Reduced mechanisms are needed for use with computational fluid dynamic codes (CFD) utilized in the design of combustors. Typically, reduced mechanisms are created from a detailed mechanism, which contain numerous species and reactions that are computationally difficult to handle using most CFD codes. Recently, it has been shown that the detailed aramco 2.0 mechanism well predicted the available experimental data at high pressures and in highly CO2 diluted methane mixtures. Here, a 23-species gas-phase mechanism is derived from the detailed aramco 2.0 mechanism by path-flux-analysis method (PFA) by using CHEM-RC. It is identified that the reaction CH4 + HO2 ⇔ CH3 + H2O2 is very crucial in predicting the ignition delay times (IDTs) under current conditions. Further, it is inferred that species C2H3 and CH3OH are very important in predicting IDTs of lean sCO2 methane mixtures. Also, the 23-species mechanism presented in this work is able to perform on par with the detailed aramco 2.0 mechanism in terms of simulating IDTs, perfectly stirred-reactor (PSR) estimates under various CO2 dilutions and equivalence ratios, and prediction of turbulence chemistry interactions. It is observed that the choice of equation of state has no significant impact on the IDTs of supercritical CH4/O2/CO2 mixtures but it influences supercritical H2/O2/CO2 mixtures considered in this work.

The direct-fired supercritical CO2 (sCO2) cycle is currently considered as a zero-emission power generation concept. It is of interest to know how to optimize various components of this cycle using computational tools; however, a comprehensive effort in this area is currently lacking. In this work, the behavior of thermal properties of sCO2 combustion at various reaction stages has been investigated by coupling real gas CHEMKIN (CHEMKIN-RG) (Schmitt et al., 1994, Chemkin Real Gas: A Fortran Package for Analysis of Thermodynamic Properties and Chemical Kinetics in Nonideal Systems, University of Iowa, Iowa City, IA) with an in-house premixed conditional moment closure code (Martin, 2003, “The Conditional Moment Closure Method for Modeling Lean Premixed Turbulent Combustion,” Ph.D. thesis, University of Washington, Seattle, WA) and the high-pressure Aramco 2.0 kinetic mechanism. Also, the necessary fundamental information for sCO2 combustion modeling is reviewed. The Soave–Redlich–Kwong equation of state (SRK EOS) is identified as the most accurate EOS to predict the thermal states at all turbulence levels. Also, a model for the compression factor Z is proposed for sCO2 combustors, which is a function of mixture inlet conditions and the reaction progress variable. This empirical model is validated between the operating conditions 250–300 bar, inlet temperatures of 800–1200 K, and within the currently designed inlet mole fractions, and the accuracy is estimated to be less than 0.5% different from the exact relation. For sCO2 operating conditions, the compression factor Z always decreases as the reaction progresses, and this leads to the static pressure loss between inlet and exit of the sCO2 combustor. Further, the Lucas et al. and Stiel and Thodos methods are identified as best suitable models for predicting the viscosity and thermal conductivity of the sCO2 combustion mixtures.

The sCO2 power cycle concept is identified as a potentially efficient, economical, and pollutant free power generation technique for future power generation. Recent work in the literature provides some strategies and best operating conditions for direct-fired sCO2 combustors based on zero-dimensional reactor modeling analysis, however there is a need for a detailed investigation using accurate combustion chemical kinetics and thermophysical models. Here, the sCO2 combustor is modelled by coupling perfectly stirred reactor (PSR) and plug flow reactor (PFR) models. The real gas effects are incorporated using the Soave-Redlich-Kwong (SRK) equation of state. Also, the detailed Aramco 2.0 kinetic mechanism is used for the combustion kinetic rates. It is found that the primary zone must be diluted either with thirty or forty-five percent of the total CO2 in the cycle to have a feasible combustor design. However, the forty-five percent dilution level at 950 K and 1000 K yielded a better consumption of CO, O2 and CH4. Also, the cross-sectional area of the sCO2 combustor can be scaled-down to 10 to 20 times smaller than a traditional combustor with the same power output. Further, from this investigation, it is also recommended to have a gradually increasing secondary dilution in the dilution zone, by using progressively larger diameter holes. This design would help retain relatively high temperature in the initial portion of the dilution zone and would help consume fuel species such as, CO and CH4. It appears that, for sCO2 combustors “lean burn” is the better strategy over stoichiometric burning to eliminate CO build up at the combustor exit. The lean burn condition at equivalence ratio (ϕ) equal to 0.9 is recommended for sCO2 combustor operation. Also, the length of the dilution zone can be scaled-down to 50% by lean burn operation of the combustor. It is also observed that the lean burn increases the net turbine power. Current work provides crucial design considerations for the development of advanced sCO2 combustors to be used with direct-fired power cycles.

The reactor residence time required for a sCO2 combustor is comparatively higher than an equal power, airdiluted conventional combustor. Therefore, the strategies to reduce the reactor residence time are very important in the design of sCO2 combustors. The current work recommends a method to reduce the residence time requirement in the sCO2 combustion chamber. Here, the combustor is modelled by coupling the perfectly-stirred-reactor (PSR) and plug-flow-reactor (PFR) models along with the detailed Aramco 2.0 combustion chemical kinetic mechanism. The real gas effects are considered by using the Soave-Redlich-Kwong (SRK) equation of state incorporated in CHMEKIN-RG. Though, the CO emission level at the exit of the primary zone of the sCO2 combustor is lower or in some cases equal to the conventional combustor, the further decline of CO in the dilution zone is identified as very poor. Therefore, very high CO levels can be expected at the exit of the sCO2 combustor compared to conventional combustors. CO from the sCO2 combustor exhaust can be eliminated by lean operation of the combustor and the excess O2 retained in the re-cycled CO2 stream due to lean operation can be mixed with primary methane before entering the primary combustion zone. This strategy is shown to reduce the primary zone residence time requirement of sCO2 combustion. However, the minimum level of O2 in the re-cycled CO2 stream is approximately 5000 ppm and the minimum required residence time in this pre-mixing chamber is around 4 ms. Also, it is observed that the primary zone residence time requirement decreases linearly with respect to the O2 level in the re-cycled CO2 stream.

No abstract