Skip to main content
Back to Blog
Sustainability

Sustainable Aviation Fuel: the cost-per-tCO2-avoided math for 2026 freight

By Yash Dhote · · 12 min read

Sustainable aviation fuel sits at roughly 0.3% of global jet fuel supply in 2025, per IATA’s tracking. ReFuelEU Aviation mandates SAF blending at 2% by 2025, 6% by 2030, and 70% by 2050 for flights departing EU airports. SAF costs 2–3x conventional Jet A depending on production pathway. Cost per tonne CO2e avoided lands between USD 300 and 800 for HEFA pathway SAF (the cheapest), and easily three to five times that for synthetic e-fuels. Aviation freight is going to use SAF because the regulation makes it use SAF, not because the cost-per-tCO2-avoided makes it the rational pick. Here is the actual maths.

I write about sustainability and compliance for EcoFreight, and SAF is one of the topics where the gap between marketing prose and operational reality is widest. Cargo customers ask me about it because they read the press releases. Airline operators talk about it carefully because they understand the supply constraint. The blunt take: SAF is real, mandates are real, the cost arithmetic doesn’t make it cheap, and SAF will not be the dominant decarbonisation lever for aviation freight before 2035. Other levers — modal shift, network rationalisation, more efficient airframes, even just better load factors — deliver more abatement per dollar through the next decade. The factor table that feeds these comparisons is published as part of the GLEC Framework v3.2 overlay we use across the EcoFreight calculator.

What SAF actually is

Sustainable aviation fuel is a drop-in replacement for conventional Jet A that meets the ASTM D7566 specification — the same spec gates whether a turbine engine will safely burn it. SAF is currently certified for blending up to 50% with conventional Jet A under ASTM D1655 Annex A1, with 100% SAF approvals in progress at Rolls-Royce, GE, and Pratt & Whitney for specific engine families. Most flights operating "SAF flights" today use a 5–30% blend; pure-100%-SAF demonstration flights have happened on specific commercial routes (United’s 2021 flight, Virgin Atlantic’s 2023 trans-Atlantic) but are not the normal operating mode.

There are seven approved SAF production pathways under ASTM D7566. The three that matter commercially in 2026 are:

  • HEFA (Hydroprocessed Esters and Fatty Acids): Made from used cooking oil, tallow, and vegetable oils. Cheapest and most commercially scaled. Roughly 85% of SAF volume in 2025 came from HEFA. Lifecycle GHG reduction vs Jet A: 65–85% depending on feedstock.
  • AtJ (Alcohol-to-Jet): Made from ethanol or isobutanol via dehydration and oligomerisation. LanzaJet, Gevo, and others operate at meaningful scale. Lifecycle GHG reduction: 75–90%.
  • PtL or SAK (Power-to-Liquid / Synthetic Aviation Kerosene): Made from green hydrogen and captured CO2. Premium pathway, near-zero lifecycle GHG, but production cost is presently 4–8x Jet A. Lifecycle GHG reduction: 90–100% when input electricity and CO2 sourcing meet RED III criteria.

ICAO’s CORSIA-eligible fuel list and the European Commission’s ReFuelEU regulation both reference the same ASTM pathways, with additional EU-specific sustainability criteria layered on top. ReFuelEU Aviation Regulation (EU) 2023/2405 sits on EUR-Lex if you want the legal text.

The supply share: 0.3% of global jet fuel

Global jet fuel demand in 2025 sits at roughly 400 million tonnes / 510 billion litres. SAF production in 2025 came in at roughly 1.5 million tonnes equivalent. That is approximately 0.3%. IATA’s tracking and the IEA’s 2025 Aviation Outlook both converge on similar numbers; the IEA Net Zero Roadmap had projected ~1.0 million tonnes for 2025 from earlier modelling, so production is slightly ahead of the lowest-case projections but still nowhere near the levels the long-horizon targets need.

ReFuelEU Aviation mandates SAF blending for flights departing EU airports: 2% by 2025, 6% by 2030, 20% by 2035, 34% by 2040, 42% by 2045, and 70% by 2050. The 2025 mandate took effect 1 January 2025, with the first year’s compliance assessment closing 31 December 2025 and reports due to the Commission in 2026. Within the SAF mandate, there are sub-mandates for synthetic e-fuel (PtL) specifically: 1.2% of total fuel must be PtL by 2030, climbing to 35% by 2050. The synthetic sub-mandate is what makes the policy aggressive in cost terms, because synthetic SAF is the most expensive pathway.

The cost per litre

Jet A wholesale prices in 2025 ran USD 600–750 per tonne, or roughly USD 0.65–0.85 per litre. SAF prices vary considerably by pathway:

  • HEFA SAF: USD 1,800–2,400 per tonne / USD 1.65–2.20 per litre — roughly 2.4–3.0x Jet A.
  • AtJ SAF: USD 2,200–2,800 per tonne / USD 2.00–2.55 per litre — roughly 2.8–3.4x Jet A.
  • PtL / synthetic SAF: USD 3,500–6,500 per tonne / USD 3.20–5.90 per litre — roughly 4.5–8x Jet A.

The price premium for HEFA has tightened since 2022 as production scaled, but it has not collapsed because the feedstock (used cooking oil, tallow) is itself constrained. UCO supply globally is roughly 6–9 million tonnes annually, of which a portion goes to biodiesel and other markets, leaving SAF with a hard supply ceiling well below what the long-horizon mandates require. AtJ ethanol feedstock is more abundant but the conversion costs are higher. PtL has effectively unlimited theoretical feedstock (water, electricity, captured CO2) but requires green hydrogen at prices well below current levels to be cost-competitive.

For an aviation freight operator looking at the 2026 cost picture: a tonne of cargo flown Hong Kong to Memphis on a 777F burns roughly 200 kg of jet fuel per tonne of cargo over the 12,760 km great-circle distance. If the operator is paying a 6% SAF blend mandate in 2030, that’s ~12 kg of SAF in the fuel mix. At a SAF premium of USD 1,200/tonne over Jet A (mid-range HEFA premium), the additional fuel cost per tonne of cargo is roughly USD 14. That looks small per tonne; spread across the air-cargo P&L it adds up to meaningful millions for a network carrier.

The cost per tonne CO2e avoided

This is the analytical number that matters when comparing SAF against other abatement options. The arithmetic:

Take HEFA SAF at a premium of USD 1,200/tonne over Jet A. Lifecycle GHG reduction vs Jet A is roughly 75% (mid-range). Burning one tonne of Jet A emits roughly 3.15 t CO2e on a TTW basis and roughly 3.85 t CO2e on a WTW basis. So one tonne of HEFA SAF substituting for one tonne of Jet A avoids roughly 0.75 × 3.85 = 2.89 t CO2e WTW.

  • Cost per tCO2e avoided = USD 1,200 / 2.89 t = USD 415 per tCO2e avoided

That number sits in the USD 300–800 per tCO2e range that gets cited for HEFA SAF in the IEA Net Zero Roadmap and the recent Argus Media studies. Different HEFA feedstocks (UCO vs tallow vs purpose-grown oilseed) and different lifecycle assumptions about the feedstock’s land-use change baseline produce numbers spanning that range. The lower end (USD 300) assumes optimised feedstock and lower SAF premium; the upper end (USD 800) assumes constrained feedstock and a higher premium.

For synthetic PtL SAF, the same arithmetic gives:

  • PtL premium over Jet A: USD 3,000–5,000/tonne
  • Lifecycle GHG reduction: 95% (assuming green hydrogen and direct-air-captured CO2)
  • CO2e avoided per tonne PtL = 0.95 × 3.85 = 3.66 t CO2e WTW
  • Cost per tCO2e avoided = USD 3,000 / 3.66 = USD 820 — or up to USD 5,000 / 3.66 = USD 1,366

So synthetic SAF runs roughly USD 800–1,400 per tCO2e avoided in 2026. That is among the most expensive named abatement options in the IPCC AR6 abatement cost curves. Direct air capture sits in a similar range; nature-based offsets with high-integrity certification sit at USD 50–150; renewable power displacing coal sits below USD 20.

Compared to alternative abatement levers

For an air-cargo operator with USD 200 million of annual emissions exposure under EU ETS Maritime-adjacent regimes, the rational portfolio of decarbonisation moves in 2026–2030 isn’t SAF-first. The cost-per-tCO2-avoided ranking for aviation freight, my working numbers:

  • Network rationalisation, load-factor improvements: Negative cost in many cases — pays for itself. USD −100 to USD 0 per tCO2e avoided.
  • Modal shift to ocean for non-time-sensitive cargo: Often saves money outright. The CO2e abatement is essentially a free byproduct.
  • Airframe replacement (787, A350, A321XLR vs older types): USD 50–150 per tCO2e avoided, factoring in fleet capex amortised over fuel saved.
  • HEFA SAF blending at 2–6%: USD 300–500 per tCO2e avoided (the lower end of the HEFA range).
  • Higher SAF blends (10%+): USD 500–800 per tCO2e avoided as feedstock supply constraint pushes premium up.
  • Synthetic PtL SAF: USD 800–1,400 per tCO2e avoided.
  • Direct air capture with offsetting: USD 600–1,000 per tCO2e captured, with vintage issues to consider.

If you let an operator allocate freely, they would do all of network rationalisation, load-factor improvement, and modal shift first, then airframe replacement, then a small slice of HEFA SAF. They would not voluntarily reach for synthetic SAF until the cost curve had moved substantially.

But ReFuelEU does not let operators allocate freely. The synthetic sub-mandate forces synthetic SAF uptake regardless of cost-curve position. CORSIA on the global side adds offset purchase requirements. So the actual 2026–2030 mix is mandated, not optimised. My honest take: this is the right political call — if synthetic SAF is going to ever be cheap, somebody has to buy the first expensive batches at scale, and forcing demand creates the supply — but the cost-per-tCO2-avoided number on the early synthetic batches is brutal. Our team has scored the airframe-replacement and load-factor moves against SAF in side-by-side spreadsheets for several aviation-cargo clients; on every spreadsheet I have run, SAF lands fifth in the priority stack at best. The slow-steaming maths follows a similar pattern in ocean freight — cheap moves first, expensive alternative-fuel pathways last.

SAF book-and-claim and what gets counted

One operational note that matters for freight customers. SAF that is physically loaded at the departure airport reduces emissions on that specific flight. SAF that is bought via book-and-claim — where the customer purchases the certificate but the physical SAF goes to a different flight at a different airport — reduces total system emissions but does not change the emissions of the specific shipment.

Under CSRD Scope 3 Category 4 and CDP disclosure rules, book-and-claim SAF certificates are accepted as a reduction in the customer’s emissions provided the certificate is retired in a recognised registry (RSB Book and Claim, ISCC Book and Claim) and not double-counted. The GLEC v4 draft — which I discussed in the GLEC v3.2 explainer — tightens this further with explicit named-registry requirements.

The acknowledged gap: there is currently no global registry standard for SAF book-and-claim, which means a freight customer can in principle find the same SAF batch claimed by multiple end-users through different registries. The industry is working on a unified registry; the IATA SAF Registry is the leading candidate. Until that exists, an auditor reading a Scope 3 disclosure that claims SAF reduction should be expected to ask which registry the certificate sits in and whether the customer can demonstrate the certificate is uniquely retired.

My opinion on where this goes

SAF mandates are real and binding. The supply will scale because the regulation forces it to. The cost curve will improve, especially for HEFA and AtJ, as production matures. Synthetic SAF will remain expensive through 2035 at minimum, because the underlying cost driver is green hydrogen pricing and that doesn’t bend down quickly.

SAF will not be the dominant decarbonisation lever for aviation freight before 2035. The dominant lever between now and then is going to be network and operational optimisation — better load factors, more efficient airframes, sensible modal shift — combined with mandated SAF percentages at the lowest level the regulation permits. Post-2035 the picture shifts: SAF blending mandates step up sharply, synthetic SAF starts to scale, and the cost curve has had a decade to bend. For freight customers building a credible decarbonisation roadmap, anchoring on SAF in the 2026–2030 horizon is paying premium prices for a relatively small fraction of the eventual abatement.

For comparison with offset-based decarbonisation pathways — particularly the question of whether voluntary carbon credits actually substitute for direct reduction — see the Verra REDD+ 2026 review.

Sources

IATA Annual Review 2025 and IATA SAF tracker for SAF supply share — iata.org. IEA Aviation Outlook 2025 and IEA Net Zero Roadmap (2023, updated 2024) for cost-per-tCO2 ranges. ReFuelEU Aviation Regulation (EU) 2023/2405 — eur-lex.europa.eu/eli/reg/2023/2405/oj. ICAO CORSIA documentation for eligible-fuel criteria — icao.int/environmental-protection/CORSIA. ASTM D7566 SAF specification. Argus Media SAF price tracking, weekly assessments through 2025. The 200 kg of jet fuel per tonne of cargo Hong Kong-Memphis figure derives from GLEC v3.2 freighter factors applied to a 777F operational profile.