Investigations on Radial Inflow Turbine for Supercritical CO2 Power Cycles

Abstract

Potential of using supercritical CO2 (sCO2) as a working fluid for the next generation of thermal power plants is being actively pursued by researchers across the world. CO2 has a critical temperature of 31.1 °C and a critical pressure of 7.38 MPa. Therefore, CO2 can easily be pressurized to its supercritical state using a compressor. In comparison to gas Brayton cycles, the compression process in sCO2 cycles consumes less power as the compression process starts near the critical point where fluid density is high. However, higher operating pressures of sCO2 cycles along with closed-loop operation result in large density variations throughout the cycle. This poses significant engineering challenges in practically realizing the cycle. Presently, sCO2 cycles are primarily being explored for small-scale power generation suitable for small modular nuclear reactors, concentrated solar power plants, energy storage, and waste heat re-covery systems. Radial inflow turbines are the natural choice for small-scale (≤ 5 MW) power gener-ation due to their efficient, economical, and robust design. Studies on sCO2 cycles suggest turbine inlet temperature varying between 400 °C and 600 °C, with inlet pressures of approximately 15–20 MPa and optimal turbine outlet pressures of 9–10 MPa. The density of sCO2 at the turbine exit is significantly higher than that of steam leaving a condensing steam turbine or the exhaust of a gas turbine. As a consequence of the high density, sCO2 turbines are smaller in size and required to op-erate at significantly high rotational speeds. sCO2 radial turbines are approximately 4–5 times small-er in size compared to similar capacity steam turbines and are required to operate at ~100,000 rpm for sub-MW power generation. Literature on sCO2 radial turbines is limited as active research in this area started from year 2010 onward. Existing literature largely relies on adopting the conventional gas turbine design prin-ciples and correlations to design sCO2 radial turbines. However, considering the distinct boundary conditions across sCO2 turbines and real gas properties of CO2 in supercritical state, existing gas turbine design principles cannot be directly applied in designing a sCO2 turbine. Additionally, the combined effect of high operating speeds and small size with narrow flow passages leads to distinc-tive flow characteristics and loss mechanisms than steam or gas turbines. Hence, the applicability of extending the existing design methodologies, correlations, and loss models needs to be examined and suitably modified (if needed) to arrive at optimal sCO2 turbine designs. Besides, the practical limits of high-speed operation force designing sCO2 turbines for low specific speeds, thus making them inherently less efficient. Therefore, it is important to identify and quantify the effect of vari-ous loss mechanisms prevalent at low specific speeds and arrive at suitable designs for optimal effi-ciencies. Furthermore, small geometries and high operational speeds impose new engineering challenges spanning multiple domains, such as manufacturing, assembly, material, corrosion, sealing, and rotodynamics. Experimental literature on radial sCO2 turbomachines reports substantial parasit-ic loss due to disk friction and leakage at the rotor backface. However, no studies have been reported to quantify or mitigate parasitic loss for sCO2 radial turbines. Therefore, additional research is need-ed to understand the influence of parasitic loss on the design of the turbine. This research attempts to address critical research gaps envisaged from the existing literature. The thesis starts with a detailed method to design radial inflow turbines tailored for supercritical CO2 applications. The engineering challenges associated with the small size and high-speed opera-tions are addressed by incorporating the limitations of manufacturing, material strength, flow phys-ics, and rotodynamics. Subsequently, a theoretical model is proposed to determine the practically feasible speed range of sCO2 radial turbines for different power scales, considering these engineering limitations. Computational fluid dynamics (CFD) tools are employed to evaluate the optimal design parameters for sCO2 applications, which reveal significant differences compared to gas turbine-based correlations. The sCO2-based design considerations lead to a 3–5% increase in turbine efficiency. Subsequently, the research evaluates the loss mechanisms in radial sCO2 turbines. The assess-ment of loss mechanisms suggests a prevalent Coriolis effect for the high-specific speed designs, whereas domination of viscous frictional losses for low-specific speed designs. The parasitic loss, which includes leakage and disk friction loss at the rotor backface, is significant in sub-MW scale turbines, resulting in a 4–15% drop in turbine efficiency. Sizable backface clearance due to small turbine size, denser fluid, and high rotational speeds are the primary reasons for higher parasitic loss. Radial labyrinth seals are employed at the backface housing, which effectively minimizes para-sitic loss. In low-specific speed designs, the viscous losses in the volute are substantially high, con-tributing to around 30–40% of the total loss. A boundary layer-based quasi-one-dimensional loss model is developed to predict the volute losses in the preliminary stage. The loss model accurately calculates volute loss within ±10% error while consuming negligible computational resources com-pared to three-dimensional CFD simulations. Finally, the work has helped bridge the gap between theoretical research and practice. The learnings and knowledge acquired from the research have aided the realization of a 60 kW radial turbine intended to be deployed in the sCO2 Brayton test loop at Indian Institute of Science.