Investigation of minimum NOx emissions for cracked ammonia combustion

Abstract

Ammonia (NH₃) is being evaluated as a carbon-free energy carrier. However, combustion of NH₃ leads to potentially significant amounts of NO_x emissions as well as flame stabilization challenges. For both reasons, there is interest in partially cracking NH₃ and combusting some blend of NH₃/H₂/N₂. Our prior work has evaluated the minimum theoretical NO_x emissions from pure NH₃ combustion, which is a useful benchmark for evaluating fundamental limits, as well as to evaluate the performance of a given combustion system relative to these theoretical limits. This work is aimed to evaluate the fundamental minimum NO_x emissions of partially and fully cracked NH₃. Significant NO_x benefits are possible with 100% cracked NH₃ – i.e., H₂/N₂ combustion – and the optimal combustion architecture is a lean premixed strategy. However, this lean premixed strategy obviously does not work for partially cracked NH₃ combustion. NO_x emissions for intermediate cracking fractions exhibit both a highly nonlinear and, in certain pressure regions, a non-monotonic dependence upon cracking fraction – in other words, NO_x emissions do not necessarily, linearly decrease with increased cracking. In general, partial cracking does provide NO_x benefits in a manner that is highly pressure dependent; for example, minimum theoretical NO emissions decrease by around 90% and 40% between pure NH₃ and 90% cracked NH₃ at 1 and 20 bar for a system with 20 ms residence time, but a 2% increase in NO is observed for the same system at 4 bar. It is only at cracking levels exceeding about 99% that major NO benefits occur, with minimum NO reaching sub-30 ppm (15% O₂ dry) values for all pressures. Moreover, these results show that rich-lean staged systems lead to optimal NO_x emissions over cracking fractions from about 0 – 99.9%; it is only above 99.9% cracking ratio that traditional lean premixed combustion strategies show comparable results. These results indicate that only if nearly complete cracking is possible, that NH₃ utilization will require retrofitting low NO_x combustors from lean premixed systems to rich-lean staged systems. The sensitivity of these results to the choice of kinetic models is also addressed in this work.