Feedstock quality gradients and net energy constraints in sustainable aviation fuel: a biophysical economic assessment of the HEFA pathway

Despite extensive policy support and climate commitments, Sustainable Aviation Fuel (SAF) remains more expensive than conventional jet fuel. Hydroprocessed esters and fatty acids (HEFA) are currently the dominant commercial SAF pathway. This paper evaluates a narrower proposition: whether differences in lipid-feedstock quality create a biophysical mechanism that could contribute to persistent production-cost differences as HEFA deployment expands. The analysis is structural and comparative rather than a time-series or causal model of observed prices. Using a biophysical-economics framework, this study compares the external energy required to acquire, prepare, and convert used cooking oil (UCO), conventional crop oils, and purpose-grown oil crops. Under midpoint assumptions for external input and liquid-energy recovery, external-energy EROI is approximately 4.06:1 for UCO, 1.55:1 for conventional crop oils, and 1.82:1 for purpose-grown oil crops. The corresponding full-combination envelopes are 3.0–5.9, 0.99–3.2, and 1.3–3.2, respectively; these envelopes combine favorable or unfavorable parameter endpoints and are not probability distributions. The UCO baseline applies a cut-off attribution convention: no original agricultural-production burden is assigned after the oil has completed its first use, although collection, transport, pretreatment, hydrogen, and conversion energy are included. This convention is not a claim that UCO is physically burden-free. Reassigning 25%, 50%, and 100% of a representative 0.29 MJ/MJ original crop-oil burden lowers the UCO EROI to approximately 3.05, 2.44, and 1.74, respectively; adding 0.50 MJ/MJ lowers it to approximately 1.23. Thus, UCO's absolute advantage is allocation-sensitive, whereas the physical energy required for collection and HEFA conversion is unchanged. The results identify a step from low-attributed-burden waste lipids to an overlapping band of deliberately cultivated oils. Eliminating the modeled midpoint conversion-stage energy would offset only about 15.6% of the representative external-input difference between UCO and conventional crop oil. This static comparison supports a plausible offsetting mechanism between conversion learning and marginal-feedstock quality, but it does not locate a deployment level or date at which one effect overtakes the other. Tallow remains volumetrically important but energetically unresolved, and a numerical petroleum comparison is withheld pending a boundary-harmonized well-to-gate calculation.