficiencies compared to the CFD results. A maximum deviation of 23% is obtained in the total-to-total efficiency at a mass flow rate of 52% of the design point attributed to the flow separation occurring at the downstream stages.

To conclude, this thesis presents detailed insights into the main mean-line design and performance analysis aspects of large-scale axial turbines for sCO2 based applications. This has considered aerodynamic, mechanical and rotor-dynamic design aspects in addition to the off-design performance analysis.

Publication Type:	Thesis (Doctoral)
Subjects:	Q Science > QD Chemistry T Technology > TD Environmental technology. Sanitary engineering
Departments:	School of Science & Technology > Engineering School of Science & Technology > School of Science & Technology Doctoral Theses Doctoral Theses

Export

Downloads