The most accurate way of testing the actual octane value is to use a Combustion Fuel Research (CFR) engine (also known as a knock engine). Currently the American Society for Testing and Materials (ASTM) uses the CFR engine method, developed in the 1920s to obtain official octane measurements. The CFR engine measures octane by combusting the fuel and physically measuring the knock that occurs. These tests have a repeatability (same operator/same lab) of 0.2 RON, and a reproducibility (different operators in different labs) of 0.7 RON. Each sample is tested for research octane number (RON) as prescribed by ASTM Method D2699. This is the method used for testing at fuel refineries. This is also the method used for all our testing, which makes them certified and independent tests that can be reproduced by anyone wanting to confirm the results.

About the knock engine

CFR (Knock) Engine

The CFR engine is basically a carburetted, single-cylinder, variable-compression engine. The head can be raised and lowered to change the compression ratio and thus increase knock intensity. By reading the knock intensity at a given compression ratio, the operator can determine the octane rating of a fuel sample. The engine has to be warmed up to maintain a 38°C to 54°C oil temperature and a 2-3Hg manifold vacuum. Air/fuel ratio is held at an elevation-corrected constant. Prior to every test, both a toluene and an iso-octane mixture of known octane are run through the engine for reference and calibration checks. The RON test is performed at 600RPM with intake and air/fuel charge temperatures regulated at 52°C. Ignition timing is held at 13°BTDC.

Why not test on a dyno or a ¼-mile?

Methods for testing fuels incorporating a chassis dyno or a ¼-mile run are seriously flawed because each uses horsepower changes to correlate to octane changes.

The problem with measuring power is that it is a function of the combustion process, which varies with ambient conditions, spark timing and air/fuel ratio; all of which are continuously varying and often computer controlled. The measurement method also isn't precise enough to establish an accurate estimation of octane gains. Most chassis dynos will have a range of variability of up to 5HP. This is enough to cover most octane boosters' gains. Though differences are measured, the amount of variability and lack of experimental controls prevent a reliable margin of repeatability. These tests also ignore the fact that to get the most out of higher octane, the engine needs to be tuned for the fuel.

On the ¼-mile even more variables are introduced. Included in these are driver error, driver bias and changing traction from run to run. These additional variables make this type of test an even worse gauge of the gains achieved by the higher octane fuel being tested.