Why Does Adding A Little Water To Diesel Fuel Cut Pollution?

Blending tiny water droplets, held in a stable mix with a soap-like additive, into diesel fuel attacks both of diesel's worst pollutants at once, and needs no engine changes. In laboratory engines the best results reach 67% less nitrogen oxide and 68% less particulate matter, though those two figures come from different studies running different amounts of water, not one engine doing both, per a 2025 peer-reviewed review in the journal Carbon Research. On real fleets the numbers were smaller: California's Air Resources Board verified 16% for NOx and 60% for particulates on the French fuel Aquazole, and 15% for NOx and at least 50% for particulates on Lubrizol's PuriNOx. Inside the cylinder, each trapped water droplet flash-boils into steam and bursts its own fuel droplet apart, a "micro-explosion" that sprays the diesel into a finer, cooler-burning mist. Both fuels have since left the market, beaten by exhaust cleanup hardware that handles the same two pollutants without a separate fuel.

Every liquid you own comes with the same unwritten rule: keep water out of it. Water ruins gasoline. Water ruins motor oil. Drip water into a deep fryer and it throws burning oil across the kitchen. So here is a strange fact. Some diesel engines run cleaner on fuel that has water mixed into it on purpose.

Not water that leaked in by accident. Water a chemist measured out and locked in place, using the same trick that keeps salad dressing from separating. Reviewed studies say this one change, no new engine, no new injectors, can cut a diesel engine's smog-forming nitrogen oxides by up to 67%. Particulate matter drops by up to 68%. Those headline numbers come from a 2025 review in the peer-reviewed journal Carbon Research, and they arrive with a condition worth stating before anything else. They are laboratory ceilings, pulled from different studies running different amounts of water, not one engine hitting both at once. When regulators later measured the fuel on working vehicles, they got a good deal less. We will come back to that.

Those are strange numbers. Water does not burn. Some engineers try the opposite trick, splitting water apart to run a hydrogen fuel cell. Here the water stays whole. It never burns at all. So how does adding it to diesel make the diesel burn cleaner?

How Does A Diesel Engine Actually Burn Its Fuel?

A gasoline engine needs a spark. A diesel engine does not.

A diesel piston squeezes the air in its cylinder down hard, into roughly a fifteenth to a twentieth of its starting size. Squeezing a gas that hard heats it up. It is the same reason a bicycle pump warms in your hand. By the top of that squeeze, the air in a diesel cylinder typically sits somewhere around 700 to 900 °C (1,290 to 1,650 °F). That is hot enough that diesel fuel ignites the instant it touches it. No spark plug needed. This is the whole trick of a compression-ignition engine.

It is also why diesel trucks rumble instead of purr, though not quite for the reason most people assume. The fuel does not light the instant it arrives. It takes a brief pause to heat up and start reacting, an ignition delay, and during that pause more fuel keeps piling in. When the pile finally catches, cylinder pressure jumps in one steep step. That jump is the clatter. Hold on to it, because water is about to make the pause longer.

Here is the detail that matters for everything ahead. Diesel is sprayed in as a fine mist of liquid droplets, not mixed with air beforehand like in a gasoline engine. Each droplet has to evaporate and burn on its own, in the split second before the piston starts back down. And a droplet does not burn evenly. It burns as a thin shell of flame wrapped around a hidden core. That structure is where diesel's two signature pollutants come from.

Why Does Diesel Exhaust Contain So Much NOx And Soot?

Picture that burning droplet again: a hot outer flame wrapped around a starved inner core. It explains both pollutants at once.

Start with the core. At the droplet's center, fuel vapor packs in thicker than the air around it can supply oxygen for. Chemists call this a fuel-rich zone. In plain terms, there is too much fuel and not enough air to go around. Some of the fuel's carbon atoms never fully burn. Instead they get cooked, or pyrolyzed in the jargon, into tiny black flecks: soot. A soot fleck that finds air later in the burn can still burn away. One that does not survive exits the tailpipe as smoke. That smoke is fine-particle pollution the EPA links to asthma and heart disease.

Now the flame front. This is where nitrogen oxides, NOx for short, come from. The cause is simple: heat. Ordinary air is about 78% nitrogen, a gas so calm it sits inside every breath you take doing nothing at all. Push its temperature above roughly 1,800 kelvin (2,780 °F), though, and that same nitrogen starts tearing apart. It joins with oxygen to form NO and NO₂. The pathway even has a name: the Zeldovich mechanism, after the physicist who published it in 1946. Below that threshold, almost none forms. Above it, a flame only a little hotter makes far more NOx.

So a diesel engine is stuck between two problems that pull in opposite directions. Cool the flame to cut NOx, and the fuel mixes with air more slowly, making more soot. Push for a hotter, faster-mixing flame to cut soot, and NOx climbs instead. For decades, that trade-off looked close to unavoidable.

Wait, Isn't Water Supposed To Ruin Diesel Engines?

A quick detour first. Search "water in diesel fuel" and you mostly get horror stories. They are not wrong.

Water that sneaks into a fuel tank by accident is a real problem. It condenses out of humid air overnight, or seeps past a worn seal. It sinks to the bottom of the tank, since water is denser than diesel. It corrodes fuel-system metal. It also feeds bacteria that grow into a slimy sludge. That sludge is a well-known cause of overloaded filters and plugged injectors, and standing water in a storage tank is what lets it get started. None of that is a myth. A dashboard warning about water in the fuel means bad news, full stop.

The technology in this article is the opposite, in every way that matters. It is not a puddle of free water at the bottom of a tank. It is water broken into droplets a few microns wide (a human hair runs about 70 microns). Those droplets stay in permanent suspension through the fuel. The trick is a surfactant, the same class of chemical that keeps oil and vinegar mixed in a salad dressing. Chemists call the result an emulsion. How long one holds together depends entirely on the recipe. The best case in the review is about 60 days, and that was for blends of 5% and 20% water carrying 3% of a surfactant called Triton X-100. Another study managed 5 days, stretching to 8 when the surfactant dose rose from 3% to 4%. Ordinary diesel sits happily in a tank for months, so even the good case is a shelf-life problem rather than a selling point. Still, it is built to be there. One difference separates an engine killer from an emissions fix: puddle by accident, or mist by design.

How Do You Actually Mix Water Into Diesel Fuel?

Water and diesel do not want to mix. Left alone, they act like oil and vinegar. Shake them and they separate again within minutes, water sinking to the bottom.

A surfactant molecule fixes this the same way dish soap cuts through a greasy pan. One end of the molecule is drawn to water. The other end is drawn to oil. So it plants itself right at the boundary between the two, like a tiny two-sided anchor. Coat enough water droplets in enough surfactant. Each one gets wrapped in a shell. That shell stops it from merging with its neighbors, or sinking to the bottom of the tank. Blending two or more surfactants, rather than using just one, tends to work best.

The finished fuel looks, to the eye, like ordinary diesel. Under a microscope, it is millions of tiny water droplets, each one wrapped and suspended in a sea of diesel. The engine needs no changes to accept it. It goes into the same tank, the same fuel line, the same injectors, as regular diesel.

What Happens Inside The Cylinder When The Water Flashes To Steam?

This is where the payoff lands, in a fraction of a second, inside every fuel droplet.

Picture that same fuel droplet from earlier, injected into cylinder air at several hundred degrees. But now it is not pure diesel. It has water micro-droplets locked inside it. The outer surface heats up and starts to burn. The water trapped inside heats up too. And here the everyday number for water quietly stops applying.

Water boils at 100 °C (212 °F) in your kettle, at one atmosphere. Near the top of a diesel compression stroke the pressure is more like 40 to 100 atmospheres, and at those pressures water does not boil until roughly 250 to 310 °C (480 to 590 °F). So the trapped water sails past every temperature you associate with steam and stays stubbornly liquid.

Then it goes further still. It superheats, climbing as much as another 30 °C above even that raised boiling point, which can put it near 330 °C (630 °F) while remaining a liquid. Two things allow that. Diesel boils at a higher temperature than water does, so the oil wrapped around each water droplet is still liquid and still heating. And the inside of a droplet that clean offers a bubble nothing to grow on, none of the scratches and specks that a kettle wall provides. A liquid in that situation can be pushed startlingly far past boiling. Measured superheat limits for pure substances tend to land around 88 to 90% of the critical temperature, which for water works out near 300 °C.

Superheated water is a coiled spring. The instant it finds a weak spot, it does not simmer. It flashes to steam almost all at once. Inside a cylinder at tens of atmospheres it swells some fifteen to thirty times over in a fraction of a millisecond, and at the pressure of an open kitchen it would be more like 1,700. That growth bursts the diesel shell from the inside out. Engineers call the event a micro-explosion. "Explosion" is not a stretch here. It is a fair name for a fast, tiny steam blast inside a droplet smaller than a human hair.

One larger fuel droplet becomes a fine spray of smaller ones. Combustion engineers call this step secondary atomization. Smaller droplets are exactly what a diesel flame wants. More surface area means faster, more even mixing with air. That shrinks the oxygen-starved core where soot forms. It also leaves the flame a touch cooler, which turns out to be the more interesting half of the story. Somewhere, an engineer had to work out how hot a trapped water droplet must get before it bursts. Then go measure it. That is a specific job to have.

Why Does This Actually Cut Both NOx And Soot At Once?

Go back to the two-pollutant problem: cool the flame and soot rises, heat the flame and NOx rises. The micro-explosion is the loophole, because it never touches the temperature dial directly. It changes how the fuel is delivered.

Finer atomization means fuel and air mix faster and more evenly. That alone shrinks the oxygen-starved core where soot forms. Water helps on the soot side a second way too, one the popular version of this story usually drops: it raises the supply of reactive fragments called OH radicals, which are unusually good at burning soot back off after it has formed.

On the NOx side, the tempting explanation is the heat soaked up by boiling the water away, its latent heat of vaporization. Run the number, though, and it is thin. In a fuel that is 13% water, boiling that water off costs under 1% of the energy the diesel releases. Add the heat needed to warm the water up first and then superheat the steam afterwards and you reach maybe 2%. That is not nothing, but it is not the main event either.

The bigger effect is that the steam is simply there. It takes up room, absorbs heat all the way through the burn, and dilutes the combustion gases, the way an extra body in a small room takes the edge off a fire. And a flame running even a little cooler sits lower on the steep part of the Zeldovich curve, so its NOx output drops out of all proportion to how small the temperature dip looks.

So one droplet, superheating and bursting, works on the mixing problem behind soot and the heat problem behind NOx at the same time. That is a genuinely elegant piece of engineering, and it is not a compromise between the two.

So What Is The Catch?

There is one, and the cheerful version of this story tends to skip it.

Start with the obvious. Water carries no energy at all. Blend 13% of it into a tank and you have handed the engine something that will never burn, so it has to swallow more total fuel to do the same work. Most studies find that brake thermal efficiency, the share of fuel energy that actually reaches the crankshaft, holds up or even improves slightly at modest water contents, helped along by the better mixing. Others find the opposite, and not by a little: one 2021 study measured plain diesel at its best at 26.99% efficiency, while a 10% water blend managed 7.19% at light load. The direction is not settled.

Then there is that clatter from earlier. Water lengthens the ignition delay, so more fuel piles up before the burn begins, and the pressure jump when it finally does is steeper. Measured at full load, one emulsion peaked at 10.11 bar per degree of crank rotation against 8.44 for plain diesel. Push the water proportion up to 20% and researchers report severe knock, especially under load. The fuel needs no engine changes to run, which is true and remarkable, but running harsher is not the same as running identically.

And two pollutants tend to move the wrong way. Carbon monoxide and unburned hydrocarbons repeatedly come out higher than with plain diesel, because a cooler, wetter burn is a less complete one. The evidence here is genuinely mixed, and a few studies report the opposite, which is why the review calls the CO and hydrocarbon reports inconsistent. But nobody claims water fixes those two the way it fixes NOx and soot.

None of that sinks the idea. It does mean "cuts pollution" is shorthand for "cuts the two pollutants diesel is notorious for, while nudging up two that get less attention."

Has Anyone Actually Put This On The Road?

This was never a lab curiosity dressed up for a headline. Water-in-diesel fuel was sold at scale, under its own brand names, and carried millions of kilometers of commuters who never noticed.

The French energy company Elf began building a water-diesel fuel called Aquazole in 1997. Elf later folded into TotalEnergies. The target was city buses and delivery trucks. The recipe was roughly 85% diesel, 13% water and 2 to 3% additives, which is worth pausing on, because "a little water" turns out to mean rather more than most people picture. By January 2000, over 1,000 vehicles ran on it daily, covering some 3 million kilometers (1.86 million miles) a month across city fleets. No engine changes were needed. NOx dropped up to 30%. Visible black smoke dropped up to 80%.

The same paper is candid about the cost: the water content causes a certain loss in engine performance. It also reports hydrocarbon consumption falling by up to 4%, which is easy to misread as free money. It is not. That is the diesel fraction going down, and of course it does, because an eighth of what is in the tank is water. Total fuel through the injectors goes up.

Then regulators measured it, and their numbers are the ones worth remembering. California's Air Resources Board verified Aquazole at 16% for NOx and 60% for particulates. It verified Lubrizol's rival fuel, PuriNOx, at 15% for NOx and at least 50% for particulates, for 1988 to 2003 model-year engines. Set those against the 67% headline and you have the whole story of a laboratory best case meeting a real duty cycle. The board also attached a caveat that is easy to skim past: its verifications do not address what emulsified fuels do to engine durability or performance.

Ride a city bus in the early 2000s in the right pilot city, and there is a real chance you rode the experiment. It sat alongside other fleet-scale attempts to clean up diesel. Some fleets tried running trucks on vegetable oil instead. Others swapped in CNG or LPG.

So where is it now? Gone, on both counts. Lubrizol announced it would stop making PuriNOx back in 2006, and its verification covers a model-year range that ended in 2003. Aquazole is no longer sold either. The reason is the quiet irony at the end of this story: while chemists were perfecting the fuel, engineers were solving the same problem at the other end of the engine. Selective catalytic reduction sprays a urea solution into the exhaust and converts NOx into harmless nitrogen and water. A diesel particulate filter traps soot and periodically burns it off. Together they clean up both pollutants without asking anyone to build a second fuel supply chain, keep an emulsion from separating, or accept a shelf life measured in weeks. Water-in-diesel did not fail on physics. It lost on logistics.

So, Why Does Adding A Little Water To Diesel Fuel Cut Pollution?

Diesel exhaust's two worst pollutants are choking soot and smog-forming NOx. Both come from opposite ends of one burning droplet. A well-hidden droplet of water turns out to reach both ends at once.

A diesel engine ignites fuel with heat alone, no spark. So every drop of fuel burns as its own small, imperfect event. A starved core leaves soot behind. It sits wrapped in a flame hot enough to force nitrogen and oxygen into NOx. Lock a microscopic droplet of water inside that fuel droplet. Hold it there with a surfactant so it survives the trip from tank to injector. The cylinder's own heat does the rest. Sealed in oil that boils higher than it does, with nothing inside to nucleate on, the water climbs far past any temperature you would call boiling and stays liquid anyway. Then it flashes, and bursts its shell in a tiny, fast micro-explosion. That shatters the fuel into a finer mist. The mist mixes faster, so less soot. The steam dilutes and cools the burn, so less NOx.

None of it needed a new engine, a new fuel line, or a new injector. It needed someone to understand exactly what a burning diesel droplet looks like from the inside, then hide one old substance exactly where it would help most. That the idea has since been retired takes nothing away from it. The fuel worked. It was simply out-competed by a catalyst and a filter bolted onto the exhaust pipe, which is an extremely boring way to lose. Water has been ruining fuel by accident for as long as there have been fuel tanks. It took real chemistry to make it help on purpose, even for a while.