Benjamin Emerson

Ammonia (NH3) is being explored as a carbon-free energy carrier, supported by an extensive global production, distribution, and storage infrastructure. Rich-staged combustor designs with sufficient residence time, so-called rich–relaxation–quick-mix–lean (RRQL), are particularly promising for direct NH3 combustion without significant nitric oxide (NOx) production, but nitrous oxide (N2O) and unburned NH3 must also be managed. At atmospheric conditions, RRQL configurations have been shown to reduce NOx to near-equilibrium levels while promoting NH3 cracking to hydrogen (H2), provided that the primary stage maintains sufficiently long residence times (τres,primary) and that thermal losses through the combustor walls remain low. However, the influence of elevated pressure on RRQL systems operating with long residence times (on the order of 102 ms) has not yet been characterized. The present work examines the sensitivity of emissions to pressures over the 1–5 bar range. Measurements of natural luminosity and OH*/NH2* chemiluminescence are used to assess pressure-dependent changes in flame structure and heat release across the rich primary zone and lean secondary zone. Corresponding exhaust data for NOx, NH3, and N2O quantify the degree of NOx relaxation at elevated pressures. Results show that NH3 slip increases with pressure in both the rich and relaxation zones, a behavior attributed to higher heat losses that inhibit effective NH3 cracking. Overall, the study highlights that staged NH3 combustion systems employing long τres,primary, particularly those operating at small laboratory scales and thermal loads, require careful management of relaxation stage heat losses to maintain low-emission performance.

Ammonia (NH3) is a carbon–free energy source and hydrogen carrier, but its fuel–bound nitrogen can lead to significant nitrogen oxides (NOx) emissions. Staged combustion strategies, such as rich–quench–lean, can achieve low NOx emissions. However, improper design may result in high NOx levels when operating with a rich head end without a sufficient post–flame relaxation time. Previous work has shown that a rich–relaxation–quench–lean (RRQL) configuration can minimize NOx emissions in the rich head end (Cole et al., 2024, “Rich Ammonia Flame Shapes and NO Relaxation: Facility Development and Characterization,” ASME Paper No. GT2024–122369). This study focuses on how different swirl geometries affect exhaust emissions and flame morphology in rich, premixed NH3–air flames, for the design of a rich relaxation head end of an RRQL combustor. Experiments were conducted using a modular swirl burner and measuring NOx, N2O, and NH3 emissions, and recording natural flame luminosity and NH2* and OH* chemiluminescence images. Swirler design affects emissions by influencing flame length (with shorter flames allowing for more postflame relaxation time) and flame–wall interactions (influencing NH3 and N2O emissions). A compact flame with minimal wall interactions allows for increased NOx relaxation, while minimal wall interactions prevent heat losses and instabilities, which minimize NH3 and N2O emissions at a richer equivalence ratio, potentially enhancing H2 production during the relaxation phase without increasing overall primary stage NOx emissions, thereby improving system efficiency. Detailed spatial evolution of NOx, N2O, and NH3 emissions further supports the RRQL as a promising combustor design for NH3 combustion, if the rich head end is designed properly.

Staged combustion systems like the rich–relaxation–quench–lean (RRQL) offer the potential for low nitrogen oxides (NOx) emissions while burning ammonia (NH3). This process involves rich premixed NH3–air combustion in a primary zone, allowing time for NOx relaxation, followed by a lean secondary combustion zone via air injection. However, improper design of the secondary stage can lead to high NOx and nitrous oxide (N2O) emissions, offsetting the climate benefits of carbon-free fuels. This study investigates the effects of primary zone length and secondary stage geometry on the stability and emissions of a lab-scale RRQL system. Experiments were conducted at atmospheric pressure using a modular axial swirl burner (swirl number = 1.1, 16 vanes), with primary equivalence ratios (ϕprimary) of 1.13, 1.15, and 1.18. Two quartz lengths (76 and 178 mm) were tested using a five holes (2.03 mm) secondary injection design. Strong flame interaction and elevated NOx were observed for the 76 mm liner. A longer chamber allowed better NOx relaxation and NH3 cracking. Additional tests with 5-, 10-, and 16-holes configurations showed that fewer holes, implying higher momentum flux ratios, yielded lower NOx–N2O emissions, especially at ϕprimary = 1.13. These effects diminished as ϕprimary decreased from 1.18 down to the optimum 1.13. Diffusion-like combustion was seen for 0.90 ≤ ϕglobal ≤ 1.10, leading to inefficient combustion marked by excess O2 values compared with equilibrium. Optimal performance was achieved with ϕprimary = 1.13 and 0.70 ≤ ϕglobal ≤ 0.90, with estimated burner outlet temperatures between 1720 and 1970 K.

Low carbon fuels face growing interest as the power generation and transportation fields embrace decarbonization. However, low carbon fuels may still introduce pollutant emissions challenges. Hydrogen and ammonia are two popular decarbonization fuels that face public scrutiny over potential NOX emissions. However, much of this scrutiny is grounded in unscientific roots, including experiments with improper control parameters or demonstrations with legacy hardware that are not appropriate comparison tests for these fuels. Therefore, EPRI and the Zinn Combustion Lab at Georgia Tech have conducted a series of chemical kinetic simulations and experiments with both hydrogen and ammonia to examine NOX emissions from direct combustion of these fuels with a target of dry, low NOX comparisons. The calculations were conducted to determine burner architectures that promise low NOX potential and fuel flexibility, as well as to predict the minimum achievable NOX (NOX entitlement). The experiments were conducted to validate the calculations and to provide first-ever demonstrations of a new fuel flexible, low NOX burner concept designed for these decarbonization fuels. This paper will present the promising calculations and experimental results. The results show hydrogen NOX emissions similar to natural gas combustion, and low double digit NOX concentrations from direct ammonia combustion.

Low carbon and renewable fuels are an active area of research due to their potential to reduce both stack and life-cycle carbon emissions. Ammonia is one such fuel as it has storage, production, infrastructure, and zero carbon emissions advantages over other fuels. Its challenges include its toxicity, its combustion characteristics, and its pollutant emissions. The authors have developed an ammonia combustor design. The intent of the design is a can annular combustor retrofittable to heavy-duty power generation gas turbines in use today. The combustor design is based on technology the research team has been investigating. The rich relax quench lean (RRQL) combustor is shown by the authors to reduce NOx emissions to levels like that of natural gas based on initial scaled testing. The authors designed a full-scale combustor to be tested in 2025 in the single can test rig at the KEPRI combustion lab in Daejeon, South Korea. To accelerate learnings on the system's design, a scaled version of the same combustor design was designed by the team. The design consists of high-temperature material additive manufacturing combustor parts. Testing of the scaled combustor was completed at the Georgia Tech Combustion Lab in Atlanta, GA. Learnings from this testing were also used to improve the full-scale design. Scaled testing has shown great promise in NOx emissions, flame stability, fuel flexibility with blends of natural gas, and ignition capability. Full-scale and subscale combustor design development, testing, results, and future plans are all discussed.