The Specification Barrier: How Fuel Standards Create Sustainable Aviation Fuel's Price Premium

1. Introduction

The aviation industry consumes approximately 7 million barrels of jet fuel per day — roughly 100 billion gallons annually — making it the world's third-largest transportation fuel consumer after diesel and gasoline.¹ Unlike road transport, which is rapidly electrifying, commercial aviation has no viable near-term alternative to liquid hydrocarbon fuels for the vast majority of routes. Sustainable aviation fuel represents the primary decarbonization pathway available today, yet its market penetration remains negligible: U.S. SAF consumption reached 287 million gallons in 2024,² less than 0.3% of total jet fuel demand.

The standard explanation for SAF's limited adoption is its high cost — typically 2–3 times the price of conventional Jet A. But this framing obscures a striking paradox. Renewable diesel, produced from identical feedstocks (waste fats, vegetable oils, tallow) using nearly identical hydrotreating technology (HEFA — hydroprocessed esters and fatty acids), has achieved price parity with petroleum diesel in California, the world's largest renewable fuels market. In some periods, RD has sold below the price of fossil diesel at retail.³ California consumed over 3.3 billion gallons of renewable diesel in 2024⁴ — more than 40 times the nation's total SAF consumption.

This paper argues that the price gap between SAF and its petroleum equivalent is not primarily a function of production technology, feedstock cost, or chemistry. Rather, it is a structural consequence of how fuel specifications are written. The diesel specification (ASTM D975) treats renewable diesel as simply diesel — if it meets the property requirements, it enters the market with no process qualification, no blend limit, and no special distribution requirements. The aviation specification (ASTM D7566, administered through D4054) treats each SAF production pathway as a distinct product requiring years of qualification testing, OEM review, and ASTM balloting before it can enter the market at a blend ratio of 10–50%.

This architectural difference has cascading consequences for processing economics, capital investment, policy eligibility, and market access. We quantify each of these effects and argue that targeted specification reform offers an emissions reduction pathway that is orders of magnitude more cost-effective than non-drop-in alternatives such as hydrogen or electric propulsion.

2. The Price Evidence

The empirical evidence for the RD–SAF price disparity is unambiguous. The Alternative Fuels Data Center (AFDC) tracks California renewable diesel and petroleum diesel prices monthly. Since the state's Low Carbon Fuel Standard (LCFS) reached maturity in the late 2010s, renewable diesel has consistently traded at or near petroleum diesel prices, and has at times been available below ULSD at retail stations.⁵ OPIS spot data corroborated this pattern throughout 2023–2024.³

SAF, by contrast, commands a persistent premium. The European Union Aviation Safety Agency estimated average 2024 SAF production costs at €1,461/tonne for biofuel pathways — approximately twice the €700/tonne price of conventional Jet A.⁶ An RMI survey found SAF green premiums of $2.34–$3.93/gallon, implying delivered SAF prices of $9–$11/gallon against a $2.50/gallon jet fuel baseline.⁷

The volume asymmetry is equally stark. Figure 1 compares California RD and SAF volumes from 2019–2024. Renewable diesel consumption grew from 480 million gallons (2019) to over 3.3 billion gallons (2024). SAF grew from 1.9 million to 78 million gallons over the same period — two orders of magnitude smaller despite identical feedstock availability and largely overlapping production technology.⁸

[Figure 1: California RD vs SAF Volumes — see figures page]

3. Specification Architecture: The Root Cause

The fundamental asymmetry lies in how the two specifications define market access.

3.1 Diesel: Property-Based, Open Access

ASTM D975 defines seven grades of diesel fuel oil by their physical and chemical properties: viscosity (1.9–4.1 mm²/s at 40°C), sulfur content, flash point, cetane number, and distillation range. Cloud point — the cold-flow property most analogous to jet fuel's freeze point — is not specified to a fixed value but is subject to agreement between buyer and seller.⁹ Crucially, D975 makes no reference to how the fuel was produced. Any hydrocarbon mixture meeting the property requirements is diesel, regardless of feedstock or process. There is no qualification pathway, no blend limit, and no special chain of custody.

3.2 Aviation: Process-Based, Gated Access

ASTM D7566 defines aviation turbine fuel containing synthesized hydrocarbons through a fundamentally different architecture. Rather than a single property table, D7566 maintains a series of annexes — currently seven — each describing a specific production process (e.g., Annex A1 for Fischer-Tropsch, Annex A2 for HEFA). Each annex specifies not only the final fuel properties but also feedstock constraints, process parameters, and maximum blend ratios (10–50%) with conventional jet fuel.

Before a new production pathway can receive an annex in D7566, it must complete the D4054 qualification process: a multi-tier evaluation encompassing bulk property testing (Tier 1), fit-for-purpose testing (Tier 2), component/rig testing (Tier 3), and engine/APU testing (Tier 4), followed by OEM review and ASTM committee ballot. The U.S. Department of Energy has estimated this process requires "several million dollars" and 5–7 years.¹⁰ Industry participants report costs of $5–15 million.¹¹

A fast-track provision, introduced in 2018, allows fuels limited to 10% blend ratio to undergo an abbreviated Tier 1 evaluation. While this reduces cost and timeline, it constrains the fuel to a fraction of the addressable market.

4. The Processing Penalty

The cold-flow specifications in D7566 — particularly the freeze point requirement of ≤ −40°C and viscosity limits of ≤ 8 mm²/s at −20°C and ≤ 12 mm²/s at −40°C — have direct consequences for HEFA production economics. Because lipid hydrotreating naturally produces hydrocarbons in the C15–C18 range (diesel-length molecules), shifting product into the jet fuel boiling range (C8–C16, 150–300°C) requires additional processing not needed for renewable diesel.¹²

Critically, producers have two distinct pathways to achieve this shift — hydrocracking and hydroisomerization — with fundamentally different yield economics.

4.1 Hydrocracking: The Legacy Approach

The conventional approach breaks C17–C18 carbon chains into shorter jet-range molecules (e.g., C18 → C9 + C9). This cracking is thermodynamically non-selective: it inevitably produces naphtha (C5–C8), LPG (C3–C4), and fuel gas (C1–C2) alongside jet-range products. Three peer-reviewed process models bracket the yield penalty:

Pearlson et al. (2013) modeled soybean-oil HEFA using selective cracking catalysts, finding 49% SAF + 23% RD + 7% naphtha at maximum jet mode, versus 13% SAF + 68% RD + 2% naphtha at maximum diesel mode — a net jet+diesel selectivity loss of 8 percentage points.^12a Zech et al. (2018) modeled jatropha oil with 90% cracking severity and found far worse results: 46% SAF + 8% RD + 28% naphtha, a 25 percentage point selectivity loss.^12b Robota et al. (2013), using bench-scale hydrocracking of algal n-alkanes at three severity levels, demonstrated that the naphtha penalty is non-linear: roughly constant at 41–44% of jet-range product up to 60% cracking conversion, then accelerating to 75% at 93% conversion as secondary cracking of already-cracked products kicks in.^12c

A central estimate: for each +10% SAF yield via cracking, RD decreases by approximately 14%, naphtha increases by 3.6%, and net jet+diesel selectivity falls by ~4%. But above ~40% SAF yield, the penalty accelerates dramatically.

4.2 Hydroisomerization: The Modern Approach

An alternative pathway — increasingly adopted by modern producers — avoids cracking entirely. Instead of breaking C-C bonds, hydroisomerization branches linear n-C17/C18 paraffins (boiling point 302–317°C, diesel range) into iso-paraffins whose lower boiling points (250–290°C) fall within the jet fuel specification. No carbon is lost; no naphtha or gas is produced.

Neste's year-long NEXBTL campaign using its ReNewFine catalyst system achieved 74 wt% SAF from total liquid product with cloud points averaging −46°C and total liquid yield exceeding 90%.^12d A recent UOP patent describes SAPO-11 molecular sieve catalysts replacing conventional amorphous silica-alumina to favor isomerization over cracking, claiming jet fuel yields >70 wt% (preferably >80%) with ≥14 wt% C18+ hydrocarbons retained in the product — a result that is physically impossible via cracking.^12e,12f

The implications for cost are profound. Isomerization-derived SAF retains the C18 carbon backbone, yielding higher energy density per gallon than cracking-derived SAF (which is predominantly C9–C12). More importantly, the near-elimination of naphtha and gas co-products means that almost all feedstock carbon is converted to credit-eligible fuel, rather than being lost to low-value byproducts.

4.3 The Persistent Cost Penalties

Even with isomerization, SAF production from HEFA imposes costs beyond those for renewable diesel:

Capital cost. Whether via a hydrocracking reactor or a high-severity isomerization unit, SAF production requires capital investment beyond a basic RD hydrotreater. For cracking-based plants, this adds an estimated $50–100 million for a large HEFA facility.¹³ Isomerization requires catalyst investment (SAPO-11 systems are more expensive than conventional silica-alumina) but avoids the extensive gas recovery and fractionation infrastructure needed for cracking.

Yield loss. For cracking pathways, the yield penalty is severe: converting RD-range molecules to jet produces 7–28% naphtha and gas that earn no environmental credits. In a RD-maximized plant, ~83% of output is credit-eligible diesel; in a cracking-derived SAF-maximized plant, only 54–73% is credit-eligible fuel (jet + diesel), with the remainder going to low-value naphtha and gas.^12a,12b For isomerization pathways, this penalty shrinks dramatically: Neste's campaign achieved >90% total liquid yield with minimal naphtha, implying >80% credit-eligible fuel at maximum SAF mode.^12d The difference — a 17+ percentage point gap in jet+diesel selectivity — represents the primary economic advantage of isomerization over cracking.

Hydrogen consumption. Both pathways consume more hydrogen than RD production alone. Cracking requires ~48% more H₂ than diesel-mode operation (4.0 vs 2.7 wt% of feed).^12a Isomerization requires additional H₂ for the branching reactions, though with no carbon loss to light gases.

Diesel-mode 30% cheaper. Zech et al. found that HEFA diesel-mode production costs are approximately 30% lower than jet-mode due to the higher fraction of valuable long-chain products and reduced hydrogen demand — a penalty that applies regardless of pathway but is substantially mitigated by isomerization.^12b

[Figure 3: HEFA Product Yield — RD-Max vs SAF-Max; Cracking vs Isomerization — see figures page and HEFA yield tools]

5. Policy Amplifiers

The specification-driven cost penalty is amplified by policy structures that, whether by design or by omission, provide less incentive support for SAF than for RD.

In California — by far the most relevant market for both fuels — the total incentive stack for renewable diesel at a carbon intensity of 40 gCO₂e/MJ was approximately $2.34/gallon in February 2025, comprising D4 RINs ($1.45, reflecting a 1.7 equivalence value), LCFS credits ($0.42), Cap at the Rack ($0.30), and the 45Z Clean Fuel Production Credit ($0.17). The equivalent SAF stack was $1.94/gallon — a $0.40/gallon deficit.¹⁵

Two factors account for most of this gap:

Cap at the Rack (CAR). California's cap-and-trade program applies to diesel fuel sold at the loading rack, adding approximately $0.30/gallon to the effective price of petroleum diesel. This inflates the price of the fossil fuel against which RD competes, improving RD's relative economics. Jet fuel is not included in CAR, so SAF receives no comparable benefit. This single policy omission accounts for 75% of the incentive gap.

RIN equivalence value. Under the Renewable Fuel Standard, renewable diesel earns 1.7 D4 RINs per gallon, while SAF earns 1.6 — a structural 6% discount that reflects energy-content calculations rather than policy intent.¹⁶

[Figure 2: Incentive Value Stack — RD vs SAF — see figures page]

6. A Framework for Specification Reform: Drop-In vs. Non-Drop-In

Rather than treating specification reform as a monolithic proposal, we organize potential interventions along a spectrum from fully drop-in (no hardware changes) to non-drop-in (requiring aircraft, engine, or infrastructure modification). Each tier has different costs, timelines, and addressable volumes.

6A. Drop-In Trace Material Reform

The lowest-cost, fastest-impact reforms involve trace material specifications where bulk fuel properties remain unchanged. The existing fleet requires no modification.

Phosphorus (≤ 2 ppm). D7566 Table 1 specifies a maximum phosphorus content of 2 ppm for the synthesized hydrocarbon blending component. This limit has significant implications for HEFA producers processing vegetable oils rich in phospholipids, requiring investment in degumming and pretreatment. However, the published literature supporting 2 ppm as a technically justified threshold is limited. The damage mechanism — presumed to be catalyst poisoning in aftertreatment systems — is better established for automotive catalysts than for aviation gas turbines, which lack catalytic converters.

The case of CleanJoule's CycloSAF illustrates the practical impact: novel pathways producing cycloparaffinic SAF from biomass face trace phosphorus contamination challenges that may require additional purification steps not needed for conventional HEFA. Whether 2 ppm is the correct threshold — versus 5 or 10 ppm — requires a systematic fit-for-purpose (FFP) evaluation. If the limit could be relaxed, pretreatment costs for lipid-based feedstocks would decrease, and new pathways facing trace contamination would have a lower qualification burden.

Other trace limits. Similar FFP investigations are warranted for other trace-element specifications inherited from petroleum refining practice rather than derived from aviation-specific testing.

6B. Drop-In Feedstock Engineering

A complementary approach addresses the root cause of the hydrocracking penalty: the mismatch between feedstock carbon number and jet fuel carbon number. Conventional lipid feedstocks (soybean, canola, tallow) produce C16–C18 fatty acids, which must be cracked down to C8–C16 for jet fuel range. What if the feedstock itself were engineered to produce jet-range carbon numbers?

Crop genetic studies have demonstrated the feasibility of engineering medium-chain fatty acid (MCFA) accumulation in oilseeds. Research at the USDA and several universities has shown that expressing FatB acyl-ACP thioesterases from Cuphea and California bay laurel in camelina can produce seed oils enriched in C8–C14 fatty acids — directly in the jet fuel carbon number range.²⁹ If these feedstocks could be commercialized at scale, the hydrocracking step could be reduced or eliminated entirely, closing much of the SAF cost gap with RD while remaining fully drop-in to the existing fuel specification.

This represents a long-term research investment (5–15 years to field-scale deployment) but addresses the problem at its thermodynamic root rather than through specification workarounds.

6C. Non-Drop-In: Wider Specifications with Engine/System Modifications

Beyond drop-in reforms, a second tier of interventions would widen specifications beyond current fleet compatibility, requiring engine or aircraft modifications.

6C.1 — ASTM non-drop-in specification development. ASTM D1655 itself acknowledges that its specifications "do not include all fuels satisfactory for aviation turbine engines" and that "certain equipment or conditions of use may permit a wider... range of characteristics."³⁰ Working groups within ASTM D02.J0 are actively exploring wider specification envelopes that could accommodate higher-viscosity, higher-freeze-point fuels — fuels that would be cheaper to produce because they require less hydrocracking. These non-drop-in specifications would apply to new or modified engine designs certified for the wider fuel envelope.

6C.2 — Engine design for wider fuel properties. How much cheaper could SAF be if engines were designed to tolerate higher viscosity (say, ≤ 12 mm²/s at −20°C instead of ≤ 8) and higher freeze points (≤ −30°C instead of ≤ −40°C)? The answer is substantial: relaxing cold-flow requirements permits less severe isomerization or cracking, potentially increasing SAF yield from ~50% to 65–80%. The isomerization pathway already demonstrates this: by branching C17–C18 chains rather than breaking them, Neste achieves 74% SAF yield at >90% total liquid yield under current specifications. Wider specifications would push this further, potentially enabling near-complete conversion of HEFA feed to SAF with minimal byproduct losses. Engine fuel nozzle and fuel system modifications to handle higher-viscosity fuels are engineering challenges, not physics barriers — they are orders of magnitude cheaper than developing entirely new propulsion systems.

A new engine program costs $2–5 billion and takes 8–10 years (Section 7). But an engine optimized for wider fuel specifications could enter the fleet through normal replacement cycles, gradually expanding the market for lower-cost SAF without requiring the entire fleet to change at once.

6C.3 — Cryogenic fuels: Hydrogen and LH₄. At the far end of the non-drop-in spectrum lie cryogenic fuels — liquid hydrogen (LH₂) and potentially liquid methane (LCH₄/LH₄) — requiring entirely new aircraft, engines, fuel systems, and airport infrastructure.

Airbus's ZEROe program — the most advanced hydrogen aircraft initiative — was officially delayed in February 2025 from a 2035 to a 2040–2045 entry-into-service target, with the company acknowledging that the hydrogen ecosystem is "5–10 years behind 2020 assumptions."¹⁷ The program has spent $1.7 billion to date¹⁸ and targets a 100-seat turboprop for regional routes — a segment accounting for less than 5% of global aviation emissions. Independent analyses estimate the total cost of deploying hydrogen aircraft for intra-European aviation at €299 billion through 2050,¹⁹ while the World Economic Forum and McKinsey estimate global alternative-propulsion infrastructure investment at $700 billion to $1.7 trillion.²⁰

Battery-electric propulsion faces an even more fundamental constraint: current lithium-ion batteries at 250–300 Wh/kg offer 40× less energy density than jet fuel (~12,000 Wh/kg). Even at a theoretical 1,000 Wh/kg, electric aircraft cannot serve routes exceeding 500 nm with payloads exceeding 50 passengers — less than 2% of global aviation CO₂.

7. Cost Comparison Across the Spectrum

[Figure 6: Total System Investment by Pathway (log scale) — see aircraft cost analysis]

The hierarchy is stark. Drop-in specification reforms costing $2–50 million can be implemented in 1–5 years and apply to the entire existing fleet of ~25,000 aircraft. At the other end, hydrogen propulsion requires $700 billion–$1.7 trillion in investment over 20–30 years and addresses less than 15% of aviation emissions. The ratio is 10,000–30,000× in cost.

8. Path Forward

We propose that aviation decarbonization strategy should pursue all tiers simultaneously but recognize their fundamentally different timescales:

Near-term (1–5 years): Drop-in spec reform. Transition D7566 toward property-based qualification. Review and relax technically unjustified trace-material limits. Raise blend limits toward 100%. Streamline D4054. These actions require minimal investment and apply fleet-wide immediately.

Medium-term (5–15 years): Feedstock engineering + engine evolution. Invest in crop genetics to produce jet-range carbon numbers directly from oilseeds, eliminating the hydrocracking penalty at its thermodynamic root. Simultaneously, develop next-generation engines certified for wider fuel envelopes, expanding the market for lower-cost SAF as these engines enter the fleet through normal replacement cycles.

Long-term (15–30+ years): Non-drop-in propulsion. Continue hydrogen and electric research for the post-2050 fleet, recognizing that these technologies cannot contribute meaningfully to near-term emissions reduction and should not divert investment from specification reforms that can.

In parallel, policy reforms should address the incentive asymmetry: including jet fuel in California's Cap at the Rack, equalizing RIN equivalence values, and strengthening production credits. But the specification itself is the foundation on which all policy incentives rest — and it is currently structured in a way that makes SAF unnecessarily expensive to produce, qualify, and distribute.

References

Draft — Revised Mar 10, 2026. ~4,800 words. Section 4 substantially revised: cracking vs. isomerization pathways, 6 new references. Pre-publication. Do not distribute.