Rigorous Self-Critique: The Three-Factor SAF Cost Framework

Feb 22, 2026. Internal analysis. Stress-testing our own arguments.

The Three Factors (Restated)

Incentive disadvantage: ~$0.40/gal less for SAF than RD in California
Yield penalty: Lower carbon efficiency when targeting jet-range molecules
Higher OPEX: More expensive to produce SAF than RD per gallon

Factor 1: Incentive Gap — Strength of Evidence

✅ STRONG: CAR exclusion ($0.30/gal)

This is a verifiable policy fact. Jet fuel is not included in California's cap-and-trade program at the rack. ULSD is. This alone accounts for 75% of the California incentive gap. It is real, structural, and could be changed through regulation.

✅ STRONG: RIN EV difference ($0.09/gal)

RD earns 1.7 D4 RINs/gal vs SAF at 1.6 D4 RINs/gal. This is set by EPA regulation and reflects energy content (BTU) calculations. At ~$0.85/RIN, this is a real $0.09/gal disadvantage.

⚠️ CAUTION: California-specific, not universal

The $0.40/gal gap is a California phenomenon. Outside California:

No Cap at the Rack → lose $0.30 of the gap
No LCFS → lose $0.01 of the gap
Remaining gap nationally: only the $0.09 RIN EV difference

For the paper: We should present the California case as the most extreme example, but be explicit that the incentive gap is primarily a state-level policy artifact, not a universal structural problem. Nationally, the incentive gap is small (~$0.09/gal). The bigger structural issues are Factors 2 and 3.

🔍 NEEDS VERIFICATION: 45Z credit parity

We assumed $0.17/gal for both RD and SAF. But 45Z is CI-dependent and technology-dependent. Some SAF pathways might actually earn more than RD from the same feedstock if the lifecycle accounting differs. Need to verify: does the additional H₂ consumption for hydrocracking increase SAF's CI and reduce its 45Z credit relative to RD from the same feedstock?

Potential counterargument: If SAF production requires more hydrogen (increasing CI), the 45Z credit could actually be lower for SAF than RD, widening the gap further. But if green H₂ is used, SAF CI could be comparable. This needs quantification.

Factor 2: Yield Penalty — Strength of Evidence

✅ STRONG: Hydrocracking reduces total liquid yield

This is well-established in the literature:

Biotechnology for Biofuels (2017): SAF yield 60–70% from five oil feedstocks
Lee Enterprises (2024): RD-max yields ~83% liquid product; SAF-max requires HDC with yield loss to gas, LPG, naphtha
NREL TP-5400-87802 (2024): "production of additional jet fuel from the diesel fraction may require additional capital expense, increase operating costs, require additional hydrogen and higher-severity operations, and reduce carbon yield"

⚠️ CAUTION: My yield numbers may overstate the penalty

I used 83% RD vs 70% credit-eligible SAF (50% SAF + 20% RD). But these need scrutiny:

The 83% RD yield is for the total liquid product (RD + some naphtha co-product). Pure RD yield may be closer to 75–80%.
SAF-max plants still produce significant RD co-product (20–30%), which earns full RIN/LCFS credits. So "credit-eligible" yield is 50% SAF + 20% RD = 70%, not 50%.
The real comparison should be: credit-eligible yield in RD-max (~80%) vs credit-eligible yield in SAF-max (~70%). That's a 10 pp difference, not 13 pp as I stated.

For the paper: Be precise. The yield penalty is real but more like 10 percentage points of credit-eligible product, not the dramatic 83→50% my Figure 3 implies if you only count the SAF fraction.

🔍 CRITICAL QUESTION: What are the co-products actually worth?

The narrative that co-products are "low-value" needs examination. See Section 3 below.

Factor 3: Higher OPEX — Strength of Evidence

✅ STRONG: Additional hydrogen consumption

Hydrocracking is hydrogen-intensive. NREL notes SAF requires "2–19× more hydrogen per gallon" depending on pathway and feedstock. This is a real operating cost increase.

✅ STRONG: Additional capital amortization

Hydrocracker + catalyst + fractionation expansion = $50–100M+ additional capital per plant. farmdoc daily (Jan 2025) confirms this.

🔍 NEEDS PRECISION: How much more per gallon?

My Figure 5 shows $0.80/gal cost premium ($3.30 SAF vs $2.50 RD). This needs validation:

Fueling the Future (Dec 2024): HEFA SAF from WCO = ~$2.75/gal; from virgin oil ~$5.05/gal
But what's the RD comparison from the same feedstock? If WCO RD is ~$2.20/gal, the delta is only $0.55/gal
The $0.80/gal figure may be too high. A more conservative estimate might be $0.40–0.65/gal additional production cost for SAF vs RD from identical feedstock.

For the paper: Present a range, not a point estimate. Perhaps $0.40–0.80/gal additional production cost, depending on feedstock, facility configuration, and hydrogen source.

Section 3: Co-Product Market Analysis (Naphtha, LPG, Propane)

What comes out of a SAF-max HEFA plant besides SAF?

Co-product	Typical yield (mass%)	Carbon range	Conventional value	Renewable premium?
Renewable diesel (co-product)	15–25%	C16–C22	~$2.50/gal (CA diesel parity)	Yes — full RIN + LCFS + 45Z
Renewable naphtha	8–15%	C5–C8	~$560–655/tonne ($1.65–1.95/gal)	Growing premium for green chemicals
Renewable propane/LPG	3–5%	C3–C4	~$0.80–1.20/gal (commodity LPG)	Small bio-premium, demand subdued
Fuel gas (C1–C2)	2–4%	C1–C2	~$2–3/MMBtu (natural gas parity)	Minimal — used internally for process heat
Water/losses	2–3%	—	$0	No

Renewable Naphtha: A Growing Market

This is the most important co-product to examine because it's the largest non-fuel stream from SAF-max plants.

Market destinations:

Petrochemical steam crackers — Renewable naphtha is chemically identical to fossil naphtha and can feed existing steam crackers to produce "renewable" ethylene, propylene, and downstream plastics. Major users: SABIC (TRUCIRCLE™ program with UPM bionaphtha), BASF, LyondellBasell. Mass-balance accounting via ISCC/RSB certification allows renewable claims on plastic products.
Gasoline blending — Light renewable naphtha can be blended into gasoline or processed through isomerization for octane improvement. Earns RINs as cellulosic or biomass-based diesel depending on pathway.
Renewable solvents & specialty chemicals — Niche market, smaller volumes but potentially higher margins.

🔍 KEY FINDING: Bio-naphtha premium is uncertain and market-dependent

S&P Global (Jan 2024): "We have not traded bionaphtha this year because of sluggish demand from petchems." Bio-propane premium dropped below bio-naphtha. The renewable premium is real in Europe (driven by RED II/III mandates) but weak in the U.S. (no equivalent mandate for green chemicals).

In Europe: Renewable naphtha commands a significant premium ($200–600/tonne above fossil naphtha) because EU RED III requires renewable content in transport fuels and creates incentives for renewable chemicals. Total market ~$600M (2024), growing at 11.5% CAGR.

In the U.S.: No comparable mandate. Bio-naphtha can earn RINs if used as fuel, but if sold as petrochemical feedstock, it may not earn renewable credits in all programs. Value is closer to fossil naphtha (~$560/tonne) without significant premium.

Impact on the Cost Argument

⚠️ HONEST ASSESSMENT: Co-product revenue partially offsets the yield penalty

If we account for co-product revenue, the effective cost of SAF is lower than a simple yield-loss calculation suggests:

Scenario	SAF yield	Co-product revenue per gal SAF	Net effective cost increase vs RD
Low co-product value (U.S., commodity pricing)	50%	~$0.25/gal SAF (from RD + naphtha + LPG at commodity)	~$0.50–0.65/gal
Medium co-product value (mixed markets)	50%	~$0.40/gal SAF (RD at full credit + naphtha at moderate premium)	~$0.35–0.50/gal
High co-product value (EU, green chemistry premium)	50%	~$0.55/gal SAF (RD at full credit + naphtha at EU premium)	~$0.25–0.40/gal

Bottom line: Depending on co-product valorization strategy, the net SAF cost penalty over RD ranges from roughly $0.25 to $0.65/gal — not the $0.80/gal I originally estimated. The renewable naphtha market is a significant variable. Plants that can sell bio-naphtha at green premiums (especially into EU markets) substantially reduce the effective cost penalty.

Potential Co-Product Markets Worth Investigating Further

Renewable aviation naphtha → gasoline blending + RINs: If light naphtha earns D6 cellulosic RINs when blended into gasoline, it could be worth $1.50–2.00/gal more than commodity naphtha. Need to verify RIN eligibility for HEFA naphtha.
Green chemicals / mass-balance plastics: Growing market especially in EU. BASF, SABIC, LyondellBasell all have programs. Premium depends on certification and customer willingness to pay.
Renewable propane → heating / export: Small volume but bio-propane earns RHI (Renewable Heat Incentive) credits in UK, and modest premiums in EU.
Internal hydrogen production: Fuel gas from hydrocracking could offset hydrogen purchase costs if reformed on-site, creating a partial closed loop.

Revised Summary: What Makes SAF More Expensive Than RD?

Factor	Estimated impact ($/gal)	Confidence	Notes
1. Incentive disadvantage (CA-specific)	−$0.40	🟢 High (CA); 🔴 Low (nationally ~$0.09)	75% from CAR; CA-specific; nationally small
2. Yield penalty (net of co-products)	+$0.25 to +$0.65	🟡 Medium	Depends heavily on co-product valorization
3. Additional CAPEX + OPEX	+$0.15 to +$0.35	🟡 Medium	Hydrocracker capital + H₂ + catalyst
Total SAF cost premium over RD	+$0.40 to +$1.40/gal		Range depends on location, co-product markets, and plant config

Revised thesis (more nuanced):

The SAF price premium over RD is driven by three reinforcing factors: (1) a specification-driven processing penalty that reduces yield and increases capital/operating costs, offset partially by co-product revenue; (2) a policy incentive gap that is large in California ($0.40/gal) but small nationally ($0.09/gal); and (3) the qualification barrier (D4054) that limits market entry and constrains blend ratios. The specification is the primary structural driver because it creates both the processing penalty (Factor 2 + 3) and the qualification barrier, while the policy gap (Factor 1) is an addressable regulatory artifact.

What Still Needs Work

☐ Verify whether HEFA co-product naphtha earns RINs when blended into gasoline (D6 or D4)
☐ Quantify the 45Z credit differential between RD and SAF from the same feedstock/CI
☐ Get actual plant-level data from Diamond Green Diesel, Marathon Martinez, or Phillips Rodeo on SAF vs RD yield splits (these may be proprietary)
☐ Model the economics of a plant optimized for SAF + high-value co-products vs. pure RD
☐ Research outside California: What's the SAF vs RD incentive picture in EU (RED III), UK (SAF mandate), Singapore, Japan?

Internal analysis — Feb 22, 2026. Pre-publication. Do not distribute.