The Last Leaded Fuel, Part 2: Why America Still Burns Lead

TL;DR: 100LL is the last leaded transportation fuel on earth, and the United States burns most of it. That is not because the chemistry is hard. An unleaded avgas that most of the fleet could burn earned a Lycoming listing in 1995 and was never sold here. A 100-octane unleaded fuel was approved by the FAA for the entire piston fleet in 2022 and is sold, four years later, at four airports. Every institution that could have changed this optimized for its next financial quarter, its next liability case, its next election or its next membership renewal, and nobody was paid to think in decades. I burn leaded fuel in an engine designed for pump petrol, add a chemical to scrub the lead back out, and change the oil twice as often. That is not a limitation. It is a choice, and this post is about who made it.

This is Part 2 of The Last Leaded Fuel. Part 1 showed how the rest of the world fuels a Rotax. This one asks why the United States still cannot.

The Argument

Every few months a Rotax owner somewhere in the United States has the same conversation with a mechanic. The oil is grey, the plugs are fouled, and the gearbox is due a look a little earlier than the book says. The mechanic shrugs, the owner adds a bottle of Decalin to the next fill, and everyone agrees that this is just how it is. It is not how it is anywhere else. In Sweden, pilots have bought unleaded avgas since 1981. In most of Europe, the question of what to put in a Rotax was settled a generation ago. Here, in 2026, I fly a 160-horsepower engine built around European pump petrol on a fuel formulated for the Second World War, and I do it because the people who could have changed that found the old fuel too convenient to replace.

The United States did not fail to find an unleaded aviation fuel. It found one twice and declined both times, and the pattern of the refusal is the same pattern that has left this country trailing the world on electric cars, on rail, on broadband, on anything that requires one generation to spend money so the next one benefits. Every actor in this story made the decision that was rational for its own next reporting period. Nobody held the thirty-year view, because in the American system nobody is paid to.

  • The refiners and the fuel chain treat avgas as a legacy line with a protected margin: nine refineries making it in limited runs, a product one analyst estimated at $200 to 300 million a year in profit, and no quarter in which a second storage tank pays back. The “one fuel” doctrine, the rule that no unleaded avgas may be sold until it serves every engine on the ramp, was born here. It is a margin decision wearing a safety badge, and in 2012 the industry got the FAA to write it into federal policy.
  • The engine makers have built a position shaped entirely by product liability. Lycoming will not test a fuel it has not approved and will not approve a fuel it has not tested, and says in writing that it “generally does not support the STC process.” Rational for the next lawsuit. Fatal for the next decade.
  • The pilot lobby did what lobbies do: it protected its members’ fuel supply this year at the cost of their fuel supply in 2032. AOPA fought the Washington State bill in 2023 and took credit for killing it, the industry lobby had California’s ban rewritten from 2026 to 2031, and AOPA championed the 2024 federal provision that forbids any airport from dropping 100LL until a replacement exists that is approved for every aircraft including helicopters. No such fuel exists, and the provision is the reason nobody is in a hurry to make one.
  • The regulator was petitioned in 2006, promised a proposal for 2017, delivered a finding in 2023 that by its own text applies no new requirements to anyone but itself, and has proposed no standard since. Seventeen years of each administration leaving it for the next.
  • The industry’s own transition plan, published by the FAA this July, still calls 2030 a goal, defers rulemaking to a later phase, and in AOPA’s own summary “assumes the market is going to decide.” The market has had thirty-five years. It decided.

Contrast the places that did it. Sweden used a tax differential, one small company and a forty-year horizon. The European Union set a hard legal exit for tetraethyl lead in 2032 and let the industry plan backwards from it. Colorado, alone among American states, pays airports to close the price gap between leaded and unleaded fuel, which is the only thing in this story that looks like industrial policy, and it is happening at the state level because it cannot happen at the federal one.

What follows is the evidence for each charge, one page apiece. Then the same chart with different labels, from the car industry. Then where that leaves me.

The fuel was never the problem. Unleaded avgas milestones above the line; the American response below it.

Claim 1: We Had the Fuel and Chose Not to Sell It

In 1981 a small Swedish company, Hjelmco Oil, put an unleaded avgas into national distribution; in 1991 it followed with 91/96UL, nationwide by 1993, and it has sold it at some seventy Swedish airports ever since, taxed below 100LL. This is not a European curiosity. In 1995 Lycoming listed the fuel in Service Instruction 1070, its own register of approved fuels, a date Lycoming’s Michael Kraft confirmed in 2012; the 2010 revision lists “AVGAS 91/96UL” by name for the O-235, O-320, IO-360 and O-540 families, the engines in most of the training fleet. In 2010 EASA cleared it for any engine whose maker approved it, no STC required. Cessna recognized it in 2012. Hjelmco ran seminars on the US West Coast that summer and said it had found an American refinery willing to make it. Nothing happened. Lycoming’s own bulletin noted the reason in four words: “no known US distributors.”

Why not? The answer is in the FAA’s own files. In 2012 the agency’s Unleaded Avgas Transition rulemaking committee, whose members were AOPA, GAMA, Lycoming, the airport and FBO trade groups and the refiners, considered a two-grade transition and rejected it. Its reasoning: high-performance aircraft are about 30 percent of the fleet but burn 70 to 80 percent of the fuel, so “the volumes of consumption and cost for dual infrastructure would prohibit any widespread availability of two grades,” and any replacement “needed to be of the highest practicable octane level.” Lycoming’s Kraft had said the same in AOPA Pilot in 2010: the market “will only justify a single aviation-grade gasoline.” Aviation Consumer asked the obvious question in 2011, why two fuels work in Europe and not here, and got the honest answer from the FBOs: the money ain’t there.

Read that again. The fuel that two-thirds of the fleet could burn was excluded from the American market, by the American industry, because the one-third that could not burn it buys most of the fuel. That is a margin argument, and it was adopted as federal policy. The committee recommended instead an eleven-year, $57 million government program to find a single fuel for everyone. Sweden had done it with one company, no subsidy and a tax break.

The Swedish fuel did reach America in the end, under another name. Swift Fuels’ UL94 meets the same ASTM D7547 standard, went on sale in July 2015 at three airports in Indiana and Michigan, and is approved for every Rotax in Rotax’s current operating-fluids instruction. Eleven years later it is at a little over a hundred airports out of some 3,200 that sell avgas, none of them in Washington State. Where it has been given a tank, pilots buy it: at Centennial in Colorado it was 21 percent of avgas sales within a year, and the tank conversion cost about $1,100. The doctrine said the demand was not there. The demand was there. The tank was not.

Can’t We Just Add Something?

A reasonable question, and I get asked it because I already do: every tank of 100LL in Roci gets a dose of Decalin TCP, a scavenger that turns lead oxide into a softer phosphate so it leaves the engine instead of coating it. That is an admission that the fuel is wrong for the engine, patched after the fact. The reverse idea, an additive that lets a big-bore Continental run on unleaded fuel, splits into two different questions.

Octane. Lead was never a lubricant in the aviation sense; it was the cheapest octane booster the 1920s had. Take it out and the base fuel is roughly 91 to 94 octane, so something else has to go in. G100UL and 100R are exactly that: a UL94-class base plus a proprietary additive package, blended at the refinery to hit 100 motor octane. The “additive for legacy engines” is not an alternative to the unleaded fuels. It is their design. The only argument is where the additive goes in: at the refinery under a specification, or in the owner’s hand at the pump. Owner dosing fails three ways. A fuel is certified as a whole blend, so pilot-added boosters produce an unapproved fuel. A dosing error in a detonation-limited engine is an engine-destroying failure, not a maintenance one. And the boosters that work are either metals, such as the manganese compound that left deposits in Phillips 66’s test engine and that Rotax prohibits outright, or aromatics in quantities nobody wants to handle on a ramp.

Valve seats. The idea that old engines “need” lead to protect their valve seats comes from the 1970s car world and the lead-replacement additives sold to classic-car owners. Certified aircraft engines have run hardened seats for decades; Rotax’s own instruction says the opposite, that 100LL “places greater stress on the valve seats due to its high lead content.” Rotax bans lead-replacement additives by name. This is the hangar myth the piece most wants to retire.

Claim 2: When a Fuel Got Approved Anyway, the System Closed Ranks

The program the 2012 committee recommended was called PAFI, the Piston Aviation Fuels Initiative. It ran from 2013, picked Shell and Swift as finalists in 2016, suspended testing in 2018, lost Swift that autumn, saw Shell’s reformulated fuel fail in 2019 and Shell walk away in 2021 saying it needed a government mandate before it would invest. A decade of consensus, and no fuel.

Meanwhile a small engineering firm in Ada, Oklahoma, General Aviation Modifications Inc., took the other road. Instead of waiting for an industry standard, it applied for a Supplemental Type Certificate, the same mechanism used to approve a new propeller or an autopilot, and in 2022 the FAA granted G100UL a fleet-wide approval covering every certified spark-ignition engine and every airframe except helicopters. For the first time in history, an unleaded fuel existed that a Bonanza, a Cirrus and a Cessna 172 could all legally burn. Its inventor, George Braly, told a House committee that the approval had sat on an FAA desk awaiting a signature for the better part of a year.

Then the industry responded. Four months after GAMI’s first approval, the alliance of pilot, manufacturer, airport and refiner associations formed EAGLE, a new consensus body with a 2030 goal, which is to say they re-established the process that had just lost to a two-man shop. Lycoming declined to test the fuel, says so on its website, calls the use of any STC fuel an “unapproved modification,” and states that it “generally does not support the STC process.” Cirrus issued a service advisory against any unleaded fuel in its aircraft. Textron said it had not approved the fuel and wanted more testing. Piper told owners it “has not evaluated any STC fuels” and that using one voids the warranty. In May 2025 a California court, with Piper, Cirrus, Textron and Diamond filing in support, ruled that G100UL was not “commercially available” because it lacks an industry consensus specification. The manufacturers refused to test it, then cited the absence of their tests as the reason it could not be sold.

Let me be fair to the objections, because some are real. Cirrus found a wing-tank sealant debonding on one bay of one test aircraft, and lab coupons showed edge effects; GAMI’s own rebuttal concedes the debonding happened and disputes the cause. A mechanic reported nitrile O-rings swelling beyond limits. Pilots at Reid-Hillview have shown paint damage. The FAA’s July 2026 transition plan lists all of these and says root causes are “difficult to pin down”; its own materials tests were due this summer and nothing has been published. There has been no airworthiness directive, no NTSB event and no lawsuit. In Sweden, 91/96UL had the same kind of questions in the 1990s, and the answer was to run the fleet on it and find out. Here, four years on, the question is open because the people best placed to close it have declined to.

STC Versus ASTM, in Plain English

Two different questions get tangled in this fight. Is the fuel safe in the airplane? That is an airworthiness question, and the FAA answers it, either through a Supplemental Type Certificate for a named list of engines and airframes, which is how G100UL and 100R were approved, or through a fleet-wide authorization, which is the route the PAFI program and UL100E are on. Is the fuel the same from every refinery, every time? That is a production and purchasing question, and the industry answers it through ASTM International, a standards body where refiners, engine makers and distributors vote on a specification such as D910 for 100LL or D7547 for UL91 and UL94. ASTM does not test fuel in engines. It defines what the fuel must be so that a distributor can buy it from anyone and an engine maker can write it into a manual.

The industry’s position is that it will not sell or endorse a fuel without an ASTM spec; GAMI’s position, until it applied to ASTM in 2025, was that the FAA had already found its specification equivalent and that the refiners who sit on ASTM committees are the same refiners whose product G100UL would replace. Both are right about something. A consensus spec is how you get a fuel to a thousand airports. A consensus body where the incumbents hold the votes is also how you keep it to four.

The scoreboard, as of this summer, from the FAA’s own plan: of the roughly 3,200 American airports that sell avgas, 212 offer any unleaded grade, and the FAA’s deputy administrator has conceded that 2030 may slip. Fuel by fuel, it looks like this.

UL94 Swift Fuels, 94 MON

Approval: STC, 2015

Covers: lower-compression engines; roughly two-thirds of airframes, about a third of fuel burn

ASTM spec: D7547

Where sold: about 100 US airports; being withdrawn in favour of 100R

Rotax 9-series: approved (SI-912-016)

G100UL GAMI, 100 MON

Approval: STC, fleet-wide, 2022

Covers: all certified spark-ignition engines; all airframes except helicopters

ASTM spec: none; applied 2025

Where sold: four airports (two in California, Mississippi, Texas)

Rotax 9-series: not listed

100R Swift Fuels, 100 MON

Approval: STC, expanding 2024 to 2028

Covers: about 1,200 engines, 56% of fuel burn; big-bore Continentals and turbos pending

ASTM spec: D8603, 2025

Where sold: former UL94 sites converting from summer 2026; 29 states

Rotax 9-series: not listed; testing in Swift’s plan

UL100E LyondellBasell / VP Racing, 100 MON

Approval: PAFI fleet authorization, targeted 2027

Covers: target is the whole fleet, with timing changes on some turbos

ASTM spec: test spec D8631, 2025; production spec pending

Where sold: not yet

Rotax 9-series: not listed

The four unleaded candidates as of November 2026. Sources: FAA Transition Plan v1.0, AOPA on the 100R approved model list, AOPA on UL100E and the plan, Rotax SI-912-016.

And here is the detail that matters to anyone reading this behind a Rotax. Neither G100UL nor 100R appears in Rotax’s operating-fluids instruction. The only unleaded avgas my engine is approved for is UL94, the Swedish fuel by another name, and Swift is retiring it. The company told AirVenture in July that it plans to cease UL94 production as 100R ramps up, its CEO said the same to Savvy Aviation, its own transition plan expects most UL94 sites to have switched to 100R by the end of summer 2026, and since November 1 its online STC orders ship with 100R placards, UL94 by phone only. Swift says Rotax testing is part of its plan. Until Rotax says so in a bulletin, the one clean fuel I could legally buy is being withdrawn in favour of one I cannot. That is what “letting the market decide” looks like from the small end of the ramp.

Claim 3: The Regulator Protected the Incumbent

The Clean Air Act gives the Environmental Protection Agency the power to regulate aircraft emissions, and in October 2006 Friends of the Earth asked it to use that power on lead. The agency took comment in 2007, issued an advance notice in 2010, and in 2012 denied the petition while promising a proposed finding “in 2017” and a final one “in 2018.” Neither arrived. New petitions came in 2021. The final endangerment finding was published on October 20, 2023, seventeen years after the request, and its own text says what it does: “These final findings do not themselves apply new requirements to entities other than the EPA and the FAA.” It sets no standard and no date. Three years later, no standard has been proposed.

This is not a small source. By the EPA’s own count, piston aircraft put about 470 tons of lead into the air each year, 70 percent of all US lead air emissions, and in 1,140 counties they are the only inventoried source. Lead left American road fuel in 1996. It left the world’s road fuel in 2021. It is still coming out of my exhaust at Arlington.

Congress, meanwhile, moved in the other direction. Section 770 of the 2024 FAA Reauthorization Act forbids any airport that sold 100LL in 2022 from restricting it until December 31, 2030, or until an unleaded replacement exists that meets an industry consensus standard and is authorized for every aircraft, with a penalty of up to $5,000 a day. The FAA’s own guidance, issued in 2025, confirms that no fuel meets the test and that none can until one is approved for helicopters, which none is. So federal law protects leaded fuel until a replacement passes a test the law itself makes almost impossible, and the people who lobbied for that provision are the same associations now running the consensus process that defines the test.

The one jurisdiction that acted was punished for it. Santa Clara County stopped selling 100LL at its two airports in 2022 and switched to UL94. In March 2025 the FAA ruled the county in violation of its federal grant obligations, found that selling only unleaded fuel created a “prohibited exclusive right,” and ordered it to bring leaded fuel back. The complaint was filed by AOPA. The states have been pushed back the same way: California’s 2024 ban was rewritten from a 2026 start to January 1, 2031 to fit the federal shield, and Washington’s 2023 bill, which would have banned leaded avgas by 2030, died in committee, as AOPA’s headline put it, with AOPA’s help. Only Colorado, with its price-gap grants, is paying to move rather than paying to wait.

The FAA’s own transition plan, published in July 2026, four years after EAGLE was formed and a year after Congress asked for it, admits that a universal drop-in replacement is “not feasible,” sets no date for rulemaking, and leaves the rest to a market that has had since 1995.

And beneath the whole structure is one factory. Innospec, at Ellesmere Port in England, states in its 2025 annual report that it “remains the world’s only producer of tetra ethyl lead for use in aviation gasoline.” The relevant risk factor in that report is titled “Decline in our AvGas business.” The company has set aside $65 million to close the plant. The largest piston fleet on earth, some 175,000 active aircraft, depends on a single supplier that has told the world it wants out, and the American response has been to legislate that nothing change until 2030. Nobody would design a fuel supply this way. It was not designed. It was deferred, one quarter at a time, for thirty years.

The Same Chart, on Four Wheels

If the pattern above feels familiar, it is because the United States is running it again, in a bigger industry, in real time. Line the two stories up and the actors fall into the same five slots.

  • The incumbents define “ready” as “every use case.” Avgas: no unleaded fuel may ship until it serves the big-bore Continentals. Cars: no electric target is acceptable until an EV can tow a boat across Wyoming in January at the price of a Corolla. In both cases the definition is set by the people who sell the incumbent product, and in both cases it means the two-thirds of the market that is ready today waits for the third that is not.
  • Consensus replaces decision. PAFI and EAGLE were industry-run bodies with goals instead of deadlines; the automakers’ trade association told the EPA in September 2025 that the 2027 to 2032 tailpipe standards were “simply not achievable” and asked for them to be revised, and got it. In neither case did anyone say no. They said “not yet.”
  • The lobby protects this year’s member at the expense of the next decade’s. AOPA wrote 100LL into law until 2030. Nearly four thousand car dealers signed a letter asking the president to “tap the brakes” on EV targets. Both are doing their job. Neither job includes the year 2035.
  • The regulator defers. EPA took seventeen years to find that lead in the air is harmful. The same agency’s vehicle emissions standards were tightened in 2024 and put up for rescission in 2025. Policy that reverses with each election is not policy; it is weather.
  • Congress removes the one incentive that was working. The federal EV purchase credit ended on September 30, 2025. US electric share had been 8.1 percent in 2024 and 7.8 percent in 2025; in the quarter the credit died it fell to 5.8 percent, down 46 percent on the quarter before, and stood at 6.5 percent this September. Section 770 is the same instrument pointed the other way: instead of ending the incentive for the new fuel, it legislated protection for the old one. Same effect.

Now the rest of the world. The European Union, with a hard 2035 date on the books, is at 22.7 percent battery-electric so far this year. China, after a fifteen-year industrial plan, hit 67 percent in September. Norway, which did it the Swedish way, with a tax differential and a long horizon, is at 98 percent. Sweden put unleaded avgas on the ramp in 1981 with the same tool: tax the old fuel more than the new one and wait. The country that invented the mass-produced car and the general aviation industry is now a laggard in both, for the same reason. Not because it cannot build the thing, but because nobody in the system is rewarded for a decision whose payoff arrives after the next election, the next earnings call or the next membership drive.

I am not arguing that electric cars are the right answer for everyone, any more than UL94 is the right fuel for a turbocharged Baron. I am arguing that a country which cannot manage a thirty-year transition for 175,000 small aircraft burning well under one percent of its gasoline has no chance of managing one for 280 million cars, and that the small case is the diagnosis for the large one.

Where That Leaves Me

Where can a Rotax engine drink? Ethanol in the pump petrol against what the airport sells, four quadrants, United States on the E10 line with a dashed arrow toward unleaded avgas and higher ethanol
From Part 1: what can a Rotax engine drink? Top right is where you want to be. The United States is the gold dot on the line, and the dashed arrow is where it is headed.

Here is the chart from Part 1 again. The United States sits on the E10 line with a gold dot, the one country where leaded avgas is cheap and on every ramp, and its only dashed arrow points to the top-left box: unleaded fuel at the airport, nothing approved at the pump, sometime after 2030. That arrow is the sum of everything above. It points there because the fuel question was left to a market that every incumbent had a reason to slow down, and because the one deadline that might have moved it was written into law as a shield instead.

The chemistry was solved in Sweden in 1981 and in Oklahoma in 2022. What was never solved is the institutional problem. Europe, with a hard date and a tax differential, got there a generation ago. The country with the biggest piston fleet on earth, a 160-horsepower engine designed for pump petrol in one of its hangars, and a bottle of lead scavenger on the shelf next to it, did not.

So here is where I stand, and it is not complicated. I am all in on whichever unleaded fuel reaches Arlington first. UL94 is approved for my engine today, and the day a truck of it parks on the ramp I will fill up and pay the premium without a second thought. If it is 100R or G100UL instead, I will be at the front of the line the day Rotax adds it to the fluids bulletin, and I will be writing to Rotax about that bulletin in the meantime. I did not build a modern airplane to run it on a 1944 fuel. I run it on that fuel because, in 2026, in Washington State, nobody has offered me anything else.

Part 1 asked what a Rotax drinks around the world. Part 2 asked why it drinks lead here. The answer, in a sentence: because for thirty years the people who could have changed it found it easier not to, and nothing in the American system ever made them.


Leave a Reply