Supercritical CO2 (sCO2) power cycles offer higher efficiency for power generation than conventional steam Rankine cycles and gas Brayton cycles over a wide range of applications, including waste heat recovery, concentrated solar power, nuclear, and fossil energy. sCO2 cycles operate at high pressures throughout the cycle, resulting in a working fluid with a higher density, which will lead to smaller equipment sizes, smaller carbon footprint, and therefore lower cost. However, the combinations of pressure, temperature, and density in sCO2 power cycles are outside the experience of many designers. Challenges in designing sCO2 cycles include turbomachinery aerodynamic and structural design, bearings, seals, thermal management and rotordynamics. According to the report from Sandia National Lab, compressors operating near critical point and turbines have received only TRL (technical readiness level) 4 and 5 out of 9. This blog discusses the impact on turbomachinery design.
Radial or Axial
The selection of radial or axial for turbomachinery is typically performed based on the operating conditions (adiabatic head H and inlet volumetric flow Q). Non-dimensional turbomachinery parameters of specific speed Ns and specific diameter Ds can be selected from NsDs charts to estimate size, speed, and type of turbomachinery. Turbomachinery types for a sCO2 recompression cycle with scales ranging from 100 kW to over 300 MW have been studied and concluded that systems below 10 MW will likely feature only radial turbines and compressors with a single-stage or low stage counts. Such recompression cycle can be simulated in AxCYCLE™ tool which is shown in Figure 1. As size increases, the most efficient configuration for the turbine and recompressor transitions from radial to axial at approximately 30 MW and 100 MW, respectively. Suitable types of turbomachinery and its components for different power range can be reviewed in Figure 2. A radial configuration for the main compressor was expected at all scales due to its lower volume flow and wider range to facilitate variation in gas properties due to operation near the critical point.
Compressor Design Consideration
Both open (with fully visible blades) and shrouded impellers can be designed for centrifugal compressors. A closed impeller improves efficiency by eliminating blade tip leakage while its performance is unaffected by axial thermal growth mismatches between the rotor and stator assemblies. Also, covered impellers are generally less prone to high-cycle fatigue failures than open impellers. Therefore, closed impellers are considered favorable for most sCO2 cycles as the high pressure and high-density fluid would require very tight clearances to keep leakages as low as possible. The high fluid density in sCO2 impellers will also affect the natural frequencies of blade-dominant modes and generates relatively high aerodynamic loading amplitudes. Thus, impeller designs should consider the dynamic stresses resulting from wake excitation from upstream and downstream stator components such as inlet guide vanes, diffuser vanes, and struts that apply periodic excitation to the blades.
From an aerodynamic performance side, there are some unique design considerations that must be understood. First, a good design should be optimized to maximize design point efficiency since the compressor power has a first-order effect on cycle efficiency. Second, the inlet conditions to the main compressor section are usually near the critical point of the working fluid to reduce the compression work and maximize the cycle output. To operate near the critical point a stage should be designed to manage a wide variation in inlet properties associated with small changes in temperature.
Through all phases of design, it is critical to use accurate fluid properties models to capture the real gas variations in the gas. The NIST REFPROP software produces accurate properties for pure CO2 near the critical point and can be coupled directly to many compressor design software like AxSTREAM®. Figure 3 is a sCO2 centrifugal compressor designed in AxSTREAM®.
The inducer should be designed to minimize either the tip relative velocity or the inlet relative Mach number. Too large an inducer will reduce the inlet blade angle beyond optimum which limits the efficiency. Too small an inducer will cause static pressure and suppressed temperature, so that static properties may move closer to the saturation line, therefore potentially leading to unwanted phase change.
Some test data showed steady operation at a variety of points across the entire saturation region with no apparent harmful effects. These results suggest that even if two-phase operation occurs, the densities for liquid and vapor phases at high pressure are similar enough to avoid harmful operation. However, until additional experimental exploration and practical operating data are available, it is good practice to maintain some margin from the saturation line during the design process.
Turbine Design Consideration
High-temperature components in sCO2 cycles operate at temperatures similar to ultra-supercritical steam turbines, and much of the work on alloys for steam applications is applicable to sCO2 cycles. For sCO2 turbines operating at high temperatures approaching or exceeding 700 °C, designs are creep-limited and nickel-based alloys are required to achieve sufficient creep strength. Different materials suitable for sCO2 are compared in Figure 4. In addition to creep properties, another critical consideration for turbine materials is corrosion performance in an sCO2 environment. There is a continued need for materials testing to determine sensitivity to CO2 purity, corrosion performance for various CO2 mixtures, and testing in a high flow velocity environment to confirm real-world corrosion behavior.
The sCO2 turbine may be of the radial inflow or axial design. In most sCO2 cycles, the turbine inlet temperature is well above the critical temperature, and the gas behavior approximates that of an ideal gas. Because of this, the turbine design can be achieved using existing design practices and tools like AxSTREAM® from steam and gas turbines for other applications. Generally, turbines are designed with the objective of maximizing efficiency in the fewest number of stages. Their blade/impeller attachments are similar to other turbomachinery applications with several key differences. One critical difference is that a sCO2 turbine has a higher power density than other types (the one exception being rocket engines turbopumps), so the pressure (static) loading on turbine blades cannot be ignored as it is with low-density applications. For axial turbines, in order to avoid large bending and tensile stress on blades, an impulse type can be chosen since a reaction type will have higher stress. However, if the bending and tensile stresses allow it, flow paths for the reaction type are proven to be more efficient than the impulse type but do ultimately lead to more stages and therefore increase the plant footprint. Also, nozzles and rotors should be designed to have longer chord and larger LE and TE radius to reduce these blade stresses. The flow meanline aero-structural optimization algorithm is integrated in AxSTREAM®. Its goal is to find the maximum efficiency value while satisfying the structural limitations (the value of chords for which actual stresses will be lower than allowable). With the optimization process in AxSTREAM® for blades which makes them suitable for sCO2 turbine, the value chords will be increased significantly to satisfy the structural requirements as shown in Figure 5. In addition, integral shrouds are typically used and improve blade dynamics, damping, and aerodynamic performance.
Wide operating range requirements and potential for condensation in the compressors and high-temperature pressure containment and compact thermal management in the turbines have been discussed above. Furthermore, the combined high-pressure, high-temperature, and high-density operating environment also bring multiple design challenges like high bearing surface speeds and loads, dense gas effects on rotordynamics and blade loading, low-leakage shaft end sealing and etc. These challenges require significant engineering to overcome before sCO2 turbomachinery can begin to displace steam turbines or gas turbines, which have been developed and refined for over 100 years. Despite these challenges, a number of sCO2 turbomachinery designs and prototypes have been successfully developed in the past decade. With existing technologies and tools, in addition to data from prototype testing, development and eventual commercialization of sCO2 turbomachines for a variety of applications are expected to succeed in the coming years.
- Musgrove, Grant & Allison, Timothy & Ames, Robin & Anderson, Mark & Bennett, Jeff & Brese, Rober & Brun, Klaus & Bueno, Pablo & Carlson, Matt & Chordia, Lalit & Clementoni, Eric & Dennis, Rich & Ertes, Bugra & Fleming, Darryn & Fourspring, Patrick & Friedman, Peter & Held, Timothy & Lawson, Seth & Moisseytsev, Anton & Wright, Steven. (2017). Fundamentals and Applications of Supercritical Carbon Dioxide (SCO2) Based Power Cycles 1st Edition.
- Allison, Tim; Wilkes, Jason; Brun, Klaus; Moore, Jeffrey (2017). Turbomachinery Overview for Supercritical CO2 Power Cycles. Turbomachinery Laboratory, Texas A&M Engineering Experiment Station.
- Moroz L, Frolov B, Burlaka M, Guriev O. Turbomachinery Flowpath Design and Performance Analysis for Supercritical CO2. ASME. Turbo Expo.