CHARACTERIZATION OF HYDROGEN COMBUSTION IN INTERNAL COMBUSTION ENGINES WITH LOW NOx EMISSIONS

Abstract

Green hydrogen, produced from water electrolysis, is a promising carbon-free energy storage medium that, when blended with natural gas, can be transported in existing natural gas pipeline infrastructure and utilized in associated power plants to reduce the carbon intensity of existing natural gas-powered systems. However, hydrogen’s high reactivity presents unique challenges in maintaining stable combustion while balancing low NOx emissions in traditional dry low-NOx combustion systems designed for natural gas fuel. Therefore, there is a need to develop retrofittable combustion technologies that can be adapted to existing gas turbine engines to permit high hydrogen fuel blend ratios. State-of-the-art Computer-Aided Engineering (CAE) tools are needed to study the impact of hydrogen within these devices and to design next-generation novel combustor design strategies to enable high hydrogen/low NOx combustion. At the heart of these CAE tools is a chemical mechanism to model the combustion reactions and species generated throughout the engine; however, the size of the mechanism is strongly tied to the computational expense. For these reasons, it is critical to have a high-fidelity and validated yet tractable mechanism that can be used in advanced computational fluid dynamic modeling approaches while also capturing the essential reactions responsible for flame stability and NOx emissions. Many chemical mechanisms exist, each designed for combustion applications and validated with experimental data under unique conditions. Here we quantitatively compare several of the most common small species chemical mechanisms for suitability in modeling at gas turbine engine-relevant conditions. A detailed report was generated from an extensive literature review comparing existing hydrogen and natural gas chemical mechanisms to a wide range of laminar flame speeds, ignition delay time, NOx emissions, and extinction limit experimental data. The predictability of existing chemical mechanisms varies from case to case, but discrepancies were observed at high pressures (>10 bar) and for blended fuels (natural gas and hydrogen). A large gap was found in current laminar flame speed experimental data at high pressures and temperatures, where the most significant discrepancies were noticed between existing chemical mechanisms. NOx flame modeling validation was also performed for several common NOx chemistry sub-mechanisms, including diffusion flame experimental data. This data aimed to find and produce a high-fidelity, low-cost chemical mechanism that best models partially premixed high hydrogen concentration combustion at engine-relevant conditions. However, the large gap in experimental literature at these conditions has led to extrapolated simulations showing inconsistencies, proving that marginal mechanism adjustments will be an unfruitful venture.To address this gap in the literature, this work proposes to investigate hydrogen/natural gas mixtures in a Laser Ignited-Rapid Compression Machine (LI-RCM). The LI-RCM can analyze combustion at engine-like pressures and temperatures (>20 bar, >600 K) using combustion characteristics like ignition delay and laminar flame speeds and can visualize combustion using high-speed schlieren imaging to view a propagating flame. The test plan includes a high concentration of hydrogen, >60%vol in methane at three equivalence ratios (lean, stoichiometric, rich) in air. This data will assist in tuning existing chemical mechanisms at operating conditions without the need to extrapolate to these conditions. Thus, allowing for a more robust chemical mechanism and optimistically reducing modeling tuning time.