Premium is harder to burn, not stronger
Premium gasoline has no more energy in it than regular. What you are paying for is fuel that is harder to set on fire. The octane number is not a measure of power, it is a score for how well a fuel resists lighting itself before the flame front reaches it, measured against blends of two reference molecules in a test engine. If your engine was not built to squeeze harder, the FTC's position is that premium buys you no more power and no more mileage.
Out now
Try Reflux Student free on Aspen Plus
Reflux Student drives your own Aspen Plus V14 from plain English. You type what you want, it opens the case, makes the change in Aspen, runs it, reads the result back and tells you what it verified. Aspen does the math. You keep the judgment. Windows, your own licence, and three free runs to start: it is a trial with a limit on it, not a free product.
Windows SmartScreen will warn you on first run: the installer is new rather than unknown. Click More info, then Run anyway.
Not on Aspen Plus? Tell us what you run and we will let you know when your simulator is supported.
The sticker What knock actually is The two tests The blend Is premium worth it Why this is a model
The number on the pump is an average of two numbers you never see
A North American pump posts the Anti-Knock Index, and the sticker says so in small print: (R+M)/2. R is the Research Octane Number and M is the Motor Octane Number, and they come from the same engine run two different ways. The gap between them is called the fuel's sensitivity, and on pump gasoline it is usually eight to twelve points.
Europe posts R alone. That is the entire reason a European 95 is not better fuel than what you buy here: a 95 RON with a sensitivity of 10 has a MON of 85, and (95 + 85) / 2 is 90.
What knock actually is
The spark lights one flame. That flame sweeps across the chamber, and the gas it has already burned expands behind it, which compresses whatever is still unburnt ahead of it. The last pocket to be reached, the end gas, gets squeezed and heated hardest of all.
If that pocket reaches its own autoignition temperature before the flame front arrives, it lights itself. Now there are two flame fronts in one chamber travelling at each other, and when they meet the pressure does not just spike, it rings, at several kilohertz. That ringing is what scrubs the boundary layer off the piston crown and breaks ring lands. Octane is the fuel's resistance to that, and nothing else.
The scale is defined by two molecules. Isooctane resists so well it is given 100. n-heptane resists so badly it is given 0. Your fuel is run against blends of the two until the test engine knocks the same amount, and the percentage of isooctane in the blend that matches it is the octane number.
The two tests, and the sticker between them
Both numbers come from a CFR engine, a single cylinder whose cylinder can be raised and lowered while it runs so the compression ratio is a dial. RON is the gentle test: 600 rpm, cool intake air. MON is the hot and fast one: 900 rpm, the mixture preheated to 149 degrees C, and the spark advanced. The hot test is harder, so it comes out lower.
The sticker, from the two lab numbers
Drag either number. The arithmetic is the whole definition: there is nothing fitted here.
A sensitivity far outside eight to twelve is unusual for a finished pump gasoline. High-sensitivity blends are common in components, not in the pool.
No refinery has a tank of 87
A refinery does not make 87 octane gasoline. It makes several streams of very different octane and blends them to a number. A reformer turns low-octane naphtha into aromatics and comes out near 100. An alkylation unit takes light olefins and isobutane and makes a branched, low-vapour pressure blendstock near 97. An isomerization unit rearranges light straight-chain paraffins and comes out around 83. Ethanol is bought, and neat it is about 109.
Try to hit 87
By volume, normalised to 100 percent of the pool, so moving one slider moves the others. This is the straight volumetric blend.
The reel names four streams. A fifth is here because the four on their own cannot make 87 without about three quarters of the pool being isomerate, which no refinery does. The high-octane streams are blended INTO something, and that something is low-octane straight-run naphtha off the crude tower at about 78. Pull it to zero and watch the number run away from you: that is the real constraint a blender is working against.
This is the linear blend, and it is deliberately not corrected, because the correction is the point. Octane does not add up by volume. Ethanol in particular behaves as though it were worth considerably more than its neat 109 when it is a small part of a hydrocarbon pool, and how much more depends on what it is blended into. That is why the recipe is an optimisation solved against measured blending values rather than a weighted average, and why a blender runs it again when the streams move.
Octane is not the only constraint, which is what turns the recipe into a real optimisation rather than a sum. The blend has to clear the octane floor, stay under a vapour pressure ceiling that changes with the season and the region, and stay under a benzene cap. Reformate is the stream that carries both the octane and the benzene, so those two limits pull against each other on the same slider.
Before 1996 there was a shortcut. Tetraethyllead did the same job for a fraction of a percent of the volume, and when leaded gasoline was banned for on-road use the octane had to come from somewhere else. The reformers and the alkylation units took its job, which is why a modern refinery is shaped the way it is.
Why this is a model, and why it is run daily
Every number on this page moves. Stream octanes move with the feed and with how hard the units are being pushed. The vapour pressure limit moves with the season. Ethanol's contribution moves with what it is blended into. The economics move with what each stream is worth somewhere else in the refinery, because a barrel of alkylate put into regular is a barrel that did not go into premium.
So the blend is not a recipe written down once. It is an optimisation re-solved against today's streams, today's limits and today's prices, thousands of barrels a day, and the answer changes. That is the layer being built at Reflux: the model that knows which constraint it is actually against, checks it, and tells you what it verified.
Run it on your own blend
Reflux Student drives your Aspen Plus V14 from plain English: open the case, change a stream's octane or a cut point, re-run, and read back what moved. Three free runs, Windows, your own licence.
Sources: the (R+M)/2 method on the pump, grades and the role of ethanol from the US Energy Information Administration's octane explainer; RON and MON test conditions from ASTM D2699 and D2700, which set 600 rpm and 900 rpm and the 149 degrees C intake mixture for the Motor method; blendstock octanes for reformate, alkylate and isomerate from Gary and Handwerk, Petroleum Refining: Technology and Economics, and the gasoline blending linear program from the same; ethanol's neat RON of 109 and its non-linear blending behaviour from Frontiers in Mechanical Engineering; the position that high octane gives no more power or mileage in an engine not built for it from the FTC and from fueleconomy.gov; and the 1 January 1996 ban on leaded gasoline for on-road use from the EPA. The sensitivity band quoted here is typical of finished pump gasoline rather than a specification.
← Back to Reflux