Fuel additives: which ones work and which ones sell hope

“Do fuel additives do anything” usually gets one of two extreme answers: “pure marketing” or “I use them and everything’s great”. Neither is very useful.

Let us go by category — some additives have a clear physical mechanism, some work narrowly, and some cannot work in principle.

What works: water absorbers

The simplest and most honest category. Alcohol that binds water in the tank and lets it pass through the system with the fuel instead of collecting into droplets.

When it makes sense: before winter, after long storage, when condensation in the tank is suspected.

Limitation: it works with small quantities of water. If litres went in, no additive helps — the tank needs draining.

On a diesel, be careful. Water in diesel is more dangerous, and a triggered water-in-fuel sensor means “do not drive”, not “add a bottle”. Covered in the bad fuel guide.

What works narrowly: injector cleaners

Here the mechanism is real, but expectations are usually inflated.

What they genuinely do: quality PEA-based detergents gradually dissolve deposits on injector nozzles. The effect is cumulative, over several tanks.

What they do not do:

• They do not clean intake valves on direct-injection engines. Fuel never physically reaches them — it goes straight into the cylinder. That is the central misunderstanding in this topic; mechanics in the GDI carbon guide.
• They do not restore a worn nozzle. Mechanical wear is not cured by chemistry.
• They do not save a dead injector. A code from P0201–P0206 is an electrical circuit, not contamination — see its own article.

How to verify it worked: compare long-term fuel trim before and after several tanks. Genuine improvement shows as a shift toward zero. Reading trims is in STFT and LTFT.

What works conditionally: particulate filter treatments

A category where everything depends on correct application.

Mechanism: the additive lowers the temperature at which soot begins to burn, helping regeneration complete when the exhaust is not hot enough.

When it helps: urban short-trip use, filter loading raised but not critical.

When it does not: the filter is clogged with ash. Ash is residue from engine oil additives; it does not burn at any temperature, and no chemistry removes it. The soot-versus-ash distinction is in the GPF guide and the DPF guide.

The key point: an additive does not replace twenty minutes of highway driving every couple of weeks. It is a supplement, not an alternative.

What works within limits: octane boosters

The mechanism is real, but the scale is small. A consumer bottle raises the octane rating by fractions of a point, not from 91 to 95.

When it makes sense: you have filled with lower octane than required and want a gentler drive home. But adding correct-octane fuel to the brim is cheaper and more effective — see the misfuelling article.

How to verify: knock correction in live data. If it has stopped permanently retarding timing under load, something worked. Mechanics in the knock sensor guide.

What does not work: “catalyst revivers”

The category with the loudest promises and the weakest mechanism.

A catalyst dies two ways: sintering of the ceramic block from overheating, and poisoning of the active layer. Both are irreversible at the material level — pouring something into the tank that rebuilds the block’s structure is not possible.

If you have P0420, the correct order is checking the upstream oxygen sensor, exhaust leaks, and whether oil is reaching the exhaust. Covered in the P0420 guide. An additive is not on that list.

What additives never do

Briefly and directly:

• They do not add power. Power comes from cylinder filling and timing, not fuel chemistry.
• They do not reduce consumption by themselves. Fixing a fault reduces consumption; a bottle does not.
• They do not cure mechanical wear.
• They do not replace oil and filter changes.

How to verify results rather than feelings

The main problem with additives is the placebo effect. Objective verification is straightforward:

1. Before: record long-term fuel trim at idle and at 2500 RPM, consumption by the “full tank, 250 miles, full tank” method, and knock correction under load.
2. Add the product, run two tanks.
3. After: the same three parameters under the same conditions.

Compare like with like — the factors that produce up to a third of the difference on their own are covered in the fuel consumption article.

In short

• Water absorbers work with small quantities; a triggered diesel water sensor means stop, not top up.
• Detergents clean injectors but not intake valves on direct injection.
• Particulate filter treatments help with soot and are useless against ash.
• Octane boosters give fractions of a point — adding correct fuel is simpler.
• “Catalyst revivers” have no physical mechanism.

Download the app — fuel trims, knock correction, and instant consumption are read with an ELM327 adapter over Bluetooth SPP on Android.