Sustainable Aviation Fuel: Part 3
SAF feedstocks and production costs
At a price of only ~$2.50-$3.00 per gallon, jet fuel is cheap. Yet airlines often still have very thin margins. Despite aggressive targets for adopting SAF, one major airline that Prime Movers Lab spoke with said it wouldn’t be realistic for them to pay any significant premium on fuel. Are any routes to SAF cost-competitive with fossil jet fuel, and if not, could they ever be?
In my last post, I introduced a few routes to produce SAF from different feedstocks (oils, alcohols, and syngas). Assessing the contributions of capital, utilities, feedstock, and other fixed and variable costs to SAF price would entail a full technoeconomic analysis. However, a simple treatment based on feedstock costs alone can give great insight into which approaches to SAF might be most promising. The post will compare the costs and other challenges of using different feedstocks for SAF, beginning with biomass before turning to e-fuels, or SAF made from CO2 and electricity.
Feedstock: Vegetable or cooking oils (HEFA)
The HEFA pathway can be used to produce SAF from vegetable oils like soybean or corn oil, used directly or collected after use as cooking oil. This is the primary production pathway for SAF now and is expected to fulfill the majority of expanded capacity through the decade to meet 2030 targets of ~5% of total jet fuel consumption. Beyond that, however, the supply of these oils will be constrained (there is only so much cooking going on at your favorite fast food restaurant), and it is already becoming so.
Used cooking oil for biofuel has been as low as $550/t in 2019, but has more recently risen to the $1100–1700/t range. Due to the inherent chemistry of, for example, Honeywell-UOP’s Ecofining process, the maximum possible yield is ~0.8 tonnes of paraffin per tonne of oil. Assuming $1100/t feedstock oil, a lower bound for HEFA price would be ~$1400/t or $4.15/gal. While energy crops like Brassica carinata are also being explored as alternatives to food oils, there does not seem to be a path for HEFA to reach cost parity with fossil jet fuel in the absence of subsidies or drastic changes in feedstock price.
Feedstock: Ethanol (Alcohol-to-jet)
LanzaJet developed a process to convert ethanol into SAF by first dehydrating with a catalyst to produce ethylene gas. Ethylene is oligomerized with another catalyst to produce heavier olefins, which are hydrotreated to yield paraffins and isoparaffins. The process was developed to utilize ethanol produced from parent company LanzaTech’s microbial gas fermentation, in which carbon monoxide-rich industrial waste gases are used as an energy and food source by special microbes, but it is also applicable to ethanol from standard corn and sugarcane fermentation. Similar chemistry is employed by Vertimass and Gevo to produce SAF, respectively, from ethanol and isobutanol.
The current market price of ethanol is ~$1.70/gal, which is the lowest it's been in 3 years but closer to historical values. A maximum yield of ~0.6 gallons SAF per gallon ethanol is possible based on the chemistry, which suggests a SAF cost lower bound of $2.83/gal. This is lower than HEFA and would seem to be near cost parity with fossil jet fuel, but we neglect any capital or operating costs. Whereas HEFA production is operationally simple, the AtJ process is multistep. Whether U.S. corn ethanol is even sustainable has also frequently been a contentious question. In fact, the Biden Administration’s recent decision to recognize an alternative methodology for calculating lifecycle CO2 emissions may be the only reason AtJ-SPK from corn-ethanol qualifies as SAF and is eligible for IRA credits.
The other possibility for a step function change in bio-ethanol economics would be cellulosic ethanol — made by breaking down more recalcitrant or woody or grassy biomass into fermentable sugars. This has been a long-standing dream, but the high price of cellulase enzymes coupled with the quantities required has made it impractical. Companies like Blue Biofuels and others hope to finally crack the code with new technology, and if they succeed it could have a major impact.
Companies like Air Company, Prometheus, and Liquid Wind are also innovating Power-to-Liquid processes that use electricity and CO2 to make AtJ precursors like methanol or ethanol. The economics would be driven by the price of electricity and/or green hydrogen, as well as captured CO2 so that discussion is grouped below along with other PtL approaches.
Feedstock: Waste Biomass (Fischer-Tropsch)
Compared with conventional vegetable oil and sugar/ethanol feedstocks, waste biomass like forest residues are much cheaper, as low as $60/ton, and aren’t at risk for a shortage. One approach to using it is gasification and Fischer-Tropsch, as employed by companies like Velocys. The feedstock is gasified by heating to a high temperature, with a low and controlled amount of oxygen and moisture. This produces a gas rich in CO and hydrogen, called synthesis gas (syngas). The H2 content in this gas is typically low, but can be adjusted through the water gas shift (WGS) reaction, where CO and steam react to synthesize H2, though at the expense of losing carbon as CO2. In this way, the syngas can be adjusted to the ideal composition for synthesis of paraffins in the Fischer-Tropsch reactor (H2:CO ratio slightly above 2).
Waste biomass feedstock is cheaper but has lower yields. Only ~0.15–0.2 tons of FT-SPK per ton of feedstock is achievable, so a forest residue price of $60/ton figures to a lower bound for SAF cost at ~$1.20/gal. This seems promising, so why haven’t Fischer-Tropsch fuels from waste biomass caught on? One challenge is the high capital costs of the gasifier and Fischer-Tropsch reactor. Both steps benefit from economies of scale, but sourcing enough consistent high-quality feedstock is a challenge. Solid feedstock is not as easy to transport as liquids like cooking oil or ethanol. The $60/ton price tag isn’t necessarily reflective of those collection costs. Plus, waste biomass is hardly pristine. It must be thoroughly dried, and dirt and leaves may cause problems in the gasifier. The raw syngas may require multi-step cleanup to remove contaminants before it is ready for use.
Other technologies like Alder Renewable’s fast pyrolysis might better fit the challenges of these feedstocks. The IEA estimates enough sustainable biomass supplies to support ~380 billion gallons of liquid fuel production if we can only find a way to make the economics work.
Feedstock: Electricity & CO2 (Fischer-Tropsch, Alcohol-to-jet)
As we’ve seen, the biofuel routes to SAF come with challenges and tradeoffs between feedstock prices and operational complexity and costs. What about Power-to-Liquid approaches that promise to produce SAF from just (low-carbon) electricity and CO2? There are slightly different paths to go about it, but they actually all end up at nearly the same fundamental calculus of inputs: In the ideal case, about 9.5kg CO2 and 56–67 kWh of electricity are needed to produce 1 gallon of e-fuel.
For example, the most standard PtL method is to produce H2 using electrolysis of water and react with CO2 to form syngas by the Reverse Water Gas Shift (RWGS). At very high temperatures, the WGS reaction can flow in reverse, favoring the formation of CO from H2 and CO2. The resulting syngas can then be used for Fischer-Tropsch synthesis. The energy efficiency of H2 production is currently ~50 kWh per kg of H2 produced, though emerging tech, like that of startup Hysata and many others, might realize ~15–20% higher efficiencies. This means ~56–67 kWh of electricity is needed to produce 1.34 kg H2, which is the amount that must be combined with 9.5 kg CO2 to make 1 gallon of liquid fuel.
There are other methods, like catalysts to synthesize alcohol from H2 and CO2 or novel CO2 electrolyzers to make CO directly. We’ll have more to say on these innovations in Part 4 of this series. But by and large, the amount of electricity and CO2 needed for any PtL approach is roughly the same, if the method is operating with perfect efficiency.
How does this translate to cost? The cost of capturing CO2 depends on source and purity, ranging from $20/t for the CO2 emitted from ethanol plants (among the purest) to likely >$200/t for “direct air capture”, where CO2 is pulled directly from the atmosphere. A reasonable benchmark for discussion is $70/t, as this is around the cost of capturing emissions from cement, steel, and power plants, and also reflects the aspirations of emerging companies aiming to bring direct air capture below the $100/t mark. With that benchmark, captured CO2 contributes ~$0.67 toward one gallon of e-fuel, and the rest of the cost depends on clean energy pricing. (In reality, the captured CO2 cost is also coupled to energy prices, as an additional ~11 kWh per gallon fuel may be associated with capturing the CO2 used.)
Comparing the cost curve of PtL from CO2 to the lower bound prices for HEFA and Ethanol AtJ, clean electricity would need to be priced at ~5 cents or 3 cents per kWh, respectively, to compete with these biofuel routes. Utility prices from the grid are certainly not that cheap, but renewables during peak hours of production may be on par.
This all assumes an “ideal” case, where the process proceeds with its maximum possible yield. How feasible is that? One technoeconomic analysis from Argonne National Laboratory modeled liquid fuel production from H2 and CO2 and Fischer-Tropsch. In the model, ~50% of the CO2 and ~30% of the H2 used in the process did not end up in liquid fuel at the end. There were a number of modeled inefficiencies, but three that stood out were:
- Low extent of conversion in the RWGS reaction, necessitating CO2 recycling (to enable multiple passes through the reactor)
- Only ~80% efficiency of the CO2 recycler/separator, so some CO2 is lost
- Only ~75% conversion of CO in the Fischer-Tropsch reactor to liquid fuels (and no way to recycle CO)
How would this “inefficient” (but maybe more realistic) process impact cost? Compare the “inefficient” and “ideal” PtL cost curves in the plot above. In the non-ideal PtL process, clean electricity must now be available at only ~1.5–3 cents per kWh to compete with biofuel routes, a much tougher challenge.
Closing thoughts: Efficiency is everything
Fossil-based jet fuel is cheap, but amongst the dizzying array of SAF production methods and feedstocks, there seem to be paths to get close to cost parity. Still, our first-order analyses made some sweeping generalizations, ignoring capital costs and assuming scenarios where processes run with perfect efficiency. We were also unrealistic in assuming that liquids output from these processes are all useable as SAF. In reality, the Fischer-Tropsch reaction and similar processes will output a range of paraffins in the naphtha, kerosene, and diesel ranges, not to mention solid waxes that need to be treated by hydrocracking to make useable liquid fuel. Perhaps only half of the liquid fuel produced will be kerosene jet fuel. To be in the business of SAF will therefore entail being in the business of sustainable naphtha and diesel as well, though these might not have equivalent value in the market.
A top-down energy perspective further illustrates how critical it will be to innovate high-efficiency solutions. If nearly 70 kWh of electricity is needed to produce 1-gallon SAF, it will take ~1700 TWh of clean electricity to produce the ~24 billion gallons of jet fuel we now consume annually in the U.S. This is about equal to the total electricity from renewables and nuclear we produce now. And that’s for a perfect process.
Fortunately, the SAF innovation space is bustling with brilliant researchers and startups who are looking to address the challenges of sustainable transportation. The next post will highlight some of our takeaways from surveying the landscape and offer thoughts on how investors might focus and discern.
Prime Movers Lab invests in breakthrough scientific startups founded by Prime Movers, the inventors who transform billions of lives. We invest in seed-stage companies reinventing energy, transportation, infrastructure, manufacturing, human augmentation, and agriculture.
