The Verification Venue · a thing everyone gets wrong
The Number That Measures a Refusal
Ask what octane means and you'll be told it's the fuel's power, or its quality, or how much energy it packs. It is none of those. The octane number measures one thing: how stubbornly a fuel refuses to explode on its own when squeezed — its resistance to knock. It is a ceiling, not a quality dial.
Here is the whole illusion in one instrument. Pick a fuel. Drag how hard the engine squeezes the charge — its compression ratio. Two things move: the engine's ideal efficiency (which is why anyone wants to squeeze harder at all), and a knock wall for the fuel you chose. Below the wall, swapping to a pricier fuel changes the efficiency by exactly nothing. Push past the wall on cheap fuel and the unburned gas detonates. Octane is not about the fuel you have — it's about the engine you're allowed to build.
Pick a fuel — the number on the pump, plus what's under it
Ideal thermal efficiency
60.2%
air-standard Otto, γ = 1.4
End-gas squeezed to
543 °C
motored, before the spark
Compression 10.0:1 on this fuel: runs clean. Ideal efficiency 60.2%.
How many times the piston squeezes the fuel–air charge before the spark. Higher is more efficient and more powerful — and closer to the knock wall.
The number that just went up on the left is the reason compression exists at all. The ideal thermal efficiency of the four-stroke (Otto) cycle depends on the compression ratio r and nothing else about the fuel:
Read that carefully: octane isn't in the formula. A higher-octane fuel does not make this engine one percent more efficient — η depends only on how hard you compress. What octane does is set how far right you are allowed to drag the slider before the unburned charge stops waiting for the flame and autoignites on its own. That is the wall on the chart. Cheap fuel, low wall; expensive fuel, high wall. The whole value of premium is that it lets you build the engine that lives further up this curve — not that it improves the fuel you pour into an engine that was never going to knock.
Why "octane" is really two numbers
A fuel doesn't have an octane. Knock resistance depends on the conditions, so every fuel is rated twice on the same 0–100 scale (n‑heptane = 0, iso‑octane = 100, the reference pair Graham Edgar fixed in 1927): RON, the Research number, measured at a gentle 600 rpm; and MON, the Motor number, measured harsher — 900 rpm, preheated intake, variable timing. For essentially all pump gasoline MON runs lower — typically 8–12 points; that gap is the fuel's sensitivity. (On the pure reference chemicals the two numbers are equal — sensitivity zero — so this is a property of real fuel, not an absolute.) The US pump prints their average, the Anti-Knock Index, AKI = (RON+MON)/2, while Europe posts RON alone. That convention explains only about half the gap between American regular's 87 and European regular's 95: US 87 AKI corresponds to roughly 91 RON (RON − AKI ≈ sensitivity/2 ≈ 4 points). The other ~4 points are real: Europe's 95‑RON regular is genuinely a few octane numbers higher than US regular — about US 90–91 AKI, what America sells as midgrade — so it is not the same fuel wearing a different label.
| fuel | RON | MON | sensitivity | AKI = (R+M)/2 | pump says |
|---|
The green column is recomputed from the two numbers beside it — it is the cross-check, not a claim. "Octane" as a single figure is a convention laid over a pair; the pair is the real object.
The check — every number recomputed in front of you
What the instrument is doing
Efficiency and end-gas temperature at the four representative knock ceilings, computed live from the exact laws:
| at ceiling r | η = 1−r^(1−γ) | T_end = 325·r^(γ−1) |
|---|
Climbing from the regular ceiling (~9.5:1) to the race ceiling (~12.5:1) buys +4.2 points of ideal efficiency (59.4% → 63.6%). That gain is the entire reason octane matters.
Octane is not energy. The two reference chemicals sit at opposite ends of the scale yet carry nearly identical energy: iso-octane (octane 100) ≈ 44.3 MJ/kg, n-heptane (octane 0) ≈ 44.6 MJ/kg — a difference of 0.67%, and the low-octane one is very slightly higher. Buying a bigger number does not buy more energy.
The scale is a definition, not a measurement: an octane-90 primary reference fuel is 90% iso-octane and 10% n-heptane by volume. Run it yourself: node research/what-octane-really-means/verify.mjs.
The lead epilogue: how modern octane got its chemistry
For fifty years the cheapest way to buy octane came in a can. In 1921 Thomas Midgley Jr., working under Charles Kettering at General Motors' Dayton lab, found that a trace of tetraethyl lead silenced knock. It was sold as "Ethyl" — the word lead left off the label deliberately. The cost was real and documented: at Standard Oil's Bayway refinery in October 1924, a lead-processing disaster killed 5 workers and left about 35 more with severe neurological injury; New York City, Philadelphia, Pittsburgh and New Jersey banned the fuel within days.
TEL found to suppress knock (Midgley, under Kettering, GM Dayton). Marketed as "Ethyl."
Bayway refinery disaster: 5 dead, ~35 severely injured. Several cities ban leaded fuel within days.
Catalytic converters — poisoned by lead — force the phase-out. Octane must be rebuilt without TEL.
Octane comes from catalytic reforming, aromatics, and ethanol (blends in at a RON around 108–109). Modern octane chemistry is the shadow of a public-health reversal.
One honesty note built into that story: the lead came out because it poisoned catalytic converters, not because higher octane is cleaner. Octane rates knock resistance — nothing about it, by itself, makes a fuel burn greener.
What's exactly true, what's idealised, and where the popular story runs ahead
Exactly true. Octane number measures resistance to knock (autoignition of the unburned end-gas), not energy or power. The scale is defined: n-heptane = 0, iso-octane = 100, an octane-N reference fuel is N% iso-octane by volume (Edgar, 1927). RON is measured at 600 rpm, MON at 900 rpm with preheated intake and variable timing; MON runs 8–12 numbers lower and that gap is sensitivity; AKI = (RON+MON)/2 is the US pump number. The efficiency law η = 1−r^(1−γ) and the compression temperature T_end = T_intake·r^(γ−1) are the exact air-standard results, and both are monotone in r.
Knock is NOT pre-ignition — this is the trap. Knock (detonation) is the end-gas autoigniting after the spark, ahead of the normal flame front, ringing the block at ~5–15 kHz and tending to shatter pistons and bearings. Pre-ignition is a hot spot lighting the charge before the spark — a different cause, and it tends to melt parts instead. Octane rates knock, the first one. (Modern turbo-GDI "LSPI" is yet a third, distinct effect.)
Idealised. The efficiency number is the ideal air-standard Otto value; real gasoline engines reach roughly 30–42%, about half of it, because of heat loss, friction, throttling, incomplete combustion and a lower effective γ. What is real and load-bearing is the direction — efficiency and power density rise with compression ratio — which is exactly why manufacturers chase it. T_end is the motored (no-combustion) adiabatic temperature from a representative T_intake = 325 K; hotter end-gas autoignites sooner, but knock is set by whether that chemistry finishes before the flame arrives (an ignition-delay race), not by crossing a fixed temperature — so don't read T_end as an ignition threshold.
Representative, not universal. The four fuels' RON/MON pairs and the compression ceilings (RON 91 → ~9.5:1 up to RON 100 → ~12.5:1) are typical values for a conventional port-injection naturally-aspirated engine, not physical constants. Better engine design moves the wall on the same fuel: with direct injection, cooled intake and knock sensors, Mazda's Skyactiv-G runs 13:1 on US regular 87 (the 14:1 variant ships only in markets with higher-octane fuel and needs ~91+ RON). The RON↔AKI offset likewise isn't a fixed constant; it depends on the fuel's sensitivity.
Required vs recommended — the "premium is a waste" line, honestly. If a car requires premium (high compression, turbocharged, timing calibrated for it), it genuinely needs it; regular will knock and the sensor can only retard timing so far, costing power. If a car merely recommends premium, regular won't damage it — the knock sensor pulls timing — but it does lose measurable power and efficiency. "Premium is always a waste" is as wrong as "premium is always better."
Energy, precisely. The defensible statement is that octane is uncorrelated with energy content — not that premium has less (nor more). The two reference chemicals differ by ~0.7% in specific energy, within scatter; real pump grades differ only marginally and inconsistently, mostly from ethanol and aromatics, not from the octane number itself.