CFD-CRN Study of NOX formatio in a High-Pressure Combustor Fired With Lean Premixed CH4 / H2 --Air Mixtures

A hybrid approach of CFD-CRN have been found useful in predicting pollutants for many combustion systems [3] without incurring very high computational costs. In this approach, flow field and major species are computed using CFD with reduced kinetic mechanism. Further, a chemical reactor network is constructed based on CFD simulation and solved with detailed chemical kinetic mechanism. In many previous researches [3], the construction of chemical reactor network depends on user experience and there are issues in solving large size of chemical reactor with detailed chemical mechanism. These are major hurdles of hybrid approach to implement in many real combustors. The currently developed Reactor network module in Simcenter STAR-CCM+ V 14.04 overcomes these difficulties by solving CFD, constructing reactor network automatically and solving large size of chemical reactor network in single framework. The main objective of current study is to etimate NO_x and CO concentrations at the exit of PSI dump combustor [1] operating at high-pressure for lean premixed methane-air mixture using CFD-CRN approach in Simcenter STAR-CCM+ . Boschek et al [1]  have also studied blending the natural gas with hydrogen or other hydrocarbon for PSI Dump combustor and found that it can be a promising technique to increase the lean blowout limit and reduce or maintain emissions at same level. Hence, this study will be extended to estimate NO_x and CO and understand the NO_x formation pathways with H₂ enrichment of methane-air lean premixed mixture. The CFD was performed by a tabulated approach for premixed combustion using a Flamelet Generated Manifold (FGM) model in Simcenter STAR-CCM+ V 14.04 with the DRM22 reduced mechanism. The wall cooling as well as radiation affects the flame temperature and correct prediction of emissions; hence, CFD is modelled with all three heat transfers modes: radiation, conduction and convection. The conjugate heat transfer (CHT) approach is used by modelling three connected domains as 1) combustion fluid, 2) quartz solid and 3) cooling air fluid, passing in annular region. The radiation is modelled by using participating media radiation. The CFD simulations are performed for five lean equivalence ratios (ϕ = 0.43-0.55, T_inlet= 673 K, V_inlet= 40 m/s) of pure methane-air mixture operating at 5 bar. The exit temperature and flame-length are compared with available experimental data [1, 2]. CRNs are then automatically constructed based on flow parameters including temperature, axial velocity and equivalence ratio distribution for all the cases studied. CRN size independence studies are performed by solving for up to 50,000 PSRs. In preliminary results, it is found that CRNs with 5,000 PSRs can provide adequate accuracy in NO_x prediction compared to CFD. The contribution of NO_x formation pathway changes from N₂O intermediate to thermal NO_x as equivalence ratio increases. Further studies are performed for two equivalence ratios (ϕ = 0.43 and 0.5) to simulate the impact of H₂ addition (up to 20 % by volume) on NOx formation pathways and CO emission.

References

[1]     E. Boschek, P. Siewert, P. Jansohn, Fuel variability effects on turbulent, lean premixed flames at high pressures, GT2007-27496, ASME/TURBO EXPO 2007, May 14-17, Montreal (Canada).

[2]     Siewert, P., Flame front characteristics of turbulent lean premixed methane / air flames at high-pressure, Doctoral Thesis, ETH Zurich, 2006

[3]     Yousefian, S., Bourque, G., and Monaghan, R. F. D., 2017, Review of Hybrid Emissions Prediction Tools and Uncertainty Quantification Methods for Gas Turbine Combustion Systems, Proceedings of ASME Turbo Expo 2017, GT2017-64271, Charlotte, USA.