HEFA-Derived Sustainable Aviation Fuel: A Review of Catalytic Upgrading, Techno-economic Viability, and Machine Learning-Guided Optimization

Abstract Commercial aviation contributes approximately 2% of global anthropogenic CO2 emissions, while non-CO2 effects, including nitrogen oxides (NOx), soot/black carbon particles, and contrail-induced cirrus formation, substantially increase the total climate impact. Hydroprocessed esters and fatty acids synthetic paraffinic kerosene (HEFA-SPK) remains the most commercially deployed ASTM-certified sustainable aviation fuel (SAF) pathway, reducing life-cycle greenhouse gas (GHG) emissions by approximately 55–92% on a CO2-equivalent basis relative to conventional Jet A-1, with variations driven by feedstock type, hydrogen source, and life-cycle assessment methodology and assumptions. However, large-scale HEFA-SAF deployment remains constrained by catalyst deactivation, limited jet-range selectivity, high hydrogen demand, and volatile lipid feedstock costs. Unlike previous HEFA-SAF reviews that evaluated catalyst development, process conversion, and economic assessment in isolation, this work provides a multiscale integration framework that systematically links molecular-level catalyst mechanisms with techno-economic performance and machine learning (ML) optimization. Within this framework, catalyst selectivity among the “hydrodeoxygenation” (HDO), “decarboxylation” (DCX), and “decarbonylation” (DCN) pathways is directly connected to hydrogen consumption and the feedstock-dependent minimum jet fuel selling price (MJFSP), whereas “hydrocracking” and “hydroisomerization” efficiency are linked to hydrogen costs and process configurations. Drawing on the literature from 2017 to June 2026, this review critically evaluates the opportunities and limitations of ML across the HEFA-SAF value chain, emphasizing that ML-guided catalyst discovery requires rigorous experimental validation before industrial implementation. To address data fragmentation across the published literature, a standardized, integrated HEFA data set framework─combining catalyst-level descriptors, process-level parameters, techno-economic indicators and a 2025–2030 research roadmap─is introduced as the primary contribution. The roadmap identifies four critical priorities: extending catalyst lifetime under realistic feeds, reducing external hydrogen dependence, establishing reliable low-cost feedstock supply chains, and developing validated data sets for ML-driven optimization. By establishing explicit linkages across catalytic science, process engineering, techno-economic analysis, and data-driven tools, this review accelerates cost-competitive and scalable HEFA-SAF deployment.