Reactor Network Analysis of a Novel Combustion Concept Based on the Heat Exchange Between Fuel Rich Products and By-Pass Air

Abstract Industrial gas turbines rely on the lean-burn concept to keep nitrogen oxide (NOx) emissions below increasingly stringent regulatory limits. The formation of nitrogen oxides increases in fact exponentially with temperature. By mixing fuel and air prior to combustion, high temperature hot spots in the products and therefore high levels of NOx emissions can be avoided. As an alternative to the lean-burn concept, known as Rich-Quench-Lean (RQL), NOx emissions can also be limited by staging the combustion in a first rich stage, i.e. mixing the fuel with only part of the air, followed by the addition of the remainder by-pass air. In fact, combustion under sufficiently rich conditions, due to the lack of oxygen, is characterised by low NOx emissions. However, the successive mixing of the hot products of the first stage with the by-pass air must be carried out very rapidly, at the cost of pressure losses, in order to avoid non-premixed combustion of the residual fuel, with very high temperatures and thermal NOx formation rates. In this paper we propose a modified version of the RQL concept, characterised by an additional intermediate stage where heat (but not mass) is exchanged between the hot gas from the first stage and the by-pass air. The effect of such heat exchange is shown to be purely thermodynamic, consisting in the reduction of the stoichiometric peak temperature of the diffusion flame resulting from the contact between the (cooled) fuel-rich hot gas and the (heated) by-pass air. Reducing the peak stoichiometric temperature helps to limit NOx emissions by allowing slower mixing of the first-stage hot gas with the by-pass air, thus reducing pressure losses. This idea can also be extended to aero engine combustion where the RQL concept is widely used today.