無碳燃料燃燒研究以氫氣與氨氣噴吹混燒及火焰光譜診斷為核心，從混合—化學耦合等基礎機制出發，發展具實際應用潛力的工業燃燒氮氧化物排放控制技術。Carbon-free fuel-combustion research focuses on hydrogen and ammonia lancing and co-firing together with flame-spectra diagnostics, working from underlying mechanisms such as mixing–chemistry coupling toward nitrogen-oxide emissions-control technologies that can realistically be applied in industrial combustion.
本方向採用氫氣與氨氣噴吹（hydrogen / ammonia lancing）配置：原燃燒器維持不變，空氣與原有碳氫燃料仍經燃燒器供應，僅將無碳燃料單獨噴吹入爐；此與傳統先將碳氫與無碳燃料混合、再分段噴注的作法不同，有利於既有設備改裝。混燃比例、噴吹位置、噴流動量與停留時間共同決定火焰結構與排放；氫燃燒可能因高溫提高熱型 NOx，氨燃燒則可能帶來燃料型 NOx、N₂O 與未反應 NH₃，因此同時量測多種排放物與局部火焰條件。以累積的反應—混合—排放資料為基礎，並結合線上火焰光譜量測與機器學習，發展可即時推估燃燒條件與排放指標的工業燃燒監診技術。The work uses a hydrogen / ammonia lancing configuration: the original burner is left unchanged, with air and the existing hydrocarbon fuel still supplied through it, and only the carbon-free fuel lanced separately into the furnace. This differs from conventional fuel staging, in which hydrocarbon and carbon-free fuels are mixed first and injected in stages, and lends itself to retrofitting existing equipment. Co-firing ratio, lancing position, jet momentum, and residence time jointly determine flame structure and emissions: hydrogen can raise thermal NOx through higher temperatures, while ammonia can add fuel-NOx, N₂O, and unreacted NH₃, so multiple emissions are measured together with local flame conditions. Building on the resulting reaction–mixing–emissions data, in-situ flame-spectrum measurement and machine learning are combined into monitoring techniques that estimate combustion conditions and emission indices in real time for industrial systems.
使用可調式燃燒器與噴吹模組，系統改變主燃料／氫氣／氨氣比例、噴吹高度與深度、噴流動量、當量比、停留時間與爐內幾何。燃燒區量測整合高速火焰影像、化學發光成像、可調式二極體雷射吸收光譜（TDLAS）與煙氣多組分分析（如 FTIR、化學發光氮氧化物分析儀），連結局部反應區條件與整體排放。再以詳細反應機理（Cantera）與 PSR／PFR／CRN 等低階模型分析溫度、氧濃度、混合時間與噴吹位置對 NO、N₂O 與燃料轉化路徑的影響，並發展火焰光譜機器學習模型，由光譜特徵推估局部當量比、溫度與排放指標。Configurable burners and lancing modules allow systematic variation of main-fuel / hydrogen / ammonia ratios, lancing height and depth, jet momentum, equivalence ratio, residence time, and furnace geometry. Combustion-zone measurements combine high-speed flame imaging, chemiluminescence imaging, tunable diode laser absorption spectroscopy (TDLAS), and multicomponent flue-gas analysis (e.g., FTIR, chemiluminescent NOx analyzers) to connect local reaction-zone conditions with global emissions. Detailed kinetics (Cantera) and reduced-order PSR, PFR, and CRN models then analyze how temperature, oxygen level, mixing time, and lancing location affect NO, N₂O, and fuel-conversion pathways, while flame-spectrum machine-learning models estimate local equivalence ratio, temperature, and emission indices from spectral features.