A lot of times fuel varied. It was too much trouble to blend
exactly to 68, or 75 or 80 octane. If they had a batch of fuel that tested to 70 octane do they try to cut it with something to knock it back to 68 or do they try to soup it up to hit 75 or do they just label it as 68 and pour it into the tanks for shipment? And move on to the next batch of 10s of thousands of gallons.
For aviation fuel the specs often said what the
upper limits of some things, like lead, could be. Like 4.6cc per US gallon was an upper limit. If refinery X hit the 100/130 performance limits with a batch of fuel that only needed 4.1CCs fine. mix up the batch, retest and ship it and see what the next batch needs.
You don't get
exactly the same crude oil even out of adjacent wells in some oil fields. Very close but not exact, now pump a number of wells into one holding tank, repump into tank cars or pipelines, then put it into ships and so one before you even get to the refinery.
As far as the octane 'scale' goes. It stops at 100. there is no 102 or 108 or other number this is despite what gasoline company X says on their pumps/advertising. You can say 100+ or 100 plus or 100>. You compare test fuel to 100 octane reference fuel. if the test fuel will tolerate higher compression in the test engine than the 100 octane reference fuel you can say the fuel being tested is greater than 100 octane. You can not say how much better unless you have a known reference fuel to test against, in the same engine on the same day (or in standard air).
Turned out that the octane scale is not liner. The difference between 99 and 100 is a lot different than the difference between 89 and 90 and both are different than the difference between 79 and 80.
Anybody who wants to argue this can argue with Sam Heron, the inventor/developer of the PN scale and since he died a long time ago........................................
Next problem is trying to figure out how to measure above 100 octane. Some people tried using 100 octane plus 1.0 cc lead and then 2 ccs lead and higher. Trouble is that you get different results from each additional CC of lead (diminishing returns) and you get different results from the same amount of lead in different types/batches of fuel. Cat cracked gas reacts much differently to lead than straight run gas. Two different slopes that actually cross over.
Same with some of the stuff used to boost the knock ratings. It may or may not act in liner fashion. It also may react differently with different base stocks. There can be up to 400 different compounds in gasoline, the vast majority it trace amounts.
But since the Allies allowed
up to 20% aromatics in different amounts, in fact the British wanted at least 20% in 1939-40 things get complicated real quick. They knew what they wanted (rich mixture response) but they didn't know how to measure it/specify it. Just specifying equal to reference fuel 100 octane (lab grade) plus 4 CCs of lead wasn't going to actually work.
They had to come up with test procedures for rich mixture response and then get a number of standardized engines (made by
Waukesha? ) sent to the test labs in both the refineries, and purchasing agencies and research labs and engine companies so everybody was testing the same thing/s. Not just something close.