(300g) Modelling and Chemical Compatibility of Super-Critical Carbon Dioxide Flow in Micro-Channels of Microvascular Carbon/Carbon Composites

Solar energy generation in the United States has greatly increased in the last decade, with a tenfold increase since 2010 [1]. Concentrating solar power (CSP) systems use focused sunlight to convert thermal energy to electricity, usually through the use of conventional steam turbines. In a central tower configuration, heliostats are mirrors that follow the sun’s position, directing the light into a receiver located at the top of a central tower. A major breakthrough in CSP systems is the use of thermal storage systems, whereby using molten salt, thermal energy can be stored for multiple hours, giving CSP systems on-demand energy generation capabilities. In light of that, central tower receivers have been developed initially to use molten salt as the heat transfer fluid.

Recently, a U.S. Department of Energy roadmap [2] has identified super-critical carbon dioxide (sCO₂) Brayton-cycles as an integral part of the next generation of CSP plants. By replacing the conventional steam turbines, higher efficiencies can be achieved, further contributing to lowering generation costs of CSP systems. Super-critical fluids have properties between gas and liquids, such as providing a lower pressure drop than liquids, while having a higher heat capacity when compared to gases. Carbon dioxide, whose critical point is at 73.9 bar and 304 K, has a higher density when supercritical, meaning that process equipment can be much smaller.

Recent research has focused on central gas receivers using sCO₂ as the heat transfer fluid [3] in place of molten salt. With this approach, the thermal losses from the heat transfer between molten salt and the sCO₂ Brayton-cycle are lowered. Additionally, the receiver can operate at higher temperatures, which generally correlates to higher global efficiencies. Molten salt receivers have to operate above 260 ËšC to avoid crystallization, and below 575 ËšC to avoid thermal decomposition of the salt. On the other hand, gas receivers operating with sCO₂ are currently being proposed to operate between 530 ËšC and 750 ËšC.

In this work, the use of a modular central receiver (Figure 1(a)) is proposed, where each module (Figure 1(b)) will consist of a novel Carbon/Carbon composite with an embedded microchannel network, where sCO₂ can flow as the heat transfer fluid. Traditional gas receivers use metallic or ceramic materials, which are known to be susceptible to thermal fatigue due to the daily startup operations of CSP plants. On the other hand, the use of C/C composites, which are strong, lightweight and highly heat conductive materials, have a much lower coefficient of thermal expansion, which greatly decreases problems with thermal fatigue. Modules with a micro-channel network are made by combining C/C composites fabrication techniques with the use of vaporization of sacrificial component (VaSC) technology, as previously presented [4].

However, the use of sCO₂ as the heat transfer fluid with C/C composites can cause mass loss of the module through a reaction that occurs between CO₂ and carbon materials above 600 ËšC. This effect may be significant, especially when considering the years of operation that the modules are expected to operate at. There are multiple proposed mechanisms for this reaction, however when it relates to the mass loss of a composite material, it is usually simplified to Equation (1). In addition, at high pressures, carbon monoxide (CO) can compete with CO₂ for active sites, changing the reaction mechanism to include a reversible step [5].

CO_{2 (g)} + C _(s) -> 2CO _(g) (1)

To predict the mass loss effects on the C/C composite modules, a Computational Fluid Dynamics (CFD) approach will be used to model the flow of sCO₂ through microchannels with a reacting surface. To perform the CFD modelling, the software Ansys Fluent is used. Preliminary results were obtained by modelling the flow of sCO₂ at high pressure (200 bar) through a straight microchannel using a 2-D axisymmetric geometry. The models and boundary conditions used for this case are as follows: axisymmetric (2D geometry), turbulent standard k-ε (viscous model), Peng-Robinson equation of state, inlet velocity as a parabolic profile with mean velocity of 5 m/s, inlet temperature of 600 ËšC, outlet pressure of 200 bar, constant wall temperature of 727 ËšC, radius of 0.25 mm and length of 10 mm.

Results from the modelling obtained thus far excludes the reacting surface, which will be used as a baseline for the development of an accurate model that describes the flow of sCO₂. The next step involves the inclusion of a reacting surface to the current model, as well as the evaluation of different geometries and operating conditions. To include the reaction surface into the model, the rate of mass loss of the carbon material can be described by Equation (2), where k is the kinetic constant, n is the pressure exponent and P_CO2 is the partial pressure of carbon dioxide.

r_CO2 = k Pⁿ_CO2 (2)

Therefore, Equation (2) can be used to describe the mass loss of the reacting surface (the C/C composite), following a similar approach from works on surface recession of graphite rocket nozzles [6]. In order to implement a reacting surface into the model, the use of UDFs (User Defined Functions) will be necessary, which consists of a function written in C programming language, that can be used to couple the Fluent solver to complex equations.

The constants k and n from Equation (2) have to be determined experimentally, and since available data on carbon materials is usually limited to graphite, they will be initially used for baseline model results. Future kinetic experiments involve the use of TGA (thermogravimetric analysis) in order to obtain the rate constants for the composites fabricated in this work [4], thus more accurately representing the conditions (rate of mss loss) at which the microvascular C/C composites will be subjected to.

An experimental analysis of the flow of sCO₂ through microchannels of the fabricated C/C composite will be done next. By combining a setup under the expected temperature conditions (600 – 1000 ËšC) with a GC (gas chromatograph) analysis, the developed model can be validated by comparing the outlet gas composition.

By accurately predicting the effect of mass loss through the use of CFD modelling, such as changes in channel geometry, operational conditions (upper temperature limit) can be set, as well the evaluation of the use of protective coatings (such as silicon carbide (SiC)) that can slow the rate of reaction.

References:

[1] U.S. Energy Information Administration (EIA). November 9, 2020, https://www.eia.gov/totalenergy/data/browser/

[2] Mehos, M., Turchi, C., Vidal, J., Wagner, M., Ma, Z., Ho, C., Kolb, W., Andraka, C., & Kruizenga, A. (2017). Concentrating Solar Power Gen3 Demonstration Roadmap.

[3] Zada, K. R., Hyder, M. B., Kevin Drost, M., & Fronk, B. M. (2016). Numbering-Up of Microscale Devices for Megawatt-Scale Supercritical Carbon Dioxide Concentrating Solar Power Receivers. Journal of Solar Energy Engineering, Transactions of the ASME, 138(6), 1–9.

[4] Cordeiro, J. C.; Argot, J., Ramsurn, H., Crunkleton, D. W., Otanicar, T., Keller, M. (2020, November). "Microvascular Carbon-Carbon Composite for Concentrated Solar Power Gas Receivers". In 2020 Virtual AIChE Annual Meeting.

[5] E. S. Golovina. The gasification of carbon by carbon dioxide at high temperatures and pressures". Carbon 18.3 (Jan. 1980), pp. 197-201.

[6] Piyush Thakre and Vigor Yang. Chemical Erosion of Carbon-Carbon/Graphite Nozzles in Solid-Propellant Rocket Motors". Journal of Propulsion and Power 24.4 (July 2008), pp. 822-833.