| description abstract | Future energy conversion devices need to operate in dynamic environments with stringent emissions requirements and various potential fuels, including hydrogen, ammonia, methane, and other light hydrocarbons from fossil or anthropogenic sources. State-of-the-art gas turbines for power generation use lean premixed (LPM) combustion to minimize NOx emissions. In achieving remarkably low emissions, they have sacrificed fuel- and operational-flexibility due to their premixed nature. Revisiting nonpremixed combustion architectures, the de facto standard before the widespread adoption of LPM, could significantly expand operational envelopes. However, any nonpremixed architecture must exceed the emissions performance of current LPM engines. This paper explores NOx emissions behaviors from the recently proposed nonpremixed, rich, relaxation, lean (NRRL) combustor architecture across various fuels and blends consisting of CH4, H2, and NH3. We analyze the fundamental minimum emissions characteristics of this concept using chemical reactor network models and compare NRRL performance to that of a LPM concept. This paper shows that NRRL architectures enable low NOx emissions regardless of fuel. These fundamental minimum emissions levels are similar or better than those of LPM given sufficient combustor resident times. Generally, the NRRL concept favors higher pressure, higher temperature, more H2, and longer residence times, while LPM systems do better at lower pressures, temperatures, and residence times. This work also shows the importance of and optimally manages the production and destruction pathways of cyanides and amines—key precursors to lean NO formation—which are unique to the sequential nonpremixed and rich zones found in an NRRL system. | |
